Publicly accessible
URL: http://nzetc.victoria.ac.nz/collections.html
Copyright 2009, by Victoria University of Wellington
Prepared for the
All unambiguous end-of-line hyphens have been removed, and the trailing part of a word has been joined to the preceding line. Every effort has been made to preserve the Māori macron using unicode.
Some keywords in the header are a local Electronic Text Collection scheme to aid in establishing analytical groupings.
Samoa is a group of islands in the
Up to the year 1900, the affairs of the whole group were jointly administered by the three powers,
American Samoa consists of the large island of
The Samoan population of
The speech of the Samoans is a dialect of the Polynesian language. As the Polynesians had no written language, the early missionaries represented the sounds phonetically by English letters as was done throughout Polynesia, The Polynesian words "malaga" and "aumaga" are accepted by the h and k. The h sound of other Polynesian dialects is represented in some words by the sibilant s and in others by f. The k sound was dropped but its place in a word is represented by the glottal closure which causes a hardening of the following vowel or almost the sound of h. In the written language, the elision of the k sound is represented by an inverted comma placed above the position which the sound originally occupied in the word. For comparative philology, it is necessary that the glottal comma be shown in the right position. The interchangeable l, r, v, and w sounds take the form of l and v. The ng sound is present but was unfortunately represented by the compilers of the alphabet as g. This usage has become official and is a source of confusion to those not acquainted with it. The important naval station in g should be altered to ng. To facilitate comparison with other Polynesian dialects, the lead given by ng sound by the letters ng will be followed in this workn sound is included in their spelling throughout this text.
Recent changes have taken place in the spoken language in the substitution of k for t and a loose mutual interchange between the sounds n and ng. The re-introduction of k in place of t is extremely interesting as it evidently indicates a Polynesian tendency not confined to one dialect. A similar change has already completely occurred in the Hawaiian dialect in which it passed through two distinct phases. Thus, in the widespread Polynesian word kumete (wooden bowl) the first phase was the dropping of the k so that the umete. In the second phase which occurred later, the t was changed to k and the word became 'umeke. Thus the lost k came back into the dialect but in no word did it reoccupy its original position. In the process of resurrection, the k displaced the t sound completely out of the dialect. In Samoa, the first phase of dropping the k had been completed before the Bible was printed in Samoan and kumete had become 'umete. The second phase of substituting the k for t is now taking place in everyday speech and a wooden bowl is now more often referred to as 'umeke than as 'umete. The talking chiefs make the change in official speeches and the retention of the t sound is regarded by the public as pedantic. It seems probable that the Samoan t like the Hawaiian t is doomed to extinction.
The interchange between n and ng has become so common that I had to constantly consult paono instead of the correct paongo, and tafangi instead of the correct tafani.
The Samoan population lives in large well-organized villages. Except for the doing away with some of the highest ranks, corresponding to that of petty kings and provocative of war in grasping at power, the introduction of a foreign culture has made little fundamental difference to the basis of Samoan society. The hereditary titles of high chiefs and talking chiefs are still conferred and supported by the family groups entitled to them. These are not inherited by primogeniture on the male line but are conferred by the group majority and hence lead to much political intrigue. The village fa'alupenga (order of rank prestige) is still jealously observed. Ancient customs connected with the drinking of kava, the distribution of food, the giving of fine mats, and much social ceremonial are still living factors in the life of the people and give pleasure and satisfaction. The pleasure derived from the exercise of native institutions is perhaps the most important factor that has led to the persistence of Samoan customs and helped them to resist the disintegration that has taken place in other parts of Polynesia. The Samoans are thus more conservative than other branches of their race and their satisfaction with themselves and their own institutions makes them less inclined to accept the changes that foreign governments consider would be of benefit to them. Their viewpoint is bounded by their own immediate horizon. This attitude of the mass of the people is expressed in the reply of a talking chief to myself after I had sketched the migrations of the Polynesians from the mainland of Asia to the remote isles of the
Persistence of custom has led to the retention of much native material culture in Samoa. Custom and ceremonial must find material expression to obtain adequate satisfaction. The ceremonial treatment of chieftainship with observance of the fa'alupenga order of precedence found public expression in the calling of the kava in a properly constituted guest house. The need for guest houses kept up the ancient guild of carpenters who, to maintain a close corporation, perpetuated the native form of architecture and technique. The guild of carpenters also kept alive the craft of building the plank bonito canoe. Connected with chiefly prestige was the continuation of the art of tattooing. The survival of the custom of giving fine mats and bark cloth at births, deaths, and marriages has created a continued need for these articles and thus perpetuated the crafts of fine plaiting and making bark cloth. The village hostess (taupou) and the chief's son (manaia) had to be appropriately garbed on ceremonial occasions and this need led to the continuation of the making of various forms of kilts, shaggy garments, and headdresses of human hair. Samoa as a result of the persistence of custom has retained a greater measure of its material culture than other branches of the Polynesians who were more adaptable and who, as a consequence of accepting more of the elements of an introduced culture, have forgotten their own.
On the invitation of the high chief Tuitele of
On the return of my colleagues, I spent two months in the
In
My thanks are due to
Where information was obtained first-hand in the field, it has not been considered necessary to refer to similar statements of fact made by other writers as no question of priority is involved. Much assistance was obtained from the manuscript account (17) of
During the six months spent in Samoa, investigation was concentrated on material culture with such customs as shed light on the degree of importance played by the various complexes in the life of the people. Owing to the active continuance of many of the crafts, it was possible to record a good deal of technique in detail. The technique may be useful to the Samoans in days to come when the broadening of their horizon will inevitably lead to the decay of their native arts and crafts. From an ethnological point of view, details of technique are necessary to form exact comparisons in material culture. The question of diffusion has been confused by comparing end products which have been arrived at by different technical processes. Unless the technical details are similar, the end products cannot well be regarded as identical. Technique may be drudgery to the student but it has a romance of its own. It indicates how different groups of people have sought to supply their material needs by adapting an old method to local material, by evolving improvements, or by inventing a new technique. Throughout each technical process difficulties occurred that had to be surmounted and human thought is expressed by the manner in which skilful fingers sought to achieve the desired end. The end product may or may not be a masterly result but the details of technique reveal the stages of evolution through which the craft has passed. Technique reveals stasis or advance. When available, it must form one of the most valuable methods of judging what culture elements have been shared in common before the separation of various groups, what elements have been improved, and what developed as new inventions to meet local requirements or express the peculiar genius of a people.
The Samoan family living in a settled community required shelter from sun or rain for the two main purposes of cooking, and ordinary living. The form of cooking in the open earth oven did not encourage the combining of the two needs in the same room. The complicated system of dividing the space under one roof into a number of walled-in rooms had not been considered. The climate did not supply such an incentive. The Samoan idea of combatting the heat prevented the use of permanent walls even around the outer margins of their permanent dwelling houses. Each roof, therefore, covered but one room. Cooking and living were permanently separated into two distinct buildings, the fale umu and the fale o'o. The introduction of benzene tins, pots, kettles and a few stoves has led to a small part of the population doing the cooking of small meals in the dwelling house. On the other hand, people of a more advanced culture have found it advantageous to follow the Polynesian custom of having a separate cooking house.
The cooking house was merely a roof propped up in the air; no provision for side walls or floor were required. The roof was originally propped up in one of two ways, to which was added an adaptation from the dwelling house. This produced three types of cooking houses.
The dwelling house from the very nature of its use required more care and detail. The problem of supporting the roof was met in two ways. Sleeping on the floor demanded that provision should be made against damp and an even surface provided for floor mats. The material requisitioned was not wood but stone. While free circulation of air was obtained by open sides, protection against cold winds and driving rain was provided by wall screens which could be dropped in any part and raised again when the necessity ceased. As the house served as both bedroom and sitting room for the family, it had to be sufficiently commodious to meet these requirements. These houses are all of the one shape, the long rectangular middle portion (itu) with rounded portions (tala) added to either end. The use of various timbers has led to a variety of names being applied but they are included under the type name of fale o'o. Besides superior timber, better material was used for thatch and lashing. In the less elaborate dwelling houses, there were no technical details that were beyond the scope of the householder.
With the development of the social side of life, a further need was created. The head of the family required a place where he could meet the heads of other families to discuss matters of common importance to the family groups in the village. This extended to the meeting of the heads of families or groups of families from other villages united by a common bond. It extended fale o'o (a long house with rounded ends). The simplest long house is the canoe shed (afolau). Any house preserving the long character of the part between the ends is referred to as fa'aafolau, after the style of an afolau. Thus the fale o'o, according to the Samoans, is a fale fa'aafolau but fa'a (like, resembling) was dropped and it became fale afolau. The term fale afolau is used as a general descriptive term but through usage, it is specifically applied to the long guest house. To meet the extra demands of accommodation and ceremonial, the fale afolau guest house was built much larger than the ordinary fale o'o dwelling house.
The increasing complexity of ceremonial that developed out of social organization, made further demands on house construction. The guest house was the place where speakers met in fono (council). In the house the Samoans speak, sitting on the ground, and do not change from the position assigned by etiquette. The middle of the rounded ends became the places of highest honor; one for the highest ranking chief of the village, and the other for the similar rank amongst the visitors. In the long house, the two ends were too far apart for convenient hearing. The objection was met by bringing the two rounded ends closer together. This was done by cutting down the length of the middle straight portion (itu) to six or seven feet. In this type of guest house, the side length was reduced to a working minimum and the name of fale afolau was no longer applicable. The sacrifice in length was compensated somewhat by increased width. The height of the roof was increased to meet other requirements that arose so that the house towered up above the others and looked larger. It was thus named fale tele (big house). The approximation of the two ends gave the house a rounder appearance and has led to the term "round house." The term is useful but the roundness was not the feature considered by the Samoans in the application of the term fale tele. The fale tele is the big house and the roundness is accepted but not specified. The idea of greatness is also conveyed in tele from the function of the house.
The greater size and ornamental requirements of the guest houses involved technical details in construction that were beyond the compass of the ordinary individual. By means of split lengths of wood, the householder could cope with the curved purlins of the rounded ends of an ordinary dwelling house. The greater arch of the guest house, however, could be formed only from short curved pieces fitted together. Expert skill was needed. The greater technical details involved led to the development of a craft of builders. Through specialization, a distinct type of building was evolved. To maintain their own status, the chiefs had to have the very best the builders could
In describing the various types of Samoan houses, it is better to follow their natural evolution and work upwards from the simplest forms to the highly organized guest houses.
To understand the difficulty of adopting the terms used in architecture, it is necessary to stress that in constructing the framework, a different order was pursued to that for which the words in English dictionaries were framed. After the ridgepole had been supported in position, the rafters were attached to it. The rafters were then supported by temporary struts and the wall plates and purlins attached to the rafters. In the guest houses, the purlins were attached to the inner and not the outer side of the rafters. Collar beams were then stretched, not between opposite pairs of rafters, but between opposite principal purlins. The collar beams thus took the main part of the weight of the roof. The wall posts were added last and were usually left to the unskilled labor of the owner's family. The framework was built in three sections; the middle rectangular section first, and then the two end sections. In the rounded end sections, the purlins had to be arched.
Beams are horizontal elements of the framework running transversely to the long axis of the building. There are two kinds: collar beams
(so'a)extend between opposite principal purlins and are lashed to them and to the middle supporting posts or the king posts, thus preventing the inward thrust of the rafters due to the weight of the roof; a tie beam (utupoto) is supported on an opposite pair of lateral supporting posts and supports the king post.An eave batten (
langolau, niufafo, oratuao) is attached to the outer side of the lower ends of the rafters. It supports the lowest layer of thatch and assists in defining the eaves.Posts
(pou)are vertical elements which support the roof. There are three forms: the king post(pou 'au'auorte'e 'au'au)extends between the tie beam and the ridgepole to both of which it is lashed; supporting posts (pou tu) support the ridgepole either directly by being placed in the middle line or indirectly by being placed laterally in opposite pairs which support a tie beam and a king post; wall posts (pou lalo) are placed round the outer margin of the house and are lashed to the wall plate.Purlins
(fa'alava). The purlins are elements placed between the ridgepole and the wall plate supporting the rafters at right angles. There are three kinds: principal side purlins (fatafata-a-fale, la'au matua, oramoamo) are straight purlins in the middle section of the house; principal end purlins(fau)are arched purlins in the rounded end sections; intermediate purlins (luangaorpae'aso) are slender rods spaced between the principal purlins in all sections of the house.Rafters. These oblique elements, which, by extending between the ridgepole and the wall plate, give slope to the roof, are of two kinds: principal rafters (
iviivi, la'au fange'aorfatunga), stout rafters which give support to the principal purlins and the wall plate;and thatch rafters ('aso), slender rods to which the roof thatch is attached. They correspond to the common rafters in architecture.Ridgepoles. The ridgepole forms the ridge of the roof. Of these, there are two: the principal ridgepole
('au'au), directly supported by supporting posts or king posts and supporting the rafters of the middle section of the house; the upper ridgepole ('au'au lunga) is smaller and runs parallel above the principal ridgepole. It is supported on the crossed upper ends of the rafters and is used for securing the ridge sheets.Struts
(te'e). These are timbers used in construction to resist thrust or lengthwise pressure. Except for a few in the long guest house, they are of a temporary nature.Wall plate (
amo pouandfau lalo). This is the horizontal element which extends between the upper ends of the wall posts giving some support, in the completed house, to the rafters. Those of the middle section are theamo pou. Thefau lalois the curved wall plate of the rounded end sections and really corresponds to a curb plate.
The carpenters' shed (fale ta), though used in connection with the construction of the highest type of houses, is in itself merely a rough leanto roof, supported by three forked stakes at the back and the front. (See fig. 1.) The roof of horizontal poles and cross pieces covered by coconut leaf sheets overlapped from behind forwards, is sufficient to afford shelter from the sun and from a moderate amount of rain. Within the shed, the older carpenters sat cross-legged trimming timber with their short handled adzes and shaping the joints of the thatch rafters and the principal purlins. The fale ta (fig. 1) was seen in use on the island of
Similar sheds seen at other places were merely covered with coconut leaves arid even the rounded end section of an old house was seen in use. The name fale ta (fale, house; ta, to strike, as in using an adz) refers to the function of the carpenter's shed and may be applied to any shape of building used for this purpose. Because of the ease of construction the leanto shed was the usual form for these buildings.
The canoe shed (afolau) provided shelter from the sun for the large canoes. With the disappearance of the large canoes, such sheds are used to fautasi boats made on European lines. The structure was built sufficiently wide to accommodate the vessel with a little clearance on either side. The problem was to support a roof without impeding the entrance and exit of the canoe and leaving the ends of the shed open. It was solved by doing without supporting posts. Two afolau that retained the old characteristics were examined and explained at Asau and Tufutafoe,
In the shed at Tufutafoe (fig. 2) the ridgepole was supported by pairs of curved poles or rafters that were lashed to a wall plate and continued downwards into the ground so as to prevent the outward displacement of the lower ends of the rafters by the downward pressure of the ridgepole and the weight of the roof. Some rafters though lashed to the wall plate did not reach the ground. Ordinary poles were used throughout. The wall posts were thicker than the main rafters and the thatch rafters much thinner. The purlins were stout poles—small intermediate poles were absent and the lowest purlin so placed as to throw the thatch clear of the wall plate. The thatch sheets of plaited coconut leaves were overlapped from below upwards and tied to the thatch rafters in the usual manner. The coconut leaf ridge sheets were pinned through below the upper ridge pole. (See p. 64.) Hau (
fau) bark was used for the lashing.
The method of supporting the ridgepole with curved rafters alone without any intermediate supporting post or king post is termed fa'fasoata. A building so constructed is termed to fa'asoata (to, to build, and fa'asoata, with curved rafters, no supporting post). The word afolau (canoe shed) is widely spread in Polynesia. In farau, is a shed for a canoe, and in the Tuamotus, horau is a shed. In halau is a long house with the end in front, used mostly for canoes. In wharau has come to mean a particular kind of long house, but also means a rough shed which included that built over a canoe. In the wharau is a ship.
The cooking house (fale umu) is the house (fale) which shelters the earth oven (umu). It is also referred to as umu, piato, and in chief's language, tunoa. It is usually situated a little to the rear of the dwelling house and is used only for actual cooking and the storing of small quantities of uncooked food. As they served no aesthetic purpose, fale umu are roughly built. There are three methods of supporting the roof:
(1.) The method of constructing a canoe shed (to fa'asoata) was used in the poorest type of cooking house. A to fa'asoata seen at
(2.) The second original type of cooking house is the to sunu'i. The term sunu'i (to thrust into the ground) refers, in connection with building, to an upright post which directly supports the ridgepole. The canoe house from which the to fa'asoata type was directly derived could not admit of vertical struts in the middle line. In the cooking house, the objection to the direct form of support vanishes, and the to sunu'i method of construction comes naturally into operation.
The
to sunu'iconsists of a middle rectangular section and two rounded ends. The middle section, constructed first (fig. 3), consists of two supporting posts(pou tu)thatcarry the ridgepole and wall posts (usually three on either side) that carry the wall plates. Three pairs of rafters ( matua) are next placed in position and across them two or three horizontal purlins and an upper ridgepole ('au'au lunga) is placed in the forks made by the upper ends of the rafters. Thatch rafters ('aso) are laid over the purlins, their upper ends first thrust through under the upper ridgepole and an eave batten is attached to the outer side of the lower ends of the thatch rafters. A thatch of plaited coconut leaves completes the building.
The natural poles used are shaped only at the wall posts, where a notch may be cut for receiving the wall plate. The woodwork is lashed together with fau bast.
(3.) The term utupoto refers to the use of a tie beam in supporting the ridgepole prop, and was not included in the two original types of cooking house. The method has been adopted from the more advanced construction of the dwelling house, and utupoto has become the commonest form of cooking house. Lalolangi of Fitiuta while maintaining that the to fa'asoata and the to sunu'i were the two original forms, referred to his own cooking house as a to fa'asoata. In it, tie beams were used. The term fa'asoata has become to the tie beam type to distinguish it from the type with the middle supporting posts, or to sunu'i. In the to sunu'i one of the supporting posts is close to the fireplace. To provide more space for the activities around the fire, the offending post was replaced by using the tie beam construction in preference to the curved rafters or true fa'asoata. The use of the tie beam in cooking houses is a backward displacement from a more advanced stage in technique. The simple form used in the cooking house indicates the stages through which the dwelling house passed in its evolution to its present form. The ground plan of the Fitiuta cooking house is shown in figure 4 and the elements of the frame of the middle rectangular section in figure 5. (See also Plate I, A.)
Across the upper ends of the two pairs of supporting posts
(pou tu)two transverse tie beams project slightly outward and longitudinal beams(tatao)are laid above their ends. From the middle of the transverse tie beams are erected short vertical king posts (pou 'au'au; props for the ridgepole) kept in position by temporary diagonal struts lashed to the tie beams. To the upper ends of the king posts is lashed the main ridgepole(au'au). Three wall posts are erected and wall plates lashed to the outer side of their upper ends. The position of the wall posts is guided by the fact that the three pairs of rafters must rest on the ridgepole andtatao. The longitudinal beam (tatao) is a horizontal element to strengthen the roof framework. It is lashed to the ends of the tie beams and rafters are lashed to it. It acts as a purlin—the first to appear on the inner side of the rafters. For convenience it is here referred to as the tie beam purlin.Two other purlins are attached to the outer side of the rafters. The upper ridgepole is held in the forks formed by the crossed ends of the rafter pairs. The thatch rafters are slender rods
('aso)laid over the horizontal elements of the frame. To them the eave batten is attached.
While thatching the middle section of a house of the to sunu'i or utupoto type, the framework of one of the rounded ends (tala) proceeds.
An end wall post in the middle longitudinal line of the rectangular section of the house and intermediate wall posts on either side are erected to form a curve. A long pole is tied at its middle to the end post bent round the outside of the intermediate wall posts and tied to the two end wall posts of the middle section to form a curved wall plate or curb plate
(fau lalo). The green wood readily takes the curve when the end post isnot too far out but in some houses, two poles were used and the ends overlapped. Three pairs of end rafters are stretched between the side rafters and the curb plate. The first and longest pair is on either side of the middle line and above the level where the highest side purlins cross the side rafter; the second pair cross the side rafter at the level of the second purlin, and the third pair cross at the level of the tie beam purlin. The upper ends are lashed to the side rafters, the lower ends are spaced about equally on the curb plate and lashed to it. To brace the rafters together a curved pole that functions as a purlin is introduced. It passes on the outer side of the rafters, is tied approximately to their middle, and its ends rest on the curb plate at its junction with the side wall posts. It is thus not only curved but runs obliquely upwards to the middle line from either side. Curved purlins in the
talaend are termedfau, and one occupying a middle position (fig. 6, 12) is distinguished asfau tu. The curved purlin is in one piece, but in order to more readily take the curve, the pole is thinned by splitting off a section on either side. Such a purlin is termed afau sasae(sasae: to tear or split off). Thefau sasaeof the cooking house is important as it formed the precursor of the very elaborate curved purlins used in guest houses.
The curved purlin not only braces the end rafters but it supports the thatch rafters which are tied to it as well as to the side rafters and wall plates (fig. 6a, 13). It thus prevents these slender rods from sagging between the side rafters and the curb plate, and hence performs a function similar to the straight purlins of the middle section. Beginning at the middle line the thatch rafters are attached more closely together. To their lower ends is attached the eave batten. Figure 6b shows that the roof slope of the tala is straight between the side rafters and the curb plate.
The other end section is constructed in the same way. The roof is thatched over with plaited coconut leaves, the curved ends presenting no technical difficulties. The floor is not paved unless low lying ground necessitates it for the rainy season. In some stony regions, such as Fitiuta, terraces on the inland side of the village are built up with stone on the downward sloping side of the hill. Stonework is here a necessity and not of such cultural signficance as where the stone has to be transported some distance. Though exact measurements are given in the ground plan, their irregularity illustrates the fact that approximations by eye were used and no checking of distances by measuring with a cord. A few inches here or there caused no difficulties in construction. The type of building was not worth any extra trouble.
The ordinary dwelling houses are usually termed fate o'o and are situated at the back of the guest houses which they serve. They are also termed fale vao from the material used. Here vao (forest) signifies any of the timbers of the forest except breadfruit which is cultivated. Dwelling houses vary in construction according to the means of the family. The better-built ones derive their technique from the guest houses built by skilled labor. The humble and presumably more ancient type are those situated near cultivations
The principal rafters were strong stout poles and the purlins were still attached to their outer side. The feature of the construction was the insertion of graduated chocks of wood between the rafters and the purlins.
The effect produced when the thatch rafters are applied, is an evenly curved roof slope. The tie beam purlins though they still assist in supporting the rafters are now
definitely separated from the thatch rafters by the addition of the lower purlins which removes the awkward curve shown in the tie beam type of cooking house. (See fig, 5 b.) The rafters, ridgepole, and the eave battens were as in the cooking house. The end sections (fig. 8) show a marked change. The straight end rafters of the cooking house (fig. 6) disappear altogether and their function is discharged by an increased number of curved purlins. The end wall post is reinforced by two intermediate wall posts on either side to support the curb plate. A number of thatch rafters are stretched in the middle line by being attached to the end side rafters of the middle section, near the ridgepole. The lower ends hang over the middle part of the curb plate. Three curved purlins of the splitfau sasaekind are now placed in position.The middle one
(fau tu)bisects the end section of the framework and its lower ends rest on the curb plate at its junction with the wall posts of the side section as in the case of the single curved purlin of the cooking house. The top purlin is midway between the middle purlin and the ridgepole in the middle line. It runs parallel with the middle purlin and its ends rest against the side rafters about midway up. The bottom purlin is midway between the middle purlin and the curb plate in the middle line. It also runs parallel with it and its ends rest on the curb plate. While the purlins keep the thatch rafters in a lateral or transverse curve, they also form a longitudinal curve from the side rafters to the curb plate. This curve is formed by tying the middle purlin in the middle line to the thatch rafters mentioned and then propping it up from below with temporary struts. The same is done with the other two purlins until the longitudinal curve is obtained. (See fig.8b.) The thatch rafters are tied to the rafters above, to the purlins, and finally to the curb plate, the additions working out from the middle line to either side. The purlins are curved in as the work proceeds and the extra lengths at the ends cut off to fit against the parts mentioned. The thatch rafters are set as closely together as three inches. The eave batten takes its usual place.
Attention must again be drawn to the use of chocks (fig. 7b, 9) between the principal rafters and the straight purlins of the sides to produce a curved roof. Compare this with the straight slope in the cooking house. The abandoning of the principal rafters in the end sections is an important step. The transference of their function to the curved purlins has been due to the desire to continue the curved slope of the roof of the middle section to that of the ends. The downward slope of the end roof in the cooking house is straight, and the single purlin used is merely to prevent the sagging of the slender thatch rafters between the stiff principal rafters. It is quite probable that chocks may have been used to raise the curved purlin above the straight end rafters. A curve, however, requires to be graduated, and an extra purlin above and below the middle fau tu eased the problem. More careful work in the dwelling house brought the thatch rafters closely together. When they were lashed to three purlins instead of one, their strength became evident. They were quite strong enough to carry the weight of thatch and brace the end section together. The principal end rafters became unnecessary and were abandoned in the dwelling house. The curved purlins were readily adapted to the required roof curve by being tied to the mesial thatch rafters and then propped up in the middle line to comply with it. The ends were purposely left long and as the addition of the thatch rafters brought them round laterally their subsequent length was settled automatically. No builder calculated out the exact length of split curved purlins beforehand but he let the work
The wooden chocks, if they were used, disappeared with the straight principal rafters that gave them support. In the middle section both rafters and chocks remained as a better solution for the curve of the side roof was beyond the means of the builders of the humbler type of dwelling house. It had to await the more affluent circumstances of the guest house.
Sennit braid was used for lashing both framework and thatch. Thatch sheets were put on more closely together, and the ends were cut off evenly below the eave batten. Sugar cane leaves replaced coconut leaves in good houses.
The median supporting posts were seen in a fale o'o at
The floor was built up of stone after the completion of the building, and smaller stones spread over the surface. Wall screens of plaited coconut leaf are used in all dwelling houses.
The guest houses occupied the front lines in all villages. They consisted of two types; the long house (fale afolau), and the round house (fale tele). They were built of the best material. Tradition has it that when the original guild was ordered by Tangaloa-langi to choose timber from the forest for the building of the first house, Malama selected the breadfruit ('ulu). This received heavenly approbation and breadfruit has been the correct timber since that incident. Thus, except for the long principal rafters and sometimes the supporting posts, breadfruit is the only timber for a proper guest house. The house is then called a fale 'ulu in distinction to the fale vao, built of any timber. The fale 'ulu type became associated with chieftainship, and one of the most disparaging taunts is to tell a man that his father never lived in a fale 'ulu.
The guest house in addition to material, size, and shape, was decorated, with ornamental lashings in sennit braid. They were built by skilled carpenters belonging to an ancient guild, and involved much expense in food and ceremonial gifts which took the place of a set estimate of cost. The chief had to summon his people to his assistance. For months beforehand the interminable plaiting of sennit braid went on and was stored in large coils. Fine mats had to be accumulated in sufficient number and extra work in the cultivation was necessary to provide food for the extra demand. The chief had to provide all material (timber, sennit braid, and fau bark for temporary lashings) and transport it to the site. He had also to feed the guild during
The long house (fale afolau) is presumably the older form as it could be more directly derived from the dwelling house. Some of the older men of B.)
The size of the house was regulated by the number of tie beams
(utupoto)desired by the owner. They range from 4 to 6. In figure 9, an average of 5 tie beams is depicted. Five opposite pairs of supporting posts are arranged in 2 rows. An advance on the dwelling houses is made by attaching longitudinal main plates to the outer side of the upper ends of the 2 rows of supporting posts. The posts are cut to receive the main plates which are set flush with the upper ends of the posts. Five solid tie beams, squared or round, are set transversely across, resting on both the main plate and the upper ends of the opposite pairs of posts. The main plate here receives the name ofamo pou, previously applied to the wall plate.Another new element appears in the form of a longitudinal squared beam termed a
tuitui, which is laid above the tie beams in the middle line. Five king posts(te'e'au'auorpou'au'au)are set up on thetuituiabove the crossings with the tie beams. In the previous houses, the king posts rested directly on the tie beams. As before, the king posts are temporarily strutted and the main ridgepole placed on their upper ends and lashed. The ridgepole is a solid piece of worked timber. The principal rafters, dubbed out of coconut wood, are hung in pairs over the ridgepole, a pair directly above each king post. Two such pairs are shown on the left of figure 9a.Two tie beam purlins
(tatao)are placed in the usual position over the ends of the tie beams. (See fig. 9b, 8.) They now take the support of the rafters from the ends of the tie beams. The positions of the wall plates on the rafters is judged on one rafter and the distance to the ridgepole measured off on a cord. The distance is then marked off onthe rafters, and the wall plates attached. Judgment is exercised in dividing up the rafter space between wall plate and ridgepole in equal distances for the purlins of which the already fixed tie beam purlins form one pair. The number is divided and the positions marked on the rafters. The tie beam purlins and the wall plates are both attached on the inner side of the rafters. The wall plates now really function as purlins, for as the wall posts are not put in until the last, their primary function of being attached to the top of wall posts to give support to rigid rafters stretched between them and the ridge-pole has been lost. The other purlins now follow the wall plates and tie beam rafters by being attached on the inner side of the rafters. The cross section (fig. 9 b) shows this important step where the wall plates, the tie beam purlins, and the ordinary purlins are all on the inner side of the rafters.The relative position of the purlins has been changed to obtain a curved slope to the roof. The method of the humbler dwelling house by means of chocks between rigid rafters and external purlins was the work of unskilled labor. Skilled labor obtained the like result by using flexible rafters of coconut wood and bending them out with temporary struts, as in the rounded ends, until the required curve was obtained. The question was how to maintain the curve when the struts were removed. This was simple for the tie beam purlins gave a guide. The external purlins were therefore changed in position to the inner side of the rafters, and from the tie beams was developed the use of the collar beams which acted as struts between opposite pairs of purlins and by keeping them apart maintained the curve of the roof. A collar beam was thus stretched between each opposite pair of purlins above the level of the tie beams. They are tied at each end to the purlins, and in the middle to the king posts in the transverse lines of which they are arranged. A transverse beam below the level of the tie beams would spoil the effect, therefore, when necessary, a short strut (
te'e) was stretched between the lower purlins and the supporting posts. When the rafters are curved by the temporary struts, their lower ends have to be tied temporarily to the scaffolding to fix the lower end of the curve. The wall posts are last of all put in below the line of the wall plate, and the lashing of the wall plate to them permanently fixes the lower end of the roof curve. The thatch rafters are set closely together. The upper ends pass under the upper ridgepole and they are lashed with sennit to each purlin and the wall plates. As the purlins are now all on the inner side of the principal rafters, the thatch rafters rest evenly on them in like manner with the thin principal rafters, and on the same plane. They can not therefore be shown in the section in figure 9b. Across the lower end of the rafters rests the eave batten which is now a proper worked batten instead of a pole. Intermediate, small purlins are a new feature and are shown in position.
The rounded ends are of similar technique to the round house. The best made long house seen was that owned by the
The better class dwelling houses are built on the basis of the above technique.
The long house has the advantage of a clear middle space. The disadvantage lies in the convenient use of the wall posts to lean against being obstructed in view by the two lines of supporting posts. In feasts and gatherings, the guests therefore have to sit to the inner side of the supporting posts while the attendants sit between the main posts and the wall posts. The people at the ends also sit in closer to reduce the distance, and are thus
The round guest house (fale tele) is essentially a fono house in which public meetings and the reception and entertainment of visitors take place. Inseparably connected with this is the ceremonial distribution of kava. In construction, two main requirements must be served; the rounded end sections brought closer together, and the side wall to the interior left unobstructed. To meet the first requirement, the long middle section of the fale afolau was reduced to a range of from 6 to 8 feet according to the size of the house. To meet the second requirement the lateral line of main posts near the front had to be removed. This could not be done without removing the rear line of main posts and thus the tie beam and king post method of supporting the ridgepole had to be abandoned. Recourse was had to the middle main post or to sunu'i method of support. (See plate I, C.)
Details of construction and technique, omitted from the description of the long house, will now be given. The processes will be taken in order, not only to draw attention to the human factors concerned, but to show how the principles of building became changed under the influence of a special building craft.
The main posts were sometimes of breadfruit, but failing this, ifi lele and pou muli were used. The long, principal rafters had to be of flexible wood, and coconut wood was always selected. The wall posts were unimportant. For all other elements of the framework, such as straight, curved, and intermediate purlins, wall plate, curb plate, and thatch rafters, the material was breadfruit wood. The old breadfruit trees that had ceased to bear well were the most suitable for building material. The thatch was of sugar cane leaf pinned together over a thin rod to form sheets. When sugar cane was not available the leaflets of the coconut palms were used with a similar technique. All lashings of a permanent nature were of sennit braid while fau bark was used for the scaffolding.
The builders were responsible for the skilled work in preparing the woodwork and erecting the frame. The erection of the wall posts and the thatching of the roof was beneath their dignity. A few wall posts deemed necessary as struts might be erected and lashed but the others were left out.
The owner (taufale) was responsible for all material and transport to the house site. One of the builders selected the trees suitable for the main posts. The family of the taufale cut them down in the forest and dragged and carried them to the site. When the posts were very large and at considerable (auamo) had to be requisitioned from the males of the entire village. For their food and refreshment, the taufale was responsible.
The women folk of the family of the taufale made the thatch sheets. They carried the leaves down from the plantations, and carried on their work while the timber was being prepared. The shifting of heavy timbers and the erection of the scaffolding was done by the party of the taufale. They also thatched the roof and completed the fitting in of the wall posts.
Where more suitable timber was on the property of another person, the builder told the taufale to ask the owner for it. The special term fa'aune was used—fa'aune mai le la'au (ask for the timber). To fa'aune involved extra expense on the taufale.
After a preliminary gift to the head builder, the builder and the taufale met ceremonially; a contract defining their mutual relationship during building operations was verbally made, agreed to, and sealed with the kava ceremony.
The builders, after being mobilized, were assigned to a house in the village for their occupation. They brought nothing except their own clothing and the carpenters' tools (fa'atufunganga). The tools consist, in these days, of hatchet heads and plane blades hafted to short handles as adzes. The plane blades range in size and all are lashed to the hafts with sennit braid in a manner similar to the old stone adzes.
The rough carpenter's shed (fale ta) was erected near the house site. It contained no equipment in the way of benches. The preparation of timber took place on the ground. Saws are rarely used, but carpenters' planes are now in requisition. Plans, paper and pencils are not used even now. The builder carries the plan in his head and it develops with the building. Charcoal takes the place of carpenters' pencils. The longer measurements are judged by eye, and a piece of sennit braid takes the place of measuring tape and rule. The builders are not restricted by exact measurements calculated out beforehand, but problems of measurement are met as they occur.
The regular meals, drinking nuts, and kava are all provided by the taufale Some of his family are constantly in attendance to administer to the creature needs of the builders. The taufale himself spends a considerable part of his time in the carpenters' shed winding the large coils of sennit braid into the smaller working hanks used by the builders, and chatting and gossiping to demonstrate his interest in the builders' welfare. It requires a good diplomat to ensure the smooth progress of the work.
Ground plan. The ground plan of Tufele's round house at (itu), the wall plate on one side is 7 feet, 6 inches, and 4 inches shorter on the other side. The two end sections (tala) also differ by itu (side). Though it loses in length it gains in width from the greater spread of the roof which is higher than in long houses. The part facing the street is luma (front), and the opposite, tua (back). The round house is always built with its long diameter parallel with the village street. The two end sections continue to bear the name tala.
Scaffolding. For larger dwelling houses and guest houses a proper scaffolding (fatāmanu) is erected before construction commences. The name is derived from fata (a raised shelf or support), and fatamanu is a contraction of fata-a-manu (scaffolding of
The scaffolding, erected by the owner's family, was made on set lines based on the construction of the roof of the to sunu'i type of building.
The position of the main posts of the projected house were marked by the master builder on the site indicated by the owner. To one side of the line to be occupied by the main supporting posts, and corresponding with the ends of the middle section, two long, forked uprights
(to'o manga)were set up in the ground. Their height was about 3.5 feet less than the proposed height of the ridgepole from the ground. (See fig. 11.) A horizontal crossbar (la'au fa'alava) was laid over the forked ends like a ridgepole. Before the scaffolding was completed the main supporting posts of the house were erected, but not permanently. The holes were dug, the posts carried to them, and the butt end of one placed at the edge of a hole. A rope was tied round the post toward its upper end, and passed over the scaffolding cross beam. The post was raised gradually at its top end by men lifting it; forked struts placed under it while a party on the end of the rope kept it taut. In this way, the posts were gradually raised into a perpendicular position and worked into the holes. When the three, or whatever the number, were raised, the scaffolding construction was continued.Three oblique pairs of timbers (fig. 11
b) were rested against the crossbar like rafters. Their lower ends rested on the ground just beyond where the wall posts would come. In the figure they are rather too upright. The oblique timbers are termedfata valainTutuila andfata sasauinManua . Horizontal cross pieces(papaniorteleteleanga)were lastly tied to the oblique timbers to act as steps or rungs from which the builders could reach every part of the framework as it was being constructed. The main supporting posts of the house were then held perpedicular and lashed in position either to the ridgepole or to a more convenient cross piece.
The scaffolding described is the set type for the middle section of long or round houses. In the long house, it is made to agree in length with the section being built. The upright forked posts, with a ridgepole, oblique rafters-like structures, and purlin-like cross pieces, is merely a roof framework within a framework. The scaffolding for the rounded end sections was not put up until it was needed. The lashings of the scaffolding were made throughout of long strips of fau bark about two inches wide. The material was provided in bundles by the owner's family. The lashing followed the set form shown in figure 12.
One end of the strip is passed round the fixed element (the oblique timbers in figure 11
b) and tied round itself with a single overhand knot. This forms a running noose or slip knot which is the orthodox commencement of most Samoan lashings whether temporary or permanent, and it will be referred to throughout as the running noose commencement. The noose is drawn taut at the exact spot required on the stationary timber, and the cross piece is placed in position. A number of oblique turns are then made in the one direction round the two wooden elements. (See. fig. 12b.) In making these turns, the strip is drawn taut with the right hand, which may be assisted by the weight of the body. When sufficiently tight, the left thumb holds the strip against the wood while the right hand passes the strip round the timber. This method is quick and effective, and demands that one end of the binding material be first fixed so that only a single strip demands the attention of the two hands. Two or three turns are usually sufficient. The last step (fig. 12c) consists of making one or more turns round the lashings between the two wooden elements as indicated by the arrows in the figure. Before the last circumferential turn is tightened (fig. 12c), the strip end is passed under it. The turn is drawn taut and further tightened by pulling the end not in the direction of the arrows, but back in the opposite direction. These circumferential turns (langolango) are always used to tighten up the lashing.
The term fa'atunga means the causing to stand upright. It applies to the erection of the main posts and marks an important stage in construction which in all important houses must be celebrated by feasting. The term fa'atunga thus applies to the ceremonial giving of food, and feasting, as well as the official erection of the main posts. I was fortunate enough to see the full ceremonial in connection with the building of the house of Misa, high chief of the island of
The proceedings opened with the special morning meal termed (lavatasi). The working party of the taufale appeared and assisted by lifting the main posts to the right level indicated by the builders. This was done by tying two horizontal bars to the post at right angles with each other. The laborers got their shoulders under the bars and lifted while another laborer shoveled stones and earth under the post, which was then held vertical while the sides of the hole were filled in and rammed tight. The builders directed operations from the top of the scaffolding. Of the three main posts, the two outer ones were fixed first. A crossbar was then laid over their upper ends. The middle post was lowered until its upper end touched the crossbar. It was held in this position until fixed by filling the hole. The scaffolding was so made that the uppermost crossbar or one of the cross pieces was in the upright line of the main posts. The upper ends of the main posts were lashed to it before the filling of the holes was commenced. The erection took place at about midday. Meanwhile the food had been prepared. The workers and builders retired to their respective parties and the fa'atunga ceremony took place.
Owing to the short length of ridgepole to be supported, it is not necessary to have the support at each end as in the to sumu'i buildings. One strong, massive post (fig. 13a) is quite common. In quite a small guest house, the single post was 46 inches in circumference near the ground. Some are forked (manga lua) with the branching close to the ridgepole (fig. 13b). A very c), while in d) was quite common. Some houses have two main posts but three (fig. 13e) was as common as the single post. More than three posts were said to be used on occasion. The main posts are sometimes referred to as to'o loto (interior prop.)
Of the three posts in Misa's house, the middle one was 48 inches in circumference near the ground and the other two, 37.5 and 33 inches respectively. They were placed 20 inches and 19 inches apart so that the ground distance covered by the three posts was 6 feet, 6.5 inches. Thus there was less than a foot of the ridgepole projecting at either end. The posts were 25 feet, 2.5 inches above the ground. The house site was an old one in which the house platform was already complete. In a new house site, the depth of the platform to be erected has to be added to the height of the main posts above the ground level. In the large round house at Iva, the main posts were 32 feet above the floor level which in turn was 6.5 feet above ground level. The main posts, including the parts sunk in the ground, must therefore have been over 40 feet in length.
Principal ridgepole. After the feast the ridgepole ('au'au) was carried up the scaffolding by one of the carpenters on his shoulder. It was a worked piece of timber, 8 feet, 4 inches in length, adzed to form five surfaces: a wide under surface, two narrow side surfaces, and two wider, upward sloping surfaces that met in a mesial longitudinal edge. It was placed in position on the upper ends of the main posts by the second in command of the carpenters, who all wore lavalava waist cloths and who had limed their hair for the occasion. The head builder (latu), who was distinguished by carrying a coconut wood walking stick (to'oto'o), viewed the ridgepole from, the ground about 10 yards away. He called a command. The ridgepole was lifted off, and one of the outer posts was chipped with an adz to lower its level. The ridgepole was again put on and adjusted so that the ends projected evenly beyond the outer posts. A wave of the walking stick expressed approval. Coconut husk fiber (pulu) was pushed into the spaces where the surfaces did not quite coincide. le sumu o le 'au'au. The method of lashing shown in figure 14, is better described with each figure in order. The object is to form the turns in such a manner that the large lozenge motive in it will show toward the front of the house. With the worker facing this aspect of the post, we may refer to the front, back, and right and left sides of the post. Sennit braid with its roughness clings better to the rounded wooden surfaces than smooth cord. It may be necessary, however, for an assistant to hold the commencing loop in position until the figure develops. The outer long stretches of sennit forming the lower borders of the large front and back lozenges are tightened by the shorter inner turns which cross them to form the upper borders of the lozenge.
The sumu o le 'au'au lashing, besides being decorative, is quite effective as a lashing. No steps are taken to fit the post and ridgepole by mortising in any way. The flat under surface of the ridgepole simply rests on the flat upper end of the main posts.
Principal rafters. The principal rafters (fatunga) are always made of coconut wood (niu). Trees are felled for the purpose and adzed into very long battens 4 inches wide and 1.5 in thickness. The straight stem of the coconut provides the length required and the flexibility to form a curve for the roof. They are made longer than necessary. The lower extra part is but throughly shaped and is kept for fear an error of judgment as to length. They are cut off when the thatch rafters are trimmed. The upper ends are dealt with in pairs. (See fig. 15.) The inner surface is quite flat, but the outer surface is generally convex, being 1.5 inches thick in the middle line and 1 inch at the edges.
The flat under surfaces of the rafters lie against the slanting upper surfaces of the ridgepole. While the upper ends are being fitted (fig. 16) assistants support the weight. Two long poles are used. (See fig. 18a, 2.) A pair is placed over each end of the ridgepole, and a pair or two pairs close together over the middle.
In the canoe shed, the curved poles were termed iviivi, and in the cooking and dwelling houses, the straight rigid rafters were termed matua and la'au fange'a respectively. The flexible coconut rafter is termed fatunga whether in guest or dwelling house.
The rafter pairs are lashed to the ridgepole where they cross and the lashing (
fausanga) is so arranged that a neatlozengepattern is worked on the under surface of the ridgepole. The lashing consists of transverse turns and diagonal turns round each element of the woodwork. The turns form the complement of the other. As a principle, it may be laid down that transverse turns round one wooden element demand a diagonal turn round the other and vice versa. The object is to form a rectangular space on the under surface of the ridgepole with two sets of transverse turns and then finish with diagonals within the space neatly crossed to form a lozenge. The builder keeps this in mind primarily, and the lashing of the woodwork together goes on incidentally. (See fig. 17.)
Such a figure is termed mamanu. The guild of builders invited me up on the scaffolding to watch the ornamental lashings being done. The braid is fixed by taking a couple of half hitches round one of the turns on the back of a rafter.
The talava. These are really purlins but much smaller than the main purlins and at the same time larger than the intermediate purlins. (See fig. 18a, 3.) In talava. In solinga. No collar beam is attached to the talava.
Wall plates. The wall plates (amo pou) are cut out of solid breadfruit wood and shaped round in section with a diameter of 5 or 6 inches. They are cut to the same length as the ridgepole, that is, about 8 feet, 4 inches. One was lashed temporarily to the under surface of the rafters at a height from the ground judged by the head builder. The lashing was the same as in the scaffolding, but a separate working hank was used on each end rafter (fig. 18c) of the same side and left hanging. A long piece of sennit was then stretched along a rafter from the talava above to the wall plate below, and a knot tied to mark the distance. The wall plate was then checked off with the measured braid on each rafter of the same side, and the wall plate properly adjusted. The distances from the opposite talava were marked off on the two end rafters of that side with a piece of charcoal. The other wall plate was then temporarily lashed in position. (See fig. 18a, 4.)
Note from figure 16 that the rafters are hung over the ridgepole where the locking of the upper ends, and the lashing, keep them in position. The weight is partly taken by the temporary struts which keep them out at an angle complying with the upper inclined surfaces of the ridgepole. The wall plates are attached and merely hung on the rafters without any support from below. amopou as there is no main plate to deprive it of its proper name. langolango.
Principal purlins. The principal purlins (amoamo, or la'au matua) must be evenly spaced along the rafters between the talava and the wall plate. The number was fixed by the taufale when he told the head builder the number of so'a (collar beams) he desired. In the round house there are no tie beam purlins as in the long house. Each opposite pair of main purlins must have a so'a collar beam between them with the exception of the lowest. Thus Misa had decided on 7 so'a, so the space between the talava and wall plate had to be evenly divided for 8 purlins. The lowest free purlin is called amoamo taunoa (inactive purlin). In la'au tauvale, la'au taunoa, and la'au sautia, Tauvale, like taunoa, carries the meaning of "inactive." The term sautia (bedewed), conveys the metaphorical idea of being open to the dew through not being covered by collar beam attachments. It is also called luanga. The height of the main posts above ground is thus considered by the
head builder from the point of view of the number of so'a collar beams and their corresponding purlins, as the purlins must not be placed too closely together. The purlins are adzed out of breadfruit wood in the same manner as the wall plates. They also are the same length as the ridgepole and decrease in diameter from below upwards, the lowest being smaller than the wall plates.
An assistant brought the measured sennit to the head builder. Holding one end in his left hand, the builder made a turn round one of the assistant's upheld hands and back to his own. He made four such turns, and handed the other end to the assistant. By working their hands they adjusted the loops to an even length with the ends of the cords at either end of the loop. Each had thus the doubled ends of four turns, which they marked by inserting strips of
faubark under one of the strands of the sennit braid. On finishing, the head builder carelessly tossed the marked braid back to the assistant.it had 8 pieces of bark evenly spaced in the length that corresponded to the distance between the talavaand the wall plate. The head builder had no pencil or paper, and he did not know the length of the cord in feet and inches. He and his guild had never needed anything beyond sennit braid andfaubark in making calculations.The assistant returned to the building, and tossed up the end of the cord to a carpenter on the scaffolding. The two ends were held against the
talavaand wall plate while a third carpenter, with a piece of charcoal, marked off against the rafter the points where the eight pieces of bark touched. (See fig. 18a.) This was repeated on the other end rafter of the same side. Another carpenter tied the free end of a working hank with the slip knot fixation round the rafter at each mark (fig. 18c) and left eight hanks hanging in the air. Other carpenters, in pairs, tied the appropriate purlins to the under side of the end rafters with temporary lashings at the positions where the hanging hanks were waiting for them. A few diagonal turns were made with the sennit already tied to the rafters. A circumferential turn was taken round the lashing between the rafter and purlin. (See fig. 18b.) In an incredibly short time the 8 purlins were attached to the end rafters on either side. The builders then adjourned as the villagers had gathered for thefuiavato complete thefa'atungaceremony.Work commenced the next morning with bending the rafters into the right curve. The head builder, with his wand of office, took up a position about ten yards away from the end of the framework. The second in command executed his orders regarding the frame from the scaffolding, and directed those below with the struts. The head builder called to lift up the rafter at the first purlin. This was done by a carpenter getting his shoulder under it and straightening his back. The second in command placed the end of a long strut under the rafter, and directed the strut bearer as to the placing of the lower end on the ground. When the weight was released, if the strut proved satisfactory, the head builder waved his walking stick in approval. In this way, 5 struts were placed commencing with a longest strut. (See fig. 19
a.) The lower ends were pushed outwards toward the rafter until taut. If too long, they were shortened with an adz stroke. I stood behind the head builder as he issued his orders. He judged entirely by eye, and left nothing to be desired. In this way the long rafter was bent to the right curve as shown in the figure. Its lower end was, of course, still fixed by the tie to the scaffolding, but it was readjusted to suit the curve.
The struts (fig. 19
a, 15-19) were fairly slender poles. There were not let into the ground or tied above. They looked flimsy and liable to slip, but both struts and rafter remained in the position assigned them. Now and again a purlin dropped out of the temporary lashings owning to the movement of the rafter, but they were quickly lashed on again. The first set of two temporary struts was readjusted and included in the five.The opposite end rafter was dealt with similarly to correspond in curve with its pair. The head builder then changed, position to the opposite end of the house, and a corresponding curve was made with the other end rafters. The rafters over the middle of the ridgepole needed no adjustment as they rested on the purlins which followed the curve of the end rafters. The two sides were checked by running a line from the base of the middle main post to the middle of the wall plate and comparing with the measurement of the opposite side. The first measurement indicated a difference of only 2 inches, showing how keen was the eye of the head builder.
Theoretically, the outer ends of the purlins correspond exactly with the outer edges of the end rafters, but in practice some do not, owing to the warping of the rafters. Inward warping does not matter, but outward warping must be corrected to bring the outer edges of the rafters into line with the purlin ends to enable a lashing to be made. In figure 19b, the left of the two rafters projects beyond the left end of the purlin. As the purlin has been cut to the right length, the two rafters must be drawn together until the outer edge of the left rafter coincides with the end of the purlin. A rope is tied round the displaced rafter, passed under and over the other rafter, and back
The temporary lashings are now converted into permanent ones. The simple but neat method is shown (fig. 20) on an end left rafter viewed from the scaffolding on which the carpenter stands. The upper row shows the view from the inner operating side and the lower row, the appearance on the outside. The edge of the rafter corresponding with the end of the purlin is preferred to as "outer," and the other edge as "inner." The temporary lashing is unfastened but the cord is left attached by its transverse turn round the is unfastened but the record is left attached by its transvers turn round the rafter. This turn is concealed by the purlin on the inner side but is seen, in the lower figures, crossing the rafter transversely. In smaller round houses with a short middle section, the two pairs of end rafters are sufficient.
Upper ridgepole. The upper ridgepole ('au'au lunga) is of worked timber slightly larger than the talava purlin. It is made longer than the main ridgepole so as to project slightly beyond it at either end. It is rested in the upper forks of the rafters and lashed to them with any combination of transverse and diagonal turns. No careful arrangement is made as the lashings are not seen.
Thatch rafters. The thatch rafters ('aso), composed of small rods split from breadfruit wood, are about an inch wide and barely that in thickness with the edges rounded off. Length is obtained by joining the short lengths together.
The join (so'o mata sai) is made with a long oblique slant as in figure 21 where the plane is broken transversely in the middle to prevent the inclined surfaces from slipping apart. Both ends of the rods are shaped by the older carpenters in the fale ta shed to the standard plan indicated by the dimensions in the legend of figure 21. It is immaterial which way they face or which end is above, as the fitting is the same.
The shaped lengths are joined together outside the shed. The lashing (fafau so'o'aso) is quickly made (fig. 22) with sennit braid, and no material is wasted. Usually one lashing over the middle is sufficient, but sometimes two, slightly apart, are used.
The joined thatch rafters ('aso) are now carried by two men to control the length, and passed upwards on the outside of the purlins. The upper ends are passed between the two ridgepoles to project a couple of inches on the other side. All lashing is done from the inside and commences on the talava. The head builder gives the spacing in finger breadths beforehand. In good houses, the space between 'aso is two finger breadths (vaelua); in others, three or more. In Misa's house, when I called up to the second in command as to the spacing, he held up two fingers. It is not necessary to wait until an upper row is completed before commencing a lower, but the individual 'aso must be lashed consecutively from above downwards so as to remove any bend or twist in its length. While some carpenters were busy handing up the 'aso and inserting the upper ends in position, one commenced lashing from the left of the talava. He spaced the first by passing two fingers of the left hand between it and the left principal rafter. Taking the hank of sennit hanging from the permanent lashing of the talava and the rafter, he lashed the first 'aso. He spaced the second and carried on the lashing with a continuous braid. After he got a start, a man commenced on the first purlin by carrying on with the hank from its permanent lashing to the rafter. As he progressed to the right, another commenced on the second purlin, and so on right down to the wall plate. Some of the 'aso were quite crooked, but the consecutive lashing from above down soon straightened them out.
The lashing of the thatch rafters to the purlins is the same throughout. As the crossings over the inner surface of the purlins are plainly viewed from inside the house, utility is combined with decoration by using a vertical pair of turns on either side of the 'aso. In figure 23 the upper row represents the outer surface, and the lower, the inner appearance on the horizontal purlin.
The parallel pairs, commencing from the principal rafter on the left are shown in figure 24. They give a pleasing effect in the guest houses. Threes are sometimes used if the taufale has plenty of sennit braid. The 'aso are parallel throughout their course from ridgepole to wall plate. An unorthodox form of decoration seen in 'aso between the wall plate and the second purlin above it. The closely set thatch rafters with the lashing to all crossing elements, while contributing largely to the stability of the roof, have been to the interest of the builders' guild in promoting decorative effect and increasing their scope of work. It is only every seventh thatch rafter or so that is actually used for the attachment of the thatch.
In Asau, 'aso are formed of long fine rods of the alamea which need no joining and no shaping beyond removal of the bark. They are pleasingly white, but if soaked in sea water to facilitate peeling they turn to a less effective dull-brown color.
Used 'aso from a dismantled house are darkened by age. A decorative effect in stripes, termed pulei, is obtained by alternating them with new 'aso.
Intermediate purlins. Horizontal purlins, the same size or a little larger than 'aso, are spaced between the talava, purlins, and wall plates. Two intermediate purlins (luanga, or pae'aso) divide each of these spaces into three while the lowest space between the free purlin and the wall plate usually has three intermediates. They are laid in position on the inner surface of the rafters and 'aso, with their ends flush with the outer edges of the end rafters. A couple of temporary lashings hold them in position while the permanent lashings are made.
Before the stage of tying on the intermediate purlins was reached in Misa's house, I had to move on reluctantly. Fortunately, however, at the next island of
The intermediate purlins are usually in one piece for the short length of the middle section, but if a join is required, the so'o mata sai of the thatch rafters is used. Every intermediate purlin is tied to every principal and thatch rafter that crosses it. This adds to the firmness of the rafters, but they are again an introduction by the builders to add decorative effect by breaking the wider spaces between the main purlins, and to demonstrate their craftsmanship. In luanga, but pae'aso (rest for the 'aso) is a better name for the intermediates.
The eave batten. While the intermediate purlins are being lashed, the eave batten (langolau, or niufafo) is put on. It occupies the same position as in the humbler houses, but is dubbed out of coconut wood into a batten. (See fig. 26.) The lashing is shown in figure 27.
Some carpenters notch the free edge of the batten. When the lashing is completed the principal rafters and 'aso are cut off close to the lower side of the eave batten. The circumferential turns prevent the turns round the 'aso from slipping over the cut ends as shown in figure 27 n. Though the 3 inch wide batten is set on edge, the form of lashing effectively keeps it in position.
The rafters are taken in their turn with the same lashing as the 'aso. In langolau, but niufafo. The batten is made of coconut (niu) and it is placed outside (fafo).
The collar beams (so'a). These are transverse beams set between opposite pairs of purlins to keep the rafters in a fixed position. They are here termed a. If the temporary struts are removed, the sides of the roof will come inwards of their own weight. The position of the struts was therefore changed by using permanent horizontal ones between the opposing sides.
The beams are adzed out of breadfruit timber and decrease in diameter from below upwards, with their corresponding purlins. The lengths, which also decease upwards, are taken by stretching sennit braid between opposite pairs of purlins and cutting the beams in the shed to the measured lengths. They are cut slightly longer and fitted with an adz after trying them in position. The beams are placed so that they rest against the main posts to which they are also lashed. With one main post, there is a collar beam on either side of it in the lower tiers and only one in the upper tiers. With three main posts, there must be a collar beam on the outer side of the outer posts in all the tiers. In the lower tiers there is also a beam on either side of the middle post, but in the upper tiers only one is used and whichever side of the middle post it is on is usually maintained throughout. Each opposite pair of purlins must have collar beams, except the lowest.
The collar beams are temporarily suspended as in figure 28 a, where one is hung in position by a cord to a thatch rafter above. A collar beam may also be directly lashed to convenient parts of the scaffolding. Any parts of the scaffolding that obstruct the line of the collar beam are removed.
The lashing of the collar beam to the purlin is ornamental as well as structural. The form of lashing (sumu) in this position is termed le sumu o le so'a (the lashing of the collar beam) to distinguish it from le sumu o le 'au'au (the lashing of the ridgepole). The distinction is one of position only,
Lashing collar beams to main posts. When both ends of the collar beams are lashed, they are lashed at their middle to the main posts. The lashings of the single collar beams to the outer side of the outer main posts are shown in figures 29 and 30.
The circumferential turns which complete the lashing enables the whole to be braced together and tightened securely. The roundness of the two wooden elements prevent the circumferential turns from being seen directly
Less care is devoted to the lashings of the upper collar beams as their decorative feature can not be seen so readily from below. The lozenge motive was usually omitted and the three distinct bands on the back (fig. 30 b, and c) were usually merged together into one broad band. The guild of builders were thus inclined to do slovenly work where it could not be observed.
The single main post or the middle post of three has a beam on either side of it. All three elements are lashed together at the same time. A running noose is made around the post. About eight transverse turns are taken around both collar beams on either side of the post. Diagonal turns are then made across the outer surface of both beams the sennit passing across from one beam to the other around the sides of the costs as in the case of the transverse turns. They are done in sequence so as to develop the lozenge pattern on the outer sides of the beams and on the sides of the post as in figure 31.
A cross section of the roof frame, with collar beams in position, gives the appearance in figure 32. Looking upwards from the floor, the perspective of decreasing collar beams gives a fine appearance to the eye, and also conveys a false impression of height. Tufele's house with five tiers was 20 feet high; Misa's with seven was 25 feet, 4 inches, and the Iva house with thirteen tiers was 32 feet.
The number of tiers not only indicates the size of the house, but also the status of the owner. There is also a variation in different villages which may have been influenced by financial considerations. The collective term for the tiers is tauso'anga to which the number is added. Thus Misa's house was tauso'anga fitu (seven tiers of collar beams).
The pair of longitudinal beams (te'e tala) described by fau tu (middle arch) of the rounded ends, is unusual. I saw it nowhere else in British or American Samoa. Its name, te'e tala, means a support for the rounded ends. The glory of achievement of the Samoan builders is that they erected the rounded ends with arches unsupported by beams or principal rafters. The longitudinal strut seems foreign to their genius. As the building was constructed for the Government, the abnormal beam, composed of what looks like sawn Oregon pine, makes it appear as if some public works official had ordered this extra strut to be put in to prolong the life of the government property. It certainly clashes with pure native art as do the many smaller struts (te'e) that seem to prevail around the region of
The wall posts (pou lalo). These may or may not be erected at this stage. There are three on either side, one at each end under the end rafters, and one in the middle. Throughout they occupy an insignificant position structurally.
With the fitting of the collar beams, the temporary struts are removed, and the middle section (itu) is completed as far as the carpenters are concerned. They turn their attention to one of the end sections while the owner's party commence thatching the completed part.
The size of the house does not depend on the length of the middle section but on the spread of the roof which in turn is governed by the height. Tufele's house at Fitiuta, with a middle section length of 7 feet 6 inches, was 27 feet 6 inches wide. The house at Iva was only 4 inches longer, but it was 45 feet 6 inches wide. Its height, however, was 32 feet as against 20 feet. Misa's house was 6 inches longer than the Iva one; owing to its lesser height it was not nearly so wide. The middle section was a structural necessity to which the rounded ends could be added.
The end sections of the long house and the round house are the same in every detail of construction. The main features are the curved wall plate which thus becomes a curb plate, the curved main purlins built up of several pieces to form arches, and the continued disappearance of principal rafters.
Preparation of curb plates and arched purlins. With the solid wood needed for large houses, it is no longer possible to bend a straight piece into a curb plate, nor is it compatible with the status of the house to use the split curved purlin (fau sasae). The arches, whether curb plate or purlin, are made of short lengths of breadfruit wood, worked to a slight curve and joined together. The verb "to lash" is fau, and because the arches are lashed together, they are all termed fau. The curved purlin name, therefore, is derived from construction. In time, the name became associated with function, and was then applied to the split purlin which is not lashed together. To mark this difference, the qualifying word sasae (split) was always used with the split purlin (fau sasae). The split purlin must have had some other older name that has been dropped or was not obtained by me.
The shaping of the short lengths is carried on by the older carpenters in the shed while the middle section is being constructed. The thickness of the pieces varies with the different arches as does the amount of curve. Beyond the fact that the curb plate must be the same diameter as the wall plate, the middle arch a little less, and the other arches graduated smaller, the workers rely on experience. If a piece of wood is spoiled for one purpose, it can be used for another and the cost of material is not borne by the builders.
Each end of the pieces is shaped for the joint except the parts that form the ends of the arches. The join is the so'o mat a sai, already seen in the thatch rafters. The curve of the wood forms no difficulty. The right-angled
The concave part looks directly into the house, totonu, and the convexity faces away, tua (outside, back, towards the thatch rafters). In Samoan, the expressions used for the upper ends of the lango pu'e and other pieces is ulupale i tua, and for the lower ends, ulupale i totonu. For the pu'e both ends are ulupale i totonu. (See Plate II, A.)
Assembling the arches. The curved pieces are assembled outside the shed in an open space that permits of the full arch being laid out on the ground.
A length of sennit is stretched between the ends of the wall plates of the middle section, and the distance pegged out on the ground as (1) and (2) in figure 35.
The first arch assembled is the curb plate (
fau lalo). As the joints are fitted, they are temporarily lashed together around the middle of the joint, or with two lashings; one on either side of the middle, withfaubark. Wedge-pointed wooden pegs (tinaormatalafi) are driven in under the lashing on the concave side to tighten them. These fitting lashings are termedu'a vale.
The fitted curb plate is tested on the marked ground. The middle of the key piece must rest against the outer side of the apical peg. The ends (
si'u fau) must then touch the outside of the side pegs. In figure 35, the curve (5) so formed makes a perfect semicircle, but it is rarely that the curb plate conforms exactly to a semicircle. In a very large house, the ends are prolonged past the side pegs, and in smaller houses they fall short. In either event the width of the ends must be the same as that between the side pegs in order to junction perfectly with the wall plates of the middle section of the house. In the large house at Iva, with the width of 45 feet 6 inches, thetalalength was 9 inches greater than half the width. In Tufele's smaller house with a width of 27 feet 6 inches, the lengths of the twotolawere 1 foot 1 inch, and 2 feet 2 inches shorter than half the width.When the ends are inside or too much outside of the pegs, the arch has to be corrected. The head builder stands outside the apical peg, and studies the curve on either side to pick out what joints should be altered. The joint or joints are then removed with the lashing still round them by unfastening the two ends away from the joint. The offending joint is taken to the shed and reshaped to lessen or increase the curve as required. The corrected joint is relashed in position, and the arch again tested. When the whole arch is correct, the lashings are termed
u'a mauinstead ofu'a vale. The termmau(fixed) means that the arch is fixed as regards shape.The middle arch (
fau tu) is next laid down. The pieces of the middle arch, which is the largest of the oblique arches, are assembled like the curb plate. It is tested by placing the midpoint of the key piece against the inner side of its peg. The ends must then touch the side pegs as in figure 35, 7. When correct, the stage is calledua u'a le fautu. The middle arch is lashed together. The curb plate is removed and set up in the frame while the assembling of the arches is continued by the experts.The remaining arches are divided into an upper and a lower series of even number which arc separated by the middle large arch, just assembled. To understand the method of assembling, it must be remembered that the middle arch is the direct descendant of the middle curved purlin seen in the rounded ends of the cooking and dwelling houses. In position, its lower ends rest on the curb plate at its junction with the end wall posts, which are directly under the end rafters of the middle section. They bisect the right angles formed by the line of the end rafters and the curb plate. The arch directed upwards and outwards divides the roof portion of the rounded section into two equal parts in the middle line. The tops of the upper series of arches are evenly spaced along the middle line, and their lower ends have to rest against the end rafters of the middle section. Thus, from below upwards, the upper series not only decrease in the diameter
of the wooden pieces used, but also in the height of the arch and the distance between the lower ends. As a result the curve of the arches narrows from the middle arch upwards. They are thus tested on the inner side of the middle arch, which is left on the ground for this purpose. The first arch of the upper series to be assembled is the one next the middle arch (upper fau vaenga). It is laid out on the inner side of the middle arch, and follows its curve (fig. 35, 8) for a certain distance, but the actual length of its sides is left for the framework to decide. It is left on the ground, and the next of the series laid down inside of it. The others follow consecutively as shown in figure 35. The apical curve becomes sharper as they narrow. The last of the series is thefau tali'aso. Within the limits of the rule laid down, experience guides, and no exact rules can be formulated.
The lower series of arches to occupy the lower section of the roof are rested on the curb plate, while their upper parts are evenly spaced along the middle line. They range along a curvature which does not contract, though their length naturally shortens as they descend. They are thus tested on the outer side of the middle arch still left on the ground. (See fig. 36, 7.) The first assembled is the one next in position to the middle arch, and it is termed the lower fau vaenga. If follows the wide outer curve of the middle arch, and like those of the upper series, the exact length is left for decision on the frame. When the lower fau vaenga is lashed as correct (u'a mau), the middle arch is taken away and set up. The remainder of the lower series are laid down consecutively on the outer side of the one preceding. Their ends are all left indeterminate.
From the above method it is seen that the curb plate and the middle arch are the two elements that are exactly determined when assembled. All that is needed with the other arches at this stage is the curve of each for setting up in the middle line. Calculations as to length were not attempted, not because it was beyond the mentality of the Samoan builders, but because it was left for the later technique to decide in a practical and certain manner.
Scaffolding. Before the erection can be commenced, the scaffolding (fata-manu) must be extended to the end section. The method of supports for the fau bark. The other side is similarly constructed. Plate II B, shows an end scaffolding.
Erecting the curb plate. The curb plate is carried into position and temporarily strutted while its ends are lashed to the end rafters immediately below the wall plates. (See fig. 38.) Some curb plates are cut away at the upper edge so as to fit against the wall plate, and bring the upper edges closer together.
Ridgepole end piece (moamoa). A short end piece is placed at the end of the main ridgepole to serve as a rest for the upper ends of the thatch rafters attached in the middle line ('aso vao). Two forms are used:
- A flat piece of wood barely as wide as the ridgepole, and from 7 to 9 inches long, was let into the under surface of the ridgepole which was cut as in figure 39. The under surface was also carved with curved lines as shown. Some are painted with dots and triangles, and others, according to
Handy (14, p. 8), "are carved in symbolic representation of the moon and stars." From above the ridgepole overlaps part of themoamoa, and the even continuous slope downwards, by breaking the abrupt ending of a squared end, affords an even slope conforming to the plane of the middle rafters.- The construction of the second type is totally different. A curved piece of wood is placed transversely against the squared end of the ridgepole, and the outer edges of the end rafters, but is not lashed against them. (See fig. 40.)
Another variety was round in section, like a curved purlin, and extended directly upwards along the edges of the end rafters, its ends coinciding with the highest main purlins (not talava) on either side of the middle section of
The middle longitudinal curve. The guide to the contour of the curve in the middle line from the ridgepole to the mid point of the curb plate, is given by 5 or more thatch rafters termed 'aso vao.
The middle arch is first lifted with ropes (
tautala) into position with its ends on the curb plate here it sets against the outer end of the wall plate. The ends may be shaped to fit the angle. Its upper arched end is placed in approximate position, and temporarily strutted in the middle line, and on either side., (See fig. 41a.)
When the arches have been placed in their correct position, they are lashed to the mesial thatch rafters that are in position in order from above downwards. More thatch rafters are added on either side of the middle set, working outwards on either side. As the work expands outwards, the arches of the lower series are successively added from above downwards so that the lowest comes comparatively late into position.
The lower ends of the arches. Attention has already been drawn to the fact that with the exception of the middle and topmost arches, the lower ends of the remaining arches are purposely left short. This feature is shown in the arches of the upper series depicted in figure 41a. Both the longitudinal and lateral curves of the rounded end of the frame are maintained by the strutting of the arches. The exact length of the incomplete arches can therefore well wait until the advancing construction decides it. As the addition of thatch rafters approaches the short ends of the arches, additional pieces are added to lengthen them. The pieces being made to plan, as regards the joints (fig. 34), can be easily fitted with a temporary lashing as if they were on the ground. In the upper series, the boundary of length is the outer edge of the end rafter of the middle section. When a piece reaches the rafter, it is held in position at the joint and the point of contact is marked with charcoal. The piece is cut off square with its long axis and attached to complete the arch. The order of completion is, of necessity, from above downwards. In the lower series, a similar method is used to make the lower ends coincide with the upper edge of the curb plate. The completed arches are parallel to each other, the line having been set by the middle arch. The lower ends of the arches merely touch the end rafters and curb plate (fig. 41b) and are in no way attached to them. They maintain their position by the lashings to the thatch rafters. The method of finishing off the lower end of the arches thus decides their length in a practical manner without any mathematical calculations which may have been beyond the scope of the builders. In some houses, the ends of the arches were observed to touch the main purlins of the sides, but with the exception of the topmost arch. However, this may be regarded as a coincidence.
The number of arches ranges with the size of the house. Four above and four below the middle arch, making nine in all, is a fair average. The large house at Iva had 12 above and 12 below, making a grand total of 25 arches.
The middle arch is both the longest and largest in diameter. The others decrease in diameter outwards from the middle arch. In the Fitiutan house described, the middle arch was 5.25 inches in diameter. The lower series of four, from above downwards, were 4.75, 4.25, 4, and 3.75 inches respectively. The curb plate was the same size as the middle arch, 5.25 inches, while the wall plate was larger with 5.75 inches. The individual arches were of the same diameter throughout from middle to ends. The ends were all cut square with the long axis, except the middle arch, which was shaped as described.
Lashing the thatch rafters. The thatch rafters are first attached at their upper ends to the moamoa end pieces in the middle line, and then to the end rafters as they diverge on either side. These lashings are the only real connection between the framework of the end and middle sections. The short moamoa of figure 39 is a very small intermediate connection between the thatch rafters and the ridgepole, while the second curved type forms no connection at all. The arches merely touch parts of the framework of the middle section. The lashing of the curb plate to the end rafters is temporary to get its right level. The stability of the house shows the strength of a combination of weak elements, such as the thatch rafters.
The pointed upper ends of the thatch rafters are passed across the inner or under surface of the end rafters to project beyond their inner edge. (See fig. 42.)
The lashing to the end rafters is the same as that of the intermediate purlins to the main rafters in the middle section (fig. 25). The sennit brait is tied round the end rafter with the slip knot commencement. (See fig. 43.)
The lashing to the end rafters is termed
fatu-o-le-ulu-'aso. Each thatch rafter is lashed to each arch it passes over. The order is again from above downwards so that any bends in the thatch rafters may be straightened out permanently by the successive lashings. On each arch, the lashing is of necessity from the middle line outwards as the thatch rafters have to be kept parallel with those first laid down in the middle line. The spacing is the same as in the middle section of the house.As the thatch rafters work outwards, the lashings come to the joints of the arches.
The technique is shown in figure 43 c. There is no technical difficulty, though a badly fitted arch may sometimes be seen gaping slightly at a join.All thatch rafters reach the curb plate and are lashed to it in the same way as to the arches.
Intermediate small purlins are added between the arches in the same way as in the middle section. They are finer than those of the sides, being simply the same as the thatch rafters with the so'o mata sai joins. They readily take the curve of the end section. They are two to each space between the arches and three between the lowest arch and the curb plate. Those above the middle arch are brought to the outer edge of the end rafters and those below to the curb plate. Their relative position is shown in figure 42 b. When a larger number of arches than usual diminish the spaces between them, there may be only one intermediate purlin used instead of two.
The eave battens are attached to the lower end of the thatch rafters by the same lashing as in the middle section. Shorter straight sections are used to take the curve.
The wall posts supporting the middle and ends of the curb plate are usually erected by builders.
The other end section is erected in the same way, but they rarely agree in exact length. When brought to the same stage as described above, the work of the builders is ended.
Variations in construction occur through the builders introducing something new to advertise themselves. At
Use of the slant join (so'o mata sai). Why the slant join should be preferred to the square join is not so obvious when the arches are in position and the framework lashed together. The slant join resists a pull or force that would tear the join apart. The weight of the roof, however, should exercise a downward pressure towards the joins, especially in those arches resting on the curb plate. The only outward pressure to separate the joins could be exercised by the wind blowing in under the roof. But the fact that the ends of the arches are not fastened in any way to the end rafters or the curb plate would make the form of the join of no extra value in resisting such a contingency. The roof would have to depend on the lashing of the thatch rafters and means of shutting the wind out. The Samoan did not enter into a scientific consideration of strains and stresses, but progressed from one obvious advantage to another. The obvious advantage of the join occurred during construction when the arches were lifted up with ropes and a), in spite of lashings, some of the end pieces would work loose and pull off. The join depicted by c) is a slight modification of the square join and does not clearly represent the slant so'o mata sai join as he intended. With the true slant join, however, the lave surfaces at right angles to the slanted surfaces formed a hook or projection, which materially assisted in preventing the lower pieces from dropping off. The same applies to the long thatch rafters after their upper ends are fixed. There was considerable weight on the lower ends, and much movement when they were spaced and stretched on the various purlins and arches. The short so'o mata sai join was used on the small thatch rafters for the same reason as in the larger arches; it prevented the lower pieces from dropping off during construction. Thus many obscure points in the completed object are simply explained by the steps in construction, but unfortunately the technical details are not always available. Failing these details, highly scientific reasons may be credited to a people who obtained them as end results in carrying out some totally different thought process that provided simple and obvious advantages.
Wall posts (pou lalo). The erection of the wall posts was completed by the people of the taufale. Any suitable wood was used. The bark was peeled off leaving rounded posts from 5 to 6 inches in diameter. The lower ends were imbedded in the ground, and the upper ends attached to the wall and curb plates. The length was thus decided by the distance the wall plates were suspended from the ground. When a high paepae platform was made, most of the posts did not reach the ground level, though they were sunk below the floor level. When the floor was put in, the height from the ground to the wall plate usually ranged from 6 to 7 feet. The house at Iva was exceptional with wall posts 7 feet, 10 inches in height. It seems probable that they were shorter in early times, for Wilkes (42, vol. 2, p. 71) describes the house occupied by the
The number of posts on each side of the middle section were three, and in a, b, c.) My experience was that the post was not usually cut square as in a, but at a slant as in d. The shaping of the post to receive the plate is termed fa'afulu lupe. The naming marks the transition to better workmanship. The shaping is not so much to give extra support to the plate as to provide better fitting. In many of the posts examined the plate did not rest on the bottom of the notch. This feature is seen in figure 47j. It draws attention again to the fact that the plates receive their support from above, from the already fixed rafters. Many of the posts arc quite loose, and can be removed without materially affecting the support of the roof. The Samoans themselves drew attention to this fact.
The lashing is simple and follows the rafter-ridgepole lashing (fig. 17) and the rafter-purlin lashing (fig. 20). Where two wooden elements cross each other at right angles the simplest ties consist of using transverse and diagonal turns. The turns are made carefully on the surface that shows so that the ornamental design may be apparent. On wall posts, the decorative surface is the inner part of the upper end which is defined into a rectangular space bounded by the Upper and lower borders of the horizontal wall or curb plate at the back. The surface is convex from side to side so that the actual corners of the space may not be seen directly from the front. In Samoan technique the diagonal and transverse turns are the complement of each other. Transverse turns are those at right angles to the long axis of the wooden element so that on the post they are horizontal while on the plate they are vertical. If transverse turns are made over one wooden element, diagonal turns must follow on the other crossing element and vice-versa. All lashings are finished off with a few circumferential turns passing around the outside of the previous turns where they cross between the two wooden elements. Of various designs used, the simplest is the single lozenge (fig. 47) common 111 the houses on a) is identical with that used on the back of the supporting post when lashing the collar beam to it. (See figure 30b and c.)
The wall posts are collectively termed pou fesisi, and also atuao on poupou'i, and to replace a rotted post with
suitu. The middle post of the right round end of the house is termed to'o matua tala. The middle posts at either end are often distinguished by being larger, ornamented with a more elaborate lashing, or even carved. A
The outer elements of the roof framework consist of the thatch rafters, 2 to 3 inches apart. As only every eighth rafter or so is used for tying the thatch to, their close setting emphasizes their use in strengthening the roof structure and adding ornamentation not only in themselves but in the extra sennit bindings round the purlins.
The thatch material consists of the sugar cane leaves (tolo) for the superior houses, and of the coconut for others. Pandanus leaves (lau fasa) are said to have been used in the remote past.
Different kinds of sugar cane are recognized, but the fatu, with narrow leaves and a dark skin, was used for thatch. The leaves are collected (fua) by women and carried in large bundles strapped to the back (fafanga). The leaves were stripped with the sheath part that enclosed the cane still attached to the leaf. In the village they are sewn together (sui) over light rods of cane (u) or the stem of a creeping plant (lafo). The leaves are pinned together with dry coconut leaflet midribs (tuaniu), the hard butt end acting as a needle point. The sewing of thatch sheets is done by a working bee of women relatives.
The rods range from 3 feet upwards. Large leaves may be used singly but small ones are placed together in twos and threes with the same surfaces facing in the same direction. (See fig. 51.)
The cane leaf is lau tolo while a thatch sheet of cane leaf is referred to as lau. Though lau is the general word for leaf, lau, as applied to thatch, definitely means sugar cane leaf.
Coconut leaflets, stripped from dried leaves and attached to rods, are used in good houses only when sugar cane is not procurable. Misa's large house at
All cooking houses' and canoe sheds are thatched with plaited split leaves called lau pola. Here nature has supplied the rod in the form of the leaf midrib, and also attached the leaflets to it. The check technique of the plait will be be described on page 169 under "Plaiting." Plaited coconut sheets are not so waterproof as the pinned leaflets owing to spaces between the leaflets, but to obviate this, more are put on. They save labor, and are quite good enough for houses that are not lived in.
The thatch (atofanga), derived from the verb ato (to thatch), works upwards in rows or sections. Assistants place the sheets in position from outside while the thatchers from the scaffolding inside lash them successively with continuous sennit braid along the upward course of every seventh or eighth thatch rafter. Thatching commences on the middle section without waiting until the framework of the ends is finished.
Ladders (atolau or apefa'i) are used to place the higher thatch sheets in position. They are also required for attaching the roof ridging. The name atolau (ato, to thatch, and lau, thatch sheet of sugar cane leaves) shows the association of ladders with thatching. They consist of two long poles with cross rungs like European ladders. They differ, however, in their extreme narrowness. One seen in to the middle pole without cutting it, and similarly to the third pole. (See Plate I, B.)
Thatching needles (lave lau) made of any hard wood are used. (See Plate III, A.) Notches are made behind the points to form hooks for picking up the braid. The point of interest in the Samoan implements is the round section of some of them, and the shallow nature of the hooks. (See fig. 52.)
The sennit braid used is not coiled into working hanks owing to the whole length having to be pulled through the thatch sheet. There are two methods of fastening by half stitches (fig. 53), and by overhand knots (fig. 54).
The overhand knot is the firmer method. For rethatching the house, the half hitch method allows the sennit to be easily removed, and thus used again.
The thatch sheets are placed transversely across the thatch rafters ('aso) with the rod edge above and the rough surface with the sheath ends to the outside. The first row is attached close to the eave batten. Each row as it proceeds upwards overlaps the preceding row, and thus sheds the rain.
The first few rows are generally very close together. The closest spacing is when the rods of the upper sheets press against those of the lower. This is termed taolafo from tao, to press down, and lafo, the rod material. A slightly wider spacing is where the level of the pin presses against the rod below. This is taotuaniu (from tao, to press and tuaniu, the leaflet midrib sheet pins). These terms, however, are especially used in overlapping the ends of the sheets on the same horizontal row. From the close spacing near the eave batten, the rows gradually space out to two or three inches, and continue so to the roof. The spacing of the sheets depends on the thatch material available. The closer the sheets, the more impermeable the roof to rain, and the longer the life of the roof; the greater also is the quantity required.
The thatchers inside work upwards on the same thatch rafter. Several thatchers have to work together as the lashing goes up on every eighth thatch rafter or so. There may be three men to lash one strip of thatch. The overlap at the ends is also attended to. The upper edges of two adjacent sheets are never on quite the same level, as the rod part is put above or below the other in order to make the necessary overlap ride easily. The men all work on the same row, and the strip or section is carried up to the ridgepole. ine'i lau. The overlapping of thatch sheets is fa'asua'i, and the ends of the sheets where they overlap each other, is su'enga, also ululau. In commencing another strip or section of thatching, the leaves of the sheets already fastened have to be lifted up to allow the ends of the sheets being added to overlap without confusion.
Thatching paddle. A special paddle-shaped implement (alai) is used. Thatching paddle is an appropriate name for it, as ordinary canoe paddles are often used when the special implement is lacking. The typical thatching paddle (alai) figured in Plate III, C, is made of a heavy wood. The blade is sharp edged, and slopes evenly back to the handle. One surface has slightly more transverse convexity than the other, and is slightly concave longitudinally. The tip is formed into a blunt point by the meeting of the blade's curving side edges. The handle is round in sections and its end is cut off square.
When the first sheet is being placed against its corresponding horizontal row, the blade of the alai is inserted under the leaves of the sheet above, between the leaves and the thatch rafters. The leaves are lifted up and the new sheet placed in position with the necessary overlap to the outside of the fixed sheet.
From the method of sticking the alai in amongst a mass of leaves already fixed has originated a saying in alai. He quotes first as his justification: "O lea tu le alai." (There will stand the thatching paddle.) The respect which Samoans have for such figures of speech enables the intruder to unburden his mind.
The ridging is attached after the middle section of the house is completely thatched on both sides. The ridging as a whole is taualunga, but the plaited coconut leaf sheets which are used as covers are termed fa'atafiti. The fa'atafiti sheets are made of single whole coconut leaves and two half leaves, plaited together so that the whole leaf midrib is in the middle line, and the split midrib some inches away on either side. The middle midrib is placed on the upper ridgepole, 'au'au lunga, and the split midribs hang down on either side parallel with it. Wooden pins are passed through from side to side above the split side midribs and below the upper ridgepole. This is termed susu'i. The ends of the sheets are overlapped, and more than one layer may be used. The upper ridgepole thus keeps the ridging pins in position, and the split midribs keep the sheets from working up over the pins. The pins must be long so as to project well on either side. The leaflets from the edges of the roof sheets hang down and overlap the uppermost rows of thatch.
As the framework of the end sections are finished, the thatch is put on
The junction between the middle and end sections is the last part to be thatched. This junction is called the pepe. When a house reaches this stage, it is well on its way to completion. In lau-a-imoa and is used in conversation to denote the advance made. The origin of the term is described in the following folk tale.
Tradition of unfinished thatchAlo-maunanae was the son of Nai-saafa, high chief of the Amoa district in
Savaii . His mother wasSina -mata-imoa, who had a rat's head. (The termimoameans rat.) Alo married Meto-tangi-vale, daughter of Punga, who lived at Leuo between Puapua andLe Alatele . He met her through a dart throwing competition, which is another story.Owing to Meto's natural importunity for a house after their union, Alo approached his parents. They referred him to Imoa-ita (Savage rat) who, however, refused to act as head builder. Alo thereupon cursed him, saying, "For your discourtesy to me, may you die by the wayside and be trodden underfoot by the passersby." (Dead rats lying by the roadside in Samoa are the result of an ill advised refusal by one of their ancestors.) Imoa-sina (White rat) accepted the commission and commenced building with his rat carpenters. Alo's mother told him he was on no account to move or make a sound in his sleep while the house was being built. The framework of the house was of
toa(ironwood) and the thatch of'ula(red feathers). All was completed except the thatch at thepepejunction when Alo moved in his sleep. The startled carpenters immediately left the work. Hence the saying,"lau-a-imoa"(the thatch as the rats left it).
In spite of this small omission, the house was evidently a good one for it is quoted in another saying when speaking of a good house, "'Ua o le fale na i Amoa, e pou i toa, ae lau i'ula." (Like the housei n Amoa, the posts were of ironwood and the thatch of red feathers).
The life of an ordinary roof of cane leaves is 6 to 7 years, but a well-thatched roof, 8 to 10 years. Old thatch is falu lau. When the thatch becomes so old that the roof leaks, the house is rethatched (ulu). It is often done in sections where the leak occurs, and the roof has a patchy appearance. To thatch over old thatch is fatu'ulu.
The last process in thatching is cutting the ends of the leaves level over the eaves (tuluiulu). In important houses, the builder came back to do this after the taufale's party had completed the thatching. The builders left after the completion of the framework of the second end section. The cutting of the eaves, however, was part of their duty as recited at the final payment feast. The part under the eaves is termed pa'usisi.
In all dwelling houses and guest houses, the floor is raised above the level of the surrounding ground surface. This necessitates the building of a raised platform (paepae). Except where the house is built on an old site, the platform and floor are always built after the house is constructed and thatched. The height and extent of the house platform depend on the status of the house. In guest houses, this also depends on the status of the owner. The platform, though built last, came into the master builder's calculations before construction commenced. It has been seen that the main posts of the house were sunk into the ground. The height of the proposed platform had therefore to be added to the height of the main posts. It also affected the level on the rafters to which the wall plates are attached. A completed house that has yet to have a high platform added is a curious sight. It hangs in the air, and emphasizes the peculiarity of Samoan house construction, the somewhat unique principle of hanging the frame from the ridgepole.
The platform material is stone. Exceptions to the stone platform occurred in the uncommon instances where stone material is too difficult to obtain. Such a condition obtains in an old village site on a tableland near Mauase,
The ground is of red earth, unsuitable for cultivation, and containing little Stone. Such stone as existed was in the form of smaller pieces that had been used to cover the floor, and had evidently been carried to the site. The tableland is the flat surface of a ridge between two streams. As several trenches ran across the ridge, the village may have been a place of retreat for defence. In places that are not of permanent or long continued occupation there is naturally no incentive to carry stone from long distances to build house platforms.
In another village site on the same property, plenty of stone is present in the ground. The house sites here have the usual stone platforms, but they are not high. Numerous trenches also exist.
In some of the present villages, the stone is transported from fair distances, sometimes by canoes to the nearest point on the beach. At Fangamalo in taufolo were prepared. Cooperation in such matters is marked by gaiety and laughter, which helps the shoulders to bear the strain. Time is of no commercial value, and days are allotted to the task from time to time to prevent too much interference with ordinary routine. It is surprising what a vast quantity of stone can be transported by primitive methods by a large number of people in a long time. I was in Fangamalo for two months and both platforms were far from completion when I left. However, there was no hurry. One thing is certain: the platforms will be completed at some time.
The platforms vary in extent. Some are confined to the immediate boundaries of the house. These are usually low, consisting of one course of stone (lautasi, lau meaning depth, and tasi, one). The face of the platform is composed of lane even sized stones, but they are not worked in any way. In low platforms, the area outside is covered with smaller stones to form a loose payment. The shape usually conforms to the sides of the house, but not necessarily so. The outer margins are defined by larger stones forming a continuous line.
In the higher platforms, the house part is made first; it is then occupied and the rest of the platform made later. In the large house at Iva, the stone platform was 6 feet, 6 inches in height from the level ground surface. The house which was 45 feet, 6 inches by 54 feet, 8 inches, had been long in occupation. For about a foot external to the wall posts, the house platform had been completed. There were four sloping approaches of stone from the front, back and two sides. The sides of the platform between the approaches were perpendicular. The spaces between approaches had to be built up to the same level to complete the platform. The matter was being kept steadily in view. With the higher platforms, the outer part was built into terraces working outward and downward from the house. The terraces varied in width, but each face was sharply defined with larger stones. The extent of these terraces again depended on the quantity of stone readily accessible.
In
It may, therefore, be stated that the stone platform to form a house floor is regarded as a necessity. The outside extensions of the platform into terraces and loose stone pavements are extra embellishments that depend on having stone on the site. The use of stone alone as filling was present in the platforms
The floor was formed by levelling off the upper surface of the platform within the house and covering it with small stones ('ili'ili). Here again the floor covering depended on available material. Broken stone with the edges unworn was used. Where waterworn pebbles of basalt were available from streams or beaches, they were preferred. At Fitiuta in
The folklore tale that refers to the uncompleted thatch also mentions a curious form of floor covering.
The tale of the unfinished floorAfter the house of Alo-maunanae had been built, his wife Meto desired that it should be floored with
lei(whale ivory). Again Alo sought his mother, who, exasperated, said, "Ola ! Se tama'ita'i e tāngi vale." ("Out upon it! This woman cries for the impossible").
The term tăngi is to cry; the a is short—tăngi vale is to cry like a child, but tāngi vale (with the a long) is the crying of an adult for something impracticable or impossible. Hence the saying, "Ua a tāngi vale ua ona o 'oe o Meto." ("You cry for the impossible as if you were Meto").
However, Alo was advised to go to his sister's husband, Tingilau, who dwelt at Faleatai in
Upolu . He was asked by Tingilau if he preferred thelei uli(darklei)orlei tumau(lastinglei). Alo refused both, and he was given thelei fanau(leiwhich reproduces itself) with instructions to paddle straight for home without turning around. On the voyage across toSavaii , Alo twice heard the sound of theleias it split and multiplied behind him. Just off Puapua, he heard it again. Curiosity overcame him, he turned to look, whereupon theleidisappeared. Deep down at this spot, white pebbles lying on the bottom of the sea represent thelei fanauthat never carpeted the floor of Meto, who cried for the impossible.
Some houses have the wall plates rather higher off the floor level than usual owing to the fact that the floor level has not been built up to the level originally intended. Thus Tufele of
The stone platform was often added to by chiefs to celebrate a further marriage. As polygamy was in vogue, the extent and number of stone terraces sometimes not always,' indicated the chief's various matrimonial alliances. Each additional one was a delicate compliment to the new wife. The new wife probably took pleasure in gazing at the other stone terraces that hers had made out of date.
More names are applied to houses than the types described. This is due to the Samoan usage of attaching qualifying words to the noun to describe material, construction, condition, and use in order to save further description. Thus a fale 'ulu is a house made of breadfruit wood, but it conveys no further information to the person unacquainted with Samoan house construction. The Samoan, however, knows that breadfruit is used in important houses; guest houses, and not kitchens or ordinary dwelling houses. Similarly; falcolamea means a house made with olemea wood, hut it is only the fine thatch rafters that are formed of olamea rods. The word used depends upon what particular aspect of the house is being discussed. As confusion is likely to occur from the number of names given in
The fale'ulu and faleolamea have been mentioned. A house of any wood that is not breadfruit is a fale vao, which name also conveys the idea that it is not a proper guest house. One thatched with coconut leaves regardless of type is a falelaupola. In the construction of a falema'o, ma'o wood is used. In a falema'a, ma'a (stone) is used; such as, falema'a.
A falefa'aafolau is a house built like an afolau (canoe house), and is thus a long house. This name has been shortened to faleafolan. The term faletele has come to describe the round guest house type. A falesae, according to fau sasae) used in dwelling houses. A plastered house is a falevali and is modern. A fale fa'alaufao is one in which the sides are closed down with thatch.
Condition: A faleaina is a house that is inhabited, and a faleofei one that has been cleaned and prepared for guests or others. A house strengthened to resist storms (afa) is a faleafa.
Use: The terms faleumu, umu, faleuli, tunoa, and paito are all used for cooking houses. Both faleto'a and faletofa are applied to the house a chief sleeps in. Palefuitui is a talking chief's house in which chiefs gather for discussion. A house in which visitors are housed is faletalimalo.
A rough house near the beach constructed for fishermen while they are faleapifangota. A faleseu is a fowler's house for netting pigeons, of which there are usually four, named according to position: falemua, faleva'ai, falematua, palalau, and fa'alele. A fowl house is a falemoa and faleui and maliunga, the latrine.
A faleta is a carpenter's shed, falelalanga, a house used by a number of women to plait fine mats in, and faleta'a, a brothel. (See
The term laulautu is applied to the house on the chief's death, and the house where the mourners gather is a faleniu. The falelauasi, a house of sandal wood, is figuratively applied to dead chiefs who are to be buried.
The terms fale'oa and faleoloa (stores), and falepuipui (prison), are of modern usage. Small houses that stand at the back of the main house are termed faleo'o and faletua. Faleo'o now implies the ordinary dwelling house, and faletua, a latrine.
Figuratively, faleupolu applies to the people and particularly the body of talking chiefs who support a high chief; faletolu, falefa, and falelima refer to various unions of villages. There are other usages where fale (house) is used in conjunction with another word in a figurative sense.
aitu, principally war gods, were honored with dwellings called faleaitu and also malumalu-o-le-aitu. They were built in the ordinary way, and there was nothing to distinguish them except that they were set apart. The double curve in the roof shown by va'a-o-taua-o-aitu tau, which he interprets as war-ships of the war-gods. They were placed in the principal malae of the village, and surrounded by a low fence. Raised stone platforms (fanua tanu) were associated with them. The buildings were raised by the village. The height and size of the platform varied with the respect in which the god was held. Priests went to their houses to consult the gods. The well-known Fale-o-le Fee was the house of Le Fee, the war god of Aana and Faleata.
There are now no special store houses for food. There seems to be no need for them. fata as a raised house for storing yams. No information was obtained.
Shelves. Associated with the woodwork of the house are certain elements which from their function, may be regarded as shelves. They differ from the accepted idea of shelves in being formed of beams or poles instead of boards. Fairly wide spaces may be present between the wooden elements. The woodwork used as shelves may be divided into three kinds; talitali, so'a collar beams, and fata.
The talitali are two special cross beams lashed horizontally to the main posts of a round house. They are placed about 2 feet below the lower tier of collar beams and run at right angles to the collar beams. Where there is more than one main post they are lashed to each post. (See fig. 55.)
The talitali is always attached by the expert builders as part of their contract. The two parallel arms of the talitali are used as a shelf on which baskets or other objects are laid transversely. Baskets may also be hung over the pointed ends. Additional talitali are sometimes seen higher up on the posts.
The horizontal so'a cross beams form important shelves upon which bundles of mats, bark cloth and house property are stored. As three main posts require a set of four collar beams, a shelf at least 6 feet in width is for provided. (See fig. 56.)
In long houses, the upper tiers of collar beams may be utilised as a shelf for articles not often required but as they form a rather narrow shelf, it is usual to place two or more longitudinal poles or beams upon them sufficiently spaced apart to form a wider shelf. There is no need to lash them together. More than one set of tiers may be used if the wealth of the family requires it.
The fata is a special platform of poles that usually stretches between the
In a round house at Fangasa, talitali at their inner ends and on the wall plate of the back wall at their outer ends. To doubly secure the outer ends, special posts were set on either side of the outer beams. A crossbar that passed beneath the beams was lashed to the two special posts.
In a long house at fata was made as in figure 57.
In both types of fata, the level was the upper surface of the wall plate and thus within easy reach. The fata was used for storing plaited food trays and the articles in everyday use. The talitali and fata were service shelves whilst the higher collar beam shelves which had to be reached by a ladder were store shelves.
Another special shelf (sasanga) was used to support coconut shell water bottles. (See fig. 58.)
For small objects, the upper surface of the wall and curb plates as they slope back to meet the thatch rafters forms a good shelf. Objects are also placed on the top of the wall posts.
Wall screens (pola). Dwelling and guest houses are not complete without sets of wall screens which pass completely round the house and can be raised or lowered at will. They are never used with cooking houses.
The screen mats are formed of plaited, half coconut leaves. (See p. 176.) Before plaiting, the leaves are cut so that their length will overlap two wall posts. The plaited depth is a foot or slightly more. Of the two long edges, one is formed by the split leaf midrib and the other by the thick three-ply braid finish of the plaiting. The mats are hung in sets from the wall plate to cover the distance between it and the floor. Allowing for overlap, it takes from six to eight mats to each set, which forms a vertical panel of the width of one mat and covers the space between two wall posts. The mats are supported at their upper edges by two continuous sennit cords spaced equidistant between the middle line and the side edges.
The method of attachment to the wall plate is shown in figure 59.
The sheets of the panel in figure 59b are shown overlapping the two wall posts on their outer side. Thus, as the wind blows in towards the house, the wall screens are let down on that side, and the overlap against the posts prevents them from being blown inside. When let down, the screens form the walls of the house. The closeness of the wall posts is accounted for by the
To raise the panels, the lowest sheet is lifted from the inside of the house by placing the hands under the lower edge and catching the lower edges of the other sheets behind as it is raised upwards. The middle cord (fig. 59b, 8) which is behind, is brought forward under the sheets and looped round its upper end near the wall plate with a couple of half hitches. It may be necessary at times to push out an adjacent panel to clear the overlapping edges. To let down the panel the middle cord is simply unhitched, and the panel straightens out with its own weight.
The adjustable screen panels make the Samoan house eminently suited to the tropics. When there is no strong wind or rain, all the panels are raised. The high thatched roof is much cooler than corrugated iron, and the open walls make them much superior to plank houses in hot weather. As sleeping-porches, they are as good, if not better, than those devised by higher cultures.
The plaited screen sheet is called pola in general, and pola sisi in particular. Sometimes a sheet works loose and falls outside (tau fafo). Anything of little importance is idiomatically referred to as a pola tau fafo, the screen which drops outside.
Fireplace. Though cooking was done in a separate house, fires were on occasion used in the dwelling and guest houses. A hearth (ta'ingafi) was formed on the floor by imbedding a number of waterworn stones on edge where the shape permitted to enclose a small square or rectangular shallow depression. This was the ta'ingaafi (from ta'i, to tend, and afi, fire), also avangalafu.
In houses not occupied for some time, a fire was lighted to clear away any mouldy smell. It was also used for lighting purposes. At a school festival in a guest house on
Lighting. In addition to the light of a fire, other forms were used that were referred to as moli, which has now come to mean a lamp. There are two forms of moli.
The term moli means coconut oil, but any doubt is prevented by using the term pôpô (mature coconut) to qualify it. Coconut oil was prepared from the flesh of the mature nut. The receptacle was formed of the half of a mature coconut with the flesh left in to protect the shell from the heat. The wick of dry coconut leaflet midrib (tuaniu) was wrapped round with a strip of thin undyed bark cloth. The midrib end was stuck upright in the flesh at the bottom of the nut.
The expression moli lama, is candlenut light. The candlenuts, cooked in small baskets on the leaf covered hot stones of an earth oven, were cracked to extract the kernels. Ten to twelve kernels were threaded on a dry coconut leaflet midrib, ana stuck in the ground or bent over a stone or half coconut shell. The end kernel was lighted and the others caught alight in turn. An attendant revolved the midrib from time to time to ensure the nuts burning evenly. Both the candlenut tree and the nut are called lama.
Floor mats. Part of the essential equipment is a sufficient number of floor mats to cover the whole floor space. When the guest house is not in use, the mats are rolled up and stacked on the shelves. When a few people enter to rest or talk, sufficient mats are taken down to cover the part that is occupied. The Samoans do not sit on the bare gravel, neither do they usually cover the parts of the floor that are not occupied. On ceremonial occasions, and when guests occupy the house, the whole floor area is covered.
Floor mats consist of three kinds: polavai or tapa'au, made of coconut leaf, papa, coarse strips of pandanus, and fala, narrower strips of a different kind of pandanus.
The polavai are laid directly on the coarse gravel floor, and the papa and fala above them. The papa mats are also called paongo from the name of the pandanus, and they sometimes take the place of the coconut leaf mats. The fala also takes its name from the fala pandanus. The manufacture is described under Plaiting.
The mats average about 6 feet in length and 3 feet in width, though larger ones are made. They are spread with the length parallel with the house walls. If a chief wishes to stretch his legs from the cross-legged position strictly demanded by etiquette, he pulls a floor mat over his feet and legs. Such an action prevents a breach of good manners.
The set of mats for a new house is plaited by the women of the chief's family (ainga), but on occasion he may get them elsewhere by taking advantage of a curious custom.
The
tu'u papaortu'u falacustom is the procedure whereby a chief may collect mats for a new house by asking for them at another village. He usually takes thetaupoumaid of his village, his talking chief, and some young men to carry home the spoil. On entering the village, thetaupougoes to the house of the women folk,, while the others enter the appropriate guest house. The village chiefs gather, and bring aua-luma aualuma 'avafor theingungawelcoming' bowl. When the visiting talking chief replies to the ceremonial greetings, he explains that the visit is atu'u papato collect mats for the chief's new house. The local chiefs then return to their houses and inform their wives. Meanwhile the visting taupou has also promulgated the object of the visit. The women get together and contribute so many each. A levy of one to two hundred is soon made up and duly presented, gratis, to the visiting chief. He usually stays the night, when alaulautasimeal contributed and shared by the local chiefs is given in the evening followed perhaps by dancing as a further expression of hospitality.The chief thus gets his quota of mats, and the few contributed by each family entails no hardship upon them.
Pratt definestu'u falaas going about begging for mats, but this conveys a wrong impression. It is a useful custom still in force, and carries no stigma. The chief gives his neighbors an honorable opportunity of assisting in the furnishing of his new house, and he himself is ready to reciprocate on any future occasion. It is immaterial whetherpapaorfalais used with the wordtu'u. A woman inSavaii who had been much annoyed by parties carrying out the introduced custom of celebrating at the New Year, determined to get something back by counter attacking with thetu'u falacustom. Accompanied by her children she made the rounds of the village, and returned home with over fifty floor mats.
Bedding. Inured to the hard floor from childhood, the mass of the people used bedding consisting of a floor mat with a cover of a sheet of bark cloth when needed. The bark cloth has given way to foreign cotton cloth sheets. For chiefs the sleeping mats consist of fala with narrower wefts than the floor mat. These are usually plaited with a twill stroke instead of the usual check. Some of these have geometrical designs introduced by using colored elements, but evidence points to this type being introduced. So also is the use of fringes of colored worsted. Sleeping mats are also made of lau'ie.
A raised bed termed ulumoenga is made by laying mats above each other to a height of a foot or so. Bark cloth is also used to raise the pile. The finer mats are placed above the coarse floor mats. Nowadays, sheets and modern pillows are used to finish off with, but mattresses are rarely used. The use of leaves and grasses under the floor mats has not been in vogue.
Pillows. The pillow of bamboo {'ali) is still much used. Thus, a floor mat, a sheet, and an 'ali serve the majority of the people. The man who uses a raised ulumoenga generally has a modern pillow stuffed with kapok which has been introduced throughout the islands. Bamboo is called 'ofe, the nodes pona, and the internodes upu. Large bamboo, about 3 inches in diameter, is used for the headrest. The length of the pillow is counted by internodes and ranges from one ('aliupu tasi) to several. The ends are always cut off close to the outer side of the end nodes. (See Pl. III, D, 2, 3.)
The legs (vae'ali) are formed in pairs from the forked branches of some light wood such as pualulu that will not split easily. In a typical pair, the legs are about 1 inch in diameter with a spread at their lower ends of 7.75 inches. The upper surface of the upper connecting part is 2 inches wide and is cut concave to fit the curve of the bamboo. In the middle line between the legs, the connecting wood is over an inch deep. The height of the legs is
The legs are lashed to the ends of the bamboo with sennit braid. In short pillows the legs are lashed close to the ends just inside the nodal joins whilst in long pillows they are placed much nearer the middle. Two methods of lashing are in vogue: round the circumference of the bamboo (fig. 60), or through a transversely bored hole (fig. 61).
The circumferential lashing does not prevent the legs from slipping around on the smooth bamboo, but as the other pair can be readily slipped around to the same position, the movement forms no serious objection to the method of lashing. The raised nodes prevent the lashing from working over the ends of the bamboo.
The lashing through holes prevents the legs from slipping round the bamboo, but it does not give such a neat appearance as the circumferential lashing. Probably it is a later development due to greater facilities for boring holes.
The "V" shaped legs of the bamboo pillow supply the "V" or chevron motive in tattooing and carving with the name of vae'ali. No corroborative information was obtained concerning the report that live snakes were sometimes confined in bamboo pillows.
An D, 1.)
Wooden pillows of the Tongan type are in use in Samoa. They were introduced and are not made locally.
People desiring to rest in the daytime call out for an 'ali. Most Samoans still prefer it to a modern stuffed pillow, as it gives firmer support. The back of the neck is rested upon it. When other people are present, it is bad manners to recline with the feet toward them, or toward the interior of the house. One may rest sitting Cross-legged against a wall post, or even stretch out the legs by covering them with a floor mat, but to slip further down into the reclining position is unpardonable. When the pillow is brought, it is turned towards the interior and the feet are thus stretched out towards the wall posts. With this proviso one may doze off and leave the conversation to others
Curtains. The rounded end of a guest house is sometimes partitioned off into a sleeping room. A cord or rope is stretched across about six feet above the floor. The ends are tied to convenient parts of the wooden framework. Over the cord a very large piece of bark cloth is hung as a curtain. The term pupuni (to shut in or enclose) is also given to the curtain.
Large pieces of bark cloth are specially prepared for curtains (pupuni). They are made long enough to extend across the width of the house, and are over six feet in depth to allow a fair overlap over the stretched cord. Nowadays the curtain is pegged to the cord. In all guest houses in which I stayed one end was curtained off to give privacy to the guest. Practically every guest house had a large pupuni curtain as a necessary part of its equipment.
Sometimes shorter lengths of cloth (potu) are used. Three or four pieces are necessary to shut off the width of the house. The whole combination forms a pupuni, but the individual pieces are potu. The part of the house screened off as a room is termed afeafe.
Mosquito curtains (ta'inamu). A tent of bark cloth has been in use from ancient times as a protection against mosquitoes (namu). References to ta'inamu in old legends prove their age. A large rectangular piece of cloth was doubled over a cord stretched six feet above the ground. The length along the cord had to be sufficient for the length of the occupant, and allow for the edges being sewed together at either end from the cord to the ground. The sewing was done by puncturing holes with a pointed stick and threading some fibrous material through them. The curtain was spread laterally by transverse curved rods passed under the cloth and above the cord which formed a ridgepole. The material had to be long enough not only to reach the floor all around, but to provide sufficient surplus upon which stones were placed to keep it in close contact with the floor. The curtains were effective but very warm and stuffy. Their use was the lesser of two evils.
The bark cloth ta'inamu has completely given away to the introduced trade article, which is in universal use. It is tied to stretched cords, and stones are used to keep the edges down.
Cords and ropes. Sennit braid and twisted or braided cords, or small ropes of fau bark, form part of the house furnishing to suspend mosquito curtains or hang clothes on when necessary.
Stones. In addition to the large gravel floor, larger waterworn stones of basalt are always to be found in the houses for use as weights, not only for
The floor gravel is also useful for expelling pigs and fowls. Dogs have the freedom of the house, but such missiles are used against them if their behavior needs reprimanding. The affection of a Samoan for his dog is psychological. He will allow it to starve and suffer from curable skin diseases, but he resents anyone throwing a stone at it.
Pegs. Straight or forked sticks are stuck in an upward slanting position into the thatch as pegs (to'otu) for use as hangers. They may be tied to the wall plate.
The tautaunga is a food-suspending cord protected from rats by a length of bamboo ('ofe). (See fig. 62.) The smooth outer surface of the bamboo prevents rats from climbing down the rope. These protectors are hung up in the dwelling house.
Food. Cooked food purposely left over from a previous meal may be hung up in its basket on the fata shelf. The plaited food platters (laulau) are kept on the fata.
Property. Rolls of floor mats not in use are kept on the supports formed by the lowest set of collar beams in a round house or the fata shelves. The property constituting the wealth of a chief, such as fine mats, fibre mats, and bark cloth, is rolled up in bundles (ta'ui) with an outer covering of bark cloth and tied with sennit or fau rope, and stored, well out of reach, on the upper tiers of collar beams of both long and round houses. This is the only method of storing, as receptacles such as boxes were not used. The wooden boxes cut out of the solid (tuluma) now frequently seen in Samoan houses, are an introduction from the fold, ta'ui bundles, the greater the apparent wealth of the family, and the greater their status. While greatly influenced by ceremonial and rank, the Samoan has a very commercial side to his nature, to which food and property strongly appeal. It is openly stated by them that negotiations for a marriage were often commenced by a talking chief for his own chief after seeing the abundance of ta'ui bundles displayed on the upper collar beam tiers of the guest house of an unmarried girl's father. The talking chief's own chief might be quite happily married, and have no personal desire for change. The girl might not even know him. Neither of these conditions affected the talking chief. The only thing that influenced him was the plentiful supply of ta'ui bundles. Material property came as a marriage gift with the bride to the husband's family. In the distribution which followed, the talking chief, from his official position of chief negotiator, established by custom, received a substantial share. Thus in social organization, the ta'ui bundles and the upper collar beam tiers played an important part. The high shelves, in full view of visiting chiefs, not only satisfied the material necessity for storing, but served the equally important psychological need for display. One wonders whether all the bundles contained what their exterior conveyed.
The tuinga headdress of human hair, and the feather 'ula, used with the headdress or as ceremonial kilts, were often stored in a split bamboo receptacle, tied up in bark cloth and hung from the roof like the food protector.
Brooms (salu) are made from green coconut leaflet midribs which, on drying, become hard and stiff. The leaflet midribs are first separated from the leaflet by splitting with the thumb nail on either side. The midrib is then torn from the leaf midrib with a sharp jerk which leaves the outer parts of the leaflet still attached to the leaf midrib. The husk of the mature coconut is also used after beating it to loosen up the fibre. Three kinds of salu are used.
A hand broom (salu lima, lima meaning hand) is formed by merely tying a conveniently sized bundle together near the butt ends. The tied part is held by the hand. An improved form is made by plaiting the butt ends of the midribs together in a three-ply braid. (See Plate III, B, 2.) In tearing them from the leaf midrib, thin strips of the leaf midrib are torn off; these enter into the braid. The braid is then neatly rolled on itself to form a convenient bundle of leaflet midribs. The lashing gives a neat finish to the butt end.
A handle (salu tu) is added to the hand broom type by arranging the butt ends of the midribs around the end of a stick, and lashing them together. (See Plate III, B, 1.)
A spider broom (salu pungaleveleve—pungaleveleve, spider) was devised for removing spider webs from the inside of the high roof. One form was made by lengthening the handle of the salu tu sufficiently to serve the purpose. A second form was made by using coconut husk fibre instead of leaflet midribs.
The first two forms are used for sweeping the floor and the adjacent surface of the stone platform. Women work very carefully over the floor in the mornings, picking up rubbish and the short fibres discarded from sennit braiding. Even the stones are picked over and replaced. Floor mats are put outside to air in the sun.
The weakness of Samoan houses is the join of the rounded ends to the end rafters of the middle section. There is danger that the wind will lift the thatch directly, and take the roof with it. To protect the thatch, heavy coconut leaves are tied together in pairs by their tip ends and straddled over the ridge. They hang down on either side, and by the weight of their midribs help to keep the thatch down. Covering the thatch in this manner is termed tanufale. (See Plate I, D.)
The danger of the ridging being blown off is guarded against in an additional manner by using the trunks of banana plants. Lengths of these are partly cut through the middle to allow of their being bent over the ridge. The weight is greater than coconut leaves, and a number of them are straddled along the ridge.
To prevent the wind blowing into the house during storms, the pola wall screens are reinforced by standing green coconut leaves closely together outside the lowered screens. Ropes are then passed around outside of them and lashed to the wall posts. The leaves may be plaited together. This protection is termed pālepoi or pāletā.
An additional protection against the wind entering the house and endangering the roof by lifting it off, is by means of a fence (tali matangi—matangi, wind). Posts are erected about six feet away on the windward side of the house and a single rail lashed to them. Green coconut leaves are set upright and closely together. They are lashed to the rail and thus form a close wall to break the force of the wind. The fence is higher than the wall plates, and prevents the force of the wind being directed against the weak walls of the house.
The craft of house-building, tradition relates, came direct from the god Tangaloa-matua. The first builders were convened by Tangaloa to consider the construction of a house. The assembly was named Sa Tangaloa (the pou tu. The question of getting a cross-piece mounted on top of the post was discussed. The builder, Manufili, erected two forked posts, placed a cross-bar over them, and then lashed oblique timbers to the main crossbar. With smaller crossbars lashed to the oblique,.timbers steps were formed up which the ridgepole was carried and placed on the pou tu. In compliment to the builder, the scaffolding was named fata-a-Manufili, since contracted to fatamanu.
The ridgepole was named 'au'au from 'au'auina (servant), as the person who suggested it was a servant of Tangaloa. The builders then searched the forest for suitable timber. Breadfruit wood ('ulu) was brought in by Malama, and accepted in preference to others. As a reward Tangaloa allowed Malama to use a two-branched thatch rafter ('aso manga lua) in buildings erected by him. He also received the kava cup title of 'aso fausia (the lashed thatch rafter), for the branched rafter was formed by lashing a shorter element to a long one to form the branch. Both Malama and Manufili became tufunga muamua (original master builders).
The principal rafters (fatunga) were suggested by Malama. The main purlins were proposed by Tangaloa-matua, and hence called la'au matua after him. Thatch rafters were made of the breadfruit wood brought in by Malama. Intermediate purlins were proposed by Luanga and received his name. So'a-fa suggested the collar beams. They were thus named so'a. Originally there were four (fa) to a house. Sauluanga proposed the eave batten. ('afa) should be used as lashing. An old song says that the builders from the sky had a discussion as to whether a house or a canoe should be built first. The decision went to the house and hence sennit was first used on a house.
In the rounded ends of the house the curb plate was set up first. The curved purlins (fau) were suggested by Malama. The fau tali 'aso arch near the ridgepole was the first erected. Then came the middle arch (fau tu) with one on either side of it called fau angai. The term angai means attendant. These angai were the original curved purlins, and hence the only ones with distinctive names. Others have since been introduced but have no individual names.
The above story was recited by pou tu) and that the house was therefore the fale tele (round house). It is difficult with traditions as now related to assess how much of the more recent has been projected into the past. Some days before, I had asked lau fasa) had ever been used as thatch. A good deal of discussion took place showing doubt, but finally it was decided lau fasa had been used in ancient times. During the recital of the above tradition I was informed that the original builders had a long discussion as to whether pandanus leaf or sugar cane leaf should be used as thatch. I could not help thinking that, though my name had not been mentioned among those at the original meeting of carpenters, my question of a few days previous had been referred back by
A matter-of-fact master builder in fau sasae). From the review made of Samoan building technique, his contention is the more probable. Even if the craft entered Samoa from beyond the horizon (langi) with the principles of technique already established, the
The Samoan tendency to rationalize, as seen in their explanation of the names of their islands by identifying them with pairs of ancestors, is present in the Tutuilan tradition of the naming of the parts of the framework of the house. So'a-fa and Luanga are used in this convenient manner to give names to the collar beams and intermediate purlins. Yet such prominent builders disappeared without being perpetuated in the name of a building society. In addition, Luanga is so unknown in pae'aso.
The first house built on earth by the Sa Tangaloa was named Faleula. Its site is to be seen at Fanga on the island of
The builders were called tufunga in general and tufunga fai fale in particular. In poua. Besides houses, they built the better classes of canoes. As the houses were held to have been derived from the first house built by the group organized by Tangaloa, so the builders themselves derived their origin as craftsmen from the first group known as the Sa Tangaloa. The original group has widened out into what
Within the guild itself, smaller groups or societies were formed who claimed origin from individual members of the original Sa Tangaloa. Originally they were direct members of his family (ainga). They became associated with particular high chiefs, and were thus called angai o tupu (the companions of kings). With increasing population and the wider spread of demands for expert housebuilding, the smaller societies increased in membership and spread from the immediate local association with particular chiefs to various villages throughout entire districts. The blood tie became practically ainga). In Samoan social organization in general, the blood tie weakened, and selection and election to various positions strengthened. Of the builders' societies within the guild, only one carries the name of a builder who figured prominently at the building of the first house. This is the ainga of
ainga of
The young men are apprenticed to experienced' builders, who are generally relatives. Thus they join the Society of the district in which they live. When they become expert they are admitted to full rank within the Society. matai or chief. The new matai cannot take his official position in the ceremonial kava drinking in his own village or elsewhere until the election made by his family is publicly notified and ratified by a meeting of the chiefs of the village. The new matai takes his place against one of the posts of the round guest house. Speeches are made, and his title name is called with his cup of kava. From then on he is publicly recognized and known by his new title name.
A relative, or someone with sufficient confidence in the new builder, gives him a commission to build a house. In such a construction, he takes the position of head builder (latu). He appoints his assistants from his own
Men of rank go into the builders' profession. In the Society they also hold their own rank in the outside world. The head builder in a contract is supreme head of his party. He is referred to ceremonially under three titles. He is the direct representative of the Sa Tangaloa, the companion of kings (angai o Tupu), and the personification of the ancestral head of his society group within the wider guild. If he belongs to the society of Sao, when he receives his cup of kava, the three titles are Sa Tangaloa, Angai o
'aso manga lua) that tradition states was personally conferred on a) forms an 'aso fausia. Hence the kava cup title of 'aso manga lua is represented in more modern times by the two thatch rafters cut out of one piece of wood. Two were noted at Annu'u Island and Vaitongi,
The trade mark of Sao, said to be a star (fetu), was placed on the moa-moa piece at the end of the ridgepole. Search for it proved vain in the Sao houses examined. A builder using the trade mark of another society was reported to the guild. The matter was discussed and the offending member fined.
A chief, seeing his way clear to assemble material, food, and gifts, sought out the services of a skilled house builder (tufuga fai fale). The builder tufuga than "carpenter." In selecting a builder, the chief considered his architectural qualifications as well as his manual ability. By accepting a contract directly from the chief, the builder became the head builder (latu) in the construction of the house. He employed the necessary number of skilled assistants from his own guild of builders and his authority was supreme. The homebuilder, or person desirous of a house, was the taufale and such he remained until the completion of, and payment for, the building. As he is constantly referred to in construction and ceremonial, the word taufale will be used for want of an exact expression in English that is free from confusion.
The taufale, on making his request to the builder, was asked, not whether the house was to be a fale afolau (long house) or a fale tele (round house), but whether it was to have utupoto (tie beams) or so'a (collar beams). Though so'a are used in the long house, in this question it is used in contradistinction to utupoto to indicate a round house which has no utupoto. The taufale having replied, the builder asked, "How many?" The reply indicated the size of the house. The builder made a mental estimation of the labor involved and probably sized up the ability of the taufale to carry out his side of the contract. If his decision was favorable he accepted the gift of a fine mat ('oloa) and the preliminary contract was thus sealed by a deposit. Nowadays its place may be taken by a preliminary deposit often dollars or two pounds sterling. Before building operations commenced, both parties met over the inevitable bowl of kava. The contract was orally recited and agreed fale angai contract.
The ordinary contract terms were recited to me by Nua of
The
taufaleand builder agree to become one family (feangainga) and live together in love and harmony during the construction of the house. The builder agrees to do everything he can to meet the desires of thetaufale. Thetaufaleagrees on his side to respect the laws and observances of the Sa Tangaloa. The question of material, transport to site, thatch, building the scaffolding, completion of the wall posts, and construction of the floor as part of thetaufale'sside of the contract has become so established that they are not necessarily enumerated. Thetaufalemust feed all the builders engaged on the work. When the food is cooked in the oven, that for the builders must be put in separate baskets. It must never be mixed with that for thetaufale'sparty attending to the builders and taking an interest in the work. Absence for any time is regarded as neglect and may lead to the builders abandoning the work. The attendant brings water, drinking nuts, and brews kava. In bringing food, the builders must be served first. After the building has commenced thetaufalemust not give food, bark cloth, or fine mats to anyone other than the builders. If any of the family or friends of thetaufalevisit him ceremonially, permission must first be obtained from the builders before he can give even food to them. If given without consulting the builders, the action is regarded as a lack of respect, and they abandon the work. In ceremonial drinking of kava between the two parties the head builder is served with the first cup, and his three titles are called. In all speeches he is addressed by the same titles.On his side, the builder promises to treat the
taufalewith all respect. He also calls thetaufaleby three titles in speeches and kava drinking. Ali'i Fai'oa (Chief of the Working Family), Taufale Ali'i (Chief of the House that is Being Built), and Ali'i Autapua'i (Chief of the Village). By mutually observing these titles, they place one another on the highest ceremonial plane. When eating in a house, if the builder's son comes to the outside of the house, the builder may give him a share of food wrapped up in leaves or in a basket. The boy must take the food home without opening the parcel. To open it at once to see what sort of food it contained would be lack of respect to thetaufale. A builder must not call from the scaffolding for anything below. He must come down for it himself. He can not use an adz on the framework. All fitting of timbers must be done in the shed or on the ground. All thatch rafters must be joined and lashed on the ground. No piece can be joined onto the thatch rafter once it has been raised to the oblique position on the frame. This does not apply, of course, to the arches. The builders must not eat or drink standing.
When the families of the taufale and builders visit the building ceremonially they bring presents of food such as pigs, talo, and kava. Their offerings (tautunga) are brought alternately. This alternate order is maintained until the house is finished.
The fale angai contract is strict. In the ordinary form of agreement, it will be seen that the taufale is in a fairly helpless position. He has to rely on the chief builder's sense of honor and that of his assistants. If he sees inferior work, he can not complain. To do so would show lack of respect to the builder who would leave the work.
To give the taufale a direct voice in the work, the fale angai agreement can be entered into. The taufale then has the right to directly watch the work and draw attention to any slovenly technique. He can tell the head builder
The head builder on his side sees that the wishes of the taufale are carried out and any mistakes rectified. But to do this he demands a higher standard of living. The food must be of the best and pork must figure more frequently. The family of a taufale are kept busy procuring fish and other foods. The builders will eat no cold food, so fresh ovens have to be made. They will drink no water, so baskets of drinking nuts must be ever on hand with an alert attendant ready to supply their demands. The prohibition of the taufale giving anything away is rendered doubly strict.
In the strict fale angai contract there is no trusting to honor, and no sentiment. The elimination of the sentiment of relationship is difficult to carry out, but is usually obviated by employing a head builder from some other village or district who has no blood tie with the taufale. The tauale gets a good house if his finances can stand the extra drain. At the finish of the work a higher rate of reward is expected than under the ordinary contract. If the taufale fails during the building to satisfy the demands of the builders, they have no hesitation in leaving the work, for in the fale angai there is no blood tie and no sentiment of forbearance. Each side is out to get the most it can. Before leaving on strike, the builders leave a sign in the house that acts as a warning to the entire guild. The Sa Tangaloa tapu the taufale and no other builders will complete the house. The only chance the taufale has of getting his house completed is to humble himself before the head builder with a substantial present of fine mats, and with the use of much ceremonial speech persuade him to resume work. The position is on a par with a civilized industrial strike for increase of wages. In Samoa, however, the strikers are in a more entrenched position than their trade union compatriots.
In a great number of villages throughout eastern and fale tala mutu. The open end of the middle section is closed with thatch and the house occupied as a dwelling. These are witnesses to the fact that the taufale's supply of food and material ran short, and that he was unable to have the building completed. He and his family usually wait until they have grown more pigs and planted extra crops. Material is again assembled and the builder reengaged to finish the work. I noticed a few fale tala mutu on
Occasionally the end section is completed by a different head builder. In a house at Taputimu, tala end section was completed by a different builder. He used the rod to mark off his work from the rest of the building. His work was much neater; he had used this device to draw attention to it and to thus advertise himself.
The building of an important house is accompanied by a number of set feasts. The food is provided by the taufale and his family. In the most important ones, the village assists with food. To be quite clear on this question, it must be remembered that in a large village there are a number of distinct families. These families may trace descent to a common ancester, but it is so far back that close connection is lost. The smaller groups connected by more recent blood ties are termed ainga. A chief may be senior in the village, exercise influence and have his position recognized in many practical ways. It is, however, to his own immediate family or ainga, consisting of brothers, uncles, sons, nephews and cousins, that he looks for sustained assistance during the building operations. They form the fai'oa (working family) from which the taufale gets the kava cup title of Ali'i Fai'oa. In an important feast the villagers outside of the taufale's immediate ainga recognize his position in the village, and their own kinship with him by bringing food contributions to the feast. They are here termed autapua'i and it is from the relationship of the taufale to them that he gets the kava cup title of Ali'i Autapua'i (Chief of the Villagers). Hence his kava cup titles have a direct bearing on the building operations, and they are used only for that period. The title of Ali'i Taufale (Chief Desiring a House) naturally lapses when his desire is realized. The title of Ali'i Fai'oa ceases when the work for which the family was mobilized is accomplished. The term Ali'i Autapua'i can not be used to express a specific relationship with the villagers after the special need that called it into operation has been satisfied.
The villagers themselves are happy to share in the festivities. Food ceremonies appeal to them especially as they get a share of the food. As far as they are concerned, the feast is like a picnic in which all the food is pooled and then redistributed. They often get better than they brought. It was admitted by my informants that they were not unaware of this. As one expressed it, a taufale gets little real assistance from the villagers as they practically get back what they bring. The whole thing is a fa'aaloalo (a ceremonial way of exhibiting in public their respect for the chief). It also forms a basis for the ceremonial speeches that add to the importance of the occasion. From this ceremonial observance the taufale gets psychological satisfaction, which balances the material obligations. The burden of supplying the food for the builders must fall upon him and his immediate ainga.
A most important feast takes place at the commencement of the work, when the main posts have been erected. I was fortunate enough to attend the feast connected with Misa's house at fa'atunga, and the feast receives the same name. The day is arranged beforehand to permit of food preparations. We travelled from the island of
The fa'atunga ceremony includes the actual erection of the posts and three meals, lavata'i, fa'atunga, and fui'ava. The lavata'i is an early morning meal cooked by one member of the taufale's family. He lights the fire, prepares and cooks the food unaided. There must be no noise of interruption. When ready, the meal is conveyed by him to the chief builder and his party, before they commence work. The lavata'i is a meal and not a feast.
After the meal the routine work of preparation is proceeded with. The three main posts are already placed in the open holes. The scaffolding has been erected. Towards noon, the posts are set perpendicular and rammed with the top ends level. The ridgepole is placed in position. By this time, the people begin to take up their positions; so the builders retire to their open shed. (See fig. 65.)
One side was formed by the working builders (Sa Tangaloa) in their shed, and the visiting builders on their left. Opposite was the taufale family (fai'oa). The left side was formed by the villagers (autapua'i). The open right side held the newly erected main posts set on the low stone platform of tufunga usu) consisted of those not employed on the building. All master builders within hail had no need to wait for a summons to attend the ceremony; it was their privilege as fellow craftsmen of the Sa Tangaloa to share in the food. Thus, they occupied the same side of the square as their working brethren. Their official title of Tufunga Usu means, "the builders who have come."
The taufale family sat in rows with the chief Misa and two others forming a front row. The villagers also sat in level rows with their highest talking chief occupying the middle of the front row. As the working builders were from ainga and invited to sit in the carpenters' shed. The whole assembly sat cross-legged on the sandy soil, with trees affording shade and setting to a picturesque scene.
The ceremonial was heralded by the characteristic Samoan yell (ailao) which came from behind the trees at the back of the taufale lines. Young men appeared carrying huge roots of green kava slung from poles and still bearing branches and leaves. The presentation of green kava ('ava mata) is the highest honor that can be paid to chiefs and is thus an appropriate tribute to the guild descended from the gods. About a dozen such roots were carried across the central space and deposited on the stone platform on the right. Then came the carriers of baskets of cooked food which were piled up in rows before the builders' shed. Accompanying them was a huge pig cooked whole, and others of smaller size. Several kegs of salt beef furnished a modern note. The food laid out in a lavish heap was fitting recognition of the angai o tupu (the companions of kings).
Then Misa, the chief of the fai'oa, taufale, and autapua'i, addressed the descendants of Sao. Misa is one of the rare combinations of high chief and high talking chief termed Ali'i taufale. In his capacity as taufale he spoke for himself as ali'i. The speech was delivered from the sitting position. The titles of the head builder were enumerated with punctilious ceremony and the greatness of the Sa Tangaloa done justice to. Apologies were made for the paucity and poor quality of the food placed before, such an august assembly. The hope was expressed in terms of the contract that nothing would man the brotherly relationship established between the two parties. Cooperation between them would result in a building that would bring joy to the owner and honor to the builder. Ceremonial phrases and ancient proverbs were repeated at length. A masterly exposition of Samoan oratory concluded with a modern invocation for a blessing on those assembled.
One of the three talking chiefs named Nua had the privilege of reply, as he was the highest ranking chief of the Sa Tangaloa. The head builder
The village then took up the oratory. Their talking chief, Lia, was of the high rank termed to'oto'o. He had been a to'oto'o when the vacant title claimed him. He repeated the honorary titles of the Sa Tangaloa, and demonstrated his qualifications for his position of chief orator of the village. When he concluded, the villagers brought forward their contribution of cooked pigs, kegs of beef and baskets of cooked food. As they were added to the pile in front of the shed, a lesser talking chief of the villagers treated them in the laulautasi method. The contents of each basket with the name of the giver was announced. He then handed it over to a receiving talking chief of the builders who placed it in the rows he was arranging with obvious satisfaction. While the baskets were being enumerated, the villagers brought the large dried roots of kava (tungase) and laid them on the ground before another receiving chief. The talking chief who received the food rearranged the pile. He then replied to the speech of the village and enumerated the articles of food to stress their quality, quantity, and variety. He spoke standing amidst the food. The third talking chief then asked the villagers to excuse them while they divided up the food.
The pigs were quickly cut up and the food divided into heaps under the direction of the talking chiefs. It is for occasions such as this that professional talking chiefs are appointed by the builders. They have to have quick judgment as to the quantity of food, and the number of people to be supplied. If any party gets less than they think they are justly entitled to, no hesitation is evinced in showing their dissatisfaction. As a protest a party may refuse to accept the share allotted to them. Such action shames the builders, who know that public opinion will condemn them as parsimonious. The food was divided up into large portions for the house family, the villagers, the visiting builders, and the Fale Tolu (ceremonial title applied to the village of
A ceremonial complication now seemed to take place. The villagers sent back a keg of beef to the builders. The Sa Tangaloa were immediately on the alert. The officiating talking chief demanded an explanation of their sign of dissatisfaction with the division of the food. He sent the keg of beef back without waiting the reply. The village talking chief replied that the portion had been returned not as an indication of parsimony, but of excessive liberality. The portion of the Sa Tangaloa had been depleted too much in their liberality to others. Once again the keg of beef traveled across the open space, and was hurled down beside the food heap of the Sa Tangaloa. The builders' talking chief called, "That being so let the surface of your minds be unruffled like a calm sea (Ia malu le vai o lou finangalo); we are the best judges as to whether we have been too liberal." He waved his hand towards the keg. One of his assistants picked it up, carried it across the open space and hurled it down beside the villagers' food pile. On this pile reposed the head of the large pig. The village talking chief, not to be outdone in spectacular liberality, waved his hand in turn towards the keg. As it recrossed the open space he called, "Enough, keep it. If you persist in arguing we will send the head of the pig over to join it." This ceremonial threat settled the argument. Honor on both sides was satisfied. The different groups set to work and divided up their shares into individual portions. This done they feasted where they sat. The fa'atunga feast was on.
The fui'ava is the ordinary evening meal preceded by ceremonial kava drinking (fui'ava a tufunga). It is here that the first cup of kava to the head builder with the reciting of his titles, gets full display. It forms a ceremonial ending to the day's proceedings with full accentuation of the prestige and status of the building craft. Unfortunately the full ceremony was abandoned owing to another ceremonial complication that had occurred. This, however, merits description as a further example of the Samoan psychology expressed in custom.
When I arrived the stage was set in groups as before. In front and to the side of the village talking chiefs, a man sat cross-legged with a fine mat spread out over his knees. The talking chief was apologizing to the Sa Tangaloa. One of the older builders with a show of indignation explained to me that the Sa Tangaloa had been insulted. The village orator went on to say that the incident was quite accidental. Everyone, most of all the villagers, knew the high traditional origin and the present prestige of the builders. He begged them to forgive the accident and not let it interfere with the smooth building of the house; to let the surface of their minds remain unruffled. The old builder here informed me that in olden
"What was the trouble?" I asked.
"It was terrible," he replied.
Here the builders' talking chief replied, pointing out that such things were not done. However, as he had their assurance that the incident was purely accidental, the builders were inclined to overlook it in their generosity of heart. They were the descendants of gods, the companions of kings and the family of Sao. Nothing of like nature must occur again or they would not be so lenient. The village apologist redoubled his protestations. The bearer of the fine mat rose to his feet, advanced across the plaza and laid the mat before the Sa Tangaloa. It was the self-inflicted fine of the village, and a recompense for guilt. Such having been publicly made manifest, the honor of the builders was satisfied. The mat was accepted and the incident closed. During this time Misa sat silent in his lines. He had personally made his apologies during the whole afternoon. The speeches took so long that the shades of evening were fast falling and the kava ceremonial was abandoned.
"What was the incident?" again I asked the old builder.
"Well," he replied touching his throat, "among the fowls that were in our share there was one that did not have this taken out."
A psychical storm requiring two hours of oratory and a fine mat to calm, had been produced by a careless cook not removing the crop of one fowl. I smiled.
"Would you eat it?" asked the old builder.
"Certainly not," I replied indignantly.
"Neither would the Sa Tangaloa," he said.
A seeming triviality endangered the building of the largest guest house in
Between the erection of the main posts and the completion of the house, a number of lesser feasts take place. They are celebrations of the completion of various stages. One such feast takes place on the erection of the middle arch (fau tu) of the rounded end.
The final feast (fa'aumatanga) takes place on the completion of the woodwork of the second rounded end section. The payment for the construction is collected in fine mats, bark cloth, and in these days, money. Pigs and other food are cooked in quantity. The house family, villagers and visiting builders collect as before. The pile of mats is placed in front of the taufale and his talking chief. The oven in which the food is cooked is termed umu sa tele. The food (suavai) from the umu sa tele is presented with the usual ceremonial. Fine mats are presented and in ceremonial language the following reasons for the presentation are enumerated: lima la vevea (the hands that were cut in the work); mata pa'ia (the eyes that were injured by chips); tafalenga (the labor done in the house); tavaonga (the work done in the forest—vao); o fanau o tufunga ua fafasi e fanau o taufale (the children of the builders who were beaten by the children of the owner); salanga o le tulutulu (the cutting of the thatch at the eaves).
The above language is ceremonial, which deals in superlatives. The more superlatives used, the greater the respect publicly shown to the builders. They did no work in the forest except look for suitable timber and indicate it to the workers. The cutting of eaves, though mentioned cannot be done until the house family complete the thatching. One or more builders return for this purpose. On that day they are fed by the taufale. The food is distributed among the groups by the builders' talking chief. The visiting builders must be carefully attended to in order that each gets adequate and equal shares of the food. The fine mats and cloth are taken away by the builders to their own lodging house, and there divided among themselves.
Within the round guest house there are set positions to be occupied by the chiefs on all ceremonial occasions. The two rounded ends are the places of honor: entering from the front, that on the right is occupied by the village chiefs; that on the left by the visitors. The middle end posts serve as back rests for the chiefs of highest rank on either side. The lesser chiefs range on either side, and each selects a wall post. Thus the middle end posts are often distinguished in some manner. It may be larger in size, carved, or with a more elaborate sennit design on the lashing to the curb plate. Two posts may be close together, or smaller posts may be close to the middle one, one on either side.
Usually the itu (middle section) has three wall posts. The principal visiting talking chief sits by the middle one on the front side (luma). His colleagues sit on either side of him. The back (tua) of the middle section is occupied by those who have to do with the preparation of kava. The village taupou maiden sits in the middle line with the kava bowl in front of her. On
In
Mention has been made of the importance attached to the number of collar beams (tauso'anga). While the number indicates the size of the house and the status of the owner, it varies in different villages without apparently impairing the status of those in the villages with a lesser number. Hence, a high chief may have fewer collar beams in his house than a lesser chief in another village because the maximum which happens to exist in his village is less. Within the village itself, however, a distinction of at least one collar beam must exist between high chiefs and talking chiefs. In the village of
The cooking house, unobtrusively situated among the vegetation at the back of the dwelling houses, was strictly utilitarian in purpose. It was used only during the time devoted to cooking, usually once a day, sometimes twice. The food was prepared, consigned to the heated stone oven, covered, and left until cooked. The cooks employed their waiting time elsewhere. When cooked, the food was placed in baskets and conveyed to the other houses. Eating did not take place in the cooking house.
The Samoan kitchen therefore contained the barest essentials; the oven site within, with its quota of loose stones and leaf covers, and the firewood without, were the essentials of equipment. A coconut grater, a few wooden bowls, and implements such as tongs, scrapers, and peelers lay on the floor. The floor was bare of mats and the walls were unscreened. Occasionally a fata shelf of poles was made to receive the baskets of uncooked food No reserve stock for more than two days at most was carried as the necessity for storing material ahead did not exist. Cooked food left over for another meal was usually hung up in baskets in the dwelling house where also hung the rat protectors already described.
In serving food, coconut leaf baskets, platters, leaves, and sometimes half coconut shells were all that was required. The food platters rested on the shelves in the guest or dwelling house. Baskets were quickly made from day to day. Leaves were selected from those used in the oven. Half coconut shells were picked up from the discards of grating. Water vessels were kept in the dwelling houses. Water was not used in cooking with the earth oven and little in the preparation of food. Fish were cooked with scales and entrails intact. Fowls were plucked, drawn, and singed. Every part of the pig, including blood and entrails, was used except the hair which was singed off. Before foreign governments laid on water supplies, water carried in coconut shells entailed too much labor to be wasted in washing food. The water carried to the houses was for drinking, to wash the hands of chiefs after meals, and those of the kava brewer as well as providing the medium for the kava decoction itself.
All Polynesians have some variant of a common myth concerning the origin of fire. Most of them state that it was obtained from a region under the surface of the earth, and after being conveyed to the upper world was stored in trees. The motumotu). He returned to the upper world and with the lighted brand, struck the tofaso, fau, and other trees, thus implanting in them the spark of fire. "Before this," say the Samoans, "man ate food raw ('ai mata), but afterwards, by kindling the fire implanted in wood, he was enabled to eat cooked food ('ai vela)."
The technical elements for kindling the stored fire adopted by the Polynesians have been termed "the fire plough." They consist of a piece of dry wood laid on the ground and sometimes steadied by the foot of an assistant. Upon this a longitudinal groove is formed by rubbing with a pointed stick. The operator clasps the fingers of both hands over the front of the rubbing stick with the thumbs behind it. The stick is inclined forward at an acute angle with the lower piece. The weight is put on the rubbing stick as it goes forward while the return movement is for position. Small particles of wood are displaced by the rubbing, and as the groove deepens they accumulate at the forward end of the groove. (See Plate IV A, 1, 2.) When sufficient particles have accumulated forward, the rubbing movement increases in speed, heat is generated by friction, and the particles begin to smoke. At that moment gentle blowing is carefully applied and the particles smoulder and glow. The dry fabric-like material from the base of coconut leaves (lau'a'.a) has been rubbed previously into fine pieces and rests beside the fire sticks. The grooved stick is quickly turned over, and the burning particles emptied upon the prepared material. Gentle blowing fans it into a flame, dry twigs are added, and the concealed fire captured from Mafuie burns as it did of old. Pieces of old bark cloth, plaited into a three-ply braid are also used to ignite the fire from the fire plough.
The lower grooved stick is termed mata si'a, the pointed active stick, ngatu, and the friction process, si'a afi. The most quickly lighted wood is tofaso, but most woods will respond if they are thoroughly dry. Carrying poles of fau, which have been in use for some time, are often used as the under piece. Even in these days, a worker in his bush cultivation, finding that he has forgotten his matches, generates fire by the ancient method often on his carrying pole when it is the driest wood available. The dry sheath of the coconut flower ('aulosoloso) may be used for the under piece.
The general name for wood is la'au, but when used as firewood, it is fafie. The best is tofaso. Next comes toi, which burns well while green, and 'o'a.
In a contest of wit between
The manner in which fire could be utilized in the cooking of food has exercised the mind of man in many ways with different results. The most obvious method is to expose the uncooked food to the direct heat of the burning wood, whether of flame or glowing embers, by direct contact or a space removed. This form as used by the Samoans is termed tunu. Any meat, and most vegetable foods, were so cooked if expediency demanded. The breadfruit was purposely grilled for a particular dish. Fowls were frequently so treated for preference.
For quantity and variety, better control of heat and a more economical use of firewood had to be devised. A medium that would retain heat and prevent the direct contact of the food with ashes, charcoal, and smoke was found in stone. Heated stones in the form of the earth oven became the accepted Polynesian method of cooking. For cooking liquid material that could not be sealed in leaf wrappers, the heated stones were dropped into a wooden bowl containing the food. This was done with the liquid of the coconut to make vaisalo, and with arrowroot for piasua. Breadfruit for taufolo was cooked uncovered on the heated stones, but otherwise the stones and the food were covered to retain the heat. Pottery was not made, and thus offered no alternative. Vessels of wood could not stand the direct application of heat for any length of time. The earth oven fulfilled all requirements. Necessity was satisfied, hence dissatisfaction could not stimulate further invention. Diffusion introduced nothing better until the coming of Europeans. Even with the introduction of metal vessels for heating water, the earth oven remains the principal form of cooking in Samoa.
The site (ongaumu) of the oven (umu) was at one end of the cooking house. The fire was built up on the level floor or in a slight hollow. Through scraping out the ashes of successive fires, the hollow became exaggerated by the elevation of the margins. The picking up of the stones and the preparation of the oven site for another fire is termed uluulu.
The langolango) which consisted of four solid logs, trunks of banana plants, or even the butt ends of coconut leaf midribs. A square or rectangular space was enclosed by laying two opposite pieces on the ground, and crossing two end pieces above them.
In building up the fire (pusa) the wood was arranged so as to offer a sufficient surface for the stones and a sufficient quantity to ensure their being thoroughly heated. The lighting was tutu. If another fire was burning close at hand, a light was obtained from it to save using the fire plough. A piece of dry coconut husk (pulu) which smoulders for a long time was used. This material is often used now to light the cigarettes of an assembly and thus save matches. Arranging the stones on the lighted wood is termed fetui. By the time the wood burns down the stones are heated (afu).
The cooking stones (ma'a umu) are black basaltic stones the size of the closed fist or larger. They are collected from streams or beaches, and from their waterworn condition receive the name ma'a ala. Porous stones of lava (anoano) are not suitable for constant use, as they crumble to pieces through the heat.
The process of removing the langolango side pieces and the pieces of unburned firewood is termed au. After this process the stones are levelled out (sasae). A stout wooden pole about 6 feet long, with or without a fork, was used for leveling out the stones. This spreader is termed sasae, or tofa umu. The further process of arranging the heated stones is termed pae le umu. The scattered stones round the margin were put back and smaller stones placed in the spaces between larger stones to form an even continuous surface and thus distribute the heat evenly.
The arrangement of the food and leaves below and above the stones is termed taufi. Leaves were placed on the stones to protect the food from burning. After the food was placed in position with smaller packages to the outer side, it was completely covered with leaves to complete the taufi process.
Every Samoan cooking house contained a pair of tongs (iofi) made by doubling a piece of coconut leaf midrib. With it, the heated stones could be picked up to place in the exact spaces required. A number were also left out at the margin of the heated bed of stones to be placed beside larger articles of food that take longer to cook. A number of hot stones were placed inside the pig. At Fitiuta, a piece of wood was laid across the oven to divide it into two compartments; one side for the pig and the other for the vegetables. The tongs were also necessary to place the heated stones in certain liquid dishes made in wooden bowls. (See Plate IV B.)
The general term for the covering leaves is tau. They are put on in three layers. 1. The layer immediately above the food (tau) is composed of green leaves not previously used. The leaves in common use are those of the banana, breadfruit, the species of talo known as ta'amu and the laufao. Women may often be met returning from the bush with bundles of short lengths of banana leaves or other leaves for oven covers. 2. The next layer (tau vela) is composed of leaves which have been used before and are thus cooked (vela).
It is well to note here that vela means cooked, not warm, as in most parts of Polynesia. Warm or hot is vevela. When fresh leaves have been used once, they are not discarded, but are kept as tau vela for the second layer. They are also used as lining and covers for the baskets to contain cooked food. 3. The outer cover (tau veve) consists of leaves which are sewn together to form larger covers. The method of attaching them together is the same as in the
Another form of cover is made by packing the fallen leaves of the breadfruit into the round type of coconut leaflet basket ('ato fili tolu). The basket of leaves is flattened down and makes an effective cover.
Earth is never used to cover over the usual cooking oven. In fact, the leaf covers of the tau veve and basket type are not often used. An extra thickness of the second layer is sufficient for most purposes. Sacks are now employed. No attempt is made to hermetically seal the oven, as the Samoans prefer some of the heat to escape through the interstices between the leaves. Otherwise the food is held to become stodgy and not to the taste of the Samoan palate. For a similar reason, the oven is not left covered for more than about an hour.
The Samoan method of cooking is by dry heat. No water is sprinkled over the stones or the food. What moisture there is, the food contains. The
The act of opening up the oven is fu'e. Pigs, which are always cooked whole, are carried into the guest house whole. The vegetables and smaller packages are gathered up in coconut leaflet baskets of the 'ato fili tasi type. The food is always carried to the other houses in these baskets which are therefore also termed 'ato fu'e umu. The leaves from the first covering layer of the oven are used to line the bottom of the basket and as a cover for the food. They are then termed afei.
Earth was used to cover up the whole oven in the umu ti in which the underground stems of the ti (Cordylinc terminalis) were cooked. It was also used in the very large ovens made during war. Gravel was used by Leutongi-tupaitea to cook the chestnuts (ifi) dropped by the bats on the rocky islet where she was marooned. This incident gave origin to the chief's title of
To understand the methods of preparation and kinds of food used, it is necessary to enumerate the actual processes employed and to give the exact meanings of the Samoan words used.
To cut:
poandtipiare the common words for cutting ordinary material. To cut up a pig into the parts established by custom for rank and status ispenga; polo, to cut up a bonito into its ceremonial divisions. The termpengamay also be applied to a bonito.To tear:
fasiis to divide a fowl with the hands. It was never cut.To peel:
fofo'eis peeling, as with green cooking bananas or cooked breadfruit, for thetaufolo.To slice:
fisimeans to slice off the outer skin as in removing the outer part of theta'amuspecies oftalo. Theta'amuwas never scraped.To scrape:
vavaluis to scrape off the outer part oftalo, breadfruit, and yam with a'asiscraper.To grate:
valuis to grate as the meat of the mature coconut with atuaigrater set on anausa'alowooden support. Note the difference betweenvaluandvavalu. The termolo(to grate by rubbing on a rough surface of coral, as withtaloforfa'ausi) is also used for grinding stone adzes.To pound or mash:
tu'iis pounding as of cooked breadfruit fortaufolo, and is also applied to the modern method of poundingkava. Tu'itu'iis pounding more gently or kneading as with the fists in preparing ripe bananas forpoi.To squeeze:
o'omiis with the hands as in makingpoi; paluwith each hand separately as inpoi; lomiorlomilomi, to squeeze lightly with the fingers as inpoi; nguti, to crumble by light squeezing as in preparingmasi.To pinch off:
fifiis done quickly as in pinching off pieces of hot mashed bread fruit in preparingtaufolo.To rub'
vavaumeans rubbing as of the immature coconut meat in thealavastrips of coconut leaf midrib in preparingvaisalo.To wring:
tatauis to squeeze, as wringing the liquid out of grated coconut with atauangastrainer of fiber; also used in preparing kava, vegetable dye, and coconut oil.To strain:
fa'amamameans strain; though wringing and straining are the same process, the words are distinct.To split:
tofiis used as in splitting uncooked breadfruit with a woodento'ipuaimplement.To stir:
tolois stirring with a wooden stirrer as in preparingvaisaloorpiasua.To wrap up in leaves:
ofumeans to wrap up small parcels in leaves for cooking astaloleaves inpalusami; fa'apapato wrap flat parcels as infai'ai talo; fa'ataisi, to wrap larger bundles as intalo fa'ataisi; afifi, to wrap up fish—the bundle is anafi; fa'alaui'ais to wrap up fish in a coconut leaf and plait the leaflets round it.To cook in oven:
taois cooking on hot stones covered with leaves.To cook uncovered:
tunuis cooking on embers or hot stones not covered with leaves.To husk coconuts:
soso'ais to husk with a pointed stick.
Containers naturally divide into three classes: small, natural vessels of half coconut shells, water vessels, and wooden bowls.
Coconut shell cup (ipu niu). The small vessels used in connection with food are plentifully supplied by the half coconut shells that have had their contents removed in preparing coconut cream. The mature nuts are divided into two transversely, and the half shells bearing the muli points are picked up from the heap of discards when needed. The other half shells are unsuitable as they perforate through one of the eyes. The cups are used in serving banana poi and raw fish such as bonito which are cut up in a bowl with fresh water. The blood-stained water flavored with some ingredient is served in the cup with the pieces of fish. Other foods of a fluid nature such as vaisalo are also served in coconut cups. The shells are not specially prepared in any way beyond tearing away any of the husk fibers that may adhere round the rim. After use, they are discarded, and fresh cups provided when the need occurs.
Drinking cups for water are provided by using the specially prepared half shell in which kava is served. This together with the kava bowl forms part
Water vessels. The common vessel for containing drinking water was the whole shell of the large coconuts. Large, mature nuts were husked and perforated through the mata eye. The liquid was poured out and the nut filled with sea water. The sea water rotted the coconut meat to a liquid consistency in a fortnight or so. The fluid meat can then be shaken out and the vessel washed out with fillings of fresh water.
A single water bottle so formed was simply called vai (water) or soni, or fua'ivai. An ancient ta'ai. Usually, however, the shells were tied together in pairs (taulua) by boring another hole through one of the depressions near the mata hole already made. A piece of sennit braid was run through the two holes and tied while the other end of the sennit was passed through the pair of holes in another shell, and the two thus looped together. Two pairs were again looped together so as to enable them to be hung over the end of a carrying pole. Plate IV D shows two sets of four hung over the short carrying pole that was usually used. In many of the more remote villages, the women can still be seen carrying home the supply of drinking water in these vessels. In most of the Polynesian islands the suspending cords of the water vessels are neatly braided, but the Samoan examples figured are very roughly tied together without any particular attention to neat technique.
tangatanga as a net to hold water bottles, but no samples were seen by me. A tulula is also a long basket to hold vessels. Gourds called fangu were also used to contain oil. They were said to be small and their use as water vessels was denied.
Large bamboo stems ('ofe) were at one time used as water vessels, the end nodes shutting off a hollow compartment while the intermediate nodes were perforated. The upper end node was of necessity perforated, and then closed with a stopper. For a recently made concrete construction in a back village in
Wooden bowls ('umete) were used in the preparation of foods such as coconut cream (pe'epe'e), poi, fa'fausi, and taufolo. For the more liquid foods cooked by dropping hot stones into the food, wooden bowls were an absolute necessity. For cutting up uncooked fish to serve raw the wooden bowl was also needed. Thus, though wooden bowls were a necessity in culinary operations, the range of use was somewhat restricted as many of the uses usually associated with bowls were not in vogue. Food was not washed. Except for the foods mentioned above, cooked food was transported direct from the oven to the eating house in baskets. From them the food was placed on leaves and plaited platters, and wooden bowls were not necessary for serving pork,
Bowls were made usually of ifilele and milo wood. The Callophyllum inophyllum wood so much used in 'umete to distinguish them from the kava bowls which are called tanoa, though food bowls are now sometimes loosely referred to as tanoa. Special names for variations in shape were not used, but they were roughly classified into large bowls ('umete tele) and small bowls ('umete laitiiti). From the shape of the opening they may be classified as elliptical, round, and acute elliptical.
In the manufacture of bowls, the external surface (lau) is shaped first of all. The bottom is left flat and for elliptical and round bowls, projections are left at either end to form handles. The sides are shaped with an even convex curve from rim to bottom while towards the ends, the converging surfaces are made to meet in a mesial longitudinal edge which extends upwards from the bottom to the under surface of the handle projections. The upper part of the wood is cut off level or in a slight, curve longitudinally to define the rim and the upper surface of the handles. The handles ('au) project outwards with a flat upper surface the sides of which are parallel or slightly constricted at the junction with the rim. The two sides of the handles are shaped downwards and inwards to an edge which meets the mesial longitudinal edge formed already. The outer ends of the handles are cut off square or they may form a slight slant inwards or outwards.
The bowl cavity (liu) is then hollowed out by commencing with sharp downward strokes to define the inner edge of the rim. A, carpenter seen at work judged entirely by eye and had no hesitation in placing his strokes. At the sides, the inner curve runs parallel with the curve of the outer edge but towards the ends it gradually leaves it to form even rounded ends in elliptical and round bowls. In acute elliptical bowls, the inner curve runs parallel with the outer curve throughout and the curves from either side meet at an angle at the ends. The shaping of the cavity defines the upper surface of the rim (laufa'i) which is horizontal and gradually widens towards the ends where it is continuous with the upper surface of the handles. As the work deepens, the curve of the interior becomes sharper to make the bowl thicker towards the bottom than at the rim. In larger bowls, the bottom is flat but in smaller bowls, the bottom is rounded. In elliptical and found bowls, the natural, flat external surface is maintained and does away with the use of legs but in
After the bowls are completed, they were left to soak for some considerable time in water holes to season the wood and prevent splitting. At Asau,
Of the three forms of bowl, the elliptical was by far the commonest. The round bowl (see figure 67c) is not quite circular, for the longitudinal
The acute elliptical bowl, owing to the sharper grooved ends acting as spouts, could be used for pouring out liquids.
The masoa bowl. The bowl for preparing masoa (arrowroot) (Pl. X D) could be classed under round bowls but the construction and use is so specialized that it is classed separately. The bowl which was obtained in
Attention has been drawn to the distinction between 'umete and tanoa which is breaking down. An old name for a food bowl in taneuli. The masoa bowl was called a tanoa by its donors, the c) was 'umete fai palusami (bowl for preparing palusami); the elliptical bowl (fig. 67 b), 'umete fai taufolo (bowl for preparing taufolo); and the acute elliptical form (fig. 67 d), 'umete fulu fa'i (bowl for preparing bananas). As palusami, taufolo, and bananas have been seen constantly prepared in the ordinary larger elliptical bowls, these terms may be regarded as general descriptive terms, and not specific names. They happened to be the particular bowls so used at the particular time by the family who disposed of them, and the terms described the use to which they had been put. Other families might apply similar terms to differently shaped and sized bowls. As an exact Samoan terminology they can not stand.
Utensils here include instruments, implements, and vessels for domestic use, grouped together for ease of reference.
An oven spreader (sasae or tofa umu), a stout, five-foot-long pole used in spreading the hot stones of the oven has been described. It may be straight or forked. There is nothing special in its technique. Any pole does; it is an implement of convenience which is in general use.
Oven tongs (iofi) are made of a piece of coconut leaf midrib, 2 feet or more in length. The side edges bearing the leaflets are trimmed off. It is doubled at the middle so that the outer wood fibers snap while the inner ones remain to form a hinge. The ends are trimmed off level if they do not coincide. A green midrib is used which dries without impairing the utility of the implement. Tongs are constantly used in shifting hot stones in food preparation, and are an indispensible article in every kitchen. (See Plate IV B.)
Knives. The long-bladed steel knife is now universally used as ax, slasher, and for general use, from cutting clown small trees to carving pork. The original sele knife consisted of a strip of bamboo split off to leave a sharp edge. Though no longer used in the kitchen it was exclusively employed in olden times for slicing off the outer part of the ta'amu talo, cutting up raw fish, pork, and the few culinary operations in which a knife was needed. It is still occasionally used by women in cutting off strips of bark cloth and weft ends while plaiting.
Peelers (fofo'e) are pieces of wood or bamboo, about 4.5 to 6 inches long and 0.5 to 1 inch wide, brought to a flat rounded cutting edge point and used for peeling green bananas and the grilled breadfruit for taufolo. The banana skin is incised longitudinally along the concave side with the point which is then used to separate the skin on either side and pry it away from the fruit. The fofo'e is used to remove the baked outer rind of the taufolo breadfruit because it is too hot to manipulate with the fingers.
Scrapers ('asi) are made from discarded half coconut shells. Part of the outer edge circumference is cut away to form a sharp edge. (See Pl. IV E.) The fingers are spread around the outside circumference of the shell with the thumb above the cutting edge and the strokes are made downwards away from the body. (See Pl. VII C.) Yams, talo, and breadfruit are always scraped (vavalu) before cooking. The breadfruit is scraped in longitudinal strips a little apart on the first round and the intermediate parts scraped on the second or finishing round.
A fair-sized hole is made through the bottom of some scrapers. The hole allows the scrapings which would otherwise adhere to the inside of the shell to pass through. The scraper is also more readily cleaned.
Mollusk shells such as that of the sele are sometimes used. When used for scraping breadfruit, it is called a sele fatu. The term 'asi also belongs to a mollusk shell used in scraping bark for cloth, but as applied to the kitchen, it definitely means the coconut shell scraper. The mollusk shell Tonna shown in Plate IV C was obtained in talo scraper. To form the grating edge an oval hole was made through the shell.
A round stick, about 2 inches in diameter, is driven into the ground just valusanga, or tu'itu'i) for the vegetables as they are scraped. The upper end which is about 6 inches above the ground is cut off level or wrapped with coconut husk (pulu) which is tied with a strip of fau bark. (See Pl. VII C.) Besides supporting the food, the rest allows the gratings to fall clear. In tu'itu'i) was seen in use. In valusanga. The advantage of the husk covering was that it prevented slippery objects such as yams from slipping off the end of the rest as they were turned. It also saved the edge of the grater from being blunted or chipped when it occasionally struck the rest in the downward sweep. Yams and talo are not washed beforehand. Scraping removes both dirt and skin so that further cleaning with water was regarded as unnecessary.
Graters. Three forms of graters are used for coconut, talo, and turmeric. Of these the coconut grater is in everyday use and forms an essential part of the kitchen equipment.
1. Coconut grater, consists of two parts; the tuai grater, and the wooden frame upon which it is mounted.
The
tuaiwas formerly made of coconut shell (ipu niu) cut into a rectangular piece about 3 inches long and 2 inches wide. The end which forms the cutting edge is cut away on the convex outer surface of the shell to form a bevel which meets the inner concave surface to form the sharp cutting edge. The cutting edge is straight transversely with a slight rounding off at each end. In the model (Pl. VA), thetuaihas curved sides and the cutting edge is concave transversely with the ends sloping around to the sides. The concave cutting edge, however, appears to be the result of bad workmanship and the straight or slightly convex edge is the normal form.The
ausa'alois selected from the trunk of a small tree or a large branch with two other branches projecting out close together. The end of the trunk or main branch with the two smaller branches forms a tripod as shown in the models (Pl. VA). The trunk or main branch resting on the ground at one end and raised towards the other by the two natural legs forms a slanting beam. To the raised end which projects from 6 inches to a foot beyond the leg attachments, is attached thetuaigrater. The end is trimmed down in large, natural frames and a slot about 2.5 inches long cut on its upper surface. The depth of the slot is sufficient to provide a surface 2 inches wide to support thetuai.The
tuaiis placed with its convex surface downwards on the slot and with the cutting edge outwards. It is lashed in position with sennit braid. In the models one lashing is accomplished by transverse turns set closely together, while in the other it consists of diagonal turns crossing in the middle line. Owing to the nature of the material the upper surface of thetuaiis concave both transversely and longitudinally. This causes the cutting edge to be directed upwards away from the surface of the woodwork. In one of the models (Pl. VA, 1) another shorter piece of coconut shell with its outer end bevelled, is placed between the wood and thetuaias a chock to prevent the force of grating from snapping it off. The full-sized frame is characterized by its large size and clumsy appearance. The main slanting beam is usually 6 inches or more in diameter. The two natural legs raise the cutting part well above the ground, thus allowing a large wooden bowl to be placed on the ground below it. The cook sits astride the slanting beam, and, holding the half coconut firmly with both hands, works the inner meat part of the nut against the cutting edge with downward scraping movements. The gratings (penu) fall down into the bowl. More modern supports consist of a single beam ofsawn plank, which is rested over a log when in use. Others have a couple of board legs nailed on near the grating end, and may even have a peg at the back end as in the form figured by Handy (14, p. 19).
Besides being made of coconut shell, tuai were sometimes made of stone. A stone tuai found at tuai hafted for E.) The tuai made of iron have now completely displaced the old forms. They are usually slightly convex at the cutting edge, which is serrated. The four-legged curved seat with a projecting arm to carry the grater, characteristic of the
2. Coral grater (lapa). The uncooked talo was grated for the important fa'ausi preparation. The coconut grater was unsuitable, so recourse was had to the oval-shaped pieces of coral of the genus Fungia. These have sharp ridged edges set close together which served excellently as a grater on which the talo was rubbed. They are called lapa. In parts where Fungia was not obtainable slabs of lava with rough surfaces formed by the bubble holes are said to have been used. Sheets of tin have now displaced the coral. Holes set close together are made with a nail from one side, and the raised margins of the holes on the opposite side form a more durable grater.
3. Sennit grater. Turmeric root (ango) was grated to prepare the turmeric yellow powder. The grated material, after mixing with water and straining, divided into an upper layer of lenga and a lower, heavier layer of malasina. The malasina was cooked as food.
The grater was made by wrapping sennit braid closely around a pole about 1.5 inches in diameter for a distance of about 18 inches. The commencing end of the braid was fixed by slanting it down in the direction of the wrapping and burying it under the turns. At the end of the wrapping three or four loose turns were made, the sennit end passed back under them, and the turns successively drawn taut. After drawing the last turn taut, the braid end was pulled to remove the slack and the sennit cut where it issued from under the turns.
Arrowroot (masoa) was also grated with the sennit grater before the adoption of the perforated tin article. Grating on coral or sennit is termed olo.
Pounders. The pounding of cooked foods in taufolo made of breadfruit and talo. Outside of cooked food, the preparation of the kava root should be expected to need a stone pounder. Kava, however, was chewed until recent times. The pounding of the root which has succeeded the old method of chewing, is done with any stone of suitable size. Any leaves or bark that were used for fish poisoning or other purposes that needed pounding, were subjected to beating between two stones. Samoan material culture thus provided nothing in the form of a worked stone pounder. An occasional kava pounder may be seen with a slight notch to fit the fingers, but even this primitive attempt is modern. Each of the two mashed foods mentioned had its own type of pounder.
The Samoan breadfruit pounder (autu'i) is surely the most primitive of food pounders. To understand its evolution we must picture the conditions that called it into use. (See Pl. VI, B.)
Beside the heated stones of the oven lies a pile of breadfruit. The young men lay the unscraped breadfruit on the heated stones and grill them until they are cooked. The charred outer rind of the cooked breadfruit is removed with wooden peelers and the cooked flesh placed in wooden bowls. The breadfruit being in the more mature stage is fairly soft. It could be mashed by pressure with the closed fists, but as the
taufolopreparation is made immediately, the heat prevented the use of the fists as pounders. Something is needed to press down the heated mass. What is more natural than an unscraped breadfruit from the nearby pile. This is done, but as the improvised pounder sinks into the mass the hot material comes in contact with the fingers. A handle is quickly formed by inserting a pointed stick into the breadfruit pounder. But one stick works loose and comes out as the pounder is lifted. The solution to the difficulty is four or five sticks which, by being held together, gripped the interior of breadfruit and made a handle that does not work loose. The improvised pounder with these improvements finishes the job. (See Pl. VI,C.)
As taufolo is only made on special occasions, an incentive is lacking to construct a permanent article for continued daily use. Each making of taufolo produces its own pounders.
The breadfruit pounder (autu'i) shown in Plate VI, B consists of four or five such pointed sticks stuck into a breadfruit around the circumference of the stalk attachment, a little distance apart. The sticks are held together in the grip of both hands which thus squeeze them together, gripping the flesh of the breadfruit. The implement is used for pounding cooked breadfruit. As shown by their stone adzes, the Samoans showed little inclination to do more working with stone than was absolutely necessary. As the breadfruit pounder proved effective, it is little wonder that no stone food pounders have been collected from Samoa. They were not made.
Talo pounder (ulu lapalapa). According to taufolo mashed preparation was also prepared from cooked talo. During a six taufolo made on numerous occasions, but did not see talo taufolo nor was it described to me under food preparation. The cooked talo is firmer and more fibrous than the breadfruit used for taufolo. Something harder was needed as a pounder. The easiest and most available material was the butt end of the coconut leaf midrib (ulu lapalapa) which was used and discarded when its use was over.
Poi pounder. The poi preparation of talo. In Samoa, the term poi is confined to a preparation of mashed ripe bananas. In the preliminary stages ripe bananas were pounded or really mashed in a wooden bowl with the closed fists. The closed-fist pounder is mentioned to show the general attitude of the Samoans towards pounding or mashing food. The fists, the uncooked breadfruit, the butt of the coconut leaf midrib—anything to serve the purpose for the particular occasion.
Strainers and wringers. In extracting liquids from chewed or scraped solids, the process of wringing and straining went on at the same time, and the material used discharged both functions. Wringers were used in the preparation of coconut cream, vaisalo, and kava. Each preparation had its distinctive wringer and strainer.
There are two varieties of coconut cream wringer (tauanga): 1. The tauanga of laufao. The proper strainer is made of laufao (Heliconia bihai), which has leaves and a stem somewhat resembling the banana. The stem, formed of layers of the butt ends of the leaves, is cut off in a length of about 3 feet. With a fofo'e wooden peeler, the edge of the outer layer is peeled off in a narrow strip. Working from this edge, narrow strips of the single layer are peeled off until a sufficient quantity is obtained. (See Plate V C.) The strips are picked up in a handful and rubbed between the hands progressively along their length. The rubbing to crinkle the strips is termed nguti or ngutinguti. One end of the bundle is then held firmly in the left hand, while the right hand sharply jerks (se'i) sections of them out. When so treated, they are rubbed and mixed together to form a tangled mass, and a sufficient quantity forms a tauanga. (See Pl. V, D, l.) The tauanga is used to express the coconut cream (pe'epc'e) from the grated mature nut (penu) prepared with the tuai grater. The wringer is spread out and a quantity of the grated nut scooped up out of the wooden bowl. The fibers are folded around and the strainer twisted in the hands until all the liquid is expressed into the bowl. The strainer is opened out and repeatedly flicked to shake off all the dry particles of coconut meat. The process is repeated until all the grated nut has been treated and only the white, creamy liquid remains in the bowl. Thus, not only is the liquid squeezed out of the gratings by wringing (tatau), but the fluid in the bowl has been fa'amama). 2. The coconut husk tauanga. Many Samoans told me that wringers were not made of coconut husk fiber (pulu). In spite of that, I saw it used in Fitiuta and other parts, and specimens are present in D, 2.) The husk of the mature nut is beaten to remove the interfibrous material and the fibers rubbed up into a tangle. It is used in the same way as the laufao wringer. The best material is laufao and coconut husk is used as an emergency substitute. Collectors of information must be careful of some of the idiomatic uses of the Polynesian language. When a Polynesian says a certain thing is not done, he sometimes means that it is not right, correct, or orthodox, and therefore, so far as the informant is concerned, it does not exist. When it is afterwards pointed out to him that you have seen the thing he denies, he shrugs his shoulders and explains that it is not the correct thing. He is an idealist when circumstances do not force him to abandon the orthodox. The fau bast is not used for coconut cream wringers.
The mincer (alava). This consists of narrow strips of the skin of the coconut leaf midrib near the expanded butt end. The edges of the strip provide a blunt cutting edge. A number are gathered together in'a mass and used in the preparation of vaisalo from the flesh of the immature coconut which, fairly soft and scraped into the bowl to join the fluid part of the nut, is picked up in the alava and rubbed between the hands (vavau). The edges of the alava strips act as a primitive mincing machine and cut the flesh into smaller pieces which may be left with the fluid. In the form of vaisalo used for invalids, after rubbing the larger pieces, the alava is used as a wringer to express the liquid. The particles of flesh are then flicked off. In this process, the alava acts as both wringer and strainer.
The kava strainer (to tau ava) is made of fau bast, and is described on page 151.
Breadfruit splitters (to'ipua or to'i'ulu) are in common use throughout Samoa for dividing large breadfruit into halves or quarters before cooking. In appearance they resemble hafted adzes and are termed to'i (adz). They are mostly made of pualulu wood, but fau is used.
A branch ranging from 1.25 to 1.5 inches in diameter is selected and the part of the trunk both above and below the branch junction is removed with the branch. The trunk portion when removed is roughly about a foot long and over 2 inches thick. The upper part of the trunk portion which forms an acute angle with the branch, is shaped into the blade, which ranges from 2 to 2.5 inches in width and deepens in thickness from the distal cutting edge to about 2 inches at the branch junction. The cutting edge is rounded with the convexity downward with the implement in the position for use. The lower trunk portion in use is turned upwards to form a heel which ends in a point. In some adzes (Pl. VII
A, 1, 2), the heel is blunt and is formed by the meeting of the plane of the blade with the line of the handle. In others (Pl. VIIA, 3, 4), the heel projects upwards beyond the line of the handle and forms an obtuse angle with it. The branch is cut off to form a suitable handle.
The action of splitting by means of a blow is termed tofi. In some implements, the handles are straight, but in others a curved or bent branch is selected to form a better angle with the blade.
Food stirrers. A strip of coconut leaf midrib split to a suitable thickness is often used for stirring preparations of piasua (arrowroot) and vaisalo (coconut) to which hot stones have been added. To stir is tolo or sa'eu, and the stick so used is tolo, or la'au sa'eu. With the introduction of the bush knife, a strip of fau wood is readily prepared, and is now more commonly used as a stirrer.
Seed flicks. In the preparation of breadfruit taufolo (Plate VI C) an assistant, seated opposite each man using the breadfruit pounder, holds a stick (i'o fatu) for flicking out the hard, immature seeds (fatu) that appear in the breadfruit. Formerly, strips of coconut leaf midrib were used, but a thin, flat piece of fau wood about a foot long, trimmed with a knife, has supplanted them. There is no special distinction in shape between a food stirrer and a seed flick, but they have received their distinctive names from the use to which they are put.
Some implements not so intimately associated with the cooking house but still used in connection with food are here described.
Coconut husker (mele'i). A stake of olamea, poumuli, or some hard wood, about 3 to 4 feet long and 1.5 inches in diameter, sharpened at both ends, is used for husking coconuts. If a hard wood is not at hand, fau is used.
The lower end with an ordinary point is driven into the ground at a slant. The upper end is cut on a single bevel to form a convex cutting edge with the round section of the stake. The side cutting edge point of the stake penetrates more readily through the longitudinal fibres of the husk. The coconut is held at the ends with both hands, and driven down on the point with its long axis corresponding with the wide axis of the point. The far side of the nut is applied to the point so that the blow forces the point through a segment of the husk with the nut towards the husker. The husk segment is torn off by levering the nut towards the body. Successive segments are removed in this way until the nut is denuded of husk. If the point of the stake does not come right through a segment on the downward blow, the nut is pressed down until the point comes through. Husking takes place in the plantation and the nuts are carried home in baskets slung on a pole. Both green and mature nuts are husked in this manner. The common name of the husker is
mele'i, butPratt (23, p. 50) also giveso'a.
Climbing bandage (aufanga). To facilitate the climbing of coconut trees to obtain drinking nuts a loop of fau bark is sometimes used. (See Pl. V, B.)
Both feet are placed in the loop so that the ends pass over the dorsum and under the instep. The climber grasps the palm trunk high up, draws up his feet and rests the climbing bandage against the trunk where it takes a grip against the rough surface. He straightens up the body, takes another hand grip higher up, and draws himself up for another bandage grip. For ordinary trees, the climber usually walks up on all fours, the hands clasping the trunk on the side away from the feet. In tall trees, the bandage enables the climber to rest on the bandage and relieve the strain on the arms.
Breadfruit pickers (lou'ulu). The breadfruit out of reach are detached from the tree with a long pole usually of fau, with a fork at the end. Three forms, the "X" picker, the "Y" picker, and the netted picker, are in use. (See fig. 68.) The "X" picker is the most common. (See Plate VI, A.)
In the "X" picker the angle included by the two arms is more acute than in the "Y" picker, making it easier to remove the fruit. If the fruit stalk is missed by the front angle it may be hooked with the arms at the back and pulled off. The netted picker consists of the ordinary "Y" picker with somewhat more divergent arms to which a bag net is attached. This is termed fangafanga and not 'upenga. The stages of manufacture are shown in figure 69. In attaching to the handle, the circumference is arranged so that the sennit braid from the last knot is at the junction of the arms. Two or three
Breadfruit pickers are easily made so they are not saved up for the future season. Though there are about three crops in one year, a new picker is made for each. Hence the saying: "O le fuata ma lona lou" (The breadfruit crop and its picker).
Carrying poles (amo). Burdens are carried slung to either end of a pole, which is balanced on one shoulder. The pole is nearly always of fau. Since bush knives are always carried to the cultivations, a pole is quickly made by cutting down a sapling and making a notch at either end. The form of notch in figure 71a is the usual one. If, owing to a basket being too full, both sides of the midrib rim will not stay together, the rim to the outer side is kept in the notch. (See Plate VII, D.)
Some poles are shaped like that shown in Plate VII B and figure 71b. This pole has two fire grooves on the under surface opposite the notch and is slightly concave longitudinally on the under surface, caused by the shaping, combined probably with the effect of the weight at the ends when the pole was green.
The coconut water bottles are carried on a much shorter pole. The poles are in common use in Samoa. Many of the men have fibrous lumps on the shoulders caused by continual use of the carrying pole. Even with one basket of food, a Samoan would sooner carry it slung over his back on a pole than in his hand.
Carrying straps (faufafa). Women carry large bundles of leaves for the oven, sugar cane leaves, and other material strapped to the back by strips of bark. To carry on the back is fafa, and the burden fafanga. The bark of the fau, fu'afu'a, or the fue creeper are used for tying on the burden. These bark strips are termed faufafa. faufili as a cord used for tying on the burden. There may be a distinction to denote plaiting from fili, to plait. Another term, avei, is also used.
The bundle is first tied together. A single strip of the tying material is passed vertically over the back of the burden on one side. The long end of the strip passes over the shoulder on that side, then diagonally downwards across the chest, under the burden, vertically upwards at the back, over the shoulder, and diagonally across the chest to meet the other end to which it is tied in front.
There is so much detail in connection with foods that it is advisable to deal with each article separately. Foods divide naturally into flesh and vegetable foods.
Flesh foods consist of pork, fowl, certain wild birds, turtle, fish, crustaceans, and shellfish. Pork forms the most desired food, and on all festive and ceremonial occasions, it is regarded as essential. Fish, however, is the staple flesh food. In olden times the dog was eaten and, on occasion, man.
Pigs (pua'a). Pigs are traditionally stated to have been stolen from so'oso'o, a term still used in calling them to feed. They were kept in enclosures of fair extent bounded by walls of loose stones and pieces of lava. These enclosures (pa pua'a) were at the back of the village or some little distance away. The walls about 4 feet high were crossed in places with stones arranged to form steps, or by a tree trunk with steps cut out of the solid. The pigs were fed sometimes with mature coconuts cut open or other available fruit. Pigs form a source of wealth to a family enabling them to make a good showing at the various functions demanding pork. They are sometimes kept for fattening in small stalls made by crossing horizontal lengths of coconut tree timbers. These are situated at the back of the dwelling house to be near any food left over from the meals. In a well-disciplined community pigs are never allowed to run about the village.
Pigs are killed by strangling immediately before cooking them. They are never killed beforehand and allowed to hang. Climate and custom are against it. They are never stabbed as that would waste the blood. The Samoan attitude is, "Why catch the blood in wooden vessels when the interior of the
The lower abdominal wall (alo) is cut off, wrapped up in leaves, and sent to the taupou as her official portion. The pig is kept on its back to save the blood which has emptied into the abdominal cavity. The fat from the sides and over the intestines (great omentum) is stripped off, shredded into small pieces, and mixed with the blood in the abdominal cavity. The heart is removed, split, and wrapped up with some fat in a leaf package. All such packages are termed ofu. The heart ofu (ofu fa'afale olo) is for the high chief. The blood mixed with fat is ladeled up into receptacles of banana leaf which are wrapped neatly around the fluid contents. These ofu are termed ofu valevale to distinguish them from the package containing the heart. There are several of them. They are for distribution among the chiefs and young men who do the work. When cooked the blood coagulates like black puddings. I was at a loss to account for the white material in the cooked packages until informed of the shreds of great omentum. The liver is mea fono (for official use) to the talking chief. The intestines and the remainder of the internal organs go to the butchers and cooks. The gullet is removed through a slit in the throat and the removal of the rectum completes the cleaning.
By this time the oven is ready. With the iofi tongs, a hot stone is placed in the slit in the throat and another in the aperture left by the removal of the rectum. In a large pig a big stone is placed in the thorax and another in the abdominal cavity. Another may be pushed down into the pelvic cavity. The abdomen is then stuffed with 'o'a leaves for preference but others may be used. The leaves give the pork a flavor and are termed lavai. The method of stealing the first pigs from lavai. This enabled them to evade the customs of the country.
Some 'o'a leaves are spread on the heated stones of the oven. The pig is placed on them with the abdomen downwards. The forelegs are bent back and the hind legs forward under the body. More 'o'a leaves are spread over the pig and then the usual covering of other leaves. The pig is always cooked whole no matter how large it is.
The cooked pigs are carried whole on poles to the guest house with the other food. In important functions they are heaped together to make a goodly show. In every community there are one or more expert carvers of pigs who delight to exhibit their dexterity. The distributing talking chief may himself carve; if not, he superintends operations. The portions into which the pig is divided have been set by usage. Each part has its name and the person or rank to which each part is allotted is also set by usage. The forequarters, legs, and head are removed and the body and neck divided according to Fepulea'i as shown in figure 72.
The ceremonial division of the pig has become such an important social event that it has had an unexpected effect. The actual cooking of the pig has become a secondary matter. If the pig is too well done, the flesh is liable to tear away and the exact boundaries of the ceremonial divisions can not be maintained. This creates adverse criticism and comment on the part of those watching. Instead of an important spectacular success, the division becomes a failure. The failure reflects on the talking chief in charge. He in turn, vents his displeasure on the young men who overcooked the pig. This, by creating a fear of overcooking, has brought about an avoidance custom of undercooking the pig. The cooks often have an argument as to whether the pig is in danger of being overcooked. They may settle it by casting lots. A stick of any length is thrust down into the leaf coverings of the oven. Commencing from below the alternate hand grips are taken on the stick by the two sides. The one desiring the oven to be opened says, "Umu vela" (cooked oven) as he takes his grip. The other grips immediately above saying "Umu mata" (raw oven), and so on up the stick. The one who gets the upper end of the stick wins.
Often the pork is merely warmed through. When the guest receives his share in the guest house he is not compelled to eat it. His correct share has
The appearance of raw pork in feasts is often adversely criticized by people of another race and wrong deductions are made. The pork is not uncooked because the Samoans like raw pork, but because it is necessary to the proper carrying out of the ceremonial division. A failure in ceremony can not be remedied, but underdone pork can be recooked.
The official portions given to high chiefs and talking chiefs are often not meant to be eaten at the time. They are mea fono (for official use). Official positions have to be maintained. The ceremonial division has been partly devised to assist them in discharging social obligations. The official portions are recooked for the entertainment of guests and visitors at a later meal. The guests may be neighbors who drop in for a bowl of kava, in which event the chief falls back on his official reserve of pork, and by sharing it with others, not only avoids the stigma of selfishness, but acquires merit. It must be remembered that these portions are not a couple of slices or chops but whole joints. Here again the supposed disadvantages of underdone pork are discounted. The official cutting up of the pig may be compared to the cutting ' up of joints in a butcher's shop. It is for distribution' primarily and not for immediate eating.
Sometimes pigs are cut up into small portions. At Fitiuta in fa'asafune (like Safune).
Fowl (moa). Fowls are grilled on the fire (tunu) or cooked in the oven (tao). For ordinary guests, the fowl forms an economical substitute for the pig. At feasts, they are provided in quantity. The cooked fowl is divided by tearing it apart with the hands. It is never cut. There is a ceremonial division. The only parts that count are the legs. In tearing the two apart, the coccygeal part that carried the tail feathers adheres to one of the legs. That particular leg is called vae ma le muli, or vae ma le no'o, and it is the correct part for the high chief or visiting chiefs. When malanga) around
Pigeons, doves, tern, and other wild fowl were netted or snared. In any ceremonial division they were treated as fowl.
The turtle (laumei). No opportunity occurred for a practical acquaintance with the turtle. lavai of leaves put in. The turtle was then cooked on its back in the oven.
The cooked turtle was taken to the guest house and divided up. The forequarters (sangamua), the hindquarters (sangamuli), the breast, and abdominal parts were removed. The stones were taken out and the fat cleared away to expose the juice (suapeau) within the cavity. The juice was dipped out in half coconut shells, and laid before the chiefs. The parts were then distributed as follows:
Fish (i'a). Fish abound within the lagoons and form the stand-by of flesh foods. The smaller fish, always cooked in leaf packages (afi) were placed unscaled and uncleaned on a banana or breadfruit leaf which was folded over them and perhaps tied with a strip of fau bark or husk fiber. Larger fish were placed on a piece of coconut leaf cut to the length of the fish. The leaflets were then brought round the fish from either side and crossed alternately over it as in the commencement of a check braid. This bundle was termed a fa'alaui'a. In serving to guests, the packages were placed before them unopened but with any tying strip removed. The guest then had the pleasure of opening his lucky package and seeing what fortune had sent him. The cooked scales are simply pushed aside with the fingers. Entrails are not wasted.
Coconut cream (pe'epe'e) may be used by pouring some into the cupped banana leaf with the fish before wrapping it up. The leaf forms an impermeable cover. Cooking curdles the pe'epe'e into fai'ai and hence the preparation is termed fai'ai i'a.
Bonito ('atu). Of the larger fish, bonito and shark (malie) require special notice. The bonito of all fish is regarded as a chief's fish. A special canoe, rod, and hooks are devised for its capture, and there is much ceremonial connected with it. It is natural therefore that a set division and allocation has become established. The parts are shown in figure 73. The head (ulu) is cut off in a vertical line that passes behind the attachment of the pectoral fin. Here again the chiefs were given important parts which they shared at a meal with guests. The bamboo knife used to cut up the bonito was called manamate.
The bonito is a favorite fish for eating raw. At Falealupo in '"Atu!" The official speech stopped and all present transferred their attention to the tautai (head fisherman) of the western end of the village who was approaching with a large bonito. A young man grasped a laulau platter from a fata shelf and laid it on the floor at the front of the house. The tautai laid the present for the guest on the platter amid a chorus of, "Fa'afetai mo le fa'aaloalo" from the assembly. The tautai went off and the platter was laid on the ground before me. The speech was no sooner resumed than the same chief again grunted, " 'Atu!" The tautai of the eastern end of the village appeared and another bonito was laid on a waiting platter. This joined its fellow. After the speeches my talking chief, on my behalf, gave one of the bonito to the chiefs present.
The bonito is cut up into small pieces in a large wooden bowl with water in it. Nowadays the juice of limes is added. Portions are then served in half coconut shells. The shells are not specially prepared but are merely those discarded after grating the contents for pe'epe'e cream. The bowl is placed at the back of the house with the server sitting behind it as in the serving of kava. Attendants quickly place a laulau platter before each guest sitting against their respective wall posts. Another attendant comes around with a basket of cooked food and places a talo and breadfruit on each platter. Then a coconut vessel containing the cut-up raw bonito with some of the bloodstained water is given to each guest. If balance is precarious a few stones are talo and the liquid drunk. It is better than it looks.
The shark (malie). The flesh of the shark is much prized for eating. The ceremonial surrounding it has also led to set division and allocation of the parts. (See fig. 74.)
The head part includes the gills (aulama). The tail is cut off from behind the second dorsal fin. The nono has its posterior boundary behind the first dorsal fin from which the portion receives its name of nono. The lower boundary is at the junction of the pectoral fin. The io takes in most of the body. It is split down the middle and each half divided into four parts. The internal parts are divided into the stomach, intestines, and liver. The stomach (tanga) and the intestines (silo) are regarded as the best parts of the fish. They are shared by the talking chief and the head fisherman, the former exercising the preference. The liver (ate) is divided up among the other shares.
When the canoes come in with a good catch the crew shout and wave their paddles to advertise the fact. The villagers gather, the chiefs taking kava root with them.
The talking chief now takes charge of the cutting up. The visitors are first advised to wait, generally in another house. The lau alofa is generally cooked at once by the young men (aumanga) for a general meal. The chiefs while away the time of waiting in the house by plaiting sennit braid, by conversation, and perhaps another bowl of kava. The allocated official parts au are ceremonial shares, not eaten privately, hut cooked and brought out at a later meal to which others are invited. Instead of a set meal they can be used at a ceremonial kava drinking with talo for supper. As in the allocations of pork they help the chiefs to maintain their position and also the prestige of the village. Where there is no taupou in a village or division of a village, the official share of the head goes to the aumanga.
The laualofa and perhaps part of the io having been cooked, a general meal is partaken of. In this the chiefs share even after shares have been allotted to them. Thus, the body and belly of the fish are for immediate and general use; the other parts for deferred and official use.
Another custom prevails in connection with official shares. Chiefs from neighboring villages, hearing of a catch, may send in and ask the chiefs of their corresponding grades for their official shares. Thus high chief sends to high chief, talking chief to talking chief, and taupou to taupou. This can be done even if visitors from the particular village have been present and have received a share. Such requests may not be denied. The only legitimate excuse is that the shares have already been eaten or given away. This custom shows how official or public the shares allocated to chiefs are and what an important part they play in social organization. The part the talking chiefs play in framing the rules is seen in the allocation of shares. The high chief receives the worst part of the fish in the tail, whereas the dorsal fin portion that falls to the talking chief is the best. Though the talking chief subsequently shares his portion, he has the satisfaction of demonstrating that in some things he exercises more pule (power) than his superior. These diplomatic workings of the Samoan mind are fully recognized and retailed by the Samoans themselves.
The fresh water eel (tuna). The tail part goes to the chief.
The squid (fe'e) is a delicacy eagerly sought after by women at low tide with sticks for poking them out of holes in the reef, and at high tide by men using squid lures from canoes.
The proper method of cooking is termed fai'ai fe'e. The dark liquid contained in the squid (taelama) is expressed and mixed with coconut cream. The tentacles are cut into short pieces and together with the cream mixture are wrapped up in banana leaf packages and cooked in the oven. Sometimes talo leaves are included in the package.
Crab (pa'a) may be cooked plain or with coconut cream in a package to form fai'ai pa'a.
Dogs (maile and uli). uli, but there is •little doubt that it should be 'uli as kuri is the u, to bite li, to show the teeth, does not seem to fit the case. He states that the native dogs were a small breed with sharp pointed ears, but he saw only one wild one in the distance.
Man. The remarks about dogs apply with greater force to human beings as a source of food. Practically all branches of the Polynesians except the
Vegetable foods form the larger part of the diet. Except for coconut, breadfruit, and banana, they consist of root crops in the form of talo, yam, and to a smaller extent, umala. Of these the talo is by far the most talo leaves formed the only kind used. The number of preparations are confusing unless taken under each food heading. The coconut is dealt with first to prepare the way for its combinations with the other foods.
The coconut (niu). The mature nut is called popo and the less mature drinking nut, niu. The outer husk is pulu, the shell ipu, and the meat 'a'ano, the outer rind of the meat tuatua, and the liquid sua. The end of the husked nut showing three round depressions is called the mata as against the other end (muli). The mata (face) is so named after Tuna, the unfortunate eel lover of sisi, while the other apart, is also the mata. A leaflet midrib can be pushed through the mata depression to form a hole through which the liquid may be drunk. This hole is also opened up to prepare coconuts as water bottles. The sisi depressions can not be perforated with a piece of midrib. The muli ends in a sharp point. This part is tapped round its circumference with a stone or knife to crack the shell, and thus enables a circular piece (ta'e) to be removed to give better access to the drinking fluid. Where no implement is handy, a nut may be cracked around with the muli point of another unopened nut. The green husk of a particular kind (utongau) is chewed for its sweet taste. The sua liquid forms a refreshing beverage freely offered to guests in lieu of, or as well as, kava. Unless trees are prohibited by particular warning marks, travellers have the right to take a few drinking nuts from any wayside trees, but they must not abuse the privilege. The flesh, both immature and mature, is used as food, either uncooked or in the following preparations.
- Coconut cream (
pe'epe'e). Thepopomature nut is husked, split and grated on thetuaigrater. The gratedpenuis wrung and strained through atauangastrainer as already described. The resultant creamy liquid is thepe'epe'eso much used and appreciated in Samoan dishes. It may be used directly as with banana poi or undergo further treatment.- Sauce (
miti). Thepe'epe'eis mixed with sea water (sami) to form a relish in which food is dipped during meals. Nowadays fresh water and salt may be used instead of sea water, and sliced onions added. The juice of limes or lemons is also added. Themitimay be kept in a coconut water bottle. For use, some is poured out into a half coconut shell or into a hollow formed with theafeileaves that accompany the food. It is used with raw mullet and cooked fish.Thickened
pe'epe'e(fai'ai). Thepe'epe'emay be cooked by itself in a wrapper of banana leaf. It thickens like curds and is calledfai'aiorfai'ai fua. Again it may be cooked with other foods such astalo, yam,taloleaves, fish and crab. To all these preparations, the name offai'aiis given with the particular food as a qualifying term such asfai'ai talo, fai'ai ufi, fai'ai i'a, andfai'ai pa'a. The exception is the preparation withtaloleaves which receives the specific name oflu.The
pe'epe'eis also heated by adding hot stones to it while in the bowl. One stone is sufficient to bring out the slightly burned smell and flavor so much appreciated. InSavaii andUpolu it is usual to place a hot stone on the grated nut (susunu ai penu) and express the liquid after. This is the stage used for thetaloleaf preparation (palu sami).- Curdled
pe'epe'e(vaisu) goes a step farther than the above. Hot stones are placed in a bowl containingpe'epe'euntil it curdles. It may now be cooked with fish to flavor them. Note thatfai'aiis curdled by the heat of the oven during cooking, whereasvaisuis curdled before the package is placed in the oven.- What is termed
niu tolocarries on from thevaisupreparation. Hot stones are added until an oil separates from a brownish cooked precipitate. The sweet precipitate (taei'a) is usually eaten by the cooks, while the oil (siu tolo) is used with thetalopreparation known asfa'ausi.- A preparation (
vaisalo) is usually made for sick people, but not necessarily confined to them. A nut, not too mature, is opened at themuliend and the fluid poured into a bowl. The circle of removed shell (ai sali) may be used instead of the fingers to scrape out the meat into the bowl. The meat is rubbed small (vavau) withalavastrips and the liquid wrung out into the bowl. When nothing but liquid is in the bowl hot stones are added until it is cooked and formsvaisalo. Atolostirrer is used while the cooking is proceeding. In some preparations, the small particles of coconut meat are allowed to remain with the fluid.- An evil-smelling, but palatable preparation (
sami lolo) is usually a by-product obtained while preparing coconut shell water bottles. A large mature nut is pierced through thematahole, and the liquid poured out. The shell is filled with sea water (sami), corked with a folded strip of dry banana leaf, and left for a fortnight. The water is then poured off, and the meat, which has become soft, is shaken out. When this is cooked withtaloleaves in a package it is known assami lolo.
The talo (Colocasia antiquorum) is the staple vegetable food of Samoa. It takes easy precedence over the breadfruit, yam, and sweet potato. Its ceremonial status is high in that it forms the correct vegetable to serve to high chiefs. The talo preparation of fa'ausi is also made the most of in serving before guests. The leaves provide the only green vegetable used. The large ta'amu species is sliced to remove the skin, and never scraped. Slicing removes the astringent and irritating properties more effectively.
Talo tao. Thetaloscraped with an'asiscraper, cooked (tao) in an oven without any individual leaf wrappings, is the general form of cooking.
Talo fa'ataisi. The scrapedtalois divided longitudinally into two or three parts according to the size of thetalo. Ti leaves are used as wrapping to form a package (taisi). The division into smaller pieces ensures their being cooked through the wrapping. When cooked, thetalois cleaner and whiter in appearance than the ordinarytalo tao.
Fai'ai talo. Coconut cream is wrapped up with thetaloin the leaf package, in which it curdles and thus formsfai'ai talo. Two preparations are distinguished according as thetalois ungrated or grated.Loloi. Theloloiistalo fa'ataisi(2) to which coconut cream is added before cooking. An alternative name tololoiisfa'alifo talo.Fai'ai valuvalu. The uncookedtalois grated (vavalu) on a slab of coral (lapa) and then wrapped up in a leaf package with coconut cream. Owing to the softer nature of the gratedtalo, the package is flatter and smaller than ataisipackage. This form of package is calledfa'apapā.Fa'ausi talo. Grated uncookedtalosufficiently moist without the addition of water, is made up into flat cakes and wrapped up in leaves to formfa'apapapackages.Kramer (18, vol. 2, p. 151) states that the leaves of thelaumapāpāfern (Asplenium nidus) are also used as wrapping. Coconut cream is not added to the packages before cooking. When cooked, the flat cakes are cut up into fair-sized cubes and put in a wooden bowl with the oily preparation of coconut cream known asniu tolo. It is served with some of the oil. The importance with whichfa'ausi talois regarded is dealt with under ceremonialobservances. Two forms of the preparation are recognized: fa'ausi malaulau, when thefa'ausiis made hot or immediately after the gratedtalocakes are cooked;fa'ausi fua fulu, when the cubes are cut from cold cakes cooked overnight.Taufolo talo. This preparation was not mentioned to me but was described in details byKramer (18, vol. 2, p. 151). The cookedtalois peeled and mashed with the butt end of a piece of coconut leaf midrib ('ulu o le lapalapa). It is served with coconut cream.
Leaves. The leaves used are those near the growing center which being folded in, are termed moemoe. The older, fully opened leaves do not soften into a pulp, but remain tough, and have an irritating astringent effect on the mouth and throat. Four methods of cooking the leaves (lu'au) and one of the leaf stalks (fa) are in vogue. ulu'fau as a synonym of lu'au. Here ulu means the end away from the 'au or stalk and lu'au is evidently a contraction.
The wrapping for talo leaf preparations is banana leaf which, by being heated over hot stones, is toughened and forms an impervious cover through which the liquid coconut cream can not soak.
Smaller packages of talo leaf come under the general term, 'ofu. The neat manner in which the leaves are prepared is described by
From the neat pile of
taloleaves one leaf is selected and placed face downwards on the palm of the left hand. While the right hand fingers pinch off the tip of the leaf, the left breaks the midrib about an inch above the stem. Thereupon the right hand pinches out the butt end of the midrib from this break to the end of the stem and throws it away. The two lower lobes of the leaf are now torn off and the three sections resulting are placed on the pile ready for filling with coconut milk.
The assembling of the package takes place in the cooking house when the oven stones are heated. The wooden bowl containing the prepared coconut cream is brought in and a hot stone is dropped into the liquid to bring out the odor and taste. In the talo leaf are laid on the palm of the cupped left hand which raises the edges of the leaves to form a receptacle for the coconut cream. A sufficient quantity of cream is dipped up with a half coconut shell and poured into the leaf hollow. The sides of the leaves and then the ends are neatly folded over so as to enclose the liquid. An assistant who has warmed a section of green banana leaf on the hot stones (lalangi lau fa'i) hands it over. Upon the side of the leaf section, the natural concavity of which is increased by the heat, the leaf package is laid. The sides and then the ends of the banana leaf wrapper are folded over the talo leaves, and the augmented package laid on the upper surface of a green breadfruit leaf. The sides of the leaf are folded over so as to overlap, the tip end of the leaf is folded in to the middle, and lastly the stalk end turned forward over the tip. The leaf stalk is pinned through the tip end by passing down through the leaf on one side of its midrib and 'ofu package is thus neatly and deftly folded and securely fastened.
When the food is placed on the prepared oven, the talo leaf packages are placed round the circumference. They require less heat to cook and must not have fish or flesh placed above them lest the distinctive flavor of the vegetable be affected.
On cooking, the talo leaf pulps into a soft mass mixed with the coconut cream. In serving, attendants remove the outer breadfruit leaf covering but the banana leaf is left as a receptacle which the guest opens out himself.
Potoa. Thelu'auleaves are cooked without any coconut cream. They are folded in a package and may be covered merely with a large piece oftaloleaf, as the impervious banana leaf is unnecessary where no liquid is used.Lu. A larger package than usual is formed oflu'au, coconut cream as treated above, and banana and breadfruit leaf covers.Kramer (18, vol. 2, p. 147) gives the name offa'afatupa'ofor this preparation.Palu sami. This favorite preparation is made exactly likelubut in smaller packages while sea water is added to the prepared coconut cream to give it a salty taste. The sea water may be added to the grated nut before the cream is expressed. A metaphorical name is'oto ma le sau(plucked with the dew) inferring that the leaves were plucked while the dew was on them.Lu'au fui. Coconut cream is not used but a little sea water is poured into the cuppedlu'aubefore wrapping with the two leaves.Fa. The stalks (fa) of thetaloleaves are peeled and wrapped in a package without the addition of coconut cream or sea water. The preparation may be made more palatable for sick people by adding coconut cream with sea water, whichKramer 's informant referred to assua palusami.
No meal is complete without a preparation of talo leaf. Of these, palu sami, is easily the most in demand. In recent times salt is often added to the talo leaves in place of sea water. The use of sea water dates from a period when other forms of salt could not be procured. The lu, with coconut cream alone, is used more in inland villages where sea water is not available.
In talo leaves is said to commemorate a historical incident that ocurred between Tangaloa and Pava in a kava-drinking ceremony. (See p. 153.)
The breadfruit ('ulu, Artocarpus incisa). The breadfruit stands next in importance as a food to talo. The name 'ulu, with dialectical letter changes as 'uru, kuru, is found throughout Polynesia. It occurs in kuru and formed one of the causes of strife in the Society group before the 'asi scraper before cooking. Medium sized fruit is cooked whole, but large ones are split in two or in thirds with a characteristic wooden splitter (to'i pua). Of the food preparations, taufolo ranks high. In masi). It is then used in times of scarcity or while working in the plantations where the pits are made.
The leaves are used as oven covers, outer wrappings for 'ofu packages, and as platters upon which food is served. As platters, the ma'opo non-pinnate variety is used.
'Ulu tao. The scraped fruit cooked without wrappings in the earth oven is the common mode of preparation.'Ulu tunu. The scraped fruit may be cooked on the embers of an open fire for one or two people to save preparing the earth oven. It is cooked on the heated stones of an earth oven without cover for thetaufolopreparation. Both forms aretunuas opposed totao.'Ulu pe'epe'e. When'ulu tunuis peeled with afofo'epeeler, it may be eaten with coconut cream. The breadfruit so cooked is usually more mature than those cooked in the oven.Taufolo 'ulu. Thoughtaufolo talohas been mentioned the termtaufolois usually associated with breadfruit. The slightly more mature fruit is cooked on the open heated stones of an earth oven, but no cover of leaves is placed over it. The charred rind is removed withfofo'epeelers and the cooked fruit placed in a wooden bowl. It is mashed with theautu'ipounder made of uncooked breadfruit, and the core (fune) and immature seeds (fatu) flicked out with a piece of wood known as ani'o fatu. See Plate VI,C. When mashed to a consistent mass, coconut cream is added to the bowl and two forms are named:a. taufolo niu, when the coconut cream has been merely heated with a hot stone to bring out the flavor;b. taufolo sami, when a little sea water is added to the coconut cream. The mass is divided up by pinching off smaller pieces with the fingers (fifi) and squeezing or rolling them into balls. The pinching off and rolling is quickly done as the mass is hot. The preparation is served onmaopoleaves with some of the coconut cream. Thetaufolofigures as a food for special occasions.
. The surplus breadfruit that has been fermented in a store pit is termedMasi masi.Pratt (23, p. 193) gives an example of name avoidance in a district where owing to the chief's name beingMasi , the fermented breadfruit was calledmamala. The pit (lua'i) is lined with coconut leaves and an inner layer of banana orlaufaoleaves which are calledafei. The mature breadfruit, often unscraped, is thrown into the pit. Fruit that is too ripe is not favored even formasi. The large fruit may be split. The pit is then covered to exclude the air. It will keep in the pit for a year or more.The fermented
masiis crumbled with the hands (nguti) and kneaded into cakes in a wooden bowl dry or with a little water. Thefunehas softened during fermentation, but the hardfatuseeds are picked out. The cakes are cooked in the earth oven with or without leaf wrappings. The latter is termedmasi afifi. Besides this plain mode of preparation, there are three others:a. masi palu, when themasiis mixed with coconut cream before cooking in a wrapper of leaves—the cream makes the dish softer (palu);b. masi penu, when themasiis mixed with grated coconut (penu) before cooking;c. masi tao 'ato, when the kneadedmasiis cooked in a basket ('ato) to provide quantity. It is hard but keeps for a long time. It forms a useful reserve food to take on journeys. The rations are broken off as required. Thismasi tao 'ato wasalso a useful provision on sea voyages.
Tradition of first masiAt Neiafu in
Savaii , a small hole in the rocks about 2 feet in diameter and 2 feet deep, is pointed out as the first pit (lua'i masi) used in Samoa. A crippled couple named Ui and Tea had a daughterSina who was taken to wife by the god, Tangaloa-langi.Owing to her parents not being able to get the breadfruit down from the trees, Sina brought them thetuaoloa(east) andto'elau(N.E. trade) winds to bring down the fruit for them. The two winds failed to bring down sufficient fruit, much to the crippled couple's disgust. In answer to their complaints,Sina sent the boisterousla'i(west) wind which effectively brought down the fruit. The old couple, at last satisfied, gathered the fruit and stored the excess quantity in the hole alluded to, where it became converted intomasi.The
to'elautrade wind ushers in the good season in Samoa and brings gladness to all. The old couple, however, were the exception, and complained about theto'elaubecause it did not serve their immediate single need. Hence the saying: "Na'o Neiafu ua mele ai le to'elau" (only Neiafu despises the trade wind).
The yam. The yam (ufi) though favored as a food, does not enter into the Samoan food complex in the same way as the talo and the breadfruit. It is much scarcer. This is partly due to the greater labor entailed in its cultivation. Its status is high, for one informant coupled it with talo fa'ataisi among the four requisites for a visiting chief's meal. In his list he retained the drinking coconut and the fowl, but left out the pig, and placed the yam before the talo in order of precedence. It is thus evident that the yam ranked high where pigs were scarce. The yam preparations are exactly similar to those of talo except that fa'ausi was confined to talo.
Ufi tao. The scraped yam was cooked in the oven,too. Small ones could be cooked whole, but large ones were cut into appropriate sizes to ensure being cooked.Ufi fa'ataisi. The same as withtalo. The termtaisi ufiis also loosely used. The coconut creamfai'aiis also used in a similar way.Loloi ufiorfa'alifo ufi, where thepe'epe'eis added to the whole or dividedufiand cooked with it.Fai'ai valuvalu ufi, where thepe'epe'eis added to theufigrated in the same way astaloand then cooked.Sofesofe. This is a variation of thefai'aipreparations. The yam is cut up into smaller pieces and cooked withpe'epe'e. It is thus intermediate between the two preceding preparations.Kramer (18, vol. 2, p. 153) terms this preparationloi ufi.
The sweet potato ('umala, Ipomea batatas), important in other parts of Polynesia, has little status in Samoa. Both breadfruit and banana are more important articles of diet. The sweet potato was said to be eaten when there was no other choice. On talo which had failed through a severe storm. They thrived in the sandy soil and were of good flavor. Even then the people apologized for having to use them. The preparations made with talo and yam could be made with 'umala but the impression given was that any such trouble was not justified.
The banana (soa'a and fa'i). The banana is divided by the Samoans into two main, groups; the soa'a (plantain) and the fa'i. Ripe bananas are eaten as a fruit. Green bunches are picked and ripened more quickly by dipping them in sea water before hanging them up or by putting them in a pit. The process of ripening in a pit is termed fa'aotanga and the oblong shallow pit, lua'i fa'aotanga, or luai fa'avevela. The pit is lined with dry banana leaves, and the bunches covered with more leaves. Sometimes a couple of pieces of dry coconut husk are lighted and left in with the bananas. Some cross pieces may be placed over the hole, and coconut leaves and earth complete the covering. Ripe fruit is also made into a thin gruel (poi) and surplus stock into masi fa'i.
As a food, however, bananas are highly esteemed as a cooking vegetable to eat with flesh foods. The best kinds are the fa'i pata and the recently introduced fa'i papalangi (Cavendish).
Fa'i tunu. The unripe unpeeled fruit is cooked on the open embers of a fire generally in the cultivations to avoid the trouble of making anumu.Fa'i taopa'ua. The unripe, unpeeled fruit is cooked in the oven on a layer of leaves to prevent scorching.Fa'i tao malaulauortao fofo'e. The unripe bananas are peeled with thefofo'epeeler before cooking in the oven. They are sometimes thrown into a bowl of sea water as they are peeled. This is the commonest form of cooking. The cooked bananas are somewhat hard, but the Samoans are very fond of them.Fa'i oloolo. The peeled unripe bananas are grated (oloolo), wrapped in banana leaf in flat packages (fa'apapa), and cooked in the oven. Coconut cream may be cooked with it. Formerly the bananas were grated onlapacoral but now perforated tin is used.Sua fa'i. Thin slices with a little water are wrapped in warmed banana leaf and cooked in the oven principally for sick people and invalids. The iron pot is now used.Loi fa'i. This is afai'aipreparation, the peeled unripe bananas being wrapped in banana leaf with coconut cream and cooked.Sai fa'i. Ripe bananas are peeled and dried in the sun. They are then wrapped in dry banana leaf (sului) and bound (saisai) closely round and round as in the modern preparation of native tobacco. The bundle of preserved banana is called asai fa'i.M asi fa'i. Themasiis made of peeled ripe bananas stored in a pit in the same way as breadfruitmasi. As with breadfruit, it is only done when the crop is so abundant that it can not otherwise be disposed of. The bananamasiI tasted inOlosenga was cooked in flat leaf packages in anumuby a working party on a cultivation. Though somewhat tough, it had a pleasant taste, and the odor was not too pronounced. It is admitted that bananamasiis probably modern and has not an ancient origin like breadfruitmasi.
Poi. Ripe bananas are peeled and placed in a wooden bowl with water. The fruit is mashed with the knuckles (tu'itu'i), squeezed between the palms ('o'omi) and between the fingers (lomi) to produce a thick gruel to which coconut cream and the juice of a lime (tipolo) are added. Before the introduction of limes, the leaves of themoengaloshrub were used for flavoring.The
poiis prepared publicly at the back of the house like kav'a by a garlanded girl or youth sitting behind the bowl while assistants peel the bananas and drop them into the bowl. Thepoiis served in half coconut shells from the discards of grating coconut cream. Each person gets an individual cup without set order of precedence or official calling of names. The attendant starts at one end and works around in the order in which the chiefs are sitting. Usually two or more young men serve as quickly aspossible, and fill other shells with which to replenish the empty vessels. Serving goes on until the bowl is empty. The poioften takes the place of an ordinary brew of kava, but never in proper ceremonial. To show attention to a guest, a chief will often bring in a ripe bunch of bananas to be made intopoiwhich he waits to share as a further sign of attention.
Plantains (soa'a) are preferred to bananas for cooking, but they are not so plentiful. They may be grilled (tunu) or cooked in the oven (tao). They are used at a riper stage than bananas and always with the skin left on. Cooked ripe soa'a are held to be good for children.
Plantains cooked with coconut cream in a package are termed loloi. The grated form of oloolo is made as with bananas. A fa'ausi of plantain is also made by serving the cooked grated fruit in the oily coconut preparation tolo niu as in talo fa'ausi.
Arrowroot (masoa, Tacca pinnatifida) is cultivated, the roots grated on a sennit grater, washed, and decanted in special large bowls called tanoa masoa. The prepared material is cooked in three ways.
Fai'ai vatia. The arrowroot is mixed with coconut cream in a bowl, made up in packages in banana leaf and cooked in the oven.Piasua. The arrowroot is mixed with water in a bowl and heated stones dropped in. It is stirred and coconut cream added. When cooked, the mass is clear and gelatinous. It is cut up into smaller pieces with a strip ofalavafrom the back of the coconut leaf midrib. It may be served with coconut cream.Vaisalo. To thevaisalopreparation (see p. 129) arrowroot may be added by crumbling it in as the heatedvaisalois stirred. The mixture, although referred to asvaisalo, is really a mixture ofvaisaloandpiasua.
Turmeric (malasina). In the preparation of turmeric from the roots of the ango (Cucuma longa) a stage is reached where the lighter lenga to form the yellow dye is separated from the, heavier part known as malasina, which is used as a food and is prepared in two ways: 1, fai'ai malasina, when the malasina is mixed with coconut cream, wrapped in warmed banana leaf, and cooked in the oven; 2, when heated stones are dropped into a bowl containing the malasina. The malasina is stirred and it thickens as it is cooked. No special name was obtained for this preparation.
Papaia (esi, Carica papaya). The papaia is eaten as a fruit but is not greatly in favor. The fact that it is appreciated by foreigners is well known. Thus during the
Fai'ai esi. The peeled papaia is cut into quarters, wrapped up in banana leaf with coconut cream and cooked in the oven.Sua esi. Thin slices of papaia with water were wrapped in banana leaf and cooked in the oven. As in the case of bananasua fa'i, a pot is now generally used. The food is for invalids, and coconut cream may be served with it. This form is described byKramer (18, vol. 2, p. 151).
Seaweed (limu). Seaweed is eaten raw. It is also cooked for invalids and old people. In fai'ai limu, the seaweed is washed with fresh water in a wooden bowl. It is then wrapped in banana leaf with coconut cream and cooked in the oven. It forms about the only alternative to cooked talo leaves.
The ti (Cordyline terminalis). The underground stem of the ti, especially the ti manu ali'i and ti vai, were cooked for the saccharine material contained therein.
In making
umu tia very large oven (umu ti) is prepared communally by the village (nu'u). The families dig up the stems and bring them in baskets to the common oven. The oven is dug out to about 10 feet in diameter. Larger stones than usual are heated on a fire made of logs. The leveling of the stones takes some trouble. Six or more men dressed in ti leaf kilts and wreaths do the work with long poles (sosofa). Sometimes a heavy log tied with ropes is dragged from side to side. The levelled stones are covered with the leafy heads of the ti. The underground stems, still in baskets, are placed in the oven in the parts assigned to the different families. They are next covered over with more leaves and buried with earth. The oven is left covered for a week or even longer. When it is opened, the families take their own baskets. The outer bark is removed. The cooked ti can be chewed direct. From its fibrous nature and sweet taste, it somewhat resembles sugar cane. No ceremonial detail was obtained, although it existed in the past.A preparation (
otai) is made by using the liquid from a green coconut, the gratedpenuof a mature nut, and cut up pieces of cooked ti.
Sugarcane (tolo). Sugar cane was primarily grown for thatch, but now some kinds are grown for their saccharine properties. Lengths of thick cane are often brought in with the food for visitors. The outer skin is peeled off and the cane chewed.
The candlenut (lama, Aleurites molucanna). Though used for lighting and to obtain carbon, candlenut does not seem to have been used as a relish, as it is in
From the detailed enumeration of the cooking recipes, an idea can be formed of the comparative values of the different foods and the uses of the various utensils in Samoan domestic economy.
The outstanding value of the coconut in cooking is evident. In some form or other it enters into combination with every vegetable food and most marine flesh food except the larger fish. The meat of the immature nut and the liquid enter into vaisalo and otai. The grated mature nut is used with masi penu and otai. But it is the expressed coconut cream that is invaluable in so many preparations. Cooked with them as fai'ai, it enters into 17 different dishes. As the oily niu tolo, it is indispensable to the most important made up dish, fa'ausi. It provides the only sauce for meat, vegetables, and puddings. Without the coconut, Samoan cooking would be resolved into its primary elements.
The talo demonstrates its position as the chief vegetable. It provides the basis of the Samoan diet, the most valued chiefly dish either as talo taisi or fa'ausi talo, and the only green vegetable.
The method of cooking in the oven (tao) is the essential one with all foods. Grilling (tunu) has a limited use. The method of heating by adding hot stones to a liquid contained in a wooden bowl is seen, however, to be important. It enters into the heating of coconut cream for palu sami and fa'ausi talo, two very important dishes. It is used in one form of breadfruit taufolo, next in importance to fa'ausi talo from a ceremonial point of view. It is the cooking method in coconut vaisalo, arrowroot vaisalo, piasua, and the turmeric malasina. Both vaisalo and piasua are invaluable foods for the sick. The use of the wooden bowl as a cooking vessel in this method must also be noted. Likewise the iofi tongs receive additional use. The use of the impervious banana leaf wrapper around food containing liquids must be stressed. It is a primitive form of broiling—a process within a process. The entirely liquid coconut cream and pigs' blood is thus cooked without a vessel even of wood.
Of the domestic utensils, the tauanga strainer is indispensable in the preparation of the extensively used coconut cream. The alava only comes into use with the preparations in which immature coconut meat is used.
The coconut grater is also indispensable. The coral grater is seen to have had more use than might be apparent at first thought. It was formerly used with uncooked talo to prepare talo valuvalu, and fa'ausi talo, and uncooked yam for ufi valuvalu, and with the banana and plantain for fa'i oloolo, soa'a oloolo, and fa'ausi soa'a. The sennit grater enters only into the preparation of the turmeric malasina.
As regards pounders, the two forms have only one use each; the green breadfruit autu'i for breadfruit taufolo, and the coconut leaf midrib ulu lapalapa for talo taufolo. Breadfruit taufolo, though important for guests, plays no continuous role in every-day diet.
The Samoans are very early risers. When busy with their cultivations they go off at daybreak without having any set 'meal. If there are any cooked talo or breadfruit left over from the previous day, these may be eaten. They return about 10:30 or 11 A.m. with a load of talo, breadfruit, or bananas. The oven for the morning meal is then prepared. This is a hearty meal and combines the morning and midday meal of three-meal races. Enough is usually cooked to provide for the evening meal. Thus, besides the vegetables, extra packages of fish and palu sami are cooked. These remain unopened.
The evening meal usually takes place after sunset. On my remarking that it was a good time, as the flies had retired, I was assured that that was one lavalava kilts they are prepared to make themselves comfortable. In these days, family prayers are held and the meal follows immediately after. With visitors in the village, a fresh meal is cooked in the evening as also when food is plentiful.
If food is plentiful, three meals are partaken of, one being cold. The people are not slaves to the conventions of time. If a good catch is taken of a particular fish that is in season, they are cooked and partaken of as soon as the fishing activities are over. On divers occasions, snacks are disposed of. Several helpings of banana poi form a small meal in itself. The women often go down fishing among the rocks, and take cooked banana or talo with them to eat with raw fish. Single dishes such as piasua or vaisalo are often partaken of without relation to a regular meal; so also is the small repast of shark or pork with the drinking of kava. The custom of two meals a day may be regarded as being influenced by food supplies and not by any dietetic rules. There is a readiness to eat and feast at any time should food be available. There is a tendency to overeat and to encroach on what was cooked for two meals. The lack of secure storing places and the availability of the cooked food baskets is against the careful conservation of food. This somewhat wasteful tendency was observed in ancient times by Le Polo of pae where he sat as the villagers returned from their cultivations. Upon the pae, he made them deposit one-half of the food they were bringing in. When asked the reason, he replied with one of those trite sayings so much admired and quoted by the Samoans: "E nanea mea mata, ae le nanea mea vela." (Uncooked food will last, but cooked food will not last). The people returned for the rest of their food on the following day. Le Polo's precaution thus prevented their running short on the second day. The stone platform sufficiently large to accommodate the villager's food is still to be seen at
The basis of the diet is a vegetable one of talo and breadfruit. With this is required an ina'i or relish of another kind of food such as flesh or fish. Fish is the staple ina'i. Equally important is the green talo leaf vegetable especially in the form of palu sami. Flesh and fish are not always procurable. A Samoan will then make a satisfactory meal of talo and palu sami. The salt and coconut cream in the palu sami afford the ina'i relish. By acting as a psychological substitute for flesh foods palu sami saves the position, and ranks as one of the most important articles in the Samoan dietary. Water is the universal beverage. Kava is never drunk with regular meals though it may precede them.
For ordinary scratch meals, women may be seen eating out of a basket. For men, the food is properly served on laulau platters. The platters are flat, oblong trays made of coconut leaflet, and form part of the essential furnishings of dwelling and guest houses. The young men or women serving the food bring down a pile of platters from the fata shelf and place them beside the baskets of cooked food at the back of the house. A platter for each guest is laid out and the cooked leaves (afei) from the oven which form covers for the food baskets, are spread over the platters as covers. The green leaves so often used on the trays for foreign visitors are an innovation introduced for them. The cooked afei leaves were the only form of leaf allowed by Samoan custom. When there are not enough platters, the less important chiefs are served on leaves alone. The food is then distributed on the platters. In an important meal with guests present, a talking chief personally superintends. Breadfruit and talo, which form the basis of the meal, are distributed on the basis of one or two whole talo and a breadfruit to each platter. The fowl, pork, packages of fish, and palu sami follow. The young men quickly place a platter before each guest, commencing with those of highest rank. The pork in ceremonial distribution may be delayed and follows after. One attendant may place the platters and others follow in quick succession bearing-shares of flesh foods which are added to the platters before the guests.
As a beverage, drinking nuts are distributed to each guest. If anyone desires a drink of water, he calls to an attendant and a coconut shell cup of water is brought to him. The fingers of both hands are used in eating, but for the soft cooked talo leaves in palu sami, a piece of green coconut leaflet midrib (su'i) is provided by the attendants. The su'i is used as a fork to convey the palu sami to the mouth. To eat palu sami with the fingers direct is not regarded as correct.
During a day meal, girls or young men fan the principal guests and incidentally keep the flies off the food. Guests of honor are particularly waited on by the taupou. The service during the meal is very good. The attendants watch the guests and are ready to anticipate every wish. At a feast, the portions of food are far in excess of what can be eaten in one meal. It is the person's just share with the appropriate portions of flesh food due to his rank. A member of his family sits behind him under the eaves with a basket. When he finishes his meal, he passes the remaining food back and it is taken home. A member of the family with whom I was staying performed this duty for me. Any choice bits were served up again later, but the rest were the natural share of the family. In an ordinary meal, the platter is pushed away by the guest and speedily removed by an attendant. There is no ceremony of waiting till the others have finished. A bowl of water is brought for washing the hands. In these days, it is accompanied by a towel.
The girls and young men who attend on the guests, never eat until the guests have finished and the food is cleared up. They may then eat at the back of the guest house or retire to the dwelling house at the back.
Food is nearly always served with the guests sitting before the various wall posts. Occasionally in a family gathering, a long setting is made by laying green banana leaves on the floor and placing the food on them. Then men and women sit down together, but the seating places are always indicated by a talking chief or one who acts in that capacity.
Food customs are very important in Samoa and the ceremonial associated with them is still observed. All those to be described were seen several times, and play an important part in the social organization of the people. The chiefs are fed on the best food available. In ancient times, high chiefs who were regarded as kings, such as
A chief's food is termed sua, and the chief's word for ordinary food is suavai. It was cooked in leaves and brought in separate baskets to that for the lesser people. What is left over from the chief's meal may not be eaten by his own family. It is sa (prohibited) to them. It goes to his talking chief or to attendants not belonging to the chief's blood family.
Another individual custom is that of calling a chief's kava cup. During the ceremonial drinking of kava before a meal, a lesser chief could call out that he desired to follow a certain chief's cup (ipu). He received a cup of kava out of his turn. When the meal was served, the chief mentioned had to give his portion of food to the person who called his name. On
The sua ta'i is the ceremonial meal brought in to a high chief to honor him (fa'aaloalo). Four bearers bring in a drinking nut, talo taisi, a fowl, and a roasted pig in that order. The bearer of the drinking nut has a piece of bark cloth siapo girded round his waist. The chief sits cross-legged before his proper wall post. The bearer as he nears him, sits down on the floor before and sideways towards him. He then passes the drinking nut sideways to the floor before the chief. He removes the siapo and hands that over also. The others follow in succession bearing the food on platters which are laid before the chief from the sideways sitting position. The whole meal is called sua and a sua ta'i from the way in which it is presented. The drinking nut is also a sua: it is punctured through the mata hole and a bent piece of green coconut leaflet midrib may be placed in the hole as a temporary stopper. The nut must be opened only in this manner. To crack the muli end as for ordinary drinking would be taken as an insult and lead to reprisals. The nut is drunk by the chief's officiating talking chief. Owing to the small hole, the fluid is drunk with sucking noises which are regarded as appropriate. The empty shell is then broken on the stone paepae of the house. This again is of high significance. In ancient days, coconuts were only broken for the gods and hence today are only broken for chiefs of the highest rank. The siapo goes to the talking chief, who now takes charge of the food. The leg of the fowl with the coccyx (le vae ma le muli), and the loins (tuala) of the pig go to the high chief with some of the talo taisi. The rest is divided up by the talking chief as he sees fit. The high chief has no jurisdiction. The pig thus presented is called a pua'a fata no matter what the size. The term pua'a fata is usually applied to a very large pig which is carried in by several bearers on a litter (fata). The talking chief puts the siapo round him, goes outside to ailao. The ailao is the custom of calling loudly to attract attention while he calls the name of the chief who gave the sua. He thus calls public attention to the fact that his own chief has been honored, and at the same time acknowledges his thanks to the chief who paid the respect.
If the chief is of outstanding rank, he alone is served on a platter. The others are served on leaves. If there are other chiefs of high rank, each may get a sua. Several sua may be brought to the one chief by various high chiefs in the village. After the visiting chiefs has received his one meal, his talking chief may do what he likes with the other sua. If he has relatives in the village, he may send them presents of pork.
The sua ta'i ceremony takes place after the ceremonial drinking of kava. It awaits the directions of the officiating talking chief. During the sua of Laloifi).
On lesser occasions when pigs are scarce, the sua may consist of the drinking nut, talo, yam, and a fowl. The yam is cooked in leaves like the talo taisi.
The si'u laulau is a small meal given to visitors immediately after their arrival in a village. The local chiefs visit the guest house and bring kava root with them. This kava is for the ingunga (welcoming bowl). The si'u laulau is just enough food to provide a laulau platter for each to form a light repast. Now it usually consists of tea and sugar. The ceremonial meal follows later when the ovens are ready.
The taunga is the very sensible custom of taking provisions when visiting another chief in the same village. The real meaning of the word taunga is surplus cooked food. When the cooking oven is opened, all the cooked food is placed in a basket or baskets. From them is taken what is needed for the ordinary regular meals. What is left over is used as a taunga and enables evening calls to be made for a number of purposes as follows:
From the above it will be observed what an important role food plays in the ordinary social life of the community. Food forms the basis of social intercourse. Hospitality is more than a virtue. It is a necessity. The taunga custom sought to relieve the burden of the host. The self-inflicted guest was sure of a welcome as he caused no embarrassment to his host.
The laulautasi is a presentation of cooked food to visitors by the chiefs (matai) of the village. The sua ta'i with pigs baked whole are given by high chiefs whose position or wealth enables them to give it. The laulautasi is a contribution in which all the lesser chiefs share. The matter of individual contribution is arranged beforehand by the village talking chief who makes a levy of three to five talo on each chief together with the appropriate accompaniment of a fowl, fish, or even tinned beef or salmon. Each chief sends his basket of food to the back of the house while he takes up a position within.
The baskets are collected in the back part of the house where the village talking chief presides over them. Each basket is handed to him in turn, and the name of the contributor is given. The talking chief calls the name saying "This is the basket of food of—. It contains five talo and a bird." As he calls the articles he lifts them up, drops them back again, and lays the basket to one side. Each basket is called, together with the articles contained, such as packages of fish or squid (afi). Some receiving chiefs make a running commentary on the food; as, "Five large talo in good condition. One bird very fat." The remarks are sometimes humorous; as, "A package of fish. If more fish were put in, the package would not feel so slack. One bird, the palu sami, fai'ai, fish, and squid. The ceremonial name of a fowl is here ta'a paepae or manu tele. At one village the receiving talking chief held up three talo and said, "These are sungalo." The sungalo is an inferior talo from a cultivation overgrown with weeds. The senior village talking chief who sat in his official place asked for them to be passed over to him. He examined them, gave a grunt of disgust and laid them aside. The donor had not anticipated that the contents of the baskets would be scrutinized. He was dealt with afterwards and fined for attempting to evade the true spirit of the custom.
The visiting talking chief acknowledges the gifts of food in a speech. He then proceeds to divide up the food into two portions, one for the visitors and the other for the villagers. The packages of fish and palu sami are usually taken out of the baskets and placed with the fowls to enable him to divide more easily. The general distribution being made, the village talking chief again gets busy and makes the individual distributions on plaited platters which the attendants quickly place before guests and hosts. All eat together. The individual division of food is called tufaanga. A larger portion given to any privileged person is a tunga. Any person with some extra portion may pass or send some of it to another. It is an act of courtesy which is never refused by the recipient.
The ta'alolo is a presentation of food to visitors given by the village, and is of a more public character than the laulautasi. Visitors of distinction, after the introductory bowl of kava, are given a ta'alolo. For the evening meal the presentation of food takes the form of the laulautasi.
The food which includes roasted pigs is carried in a public procession and deposited in a heap on the house platform. The procession is headed by the taupou in full regalia, accompanied by buffoons, who by their antics, direct attention to her. The food is then handed over by the village talking chief with a speech. In the speech, he addresses the visiting chiefs by their ceremonial titles and names of the districts they came from, indicating that the food offering is a token of respect paid to them by the village, and enumerating the articles of food.
A chief representing the visitors steps out on the house platform beside the food. He is sometimes accompanied by others to show greater respect. With the lower end of his to'oto'o staff placed between the first and second toes of his right foot, he draws the attention of his own party to the food offering. The formula used is a variation of the following: "Silasila ia lau susunga a Laloifi, ma Falesau ma Faleomavaenga, Tuitele ma oe le matua ma lenei malanga i le fa'aoloalo" (Behold O presence of Laloifi and Falesau and Faleomavaenga, Tuitele and you the father and this company of travellers,
In connection with ta'alolo, it was usual to serve some special dainty such as fa'ausi or taufolo. They were always brought in separately from the other food.
The fa'ausi of baked, grated talo in the oily niu tolo preparation of coconut is always brought to the house platform in a wooden bowl. Young men garlanded with leaves bring the bowl from the cook house with ailao yells of "Mua o", to herald their approach. This announces to the visitors and the village that special attention is being paid to the guests. Green leaves of the maopo breadfruit are used to hold the individual portions. The bowl bearer sets the bowl down on the house platform, an assistant cups a leaf in his hands, and the portion of fa'ausi cubes with a share of oil is ladled into it with a half coconut shell. In placing the portion before the guest, a hollow is usually made for the leaf platter by scraping away a few stones from the gravel floor in front of him. For high chiefs, the leaf is first placed on a mailo platter of coconut leaflets before the portion is served. In olden times, a mailo platter was used for high chiefs alone. The lesser chiefs were served on the bare leaf. Now everyone gets a mailo platter.
The taufolo preparation of pounded breadfruit is brought in the same way by garlanded attendants making the ailao yells. The already pounded breadfruit mass in a bowl is placed on the house platform. The bowl bearer sits before it while attendants pour in the coconut cream to which salt water has been added. The operator proceeds quickly to pinch off the pounded mass into smaller pieces which are rolled or pressed into rounded lumps. The portions are placed on maopo leaves with some of the cream and conveyed by attendants to the guests. Banana leaf may be used instead of maopo. The usual hollows are made on the floor if necessary. The taufolo is served to all alike on the bare leaves, and no mailo platters were used ceremonially in the past, though they may be used incorrectly now. While greatly appreciated, taufolo has not got quite the status of fa'ausi. For one thing there was no oil which the system required and had few opportunities of obtaining.
Talinga. Similar to the preceding ta'alolo ceremony is the women's ta'alolo, or talinga. This ceremony has gained ground and became separated from the men's ceremony owing to the organization of women's committees in the various villages. In those given to us at the
The women, arrayed in bark cloth and fine mats, assembled at some little distance from the guest house and then marched in mass. Each had some present of
siapo, shell necklaces, or food. The food usually consisted of fruit such as bananas, papaia, and drinking coconuts. Sometimes fowls were brought. In front was the village maid arrayed in fine mat, dancing skirt of special make and thetuingaheaddress of bleached hair. She carried a weapon, now usually the modern steelnifo 'oti. She advanced at a slow trot, every now and again making a curious little side step as if changing step and bending the body sideways towards the visitors. With the club, she made appropriate gestures. Associated with her, were one or two old women who by wild grotesque antics drew attention to her. They were dressed in an unusual manner to attract attention, and armed with sticks, ran about hitting trees, striking the ground, and pointing to the taupou. Their efforts were accepted as a matter of course and regarded as of a humorous nature. The village maid came right down to the guest house, paraded in front of it, and trotted back. During the whole time, the main body kept up a chant and beat time with their hands. Owing to the deliberate and slow approach of the main body, the taupou was enabled to make two or three excursions before the women got close to the house.The women came forward, laid their offerings on the house platform, and retired a few paces. One of their leaders then made a speech welcoming the visitors in the usual ceremonial fashion. A visiting chief replied, thanking them for their tokens of respect (
fa'aaloalo). Another chief replied also with a running commentary on the presents as he picked them up and bundled them into baskets. On these occasions, it was usual to let the younger or lesser chiefs reply on behalf of the party.The women meanwhile had arranged themselves in rows usually on floor mats which were brought forward. They then performed a
sivaaction dance in unison and in perfect time. Time was beaten on folded mats with two sticks and the bent knees in the cross-legged position kept time to the beat. The motions were made with the hands, head, and upper part of the body. Now and again the whole lines changed position on the mats so that they all faced to the side or back. In some movements, they rose to their feet and after some action subsided to the ground again. The taupou occupied the middle position in the front row. The old women remained standing, and by exaggerated gestures continued their humorous role.After the massed
siva, individual posture dancing in time to the beating on the mat was indulged in by two or three relays. An individual dance by the taupou usually concluded the performance.
The talo pa'ia is a form of ta'alolo, in which a whole district takes part in presenting food. In olden times, they were difficult to arrange owing to quarrels that arose between villages through questions of precedence. Nowadays, they are sometimes arranged for government officials.
Districts where precedence has been sharply defined and maintained do, on occasion, give a talo pa'ia. Such occur in the southern district of pule (rule). In Safotulefai there is a hereditary heraldic title of Tangaloa-tea. Four chiefs who have tulafale ali'i positions, combining high rank (ali'i) and the powers of talking chiefs (tulafale), exercise the pule over the district. They meet and decide as to whether the ceremony shall be performed for a specific occasion. They then call a meeting of chiefs and orators, who ratify the decision. The herald is sent to the different divisions of the district to notify them and to announce the number of tuinga expected from each division. The tuinga is the head-dress worn by the village maids and chiefs' sons. Only certain chiefs' families tuinga. In some villages, there may be more than one family so entitled.
The divisions or villages proceed to make preparation. The talo is always the basis, but arrangements have to be made about the meat, mea, ina'i or 'i'i, to go with it. The village talking chief rules that there must be talo 'u'u, that all must contribute talo. He rules 'i'i le pua'a (the meat must be pork). He allocates the providing of a pig to certain sections within the village. The matai (lesser chiefs) of that section decide whose turn it is to contribute the pig. They take turns to supply the various calls made. The term 'i'i le pua'a literally means to make the pig squeal.
The official herald goes through the village proclaiming, "Sao le ta'alolo" (Collect the food for the ta'alolo). The food is gathered together and the shell trumpets sound "Tatou tu'u" (Let us go). When the collecting village of tuinga headdresses in front. The tuinga, are thus in line in the advance to the collecting place for depositing the food. They are all equal and no trouble occurs. If they advanced in column of sections with a tuinga heading each section, there would be trouble as to which tuinga should lead the first section.
The open line advances on the assembly house, each person carrying some food, the wearers of the tuinga in front, going through appropriate actions as in the women's ta'alolo. The main body as they advance, indulge in the songs and dances which have been previously rehearsed.
The food is placed before the assembly house and the bearers retire to the sides or to a neighboring house. The official talking chief of the party makes a speech in the open termed fa'afiti le mu'a. No matter how large the contribution, he apologizes for the small supply brought.
The receiving talking chief steps out in front of the assembly house, and in ceremonial language thanks the visiting village for their contribution. The food is now termed folafola le ta'alolo. The talking chief of the malanga (visitors) now has pule (power) over the food of the ta'alolo. He distributes it in the usual way. Each share is here called a tu'unga.
An interesting custom exists with regard to the rights of a stranger to demand hospitality at a feast. A stranger who comes as a guest is, of course, provided for in the usual way. The stranger who is unheralded and unknown has the right to the stranger's post. Entering" from the front, it is on the" left in the guests' section (tala). Being careful to eliminate the wall posts of the middle section, he commences counting from the post nearest it. The fourth wall post is the stranger's post. If the stranger is of chief's rank, he takes the fourth post on the front, but if beneath that rank, he takes the corresponding fourth post at the back. No matter what chief is sitting before
Kava (Piper methysticum) is cultivated not only for ordinary use but as an essential element in social usage and ceremonial. It grows best in stony around and is to be seen planted close to the houses where such conditions exist. Different varieties are recognized. The Samoan form of the kava name is 'ava.
The Tutuilans maintain that the first kava was found at Fanga-fue on the north coast of 'ava va'ai (wild kava), which joined together again immediately it was cut. The medical use of kava has been summarized by
The plant was suitable for use after four or five years growth. The whole plant was dug up and the tops removed leaving part of the stems attached to the roots. The mass of roots was split up into convenient portions. Two internodes of stem were left attached to large pieces of root to form the tungase for presentation to visitors of rank. (See Plate VIII A, 1.) The stem part formed a handle for carrying the tungase. The earth was beaten off and the roots washed, scraped, and dried in the sun. Roots not suitable for tungase had the stems cut off, and were divided up into short lengths to form the fasi 'ava for ordinary use. The chiefs' name for this form was 'ava fele fo'e. The long liner roots made the best kava and were not scraped. (See Plate VIII A, 3.) They were also used in the 'ava uso ceremony (uso, brother). A chief had to have a stock of dried kava always on hand not only for his own use, but for the innumerable calls made upon him by ceremonial custom. Those below the rank of matai (head of a small family group) were exempt from such calls. On election to the rank of matai, one of the first things to be done was to lay in a stock of kava. Failure to comply with the customary presentation of kava led to loss of prestige and probably deposition from a position which the holder failed to maintain with dignity.
The materials necessary for preparing the beverage also termed kava are the dried kava root and water. The utensils used in preparing and serving are the bowl, strainer, and cup. In former times, the root was broken up into small pieces and chewed, but now it is pulverized mechanically. Mortars and
Kava bowls (tanoa 'ava) may be divided into two main types, round, and elliptical. All are cut out of a solid section of the tree.
Round bowls are practically circular at the rim with an inverted dome-shaped body set on legs and furnished with a perforated lug by which to suspend the bowl. They range in size from the small bowls for family use to the very large bowls used on ceremonial occasions. The cavity (liu) is evenly concave and fairly shallow. The proportion of depth to diameter ranges in the examples figured (Pl. IX, A and C) from 26 to 30 per cent. This forms a marked contrast to two large but proportionally much shallower Fijian kava bowls in
The rim in all bowls has a level upper surface formed by the horizontal space between the inner and outer surfaces of the bowls. When the inner and outer surfaces are approximately parallel near the rim, the rim surface is but slightly wider than the thickness of the wood below which is accounted for by the angle formed by the upper surface. The outer surface of the bowl is sometimes inclined outwards from its normal plane near the rim and may make the upper surface of the rim as much as 0.5 inches greater than the thickness of the wood below it. (See Plate IX, A.) Again, the outer surface may be trimmed inwards which makes the rim upper surface narrower than the wood thickness below it. (See Plate IX C.) The upward directed surface of the Samoan rim is emphasized because in the two Fijian bowls mentioned above, the cavity has a concavo-convex curve which directs the surface of the rim outwards. The Samoan rim surface is usually perfectly level and plain but two bowls in B). The rim is rarely quite circular, cross diameters usually varying about 0.25 inch and in some bowls as much as 1 inch.
The legs, normally consisting of four, are attached to the sides of the bowl usually 1 to 2 inches below the rim but sometimes as close as 0.5 inches (Pl. IX, B and C). The attachment to the bowl is oval with the broad end internal and the long axis radiating from the center. The curve of the outer part of the attachment is blunter when the legs are rounder in cross section. The legs taper towards the lower ends where they are elliptical or round in cross section. The amount of taper varies, ranging from legs nearly cylindrical to those that almost appear pointed. In some legs, the outer line from the junction to the lower end instead of being straight may be concave giving the false appearance of a curved leg. The lower ends may not be cut level,
The spacing between legs is not exact, no two pairs being the same distance apart. With the lug towards the observer, the legs are wider apart from side to side than antero-posteriorly. The length of the legs, measured on the inner side, range in the series figured from 3 to 4.25 inches and all give good clearance to the bottom of the bowl.
The suspension lug, a constant feature in all bowls, projects from the outer side of the bowl midway between a more widely spaced pair of legs and approximately on the same level as the leg attachments. Two forms prevail; the "V" shape and the "T" shape. The "V" shaped lug may be open or solid. In forming the "V" shaped lug a solid triangle with the apex towards the rim is left in position on the outer surface of the bowl when shaping it from the solid. The base of the solid triangle is subsequently cut out to define the two sides into the arms of the "V" when making the open form. The arms range up to 2 inches in length and from 0.5 to 1 inch in depth or outward projection from the bowl surface. The thickness is less at the outer ends and may reach 1 inch at the junction of the arms. The surfaces of the arms are sloped towards each other to form a sharp edge, or the edge may be rounded off. The free edge of the arms may be trimmed down towards the free outer ends, making them lower there than at the apex. The free ends may be left vertical or cut at a slant. A hole is bored transversely through each arm on either side of the apex and on the level of the surface of the bowl.
In the solid "V" form, the wood between the sides of the triangle is not removed. Of the three sides of the triangle, the base is vertical but the sides are sloped inwards towards each other. The plane of the external surface is altered by cutting at a slant from the apex towards the base so that the apex is deeper than the base. Two holes, one on either side of the apex, are bored through the solid triangle to emerge through the base surface. (See Plate IX, C.)
In the "T" shaped lug, a triangular piece of wood, with the base towards the rim, is left on the outer surface of the bowl in shaping from, the solid. This is subsequently shaped into a raised "T" with the crossing limb formed by the base running parallel with the rim and the stem running radially towards the center. A hole is bored through the cross limb on either side of the stem junction.
One or more cords of sennit braid are passed through the holes and tied into a loop to which a thicker cord is tied, or longer cords are drawn through to their middle and plaited together into a thicker three-ply braid.
One type of Samoan bowl has many legs, rounded, or sometimes square, set close to the rim of the bowl. The extra number of legs was supposed to D.) Such bowls are in common use but are particularly made for tourists who are charged so much a leg. The method of incised carving of the rim and inlaying with lime seems to have also been stimulated by the tourist traffic.
Elliptical bowls are small and only suitable for a few persons. The cavity at the rim is elliptical but the outer edge of the rim is prolonged into points at the ends to form an acute ellipse. The bowls have four leg's and a suspensory lug. The outer surfaces from the sides meet in a mesial longitudinal edge which extends from end to end on the under surface. The legs are round but somewhat elliptical at their junction with the body clue to the curve of the outer surface and are further away from the rim than in round bowls. The suspensory lugs, doubly perforated, are placed in the middle of one of the long sides, at a slightly higher level than the leg attachments. In the two bowls figured, one (Pl. X, B) forms a widely open "V," whilst the other (Pl. X, A) forms a roughly made straight projection parallel with the rim.
In
Some of the large round bowls belonging to high chiefs were named and it was considered an honor to have the privilege of drinking kava prepared in them. Such a bowl belonged to the
The drinking cups (ipu'ava) are made of the muli end of the coconut cut transversely. The transverse direction is specified because the nut may be cut longitudinally or obliquely as was done by the Hawaiians. Through the projecting muli, a transverse hole is bored through which a loop of sennit braid is usually attached. The cups vary in size, not only from the size of the nut used, but from the segment cut off. Regarding the mata and muli ends as the poles, the smaller nuts are usually cut off on the mata side of the greatest circumference thus forming a small but deep cup. With large nuts, the division is made on the greatest circumference or even below it, thus
In some kava cups the outer surface is not polished but left in the rough, natural state except for a depth of 1 to 1.5 inches from the circumference of the rim. Such a cup was that owned by the late
Some cups, from the polishing both inside and outside were very thin. An ordinary half shell from a drinking nut is used in the kava naming ceremony (See Plate VIII B, 4.)
The strainer (to tau 'ava) and wringer is made of strips of fau bast and hence also called fau tau 'ava where fau refers to the material, whereas to denotes the actual article, and tau, to wring. A strip of fau, bast to serve as a connecting cord is tied by one end to the big toe. Strips of fau ranging from 3 to 4 feet are doubled and the loop passed under the fixed strip, doubled over it, and the finger passed down through the opened-out loop to pick up the ends which are pulled up through the loop, and drawn taut. In this way, strips are attached close together along the fixed strip in exactly the same way as in making one form of bast kilt. When a number of strips have been so added for about a foot along the fixed strip, it is unfastened from the big toe and the two ends tied together. (See Plate VIII C, 1.) The process is called fifi fau.
To render the new strainer fit for use, the woody particles from a kava brew instead of being flicked off outside are collected on a mat. After the prepared kava is drunk, the woody refuse is put back in the kava bowl, mixed with water and the new strainer soaked in it. The damp strainer is then wrapped up in a green banana leaf or put into the kava cup and left there until needed. The process termed fa'amata takes the newness out of the strainer and prevents it from tainting the taste of the kava. Even with a used strainer, the kava is better if the strainer, instead of being hung up to dry, is put damp into the cup and left there until needed. In the course of time the used strainer gets a tangled appearance. (See Plate VIII C, 2.)
The dried kava root in olden days was easily broken into smaller pieces by beating with a stone on one of the larger stones of the house platform. The pieces were distributed among young women with good teeth and chewed. The chewed material was placed on leaves, and when a sufficient quantity had been so treated, it was collected and placed in the bowl. Young men or boys were also requisitioned for this purpose. The method of chewing prevailed until quite recently. There are many men living, slightly past middle age, who chewed the kava root in their youth.
Chewing was abandoned as a result of European culture influence, and the root is now always pounded on a stone with a smaller one. There are thus no special stone implements belonging to ancient Samoan culture that were used in the preparation of kava. Since pounding on stone commenced, every household has selected some stone to serve as an anvil. Some are flat stones incorporated at the edge or corner of a terrace in the house platform. The most suitable of all, however, are large portable stones that were originally used for grinding stone adzes. The hollows formed by the grinding of the past furnishes a convenient receptacle for the small pieces of kava root. (See Plate VIII A, 5.) Here they are pounded with any rounded, water-worn stone of suitable size and weight and the powdered material is thus prevented from falling over the edges onto the ground. Such stones have been transported from the edges of streams where they were formerly used because of their nearness to water, which was essential to the grinding process (olo). Some of them simply lie on the house platform or have been fixed on the edge of one of the courses, while others lie loose at the back of the guest houses. Many foreigners, and even Samoans, think that the hollows have been formed by pounding the kava root, but though in many the hollows have been roughened and chipped by pounding, careful scrutiny will in most cases reveal parts of the smooth surface originally produced by grinding. Such stones have become regular anvils for pounding kava but the original use which produced the hollow concavities must not be overlooked. They are now in use throughout Samoa.
The stone pounder now used is as stated an ordinary waterworn stone which is kept beside the stone anvil for use in connection with it. A few seem to have been slightly chipped to afford a better grip, but this is unusual. There are no worked stone pounders for this special purpose as any incentive that might have led to their production is of recent occurrence. It has been shown that no such incentive existed in regard to the preparation of foods. Judging from the rough types of the adzes, the attitude of the Samoan toward working stone seems to have been a conservative one. So long as the stone served its immediate purpose, why go to further trouble for aesthetic reasons? The small stone in Plate VIII, A, 4 has been actually chipped out to form a
The kava root was always pounded outside where the anvil had been located. The pounded material was scooped out of the anvil hollow into a section of banana leaf or a breadfruit leaf, brought in at the back of the house, and handed to the taupou or whoever presided over the kava bowl. The pounding was done by one of the young men connected with the household to whom the chief threw out a piece of root, calling out, "Tu'i le 'ava (pound the kava)."
On important ceremonial occasions, the taupou makes the kava as part of her official duties. For lesser occasions an unmarried girl of the family presides over the bowl, and for ordinary drinking, a young man, irrespective of rank, may perform the duty.
A mat is spread at the back of the middle section of the house. The taupou seats herself often with a female companion on either hand. The one on the right attends to the water. The bowl with the strainer and the coconut shell cup are placed before the taupou. The attendant on the right takes the cup. The taupou always turns the bowl around until the suspensory lug with the hanging string is towards her. She dusts out the bowl with the strainer. The kava root which has been pounded outside is brought wrapped up in a leaf and handed to the taupou who pours it into the bowl. The taupou turns to the right rear, holds out her cupped hands to the girl on her right who pours some water into them. The maid never omits to wash her hands in this fashion. The attendant then pours some water into the bowl. Formerly this was done from a coconut water bottle, now it is dipped from a bucket with the cup. The maid proceeds to work the material with both hands to promote the dissolving and mixing of the powdery part of the pounded root. The fau strainer is spread out along the near circumference of the bowl and allowed to become soaked. It is then pushed clown along the bottom and the sides to scoop up the undissolved material. At the far side of the bowl, the ends are brought in to overlap and enclose what has been caught up. The scooping movement is called ao. The strainer is brought up out of the fluid on the palms of both hands directed upwards. It is doubled over and twisted with both hands to wring out (tatau) the fluid. All the fluid being expressed, the maid glances back and throws the folded strainer with precision between, the wall posts to an attendant waiting outside on the house platform. The young man catches it in the air, opens it out and by successive flickings gets rid of the pieces of wood. The strainer is tossed back, caught before it can touch the floor and the maid continues the ao process. A good deal of style is imparted to the wringing by tossing up the folded strainer to reverse its ao movements are changed to fa'apulou (fa'a, like; pulou, head covering). In fa'apulou, the opened-out strainer is dropped on the surface and pressed down to the bottom to pick up the powdery parts that remain undissolved. The outside attendant is no longer needed as the maid after wringing the strainer turns her body to the right rear, opens out the strainer and flicks it herself from where she is seated. She also imparts style to her movements in wiping the sides and upper edges of the bowl with quick upward sweeping turns at the end of each movement. She also wipes any pieces adhering to her hands before she flicks the strainer. The chief who supervises the distribution of the kava, if the speeches are over, may shorten the period of the ao movements by ordering the maid to go on with the fa'apulou by calling, "Fa'apulou le 'ava."
All the undissolved material having been strained off, the maid drops the strainer into the bowl and lifting it high allows the fluid to stream down. From the color, the strength is judged. It is stated that the experts judge by the sound of the falling liquid. If too strong, more water is added.
An attendant now comes forward with the kava cup and stooping down holds it over the bowl. The maid lifts up the saturated strainer above the cup and allows the kava to stream into it until it is about half full. The attendant must always be stripped to the waist in formal kava drinking. He straightens himself up, and holding the cup about shoulder high, awaits directions. The chief who distributes (fa'asoasoa) then calls (tapa) the name of the chief to whom the cup is to be served. If the attendant hesitates, it is usual for the chief named to clap his hands, slap his thigh, Or say something to indicate his position to the cupbearer. The cupbearer should always take the longest way around to the chief whom he serves. Thus, if the chief named is on the left side of the house, the bearer walks around the right side of the central supporting posts of the ridge and then towards the chief, to approach by a short cut from the same side of the supporting posts is poor form and shows that the cupbearer has not been properly brought up. As the cupbearer approaches the chief, he raises the cup head high. The manner of presenting the cup depends on the rank of the chief. To any but a talking chief, he stoops and bringing down the cup with a sweep from the right as low to the ground as possible, he presents it with the palm of his hand towards the chief. To a talking chief he turns slightly to the left and brings down the cup with a backhand sweep from the left so that the cup is presented with the back of his hand still towards the chief. The distinction between forehand and backhand is important. When the chief takes the cup the bearer steps backwards two or three paces and stands still at attention whilst the chief drinks. He then
Receiving the cup. Proper etiquette must be observed in receiving the cup. A chief, other than a talking chief, takes the cup by hooking his forefinger over the rim. A talking chief receives the cup in the open palm without hooking his finger over the rim. If his neighbor is a high chief, he holds out the hand on the side away from him. If there is a high chief on either side of him, he compromises by taking the cup in both hands.
Drinking. On receiving the cup, a few drops are poured out on the floor before drinking. This is termed sa'asa'a and is usually held to be a libation to the household gods. It is sometimes said to be a way of deflecting mischievous spirits while the guest drinks in peace. The people of
Tangaloa-ui while drinking kava with Pava near
Saua on the island ofTau killed Pava's small son owing to the noise and disturbance he created during the ceremony. To the angry parent, Tangaloa said, "Liunga lua le taeao," interpreted as meaning, "There will be another bowl of kava later," or, "The second bowl of kava is the better." The two passed on to Namo, taking the corpse with them, to a house beside the stream which was flooded. There Pava went out, tied sometaloleaves around his head and floated down past the house. Tangaloa, however, not deceived by the floatingtaloleaves called Pava into the house. By this time the kava was ready. Tangaloa again said, "Liunga lua le taeao." When the cup was passed to him, he asked Pava for ataloleaf from his headdress. Taking the leaf and tearing it in two, he poured some of the kava from his cup upon it. Turning to Pava he said "Just as water will not soak into ataloleaf, so death has not penetrated the body of your son." The boy quivered and sat up. Thus for Pava, the second brewing of kava was the better.
In memory of that event, say the Manuans, a little kava is always poured out on the floor ere drinking and the tip of the talo leaf is pinched off before cooking. The usage has been rationalized by the Christian community who as they pour out the kava repeat some formula, such as, "May God bless this gathering."
The mats are usually separated so that the drops may be poured on the gravel floor, or the guest may turn slightly and pour the libation out on the bare gravel near the wall posts. Some pour a few drops on the mat in front of them and others merely go through the motion of tilting up the cup without actually spilling any kava. The small quantity poured out is termed sa'asa'a. The guest then raises the cup to his lips with the salutation of "manuia." Some continue the invocation into a salutation to those present, especially the guests. The salutation takes some such form as "la manuia le afionga o Langi Filoa ma le faletua. A 'ia manuia lo latou fa'atasianga" (Good fortune attend the presence of
Some Samoans do not cafe for kava. They either call out their thanks and refusal when their names are called, or accept the bowl and, after raising it with a salutation, return it untouched to the cupbearer. The bearer on returning pours it back into the bowl and gets a refill for the next call. Others retain the kava in the mouth, hand back the cup and after gargling the mouth reject the fluid between the wall posts to the platform outside. This was often done, and is no breach of manners either in sound or procedure. The cup generally contains a little sediment at the bottom. After drinking, the dregs are tossed out to the back (sasa'a). The cup is always handed back to the bearer and not tossed back as in
Seating. The position of guests and hosts within the guest house during the formal drinking of kava is the same as at a feast. The kava is always introductory to other activities. The position of the person preparing the kava and the assistants has been described. The remaining special position is that of the person who presides over the distributing of the kava. On ordinary occasions he usually sits on the right, next to the attendant with the water. When specially appointed, he may be on the left near the bowl.
Distribution (fa'asoasoa). On ordinary occasions, one of the lesser talking chiefs of the family group, sits down beside the bowl as a matter of course. Where guests or others are present, the talking chief on the front side (luma) of the house may call out to a particular person, "Alu fa'asoasoa le alofi" (Go and distribute the kava). Alofi is the honorific title of the prepared kava. On ceremonial occasions a special appointment is made from the higher talking chiefs. The appointment is really one of the perquisites of the talking chief's office. If appointed by the village high chief, the village chief usually rewards him with a piece of barkcloth. The appointment, however, is usually delegated to the visiting high chief. The village talking chief asks who is to distribute the
The distributor, besides calling the names of those present in correct order, has a certain formula to call as the making of the kava nears completion. After taking up his position and watching the steps of the preparation, he calls in a loud recitative tone the following announcement or some variant of it.
As it is nearly strained at this point, he commences to clap his hands together (tapati) and all those present in the house follow suit.
The first cup is then called and so on in ceremonial order of precedence. The most important cups are the first and the last. The distributor watches the bowl to see how far it will go. If there is not enough kava to go round a large assembly, he leaves some of the lesser chiefs out. He takes care to call the last cup before the bowl is empty. He is then sure of getting a drink himself if he has not called his own name. If there is plenty of kava to go round, the remains after the last cup called may be drunk without announcement by the lesser chiefs seated near the bowl. The call with the last cup is an announcement that the bowl is empty. There are many variants of the following call.
The last official cup is then taken to the chief named. The visiting chief then gives the distributor a present of a piece of bark cloth. Money may now take its place; half a dollar being the equivalent of a piece of bark cloth. The coin is tossed over with the words, "Here is your siapo (bark cloth)." The distributor touches his head with the coin or the cloth as a mark of respect. He then goes out on the house platform and calls the names of the visiting chiefs, finishing up with fa'afetai (thanks): "Laioifi e ! Falesau e! Falemavaenga e! Ie'ei! Fa'afetai!"
Order of serving. The order of serving the cup is definite. The senior visiting chief gets the first cup and the local chief the next. Then it alternates between the two parties according to the order of precedence on each side. Next to the first two chiefs are the senior talking chiefs of the visitors and then the senior talking chief of the village. On our expedition some of the local high chiefs to pay us a special compliment had our three cups served first before allowing their own names to be called. One talking chief at
A lesser talking chief who would come well down the list may get a turn before it is due by calling the name of a high chief or visitor, saying, "O le ipu o Laloifi o lea 'ou alu ai" (The cup of Laloifi is that which I follow). He is served with the next cup without the distributor calling his name. If the distributor has a sense of humor, he sees that the cup is filled to the brim. If the recipient cannot empty it, he feels rather ashamed as humorous remarks are sure to be made about him. In
Prompting. The distributor knows the order of precedence of the local chiefs but he may not be sure of that of the visitors. Many chiefs have kava cup titles which must be called officially. To avoid error, recourse to prompting (taulalo) is necessary. The visiting talking chief sees to it that the chiefs of his own party are fittingly announced by telling off one of the lesser chiefs to act as prompter to the distributor. He calls out, "Sau se isi e taulalo i le alofi" (Come some one to prompt the kava). On our malanga, the young chief, Samonga of taulalo i le alofi. He told the distributors our newly-conferred cup titles to enable them to make the correct calls.
In olden times in ipu was confined to the taumafa (food) and the call was "Aumaia le taumafa a" (Bring the food of). In the other islands of Samoa this monopoly of the term ipu did not exist. Prominent chief's, however, had a special title that was called at kava drinking in place of their personal names. Where such titles had been conferred or inherited, to call the personal name or chief's title was a serious breach
Some titles include a preliminary phrase before the actual call for the cup. The phrases sound well and have interesting meanings as shown by the following examples.
Tala i malo tu. Speak of unshaken authority.
Fa'avae mai le vavau. The foundation, firm from ancient days.
Toto'a mai le vavau. The threshold from olden times.
The first preliminary phrase, "Tala i malo tu, aumaia le ipu a Laloifi," was used in the cup title given to
Another variant consists in using two phrases and dropping the proper name. At Aoloau, such a title was given to 'ava uso ceremony.
Various cup titles run in certain families which have the right to confer them on others. The cup titles given to the
At Nua,
When we were seated in the guest house yells were heard from the bush at the back. A man came out of the bush with a stick which he flourished as he capered about with exaggerated movements. He worked around to the front of the guest house calling in a wailing cadence, "Mua o—." A group of young men carried in a freshly dug-up kava plant which was cast on the front of the house platform. People bedecked with leafy garlands also filed in from the sides answering the soloist's calls with a similar call in unison. They also worked round to the front where they sat down cross-legged under
the trees in lines facing the house. A kava bowl was set down in front of the middle of the front row. A speech of welcome was made from the open space and a tungasesent up. Our senior Samoan talking chief took charge of it. Kava was prepared and more speeches made.One of our own party entered the house and sat down cross-legged in front of us with his back towards us. Another of our attendants stood just outside the house. Both were young chiefs. The distributor of the kava sat outside near the bowl. The kava cup title of Laloifi was then called. The cupbearer brought the cup and handed it to the attendant outside. He entered the house bearing the cup and marched around to face the second seated attendant. The latter held a new cup made of the half of an ordinary drinking nut. It had to be new also in the sense of never having been used for the drinking of kava. Holding the cup of kava high, the first attendant poured the contents from this height into the cup held low by the seated attendant. It mattered not that some was spilt as long as a certain amount was caught in the lower vessel. The seated attendant, without turning, passed the cup behind his back to Laloifi. Laloifi took the cup with crooked forefinger, poured out a libation, raised it with the customary salutation, and drank. The seated attendant changed his position to the front of Falesau with his back towards him, while the first attendant went outside and handed his cup back to the cupbearer. Falesau's cup title was called and the same procedure gone through. A third time it was repeated for the cup title of Faleomavaenga. The people outside, led by the distributor Noa, clapped hands in unison, first high up and then low down in a number of alternating movements. So far as we were concerned, our cup titles had been officially confirmed by ancient ceremonial. Our two official Samoan talking chiefs were served in the ordinary way. The high chief Tuitele who accompanied us should have had a fresh bowl of kava made for him to accord with his rank, but he dispensed with that part of his privileges to shorten the ceremonial.
The various phases of the kava drinking ceremony have been described. It is necessary, however, to describe the order of events to get an idea of its place in the social usages of the people.
Nothing of any importance can be commenced in Samoa without a preliminary bowl of kava. Visitors of any note must be welcomed not only with a bowl of kava but also with some pieces of the dried root. Thus when visitors enter a village, they are allowed a little time in which to rest and compose themselves after their journey. The local chiefs usually gather in a nearby house until they see that the visitors are ready. Each chief of any standing brings with him some dried kava, generally a tungase, or failing that, an ordinary piece without stem (fasi 'ava). Apart from the presentation aspect of the question, he owes it to his own social position to bring something. Otherwise he does not exist in the fa'alupenga (chiefly list of the village). By the time they enter the guest house, the visitors have assumed their correct positions besides the wall posts that mark their rank. The local chiefs may shake hands as they enter. They may drop their pieces of kava root in front of the visiting talking chief. There is no doubt as to who or where he is. He is sitting by the middle wall post of the front side of the middle section of the house. On the other hand they may take their kava root with them and pass directly to their own positions in the house. In strict ceremonial the fa'alupenga of the village and district to which the visitors belong. To mark respect for the highest rank, afio is used instead of susu, and afionga for susunga. Each chief repeats it, often in a singsong voice. The visitors then reply in the form demanded by the local fa'alupenga. The senior village talking chief has the kava root contributions passed along to him. He collects them in front of him and then makes a short speech enumerating the number of tungase and other pieces that the village has contributed as a token of respect to the visitors. He calls an attendant who brings the kava to the visiting talking chief. The visiting talking chief acknowledges the respect paid them. He then holds up the pieces of kava root and calls the attention of the chief of his party to them by saying, "Silasila mai" (Look). He names his chiefs in order of precedence and also enumerates the number of pieces of kava root. He then proceeds to pule (rule) over the distribution of the pieces of kava, allocating certain ones to certain chiefs and taking out his own share. Some of the smaller pieces he may throw on the mat before him, calling to one of his own attendants to take them away to beat for a bowl of kava. In this he may have been anticipated by the local talking chief. It matters not, it serves for the next bowl.
While the kava is being pounded outside, the taupou takes up her position behind the bowl and the village talking chief commences his official speech of welcome on behalf of the village. Meanwhile the pounded kava is brought in and the preparation proceeds. The visiting talking chief replies. In both speeches the orthodox fa'alupenga is again gone through as the correct commencement. The reply generally finishes before the kava is ready. If not, the village maid prolongs the straining. The distributor has taken his place on ordinary occasions, or is now appointed if the meeting is a special one. He gives his, first call irrespective of who is speaking and this temporarily ends the speech. The kava is then served. The serving over, the village maid usually retires and the speeches are continued. After the speeches, the food is brought in.
On ordinary occasions though few may be present, someone always does the calling, the hands are clapped and a cup bearer serves the kava. In small family gatherings I have heard the chief's wife calling the kava. Walking through a village, a stranger is often accosted by a local chief calling from his house or coming out into the road to invite him in to rest and partake of the 'ava'ona. Kava is the universal medium of hospitality. It corresponds to the tea, coffee, or alcohol of higher cultures as a means to ordinary social intercourse and it forms the introduction to all ceremonies and matters of great pitch and moment.
Forms of kava. A number of names are given to kava for special purposes and various usages. The following list is not exhaustive.
The 'ava oso is the kava root taken by people on a tour or journey to give to the chiefs of the villages they visit. During the
Early morning kava (pongipongi) is usually prepared for people on a journey as they leave very early to avoid the heat of the day. On our malanga (tour), the white chief of the expedition preferred early morning coffee, somewhat to the disappointment of our senior Samoan talking chief, who was an inveterate kava drinker. At one village, the latter was heard asking the local talking chief why no 'ava pongipongi had been prepared. On being told that it was through fear of waking the visitors at the early hour of 4:00 a.m., our thirsty talking chief severely admonished the local orator for departing from ancient Samoan custom. The 'ava pongipongi was the only refreshment possible in the early morning. Visitors left after it and relied on getting breakfast at the end of the journey.
The 'ava mata is green kava freshly dug up as a present of the highest respect to visitors. If an important malanga does not receive green kava, it considers that it has been treated with lack of respect. The kava with stems and leaves attached and earth still adhering to the roots is carried in by bearers uttering yells (ailao) as they emerge from the trees at the back. They march round the guest house and deposit the burden on the house platform in front of the middle section of the house with the call of "'O le lau 'ava" (The leaf of the kava). The kava is given by high chiefs. The clumps are afterwards trimmed, cut up and taken away by the visitors to be dried at home. The 'ava mata was given to our malanga at
The 'ava uso consists of a number of long, thin roots which are bundled together. It is presented to a visiting relative who has not been seen for a long time. The term uso means brother and 'ava uso signifies the uniting together again of long separated relatives. The custom was seen at
At Vailoa, 'ava uso was given to our malanga by the chief Satele and his people. On the malanga our leader, 'ava uso instead of the 'ava mata. As we sat in the Vailoan guest house awaiting our hosts, our two Samoan talking chiefs saw the local chiefs in a house near by tying the long thin roots of 'ava uso together. They therefore anticipated things by changing my 'ava cup name to Fetaia'i-ma-uso (the meeting together of brothers). The 'ava uso was duly presented with the other kava root. During the serving of the kava shortly after, the calling of the cup name surprised and pleased our hosts. In giving the 'ava uso, the singing of a special song referring to it usually forms part of the ceremony.
Kava prepared for warriors before going to war is 'ava mua au. Fe'epulea'i 'ava mua an.
Special ceremonies with varying details of procedure were built up around the installation of high chiefs and different districts seemed to have vied in elaborating a form of ceremony peculiar to their particular title. This was particularly so in connection with the high titles that have been designated as tupu (kings). Thus, in the high ceremonial connected with the
Kava was the only beverage the pre-European Samoans had besides water and coconut water. It is cooling, refreshing, and stimulating without being intoxicating. Its medicinal properties are recognized by the medical profession and some of the well-known drug firms make up preparations of kava used in moderation, it is probably the best drink for a tropical climate. Continued use creates a habit, and drunk to excess it may affect locomotion but does not cause intoxication like alcohol. Continued use of large quantities results in a scaly skin affection and chronic conjunctivitis (mafuna), to be sometimes seen in old talking chiefs, who by reason of their office drink more kava than others. (See
To the Samoan, kava was far more than an ordinary beverage for tungase dried root also formed the official medium of paying respect to visitors and titles. Some of the large tungase were never meant to be pounded but passed from one recipient to another as each in turn paid his respects at subsequent ceremonies. Kava thus ranks as perhaps the most important material element in ceremonial and has become an integral part in the social fabric. If the ceremonial calling of kava were dropped, the value of titles would depreciate considerably as they would lose the most active factor that keeps them prominently before the eyes of the Samoan people.
Plaiting forms one of the most important crafts of the Samoans. Plaited roof sheets are used in the canoe sheds and cooking houses, plaited ridge sheets in all houses, plaited carrying sheets for bringing the sugarcane leaf for the roofing of the better class of houses, and plaited wall screens for all dwelling and guest houses. The furnishing of houses is still incomplete without the plaited floor and sleeping mats. In connection with food, plaited baskets are a necessity for both transporting and storing cooked and uncooked food, while plaited trays are used in serving the meals. In providing clothing, apart from simple kilts and bark cloth, the technique availed of was entirely plaiting.
To appreciate the part played by plaiting in material culture, it must not be confounded with weaving. Plaiting is an older and simpler craft than weaving. As it dealt primarily with wider and stiffer strips of material than those usually used in weaving, the fingers were quite able to deal with the problems of technique encountered without seeking mechanical assistance. Plaiting in a technical sense means the interlacing of two sets of elements to form a continuous surface. The fabric or object made is usually of some extent. It may be kept flat as a sheet or mat, slightly concave as a tray, or the directions of plaiting may be so arranged as to form a receptacle or basket. The term plaiting is also applied to braiding. In braiding only a few elements are used and during the process of interlacing they are turned in successively at the sides to restrict the width and produce length.
Weaving resembles plaiting in using' two sets of interlacing and
Weaving is a technique developed in dealing with soft, pliable elements. The difficulties involved in fine plaiting were obviated by dealing first of all with one set of the crossing elements. This set of elements (warps) was fixed to some support or supports at one or both ends. The other set of crossing elements (wefts) was attended to as a single element at a time. Hence at the commencement edge only one set of elements was fixed together vertically or longitudinally to the worker. These were diverged into two sets, according to the stroke used, and one weft element carried across between them at right angles to the warp and transversely to the worker. Apart from he use of mechanical contrivance and the nature of the material, the fundamental difference between plaiting and weaving begins at the very commencement edge of the article in process. The above recapitulation is necessary to a clear understanding of the technique of Samoan textiles.
The terminology used here is the same as that in the work on the
Check: each weft passes alternately over and under each consecutive crossing weft.
Decoration: the addition of elements for the purpose of ornament; a, b, structural, where the elements form an integral part of what they adorn.
Foundation wefts: a structural weft that is essential to the construction.
Overlaid plaiting: wefts not essential to the construction that are laid on the foundation wefts and are plaited with them to form applied decoration.
Shed: the space formed by the separation of the two series of alternate dextral wefts for the reception of a sinistral weft, or two series of sinistrals to receive a dextral.
Stroke: the passing of a weft over or under one or more crossing wefts to produce the pattern and technique of the plait as in check or twill.
Twill: each weft passes over and under more than one crossing weft. In twilled twos, it passes over and under two crossing wefts; in twilled threes, over and under three crossing wefts. Combinations may be used with a check, as three and ones, or with another twill, as threes and twos.
Weft: a technical factor in plaiting. A weft may consist of a single strip of material or more than one strip as in overlaid plaiting or in narrowing where two wefts are brought together.
Double weft: the term "double weft" is used where two separate weft strips are placed together for a special purpose in technique but they are treated as a single factor in the plaiting strokes and movements.
Dextral and sinistral wefts: the confusion between using the terms warp and weft which belong to weaving is obviated by terming the strips which lean towards the right, dextral wefts, and those towards the left, sinistral wefts. When a weft is turned in the opposite direction, as in defining the edge of the fabric, it is renamed from the direction in which it functions. If distinction is required, it may be termed an acting dextral or sinistral as it may be.
In actual plaiting, single strokes in check or twill are not made except in special circumstances. To facilitate the work, as many dextral wefts as can be conveniently handled are treated by separating them into two alternating series to form a shed into which the appropriate sinistral is placed. (See fig. 75.)
Movement: a technical movement in plaiting consists in placing a sinistral weft in the shed provided by a previous movement, lowering and raising the two series of working dextrals to complete the plaiting strokes and, as a result, prepare the shed for the next sinistral.
As used here, the definition differs in the order of the two parts of the completed movement to that enunciated in my work on the
Working dextrals: the dextral wefts raised and lowered at the working edge to provide a shed for the sinistral weft.
Working edge: the oblique edge formed by a sinistral weft and the working dextrals. The working edge varies in length according to the number it working dextrals being used.
The plaiting of a deep surface is therefore carried out in a series of sections. The width of the section is the full width of the article being made, as
The distinction between plaiting and weaving is now more obvious. The dextral wefts in plaiting correspond in function to the warps in weaving for they are the elements that are separated into two series to provide a shed for the sinistral weft in plaiting and the weft in weaving. In weaving, all the warps along the commencing elge would be separated to form a shed extending for the whole width of the fabric, whereas in figure 76 it is only the relatively small number of working dextrals at the short oblique working edge that do this in plaiting. This goes back to the fundamental arrangement of the elements at the commencing edge.
The material used in plaiting house accessories, food platters, and baskets, and the commonest form of floor mats, is the leaf of the coconut palm. For other baskets and mats, recourse is had to the pandanus.
The coconut leaf. The coconut leaf is termed lau niu (lau, leaf, and niu coconut) while the leaf midrib is lapalapa. The upper surface of the leaf is the alo (front) and the lower surface the tua (back). The leaf midrib is four sided for about half its length. The lower surface is wider than the upper. All surfaces diminish in width towards the tip end. At about halfway, the converging edges of the upper surface run together and continue on as a raised ridge, while the two lateral surfaces and the under surface persist, thus making the tip end half of the midrib triangular in section. The leaflets are symmetrically arranged on either side and spring from the lateral surfaces of the midrib. They run obliquely outwards and towards the tip end of the leaf. Each leaflet has a strong midrib which is attached to the lateral surface of the midrib just below its upper edge. At its attachment it runs a little backward towards the butt end and then merges in the leaf midrib. The leaflet midrib is termed tuaniu and is used for various purposes, such as needles and for sewing roof sheets of sugarcane leaf. The leaflet material on either side of the leaflet midrib come together at the leaf midrib and are attached to its lateral surfaces in a vertical line extending downwards from the leaflet midrib. In certain forms of sheets and baskets, the leaflet midrib is pressed down with the thumb to open out the leaflet to its full width. In technique, this will be referred to as the "open leaflet," In other forms of sheets and baskets the leaflets as wefts are not opened out but kept doubled throughout with their midribs forming an edge of the weft. For distinction, this will be called the "closed leaflet." The young leaves in the middle of the moemoe (sleeping). The young, soft leaflets become very white when dried and are used in making the white ilitea fans.
The formation of the leaf influences technique. The leaf midrib readily splits longitudinally and excess material is split off from the lateral surface strip bearing the leaflets. The midrib strip forms a natural commencing edge with the wefts already fixed in position. With the alo surface upwards and the midrib strip towards the worker, the strip from the left side of the leaf (itu lauanga vale) has the butt end towards the left and the leaflets incline towards the right, thus forming natural dextral wefts. With the strip from the right side of the leaf (itu taumatau) in the same position, the butt end is towards the right and the leaflets incline towards the left, thus forming natural sinistral wefts.
The plaited materials used in connection with actual house construction are thatch, carrying sheet, ridge, and screen sheets. They are all made of coconut leaf and illustrate various phases of simple plaiting.
Thatch sheet (laupola). The coconut leaf sheet is used for thatching, canoes, cooking, and rough, temporary houses. The leaf is cut off into convenient lengths of 6 or 7 feet and split down the leaf midrib. Excess midrib material is split off. The position of plaiting is with the midrib strip towards the plaiter and the natural upper surface upwards. The open leaflets are used as wefts and the plaiting stroke is the check. Each half leaf is dealt with differently.
Left half leaf. The left half leaf, forming natural dextrals, is plaited from left to right. It is obvious that in a check plait, alternate leaflets must be bent to the left to provide crossing sinistrals and that the leaflets will be dealt with in pairs. It is characteristic, however, of Samoan technique that the first leaflet is left out of the pair groupings. (See fig. 77.)
Pinning the outer edge. A neat method of keeping the leaflets in position at the far edge of the plaiting is shown in figure 78.
This is quickly and easily done, here and there, along the free edge of the plaiting and, though not absolutely essential, helps to make a well-balanced sheet.
Right edge finish. When the plaiting reaches the right end of the leaf Strip, the right edge is formed in the same manner as on the left but it is the dextrals which project beyond the last sinistrals that are turned in. (See fig. 79.)
Right half leaf. In the right half leaf with natural sinistrals, plaiting technique departs from the usual and commencing on the right works towards
The left edge of the sheet is turned in in exactly the same way as the left edge of the left half leaf.
By keeping the same number of working wefts at the working edge, the completed plaited edge of the sheet is level throughout and is left thus without further treatment. The plaiting is enough to rearrange the leaflets by crossing them and keeping them in that position. Before thatching, the sheets are usually left out exposed to the sun for a day or so to partially dry and shrink them. A right and left sheet are placed together and treated as one element in tying to the thatch rafters. (See Pl. XI, A.)
Carrying sheet (laupolapola). The carrying sheet (Plate XII, E) is used as a broad band upon which material such as cane leaves is placed. The ends of the sheet are brought round and tied over the bundle with a strip of bark. The burden (avenga) so formed is carried on the back by women in the fafanga method.
A section of the coconut leaf about 2 feet long, split in two, forms the material. The two half sheets are plaited separately in check with the open leaflet in the same way as the preceding roof sheets with this addition. (See fig. 81.)
Joining the sheets. The two sheets are now joined together, at their plaiting edges with a three-ply braid which is plaited in two courses. The two sheets are therefore placed longitudinally in front of the worker with their plaiting edges close together and the first course of the braid commenced. (See fig. 82.)
The first course of the braid being completed, the ends of the sheets are reversed to place the braid tail at the far end and the second course commenced. (See fig. 83.)
An alternative technique consists of plaiting the leaflets on the two sides of the leaf section without splitting the leaf midrib. After plaiting each side, the two plaiting edges are joined together by the two-course braid and the leaf midrib split last of all to open out the sheet.
In use, the true under surface of the leaflets are turned upwards as this surface forms a natural concavity. The material carried is longer than the carrying sheet but the broad band is quite effective in keeping it together. Bark strip ties are passed between the midrib edges which need not necessarily meet.
Roof ridging sheets.Taualunga is the general term used for roof ridging. It includes the ordinary unplaited coconut leaves which are placed horizontally fa'atafiti which form only a part of the taualunga.
The fa'atafiti ridging sheet is made of two coconut leaves, one of which is split down the midrib into two halves. Plaiting takes place with the true under surface turned upwards. The whole leaf occupies a middle position. The half sheet whose leaflets cross naturally with those on the right side of the whole sheet is placed on the right with the leaf midribs parallel and about 8 inches apart. The crossing leaflets are plaited and then the remaining half sheet is similarly treated on the left. In plaiting, two adjoining closed leaflets
The completed ridging sheet is shown in Plate XI B. The unplaited tip ends of the leaves are not cut off but are left to be covered by the plaited parts of other sheets. The sheet when turned over has the appearance in Plate XI C, and it is this surface that is to the outside when the sheet is fixed on the ridge of the house. The outer split midribs thus keep the projecting leaflets down. The middle midrib lies along the top of the ridge and the split midribs hang down on either side. The wooden pins are passed through the sheet below the upper ridgepole and above the split midribs on either side, which thus prevent the sheet from working up over the pins. The ridging may have to be renewed before the thatch.
The wall screens (pola sisi) used to close the sides of the house show a further advance in plaiting technique. The crossing elements are supplied by two strips of midrib from opposite sides of the leaf; twill strokes are used instead of check and the finishing edge opposite the midrib commencement is formed by a three-ply braid.
The leaf sections are cut slightly longer than the spaces between the wall posts in order that the sides of the screens may overlap the posts when hung. The leaf sections are then left out to dry for a clay or two. Before plaiting, the leaf sections are split down the midrib which is then pared in both halves. The length is checked with a measured rod. The right half leaf is placed above the other so that the lower layer of leaflets consist of naturally directed, dextrals and the upper layer of natural sinistrals. There is thus no bending of alternate leaflets at right angles to their natural course for the crossing elements are provided naturally. The leaflets are kept closed. In the upper layer of sinistrals, the leaflet midrib forms the left margin of the wefts and in the lower dextrals, the right margin. The commencement (fa'a'au) is made by plaiting the crossing elements together with one row of twilled-two strokes (su'i lua) along the length of the midrib strip. The technique is shown in figure 85.
In actual technique, the right hand holds a number of sinistral elements curved away to the right to display the dextral layer below. As the left hand picks up the new number of the pair, the right hand simply lets go the left sinistral of the bunch and it falls into its natural position. When the whole row is plaited, the leaflet ends of the two series are pulled so as to bring the midrib strip closer together and tighten up the plaiting. The whole process of this stage is termed the fa'a'au. As it is a common commencement, it will be referred to as the "twilled-two" commencement.
The side edges. Before the next section of the plaiting can be continued from the commencement plait, the left edge of the sheet must be defined. This is done as in previous sheets by turning in the projecting sinistrals successively at right angles to function as dextrals, but with this difference. In the roof and carrying sheets made with the open leaflets, the leaflet is given a half turn which exposes the other surface. In the closed leaflet wefts, the angular turn at the edge is made without turning the other surface upwards. The leaflet is simply bent in and the leaflet midrib still remains as the left edge of the wefts though the leaflet part is somewhat crumpled. This method of defining the edge is termed afeafe. (See figure 86.)
The question of whether the turned in sinistral passes above or below the sinistral above it, depends on the craftswoman, the method of beginning the plait, or the pattern to be followed. Thus in the screen figured in figure 86, the first dextral (1') was brought up from under one sinistral (1) instead of two. The plaiter subsequently formed the left edge by turning the sinistrals under the ones above it. Before the uppermost sinistrals are turned at the edge, their position in the pattern between the twill of the commencement and the left edge must be attended to.
The right edge of the screen is defined with the completion of each section, by turning in the projecting dextrals successively to function as sinistrals with the afeafe turn.
The body. A decorative effect is produced by varying the plaiting stroke. There are three usual forms: a, horizontal lines of twill are simply a repetition of the technique of the commencement. The plaiting edge is defined by a horizontal line of dextral twilled twos shown by the arrow in figure 86 a. If twilled threes are used, three dextral wefts are held up instead of two. b, Vertical lines of twill are shown in Plate XII A, 1. The usual commencement marked by the horizontal line of dextral
c, Two forms of triangular geometrical patterns (mamanu) are in common use. (See fig. 88.) Smaller triangles (Pl. XII, A, 3) are worked in figure 89.
Finishing braid (fili). The depth secured, the transversely placed plaiting edge is turned longitudinally, and braiding commences at the far end. Both the dextral and sinistral projecting free ends are alternately plaited into a three-ply braid to finish off the far edge of the sheet. (See fig. 90.)
At the near end, when all the wefts have been included in the braid, the plaiting is continued on as a free tail for about 6 inches and the end knotted. The tail is doubled obliquely along the under surface and passed through under two crossing wefts. (See Plate XII, A, 4.)
A general average for wall screens is 3 feet in width and 12 or 13 inches in depth.
The name pola appears in three of the sheets. In general it means a plaited coconut leaflet mat. When the word is used by itself, the pola sisi wall screen is usually meant. The wall screen is also called pola taufafo (to pola taufale (in the house) which, though pola of the house, was explained to me as the coconut leaflet floor mat also called pola vai. Pola taufafo also alludes to strangers or intruders, for when a wall screen breaks away, it falls to the outside of the house. My informants distinguished the roof sheet as laupola from laupolapola (a carrying sheet), also termed polani. When a plaited leaf is used as a canoe cover it is polava'a.
Platters for serving food consist of temporary articles quickly made for one meal and more carefully made platters for continued use. The temporary class contains two forms; the mailo and the laulau. The permanent class consists of one form termed laulau but distinguished from the temporary laulau by technique and use.
The mailo, also termed ma'ilo, consists of two forms; the unplaited, and the plaited. The unplaited mailo is quickly made by splitting off a strip of midrib carrying about 10 leaflets, separating the leaflets in the middle into two equal sets, pulling the leaflet ends together as two plies, and tying them in a reef knot. The under surface is turned upwards and forms a slightly concave triangular platter upon which food is placed. The plaited mailo (P1 XII, C, 1) is made from two leaf strips from opposite sides, each bearing 6 or more leaflets which are plaited together with a check stroke. (See fig. 91.)
The temporary laulau (Pl. XII, C, 2.) is also plaited from two strips from opposite sides, each carrying 6 or more leaflets. It is commenced in a similar way to the plaited mailo but after the wefts have all crossed each other once, the plaiting is continued by bending in the leaflets to form side edges. (See fig. 92.)
Of the two forms of temporary plaited platters, the mailo is the much more important. It has come to assume an importance out of all proportion to its technique owing to custom having made it the correct platter upon which the fa'ausi preparation of grated talo must be served to high chiefs. Strict observance restricts the use of the mailo to those of the highest rank while chiefs of lesser status are served on breadfruit leaves alone. Fa'ausi is never served on the temporary laulau which, though better made and sometimes used mailo. Here a cultural convention takes precedence over improved technique.
The permanent laulau (Pl. XII, D) marks an advance on the technique of the wall screen. As in the wall screen, the closed leaflets are used as wefts, but, whereas the screen wefts average about 0.8 inches in width, the platter wefts are reduced to half that width by splitting off the free edges of the leaflets. The reduction in width, which would make the plaiting too open, is corrected by using two strips instead of one to supply each set of wefts. The method of twisting the leaflets to keep the two midrib strips together to form one set of wefts is shown in figure 93.
The commencement is made by placing two double strips together with that bearing the sinistral wefts above the other as in the wall screen. The fa'a'au technique of running a horizontal row of dextral twilled twos across the sinistral wefts is exactly the same as in the wall screen. The craftswoman may please herself as to whether the first dextral shall pass over one, two, or
The braid finish (fili). The far plaiting edge of the platter is finished off with a three-ply braid which differs in technique from that of the wall screen. Of the two sets of projecting dextrals and sinistrals, one set is braided on each surface of the platter. The set plaited on each surface is that which leans towards the plaiter, and the wefts are added to the middle ply of the braid from the right or plaited side. The back ply of the working pairs is always brought to the middle position by passing under the ply in front instead of over. To avoid confusion, the front ply of the working pair is always doubled back over its own course and the back ply brought into position with a half turn twist. On the right side, the additional weft from the plaiting edge is pulled up into position first before the back ply is twisted to the middle. This makes a further difference to the screen sheet braid in which the ply is twisted before the weft element. Another point of difference consists in dropping one element of the back ply when it enters the braid from the right. Braiding
The permanent laulau is important in domestic economy. It is still in common, everyday use. All meals are served on plaited laulau. As they are effective and easy to make from material that is always available at no expense, they have rendered unnecessary the use of wooden vessels in the past and restricted the adoption of foreign crockery in the present.
The baskets made of coconut leaves that are in common use may be divided, into two main classes: the 'ato, and the 'ola. In addition there are some smaller types and a large round basket the origin of which is uncertain. A satchel is also made from pandanus leaf and a rare type of basket from coconut fibre. Round baskets made by coiled weaving are extensively made but the technique has been introduced.
The 'ato baskets are those in common everyday use for transporting and holding cooked and uncooked food. They are made from a single strip of coconut leaf midrib carrying leaflets running in one direction. The crossing elements have therefore to be provided by bending the alternate leaflets in the opposite direction. The open leaflet and the check technique is used. The midrib strip forms the rim of the basket and the bottom is closed by a three-ply braid. The manner in which the three-ply braid is plaited results in two varieties of 'ato named according to the braid technique; 'ato fill tasi, or 'ato fili tolu. (See Plate XII, B, 1 and 2.)
1. In the 'ato fill tasi a midrib strip twice the length of the required basket is torn off the side of a leaf, not too near the tip or the base in order to obtain even leaflets. If only one basket is to be made, the plaiter selects the left side of the leaf, as it will have natural dextrals and the plaiting will proceed from left to right. With a right strip, the natural sinistrals are plaited from right to left as in the roof sheets.
On a left sheet, the bending of sinistrals and the check plait are exactly the same as in the roof sheet; the only difference is that the leaf midrib is pared down more so that it is pliable and may readily be bent around to form the circumference of the rim. The technique of leaving the first leaflet on the left out of the first pair is also observed here. Thus, the third leaflet forms the first sinistral to be turned back and from there on the right element of each succeeding pair is turned back as a sinistral over the first element of the pair which remains in its natural position as a dextral.
When the required depth, which is greater than in the roof sheet, is obtained, the dextrals are successively bent downwards at right angles to rest on the sinistral which has been laid in the shed formed at the working edge. Each dextral is then crossed by the top working dextral as in the technique of the carrying sheet. (See fig. 81.) The reason in each case is to clear the way for the braid join. On reaching the right end of the strip the odd leaflet is left after the last pair and the midrib strip is cut off beyond it.
The two ends of the midrib strip are brought together to form the so'o
Closing the bottom. The edge of the closed plaiting formed by the midrib strip margin remains unchanged as the upper rim of the basket so the other edge at which the plaiting ceased must be closed to form the bottom of the basket. From the plaiting edge, the free ends of the sinistrals project upwards and the ends of the dextrals are tucked down out of the way. The closing commences usually above the so'o join of the rim so the basket is rested on its rim and turned until the so'o join marked by the crossing of the leaflet round the rim, is away from the plaiter. (See Plate XIII, C.) The basket takes its name from fili (to braid) and tasi (one). The free ends of the wefts are plaited into one line of three-ply braid which, however, is done in two courses in the same way as the join of the carrying sheet. The first so'o join, works towards the plaiter. See fig. 98.)
The braid having reached the near end, the plies are plaited on for an extra turn to project as a free tail. The ends of the basket are reversed, the
Though exact wefts have been figured at the doubling over of the braid to commence the second course, the plaiter uses her judgment as to what wefts fit in with the direction of the plies. Mathematical accuracy is not attempted in practice and is not necessary.
In a basket plaited from a right hand leaf strip, the sinistrals are turned down and the dextrals project free above the plaiting edge. Though the directions are opposite to those described, it need not confuse the closing of the bottom, if the principles described above are observed. Hence, after
2. The 'ato fili tolu, differs merely in the method of closing the bottom. Of the two sets of free weft ends at the plaiting edge, one set is used to form a marginal three-ply braid round the plaiting edge. This forms a circular opening which, however, owing to the tightening up of the plies, is very much smaller than the rim opening. When the first course of marginal braiding reaches the commencement point, the other set of free wefts are added to the three ply braid from either side in such a way as to bring the braid down the mesial line. Owing to the pulling in of the wefts of the second fili tolu (three braids).
The type described (fig. 100) was made from a right hand leaf, not from preference, but because the left hand leaf strip had been used to make the previous type. It is described here in preference to the normal left hand strip to demonstrate the adaptability of the technique. The first course commenced above the so'o join at the rim.
The so'o join. Reference back to figure 100 f shows that the turned down weft 3 was the first used in the braid. The dotted course shows the position it previously occupied. Above it are the turned down wefts (1') and (2') which are turned down on the course of the two free wefts (1) and (2) which form the first two plies of the braid. Also the free weft (4) is the first of the series of free wefts that were pulled out over the braid. (See figure 101.)
The 'ato baskets are made of green leaves and discarded when dry, as they become brittle and weak. The green basket is termed 'ato mata (mata, green) and the old dry basket, 'ato 'ato. The single braid basket is termed 'ato fili tasi from its technique and 'ato fu'e umu from its important use as a receptacle for the cooked food after the oven is opened (fu'e umu). The elliptical bottomed basket is 'ato fili tolu from its technique and 'ato 'ave 'avenga from its principle use in carrying burdens ('avenga) from the cultivations. It is also termed 'ato toli 'ulu from its particular use in carrying breadfruit (toli, to pluck, and 'ulu, breadfruit). The technique of the bottom is said to have been introduced into Samoa from 'ato fa'aniue (made after the 'ato baskets are also termed 'ete [Cf. kete] but 'ato is in more common use.
The 'ato types of baskets have fairly wide interstices between the wefts which become wider as the material dries. For fishing purposes, closer wefts and a more permanent use required a different technique. Such baskets (ola) were in ola malu, ola tu, and si'u ola.
1. The ola malu of fau bast is run through between the wefts close under the rim of one side and the ends are tied round the waist. Both hands are thus freed. Three varieties range from 8 to 12 inches wide at the rim to 10 to 12 inches in depth. Leaflet-bearing midrib strips are used and the plaiting stroke is the check. The midrib strip ranging from 16 to 24 inches in length is split off as thin as possible. The A, 1), the two wefts thus provided by each leaflet are 0.4 inches wide. In the larger, coarser basket (Plate XIV A, 2.) the wefts are 0.6 inches wide. By using two midrib strips a closer plait and a narrower weft is made possible. The baskets figured have each a different method of commencement at the midrib strip.
The larger ola malu from the south coast of A, 2.)
In the next variety a closer plait is obtained by using two midrib strips and a narrower weft is thus possible.
In the basket (Pl. XIV, A, 1) from the north coast of
A third ola malu has also two midrib strips with narrower wefts. It commences in the same way as the form shown in Plate XIV, A, 1, but half way along the strip a change in arrangement takes place as shown in figure 104.
The depth of the basket body is in marked contrast to the width. It forms a deep bag more suitable for holding objects when the carrier is continually stooping down and groping amongst the stones. Depth is secured by adding 'ato type, owing to the wide wefts, the depth of the basket was secured by the first working section.
The bottom is closed by a double braid as in the 'ato, but, instead of doubling over the end of the first braid course, it is continued on in a free tail.
The ends of the baskets are reversed and the second set of free wefts commenced as a fresh braid. The technique after the three initial plies have been established is exactly the same as in the second course of the 'ato braid. At the near end, the braid is continued as a free tail.
These baskets have thus a free tail at each end of the bottom. The braided tails are made extra long. In the smaller form, they are longer than half the depth of the basket. They are turned up along the outside of the ends for a few inches, pushed through to the inside, then brought outside again and the knotted ends are left on the outside. In the larger basket the tails are not so long and the knotted ends, after a short course outside, are pushed through to the inside. (See Plate XIV, A.)
2. The ola tu are large leaflet baskets made with two midrib strips from opposite sides of the leaf to supply natural crossing wefts. The plaiting stroke is a twilled-two changing towards the bottom to lines of check. The sinistral bearing strip is placed above the other and the closed leaflets are used. The free edges of the leaflets are trimmd, but the leaflet midribs are left intact. The baskets are thus stronger and stiffer than the malu type which, owing to the removal of the leaflet midribs, are soft and pliable.
The fa'a'au commencement by horizontal twilled twos is exactly the same as in the wall screens. The first dextral on the left is, however, lifted after the first sinistral. This results in marked differences of appearance in the plait along the rim. Where the two midrib strips have worked a little distance apart, the transverse row of sinistrals is deep and marked. Where the midrib strips are close together only the attachments of the sinistrals to the outer strip can be seen and the dextral horizontal row appears prominent near the rim. The appearance of a change of technique is given but it is not so. (See two ends of the rim in Plate XIV, B.)
As in the ola malu no special weft join is used when the ends of the strip are brought together. Any long ends that may be left on the midrib strip are turned up with the wefts and plaited in.
The horizontal twilled twos are continued but here and there twilled threes have been introduced to level the line of plaiting which has been affected by a curve in the rim. Narrow strips are also torn from the free edges of the wefts to narrow the wefts and thus narrow the basket towards the bottom. Thus wefts which are 0.8 inches wide are 0.6 inches towards the bottom. Within 6 inches of the bottom, the twilled twos give place to check.
The bottom is closed with the two-course, three-ply braid each ending in a free tail as in the ola malu. The knotted tails are pushed through to the inside close to their origin at the bottom and after a course inside, the knotted ends are pushed through to the outside.
The ola tu is used by women to catch fish which are driven out of the heaps of stones in the lagoon that have been erected to attract them as resting places. These baskets have been termed fanga i'a (fish trap) which, however, is a functional and not a generic name. They must not be confounded with the true fanga i'a trap made of vines with a single-pair twine.
3. The si'u ola forms a receptacle for the catch of fish either on the canoe or after it is brought in to shore. Like the ola tu, it is a twilled basket made with the closed leaflet wefts and finished at the bottom with a two-course braid each braid with its own tail and knot. The crossing wefts are also supplied by two midrib strips from opposite sides of the leaf. The basket, however is much smaller and the strip receives special treatment.
The strip from the left side of the leaf has each leaflet twisted forward in its natural direction over the leaflet in front of it. (See figure 105.)
In the si'u ola figured (Plate XIV, C) the slight narrowing at the bottom is due to the natural narrowing of the wefts and not to special paring of the wefts as in the ola tu.
After establishing the horizontal twilled-two commencement, the twilled twos are worked in vertical rows on the body and end in check at the plaiting edge. The bottom closing technique is identical with the ola tu but the two knotted tails were pushed through to the inside and left there. One tail in the ola has been pushed through so close to its origin that it is not obvious from the outside and at first gives the impression of the doubled-over technique of the 'ato ending in one tail.
In ola is made with the outer sinistral leaflets twisted on the strip whilst the inner dextral leaflets were added untwisted. ola malu but differ considerably from the ola malu of
The best twilled basket seen was given to C, 3). Both the dextrals and sinistrals were supplied by two midrib strips treated by the double twisting method described with the laulau platters. (See fig. 93.) Closed leaflet wefts were used and after the orthodox horizontal twilled twos of the fa'a'au commencement, sinistral threes and dextral twos were used and then vertical rows of twill ending in horizontal rows of twill.
The closing of the bottom presents a new ornamental motive. The first course is braided in the usual way, finished off with a free tail, and knotted. In the second or outer course, plaited with the remaining free weft ends on either side, a five-weft plaited strip is made instead of the usual three-ply braid. The wefts, which started off with a width of 0.4 inches, by the time they reach the full depth of the basket are 0.3 inches in width, and it is thus possible to make a neat braid with the narrower material for the turns and twists. The five-ply braid is shown in detail in figure 106.
When the braid reaches the near end of the basket, the weft elements left are continued on into a three-ply braid tail and knotted. Both tails are carried up on the outer side of the basket ends, pushed through to the inside and then the knots pushed through to the outside.
The basket described (Plate XIV, C, 3), though it has the technique of a superior ola, was used by chiefs to hold sennit braid that had not been hanked. 'ato ofa (basket for sennit). The name is thus merely a functional name describing" the use to which it was put by Tuitele. It could be used for other purposes but was too good to use for fish. Tuitele valued it highly and said the type was only made for chiefs.
The ola baskets used for fishing were also called ola fangota from their use. When used for carrying coconut shell water bottles, they were called ola 'ave vai. A basket used for keeping oil bottles was also called tulula.
Round baskets (Plate XV) made of closed leaflets with twilled strokes are made throughout Samoa but doubt exists as to whether the form has been introduced or whether it forms part of their own culture. Except for the idea of forming the round bottom there is nothing in the technique that marks any gap between it and the processes already described. As, however, the elliptical bottom of the rough coconut leaf basket has been attributed by the Samoans themselves to influence from
The size of the baskets ranges from the neat size figured in Plate XV to very large ones used for holding clothes. Their use as clothes baskets has led to their receiving the name of 'ato lavalava (lavalava, kilt or skirt).
The rim is formed of two pairs of midrib strip, each pair supplying the opposing wefts after the leaflets have been twisted as in the laulau platters and the chief's basket. The commencement is the usual fa'a'au horizontal lines of twilled twos with the sinistral bearing strip to the outer side. Instead of horizontal twilled twos, the commencement may be made by lifting up three dextrals from behind three sinistrals and then dropping the dextrals alternately over the sinistrals. From this vertical lines of twilled threes are readily developed. The closure of the ends of the midrib strip at the rim is formed by the so'o join used in the si'u ola baskets. (See figure 105.) When the depth is reached, the plaiting edge is finished off evenly with a horizontal line of dextral twilled twos or checks.
The bottom rim. At this stage there is no deviation from the technique of the better class of ola baskets. If the sides were flattened in, and a two-course braid used to close the bottom, an ola would result. To form a round bottom, however, all the wefts must be turned in at right angles to the level plaiting edge. This is done by twisting each sinistral weft with a turn towards the plaiter so that it stands on edge with its midrib edge uppermost. It is then passed to the left round the next sinistral from the near side outwards where it receives another slight twist that againt flattens it with its midrib edge to the left. (See fig. 107.)
The bottom is continued in the horizontal plane round the inner circumference of the bottom rim. The contraction occurs naturally owing to the narrowing of the wefts which have changed from 0.4 inches on the sides to 0.3 inches and less on the bottom. Here and there if necessary two wefts may be run together to insure narrowing but little of this is necessary. In a basket 15 inches across the bottom, the plaiting was carried on in the usual twills for about 5.5 inches from the bottom rim circumference, leaving a circular gap of about 4 inches in diameter to fill in.
Closing the bottom. The free projecting wefts are then divided into four equal divisions. Each of these is plaited into a free three-ply braid tail which is knotted. Two braids from opposite sides are then drawn across the opening on the inside of the basket and parallel with each other. Each braid is passed through under the wefts that form the near ply at the base or beginning of the opposite braid. The wefts are the long outer wefts of the ply and easily admit a braid being passed up under them. The ends of the braids are then carried along the inside of the bottom and the knotted ends passed through under a couple of wefts towards the bottom angle, the knots remaining on the inside of the basket. The basket is turned over and the other two braids crossed side by side over the two fixed braids. Their ends are pushed down through the openings left. The ends are pulled up on the inside through appropriate wefts near the margin of the hole and then carried out to have the knots fixed under a couple of wefts near the bottom margin. (See Plate XV, A, 1.) In this manner the bottom is effectively closed and there is no danger of the braids giving way. The method of adapting a three-ply braid to close the bottom is quite in keeping with Samoan technique though some think that the round basket may have been introduced.
Variations in closing. Two variations from the above common method were noted: (a) the four braids instead of commencing at the margins of the central opening were commenced near the bottom rim and the wefts from the plaiting simply added to the braid on either side as the plaiting proceeded b) a large basket made in B.)
The breadfruit cover (pulou 'ulu) was demonstrated by an old man at Fitiuta, who maintained that it was an old technique which he alone in the district remembered. The name (pulou, head cover and 'ulu, breadfruit) denotes its use in covering any special breadfruit to protect it from the flying-fox. pulou 'ulu referred to something else. The cover (Plate XIV, D, 1) is made of two sections of coconut
The plaiting edge forms the bottom. Here we have the dextrals all turned down and the sinistrals projecting upwards, the position similar to that in the 'ato baskets. Making the leaf midribs form the sides, the plaiting edge is laterally compressed and the bottom filled in by the doubled-over, two-course braid ending in one free tail, as in the one-braid 'ato basket.
The cover is readily slipped over the breadfruit selected while still growing on the tree and is left there as a protection.
In D, 2, 3) were used by children for carrying fruit. They were made of three and four midrib sections, each carrying three leaflets on either side. The plait and closure of the bottom was identical with the Fitiutuan cover. They were held to be a recent innovation, which they probably were, as there seems no particular use for small parcels in Samoan life, while larger ones are met by the use of the 'ato basket. The pulou 'ulu, however, seems to have had a specific use supported by myth. Larger baskets were made in the
The basket figured in Plate XIV, D, 4 resembles the ohini, said to be derived from the oini. The technique is somewhat different to the usual type, but again local authorities maintain that they are new to Samoa.
Baskets are now plaited of pandanus leaf with the technique derived from floor mats. The check stroke is the usual one. The baskets are small and the commencing edge which is to form the bottom must have the weft ends projecting towards the plaiter sufficiently long to form the other side of the basket. The commencing edge is plaited in check for the width of the basket and continued in depth until the plaiting runs out in a triangle. (See fig. 109.)
When the depth of the basket is secured, the edge to form the rim may be turned down with a plain straight edge or a serrated edge may be made. Both finishes are described under sleeping mats. The basket is then turned inside out when the true outer surface of the wefts is exposed and the turned-down finish of the rim concealed on the inside.
The handles (avei) are made of strips of pandanus, forming a foundation around which four strips of pandanus are plaited in check so that the
Where the serrated edge rim is used, the knotted ends of the handle loop are concealed under the plaited finishing band of the rim.
The body of the basket may be decorated with a strip of soa'a plantain bark laid along a weft course and inserted under the crossing wefts.
Samoans hold that such baskets are really a modern development made since women began to smoke and carry small articles not available in olden days. In pre-European days there was little use to which they could be put, though 'ete li'i (finely made baskets) and 'ete mamanu (ornamental baskets) in his dictionary, they could still be a fairly modern development. Nowadays they are made for sale. (See Plate XVI, E.)
Large baskets made of pandanus leaves, either fala or paongo, are said to have been made somewhat square or oblong and with a cover for use in storing kilts and skirts. They are called tanga but their manufacture had been forgotten according to my informants. The basket figured in Plate D was made with double wefts of pandanus, as the bottom part shows. The sides have been covered by thin white and colored pandanus material split into narrow wefts to work decorative designs in twill. The decorative part is elaborated as a result of trade influence, as is also the small size of the baskets. The thick, coarse foundation wefts beneath the decoration follow the check plait shown on the bottom and at the rim. The wefts cross not obliquely but at right angles to the edges, as will also be seen later in the baby mats (tapito). It seems likely that the large, square or rectangular tanga baskets followed the technique of the model basket shown in Plate XVI, D, but without so much overlaid decoration. Turner (40, p. 275) figures a basket evidently of the tanga type with a cover and some oblique lines of decoration. The thick, double wefts in the models make a stiff basket that in a larger size would do excellently as a clothes basket for pieces of bark cloth.
The introduction of wooden chests has displaced the use of the tanga basket and the modern development of coiled work has extinguished any chance the model rectangular and four-cornered types may have had of surviving for trade purposes.
Baskets made of sennit ('ato to) were used in olden days and, though the technique is not plaiting, they are grouped here for convenience. They were used by tufunga (builders) to carry their adzes and tools for making houses and canoes and were thus called 'ato to (to, to build). They are very scarce but one was obtained at 'ato fa'apaupau and quoted as his authority the saying applied to wicked people likening them to an 'ato fa'apaupau. Paupau means ancient and has come to signify heathen to the Christianized Samoan.
The example basket shown in Plate XVI, B, is made of three-ply sennit braid but two-ply twisted sennit is said to have been also used. The technique has been forgotten for baskets but fortunately it survives in the scoop nets made for the small ingana fish in
In the smaller Ofuan basket the technique is the same in all respects but the braid is smaller. It was described to ola fangota (fishing basket) for carrying bait. The use of bait cannot be old. It was probably used for containing trolling hooks (pa) and line and the functional use of the name was loosely applied for the true ola fangota is made of coconut leaflets. The smaller basket obtained by
Coiled basketry was introduced into Samoa, and round baskets in various shapes and sizes are now extensively made for the tourist traffic. Though foreign, the technique has been adapted to local material and merits a brief description. The foundation of the coil is coconut leaflet, midrib or strip of the leaf midrib (lapalapa) which may be varied in size according to the position in the work, being narrower and thinner towards the centre of the coil where work commences. Only one foundation strip is used, forming what Otis
Baskets commence in the center of the round bottom and the coil is worked from right to left. The foundation strip is wrapped in a close spiral with the pandanus strip. The turns on the outer side towards the worker run
On the bottom the connecting stitches are made from the outer coil round the inner and on the sides with the bottom upwards, from the lower to the upper. The stitch is made from the coil that is being wrapped to the adjacent part of the coil already dealt with. (See figure 110.)
At the commencement the stitches are made close together but after the continuous coil is established the stitches are spaced more widely. They are now arranged to form various patterns as shown in Plate XVI, A. Round or oval platters are also made and the baskets may have a single long handle, or flat covers with a rim. Round baskets with the same stitch are present in the
The roughest type of floor mat was made from coconut leaves, hence the name polavai. In olden days it had considerable status, as it was the proper type of mat upon which to place the gods during religious ritual. The name lapa'au or tapa'au vai is connected with the ancient religious aspect of its use. The polavai are still in common use as the undermost layer in the carpeting of a house with floor mats. They are easily and quickly made and by taking the direct friction against the gravel of the floor they save the upper mats of pandanus leaves. See Plate XVI, C.
Younger leaves with more perfect leaflets are selected. Two six-foot lengths are selected to provide pairs of right and left midrib strips. The pairs of a kind are placed together so that their closed leaflets coincide, and a pair of leaflets used as one element in arranging the commencement. They are laid clown with the butt ends towards the worker, the under surface of the leaflets turned upwards, the midribs to the middle line and the strip from the right side of the leaf thus on the left. (See figure 111.)
Apart from neater workmanship, the floor mat is the opposite of the carrying sheet in which the outer edges are formed by the midrib strip and the two sides joined together by a three-ply braid in the middle line.
The polavai is the pola placed inside the house and, according to some Samoans, is the polataufale (pola within the house) in distinction to the polataufafo, the wall screen that hangs to the outside.
Material. In mat making the Samoans use three kinds of pandanus, distinguished as paongo, fala, and 'ie. As the leaves are used the material is generally referred to as laupaongo, laufala, and lau'ie. All three plants were seen growing in cultivation, where they were tended by keeping the larger growth cleared. In cultivation the trunks were a few feet high, ending in a single head of leaves. The old, dry leaves were removed and frequently the lower, outer leaves of the 'ie were bound round with a strip of dry leaf to prevent them from falling down. Leaves from uncultivated plants may be used as an expediency, but, for her mat material, the Samoan woman goes to per cultivated plants.
Of the three plants the paongo has the longest and widest leaf. Some of those measured that were lying out to dry were 10 feet long and over 4 inches side. The edges of the leaf are serrated with sharp spines and the under surfaces of the midrib are also armed with long spines. From the paongo are made coarse floor mats (papa or paongo) with very wide wefts. The correct time is paongo, as given by ng sound into n, refer to the plant as paono.
The fala has a much smaller leaf, dried specimens measured being 5.5 to 6 feet long and 3 inches wide. It also differs from the paongo and the 'ie, in saving smooth edges to the leaf and but occasional spines on the back of the midrib. From the fala are made floor mats with a narrower weft, sleeping mats, and the small mats made for babies, which receive the general name of fala. The term fala is also applied to the fruit of the pistillate plant of all species, the keys of which are used for making necklaces ('ula fala).
The 'ie has a leaf much the same size as the laufala. A cooked specimen measured was over 5 feet in length and its width near the trimmed base of the leaf was slightly over 3 inches. It differs from the fala in having serrated edges with closer, smaller spines than the paongo and a close set of spines along the under surface of the midrib. From the upper layer of the horizantally split leaf, the 'ie tonga fine mats used as garments are made and also, in occasion, some finer sleeping mats. The 'ie must not be confounded with the 'ie'ie which is a Freycinetia and from the roots of which fish traps are made.
paono, recognized the plant as Pandanus Martelli.
fala to Pandanus tectorius and its two varieties of P. saviensis and P. upoluensis described by Martelli. His specimens, however, were from uncultivated plants. 'ie leaf with an intact tip end to specimens of P. tectorius from other regions, was of the opinion that the 'ie belonged to P. tectorius, but there was not enough material available to locate the variety. From my description of the cultivated fala he was of the opinion that it closely resembled P. tectorius var. laevis, which is also cultivated for mat material in the Society,
The adult, uncultivated tree pandanus with many branches and aerial roots is known as fasa and the efflorescence as singango. When there is a fruit, the tree seems to take the name of fala as well as the fruit. fasa and the pistillate, lau fala. In lau fala, lau, of course, refers to the leaf.
Preparation. The laupaongo are trimmed along the midribs to remove the sharp spines, and the spinulose-dentate margins are split off. This is termed autala. The leaves are spread out in the morning to dry in the sun (fa'ala) and gathered in at evening for a period of four or five days. Each individual leaf is then rolled round the left hand to open out the curled-in margins, thrown in a heap, and then rolled round the hand and other leaves added until a fair-sized circular roll one leaf in width is produced. The process and the roll is called fa'amasina (to make like a masina—a moon). The roll is kept together usually by four radial ties of fau bast or pandanus leaf strips which pass through the hole left on removing the hand.
The laufala are trimmed (autala) in the same way to remove the few midrib spines and smooth margins. Each leaf, commencing with the tip end, is then rolled round the left hand. The butt end is then doubled into the coil and at the same time the right hand draws out some of the middle loop, the left hand holds the tip end against the bent-in butt. The drawn-out turns form a loose spiral and the process is termed sapine. A number of leaves are attached loosely together by passing a strip of fau through the end loop of the leaves and knotting the ends of the strip together. The process is tui laufala (tui, to thread together) and each bundle is a taulanga laufala. They can thus be carried out and brought in more readily by means of the fau strips. The leaves are exposed to the sun (fa'ala) for 4 to 7 days. Straightening out both the leaf and its margins is termed fa'amafoloa. The leaf is rolled round one hand and then reversed by rolling it around the other hand. fa'amasina bundles (also termed ta'ainga) and then stored.
The preparation of the lau'ie is described on page 275.
Before plaiting, the rolls are untied and each leaf straightened out by scraping along the back or outer curve with a shell. The leaves may be further straightened and smoothed by rescraping with the shell. Each half leaf is then split from either side of the midrib which is discarded. At one time the splitting was done with a sharp point of bone or the spine of a porcupine fish but now a knife is used. The plaiter then splits the half leaves into the required widths for the wefts with the sharp point. The process is called totosi. A point was also obtained from the shell of the 'u'u, a species of mussel. The splitting usually commenced a few inches from the butt end
The wefts in the coarse papa mats (Plate XVII, A, 1) made of lau-paongo were an inch or more in width. Two strips of leaf were used as double wefts and were reversed so that a tip and butt end coincided to insure an even thickness in the double weft. Where a half leaf produced two wefts, the undivided butt portion was run out to the end to separate them. The right forefinger and thumb held the butt end of one strip while the butt end of the other was reversed between the second and third right fingers. The left hand seized the butt end of the second strip and the two hands were drawn apart, thus reversing the ends of the second strip and forming the double weft.
Plaiting commences by defining the corner (fig. 112) and is extended from the corner by adding double wefts to the left. (See fig. 113.)
As each subsequent weft is added on the left, the check technique is maintained with the double wefts but not with their individual elements except at the marginal turns forming the edge which are a repetition of the technique shown. When sufficient wefts have been added to get a working edge of dextrals, the plaiter twists the plaiting into a convenient position so that she may work in orthodox fashion.
The papa mats may be made in any size, but the common form is a little over 2 feet wide and 6 to 7 feet long. It is usual to plait one short end and then plait sections across the length so that the wefts may be turned in at the sides, thus employing fewer wefts on the shorter sections. After forming the first corner, the corner is turned to the left and a working section carried along until the 2 feet width is obtained. The next corner is then turned as in figure 114.
When the various working sections which have the left and right edges defined by the edge technique have resulted in the mat nearing the required length, the left edge and left corner are completed. The far edge and far end of the right edge are reached by turned-in wefts from both margins. They are simply left on as extra thick wefts and instead of turning single elements at the remaining parts of the edges, two strips of material may form each turning element of a weft. In some mats, the extra strips are cut off before reaching the edge turn. These extra strips are then turned individually at right angles and run back by being pushed under crossing wefts already plaited. They give the false appearance of a double edge. The true edge formed by orthodox technique is distinguished by the turning wefts forming triangles. On turning the extra wefts forward, they cross not only a triangular space but the next full weft space as well before they pass under a crossing weft, as will be seen in the upper left corner of Plate XVII, A, 1.
At the last corner where the working edge runs out, the turned-back wefts have to be pushed individually under a sufficient number of crossing wefts already plaited in position to fix them.
New wefts are added by simply laying them on the shortening wefts and plaiting them together for a number of strokes. The shortening weft is then dropped and subsequently cut off close under a crossing weft to hide the end.
Sometimes papa mats are marked by a woman making two or three nicks, or steps, at a corner, or putting in a short row of twill on the body to distinguish her own property.
Mats made from laufala have received the name fala from the material. The name has come to be used generally and may be carelessly applied to mats made of paongo. They may be divided into floor mats and sleeping mats for which a different technique is used. With sleeping mats may be included the small special mats made for babies.
Floor mats. The fala material does not admit of such wide wefts as the papa. An average weft is 0.5 inches. They may be made in the same sizes as the papa or much larger. Very large mats have been made to carpet the rooms of foreign houses but such forms were inconvenient for strictly Samoan use. The smaller sizes are more convenient for rolling up and stacking away on the house shelves, and on ordinary occasions only such parts of the house were covered with mats as were in use. The smaller sizes were also more convenient for pushing aside to expose the gravel floor on which a kava libation could be poured or a round-bottomed receptacle set by scraping away the stones.
Double wefts with a check stroke and the individual element turn at the corners and edges were used as in the technique of the papa mats. There is no difference except in the nature of the material and the width of the wefts. (See Plate XVII, A, 2.)
Sleeping mats (fala moenga). In the sleeping mats the single narrower wefts were used, averaging about 0.25 inches or narrower. Though the check formed the basis of the technique, twills were commonly used and change of stroke as in the wall screens and laulau platters were introduced as structural decoration. Applied decoration, by overlaying with the black strips obtained from the outer skin at the base of the plantain trunks (soa'a), was used sparingly or probably not at all in olden days. So also the structural decoration by means of wefts dyed black with lama, to be seen on mats with or without introduced worsted fringes (Pl. XVII, C, 2) are probably as modern as the colored wool material. Tourist traffic has given a greater incentive to the use of color.
Plaiting (lalanga), qualified by the number, supplies the name for the plaiting strokes used. Thus lalanga tasi denotes check and lalanga lua, tolu, fa, lima gives the twills from two to five. A combination of strokes may be lalanga lua ma lalanga tasi (twilled two and a check). The term si'i (to lift) is also used with a number to denote twills as si'i lua (twilled two), si'ifa (twilled four). Instead of lalanga, lau (weft) may be used as lautasi (check) and laulua (twilled two).
The half leaves of prepared fala, usually about 1.25 inches in width, are split (totosi) into five wefts, commencing at a few inches from the butt end and running out at the tip end. (See fig. 115.)
The Samoan commencement corresponds to the hatu rua (39, p. 110).
The side edges are usually formed by binding in the marginal projecting wefts directly without a half turn as in the afeafe turn of the laulau platter. In some mats, a half turn is made as the projecting sinistrals on the left are turned in at right angles to function as dextrals or as the projecting dextrals on the right are turned in to act as sinistrals. The side edges formed by the above two methods are plain and simple. (See fig. 116.)
Adding fresh side wefts (fa'aulu): In some mats, in order to afterwards make more elaborate side edges, the wefts are left projecting at the side edges for subsequent treatment. The plaiting, however, cannot advance in a c, by adding fresh butt strips and adding their weft strips successively from below up as the left margin is built up to start another working section. Similarly, on the right, fresh sinistrals have to be supplied as each working section reaches the right in order that the right margin may be defined in closing up the section to the level of the plaiting edge.
The finishing edge. The Samoan method of finishing off the far edge of the mat is simple. The finishing technique is carried out along the plaiting edge of the last working section, by turning down the top dextral which would otherwise be dropped out of the working set of dextrals. The top dextral is turned down on the working sinistral that is placed in the shed prepared for it and included with the sinistral between the crossing dextrals as the movement is completed. (See fig. 117.)
On the right end of the finishing edge, the last few wefts which have no depth of working edge in which to be fixed, are plaited on as a free three ply tail which is knotted, doubled back on the under surface, and fixed under a couple of crossing wefts of the completed plaiting.
The object of plaiting the mat with the under surface turned upward is now apparent. The turned-down wefts with their cut-off ends are on the under surface of the mat and prevent any irregularities from appearing on the true upper surface when the completed mat is turned. If a mat is plaited with the true upper surface upwards in order to work out some geometrical designs, it is always turned under side upwards to plait the finishing edge.
When the turns of the top dextral are made to define the finishing edge, the top dextral is raised and then dropped on the sinistral placed in the shed. The half turn made by the dextral is thus above the sinistral which projects
The commencing end finish. The commencing end of the mat with two sets of crossing, unsplit butt strips is now turned to the front, still with the
The varieties of the technique used in turning mat edges that were observed in the field, and on mats acquired by
Direct bend. This is the afeafe bend used at the side edges where the wefts are turned in successively at right angles to their course without changing the exposed surface of the weft. (See fig. 116 a.) Owing to the difference of color and appearance of the true upper and lower surfaces of the weft material, the true upper surfaces of the weft must be kept to that which will form the true upper surface of the mat. A half turn at the right and left edges would cause the under surface of the weft material to appear on the true upper mat surface and spoil the appearance. This technique is used in wall screens and laulau platters for a similar reason.
Half turn. The half turn which exposes the other surface of the weft b and 117.)
The half turn is used at side edges when the wefts contain two elements. Double wefts are arranged back to back so to speak so that the smoother, more shiny true upper surface faces outwards on both surfaces. In such cases, the half turn makes no difference in appearance to either mat surface. This is evident in the paongo and fala floor mats and in the baby mats made with this technique. Being used at the side edges, where the projecting weft of one series is converted by the half turn into the commencement of the other series, there are no projecting ends left to dispose of and the method thus forms a complete technical process. (See figures 112, 113 and 114.)
The cut-off sinistral edge. This is the end finish described which forms the complement to the half turn method adopted with functioning dextrals. Two forms are in use. (See fig. 118 a and c.)
The tucked in method applies to the finishing and commencing ends of mats. (See fig. 118 d.) As it is somewhat tedious, it is usually reserved for baby mats or for more ornamental sleeping mats in which color has been introduced. It can also be applied to the side edges, when wefts have been added to the sides to form side plaiting edges.
The two-side three-ply braid. In a baby mat, the three-ply continuous braid made along one surface with the projecting free sinistrals and continued on down the other surface with the free dextrals, as in the laulau platter finish, was observed. (See figs. 95 and 96.) The narrower wefts, however, were more neatly brought in with a flat, half turn and instead of being finished off with a free, braided tail, the end wefts were pushed through the commencing wefts to continue the braid which is thus continuous without any apparent end. The weft ends, discarded after a course in the braid, were neatly cut off.
The serrated edge (fa'atalatala). The serrated edge (talatala, with points) is formed of sets of three or more crossing wefts plaited to form triangles of which the base is formed by the original plaiting edge and the sides by free edges obtained by doubling back the wefts which project beyond the marginal wefts defining them. In figure 119, sets of four are used.
The second part of the process consists of dealing with the free ends of the wefts used in making the triangles. The points of the serrated edge are now turned towards the worker, as in figure 120. The technique is simply that of the finishing edge. (See fig. 117.) The method forms a separate, distinct band of plaiting on the plaiting side of the serrations, distinct from the plaiting of the body, and has a free edge formed by the dextral turns. This is the common form, but in some better mats, where the sinistrals have
The sleeping mats with serrated edges usually have the serrations carried around on all four edges. (See Pl. XVII, C, 1.) In forming the side edges, new wefts have therefore to be added at the sides (fa'aulu) to provide the crossing elements, and the direct bend or half-turn edge can not be used.
Baby mats. Small mats averaging 38 inches in length and 24 inches in width are made as sleeping mats (fala lili'i) for babies. The term lili'i (small) applies to the size and also to the narrower wefts but the name of fala pepe is now usually used. The Samoan term for "baby" is pepe. The technique for ordinary sleeping mats is used, but with more care and variety of stroke. A mother or grandmother demonstrates her regard for the baby by care in the selection of material and skill in the use of the more uncommon forms of decoration and edge finish. The wefts may be as narrow as 0.2 inches. The variety in technique is demonstrated by three such mats in
The body is made of double wefts of the somewhat wide width of 0.4 inches. The commencement was made along one of the wide edges and the technical side edges form the shorter ends when in use. The double wefts are turned in at the sides with the typical two half turns described with floor mats. (See fig. 114.) The stroke of the body is check, but at intervals the monotony is relieved by rows of dextral and sinistral twilled twos. About 2.25 inches from the finishing long edge, the double wefts are separated into two layers and each weft element split into two. The plaiting of one layer is continued with the narrower wefts in twilled twos for a depth of 3.5 inches and the plaiting edge finished off with the turned-down dextral and cut-off sinistral technique. (See fig. 117.) The plaiting is continued around the end, when the mat is turned and the other layer of narrow wefts plaited. The two B, 1.)
In Plate XVII, B, 3, single element wefts of two differently colored fala materials were used, the darker color forming the dextrals and the lighter, the sinistrals. The commencing edge was again one of the long edges and new wefts were added at the side edges. With the two colors, which were not dyed, a geometrical design was worked in twill with combinations of four wefts and then spaced with lines of check forming square boundaries. All the edges were then finished off with the tucked down sinistral edge. (See fig. 118 d.) Owing to the use of single element wefts, only one surface could be used as the upper surface.
In Plate XVII, B, 2, double wefts in two natural colors were used and differ in arrangement from the previous mat, in having the alternate wefts, both dextral and sinistral, of one color. The commencement was again at one of the long edges. The side edges were formed by the simple half turn, without interlocking. Here, owing to the double wefts having the bright surface outwards, the reversal of surface by the half turns at the side edges made no difference in appearance.
The body was plaited throughout in check but so arranged that the two colors were worked into single weft oblique lines arranged to form squares. About an inch from each long edge, a series of narrow open slits were formed by giving two wefts about to cross, a half turn to deflect them away from each other. (See fig. 121.)
The finish of each long edge is by the two-side three-ply braid of permanent food platters. (See figs. 95 and 96.)
In each of the three mats, a different technique has been followed in the plaiting of the body, and the finish of the long and short edges. These small mats are a distinct type created for a specific use and must not be confounded with samplers made for tourists.
The baby mats described have followed the rule, laid down for plaiting, that both sets of crossing wefts are included at the one commencing edge and thus all wefts run obliquely to all edges. At Fitiuta in D). If the old tanga baskets were made with this plait, it is possible that the technique may have been applied to small baby mats, originally without diffusion from a foreign source. A), together with the Samoan terms used. After conversation with her, the technique was recorded, but judgment is reserved as to whether it is original or due to more recent diffusion. The statement of the present old Samoans as to the age of a technique cannot be accepted without reservation in doubtful instances. An examination of reliably old material in other museums, if present, may clear up the doubt that exists. To distinguish the two general techniques, the mats with the orthodox oblique wefts are termed fa'aalo and those with the right angled plaiting, tapito.
Tapito mats. The commencement is made with unsplit butt strips bearing 5 or 6 wefts. The arrangement is shown in figure 122.
The usual twilled twos in parallel oblique lines (fig. 122 b) retains the twilled two technical term of lalanga lua, but when arranged to form zigzag lines (Pl. XVIII, A, 3), the technique is termed fa'api'opi'o (crooked), but the alternate oblique panels of plaiting so obtained are termed fa'afatuamanga (Pl. XVIII, A, 4.) Geometrical designs (fa'asumu) may also be worked. By tilting the work with one corner towards the observer, the technique by which geometrical motives are produced in the orthodox plaiting of fa'aalo mats is seen to have been followed in the right angled plaiting of tapito mats.
Another decorative effect is obtained in check by using alternate wide wefts unsplit and splitting the other alternates into three narrower wefts. The check plait is made regardless of the width of individual wefts. (See Pl. XVIII, A, 2.) Thin strips of lau'ie are overlaid on the narrow wefts to add a color contrast. The technique is termed lalanga atoa (atoa, wide).
The edges of the tapito mat are finished off by doubling the wefts that project beyond the marginal wefts round their outer edge back on the under surface and pushing them through under a number of crossing wefts on the body of the plaiting. The two edges with the butt strips are treated in the same way after first splitting the undivided butt strips into their component wefts. fa'atala serrated edge was also formed by plaiting sets of five projecting wefts to a diminishing point.
Mats made of lau'ie are mostly of the fine mat type known as 'ie tonga. As they are worn as skirts, they are described under clothing. Some not so thin, or as elaborately decorated with red feathers, are used as chief's sleeping mats termed 'ie moenga.
One donated to 'ie tonga. Though the wefts average 13.5 to the inch, it is thicker than the usual fine mat and Tuitele states that it was used as a sleeping mat. The mat, like so many owned by high chiefs, bore a name. The name of the above is Tautala ma le amonga (Speak with a burden of food). The idea amonga). This was deposited outside the house and the bearer or the bearer's employer was privileged to speak (tautala) with his high chief. As the speech usually took the form of a request, the burden of food assisted materially in obtaining favorable attention.
Check. In close check plaiting each weft stroke forms a square with its sides equal to the width of the weft. Without change of color, each square looks the same whether it inclines to the right or the left. The only possibilities of pleasing the eye in this technique are the regularity and fineness of plait obtained by splitting the wefts evenly and getting them as narrow as the material will allow. With objects of ordinary use, utility comes first. Wide wefts are easier to handle and the object is more quickly made. When we pass from floor mats to sleeping mats of the same material, the wefts narrow until they diminish to 0.2 inches. Further advance necessitated a special technique in preparing material to render it possible to obtain narrower wefts. Fine weaving thus reached its highest development in the check technique of fine mats where, owing to the lau'ie material being split into two layers, it was possible to split wefts as fine as from one-fourteenth to one-twenty-second of an inch. The aesthetic value of a fine mat is due to narrow wefts, but the technique, though difficult to work with the fingers and trying to the eyes, remains the simple, common check in universal use. In the use of the check, Samoan technique specialized in the direction of increasing the fineness of the weft and the alternate decorative possibilities to be obtained by using change of color was not exploited to any extent.
Twills. With twill strokes, it is possible to get change of appearance with materials of one color. Aesthetic value was not primarly influenced by reducing the weft width as in check, and different patterns appear early in material made with coarse coconut leaflet wefts such as wall screens. Structural decoration due to change in the strokes made with the foundation wefts appears as an early form of decoration.
In joining together two or more coconut leaf midrib strips to provide two distinct sets of crossing wefts, the Samoans found that a commencing line of horizontal twilled twos was the easiest and quickest form of technique. Horizontal lines of twilled twos were continued up on the body and are probably the oldest pattern. Any departure from horizontal lines created a new pattern and a further decorative effect. Vertical lines of twilled twos thus became popular on the wall screens and laulau platters as a departure from the common pattern produced by ordinary plaiting.
In the ola fish baskets it was easier to continue with the ordinary technique of horizontal twills but in the more pretentious round baskets vertical twills were preferred. Twilled twos also developed into twilled threes, fours, or fives and combined with check or lesser twills as the desire for variety dictated. In coarse floor mats the working check was retained but in sleeping and baby mats the more decorative twills were used. In fine mats the check was retained, as aesthetic value was supplied by the fineness of the plait and a change to twill would have complicated an already difficult technique as regards manipulating the fine wefts. Changes in the direction of the lines of twill led to zigzag patterns termed fa'api'opi'o (crooked).
Further advance in structural decoration was made by raising groups of three or more dextrals and altering the order in which they were dropped to cross with a similar number of sinistrals. This has been seen in the wall screens and has become associated with them in name, as lalanga fa'atalafale (plaiting like the wall of a house). The geometrical figures produced mamanu of sumu, names also applied to sennit designs worked on wall posts, beams, and canoes. They are regarded as important forms of decoration. There is pride in the voice of a craftswoman when, as she starts a combination, she says, "Mamanu." Similar pride is expressed by a man when he points out the sennit design on a wall post or beam, and says, "Mamanu." The pride that a Samoan woman takes in the geometrical pattern of a wall screen made out of coconut leaflets shows that she obtains pleasure from the structural decoration which is unassisted by change of color. Hence, when she adopted the geometrical patterns on floor mats where color could have been easily employed, she was so satisfied with the plain structural decoration that Samoan plaiting remained in that stage of decoration and advanced little, if at all, until more recent influences opened up the possibilities of employing colored elements to enhance the appearance of the geometrical motives that had come into use.
Structural geometrical motives. The motives used in the mats examined were the two used in the wall screens with modifications regarding the bounding of them with lines of check or twill, and one other, also probably used on wall screens. (See fig. 123.)
Colored wefts. A restricted use of colored wefts is now in vogue but owing to the satisfaction derived from structural twill decoration in the color of the natural wefts and the high value of fine check plaiting, it is probable that the technique is of fairly recent development. Color contrast may be obtained by using two shades of the weft material or a distinct color from some other material, such as a dye.
The method of using two shades of the same material is the most common and is probably as far as color use went in olden days. The laufala has a natural brown shade which gives the general color to all mats. If some of the
One set of wefts (dextrals) is provided from one shade of material and the crossing elements from the other shade. In check plaiting, a draft board pattern is produced in small squares the width of one weft. In this pattern all the white elements run in one direction and the dark elements in the opposite. In the baby mat described with the longitudinal open slits, the change in direction of two wefts on every third sinistral changed the direction of every alternate dextral and every alternate sinistral. The pattern then changed from colored checks to thin oblique lines running to the right (see Plate XVII, B, 2), which were formed by the alternate arrangement of color throughout, resulting in the sinistrals of one shade crossing over the dextrals of the same color. After a sufficient depth of this pattern, the direction of the oblique lines was reversed by plaiting a single horizontal row of sinistral twilled twos. Any twilled two stroke will change the direction of the line. By using twilled twos to change direction and mark the angle of the change consecutive squares may be formed.
The slits that may be regarded as part of the decoration were really used instead of a twill stroke to change the pattern. In twill, the horizontal lines of twill naturally arrange into the different colors of the dextrals and sinistrals. Vertical colored lines and zigzag lines can be arranged by simply following the stroke technique described. It is, however, with the geometrical figures that the two colors are principally used to accentuate the design, as in the baby mat (Plate XVII, B, 3, and fig. 123 c). Here the lines in twill and check are used as bounding motives to the more elaborate geometrical motive.
The use of the black outer skin from the base of plaintain leaves (soa'a) is claimed as old. If so, it is not so extensively used as in the soa'a is stripped off and scraped on the inner surface to make the material as thin as possible. It is then split into wefts soa'a material, the black color is on the outer surface while the inner surface is brown.
The method of using soa'a wefts is by overlaying them on the foundation wefts with the black surface upwards. Two methods of overlaying may be used: 1, Structural overlaying consists in laying the soa'a element on the sinistral foundation wefts as they function at the working edge. When the working dextrals complete their movement over the sinistral in the shed, the soa'a element is fixed on the working sinistral and the parts not covered by the working dextrals which cross over it, show up. Thus each sinistral so treated is black on the upper surface but no color shows on the under surface. Structural overlaying is not much used by the Samoans. The only specimens seen formed horizontal lines of twill. 2, Inserted overlaying is more common than structural overlaying but is little used. The completed structure of the mat is then decorated by pushing single strips of soa'a under crossing wefts over the course of selected wefts. Thus in the sleeping mat (Pl. XVII, C, 1) zigzag lines in black run close to the serrated edge, while further in, lines are crossed so as to form a series of rectangular figures, tilted on one corner, and merely outlined at their margins.
Though well acquainted with the use of native dyes in coloring bark cloth, dyed pandanus material did not figure much in the craft of mat making. It is doubtful if it did until after the more extensive use of foreign dyes in making kilts for tourists. The stimulus given to dyeing kilt material extended to mat material, not for trade but for their own use. The elaborate mat in Plate XVII, C, 2 has all the sinistral wefts dyed with the black native lama dye. The same geometrical motive used with the two natural shades (Pl. XVII, B, 3) are used here with effect and interspersed with squares worked in twill arranged in four triangular segments (fig. 123 c). Though plaited by Samoans with motives that could easily be derived from their own craft, one feels that the idea of such elaboration is as foreign to their own native culture as the colored worsted fringes around the edges. In Samoan plaiting, therefore, the importance and value of the fine mat ('ie tonga) directed the attention of craftswomen in the direction of fine check plaiting and restricted the development of other forms of decoration.
Eye shades (taumata) are plaited with coconut leaflets for the use of bonito fishers to protect their eyes from the glare of the sun. They are roughly and quickly made. (See fig. 124.)
The plants which supply material for cordage are the fau or fau tu (Hibiscus tiliaceus), fau songa or fau olonga (Pipturus propinquus), mati or matiata (Ficus sp.), and the coconut. The bast of the breadfruit (Artocarpus incisa) is used for a particular net and sometimes the bast of the paper mulberry. Except the coconut, the inner bark or bast of the plants is used. It readily splits off from the outer bark and for finer lines, it is scraped on a board with a shell to remove the coloring matter and mucilaginous material.
The fau grows practically everywhere and supplies the material for ordinary ropes. The whole bark is used in wide strips for tying scaffolding and in narrower strips for other minor purposes, such as tying baskets. For common everyday use the fau is invaluable.
The fau songa is called fau olonga in olonga, fau tu, it does not grow in such quantity and the better fibre is thus restricted to finer cords. After being scraped it is usually braided together (Pl. XVIII, C, 1) and rendered whiter by soaking in sea water, rubbing in sand, and bleaching in the sun. The extra labor in getting sufficient quantity prevents its being used to make ropes.
The matiata supplies strong fibre in the bast of the long, slender rods which characterise the plant, and is used as cordage for shark nets.
The coconut supplies strong fibres in the husk surrounding the fruit. The large quantity of interfibrous material is separated by a special process. The much-used three-ply braid is made from it as are also the strong ropes.
The fau supplies the material for ordinary rough use, fau songa, the strong lines and cords not requiring too much in the way of quantity, while, for quantity and strength, coconut husk fibre supplies the 'afa braid and strong ropes.
The term "plaiting" has been used so much with cordage that, as a general term, it need not confuse us. There are, however, two distinct methods: twisting and braiding. Twisting again is divided into two forms of technique according to whether the quantity of material used is small or large. For cords and lines, the plies of the cord are twisted around each other on the bare thigh, which process is termed milo. Here milo refers to both twisting the individual strands or plies separately and then over one another. The one follows the other as a matter of course. In rolling coconut fibre together on the bare thigh, the process stops short at the rolling of the individual strands. They are afterwards braided. The term milo could not be applied, so the term fa'ata'a is used. In twisted cords and twisted ropes of coconut fibre, the twisting of each ply separately and then over each other was done by hand, the milo process on the thigh being unsuitable or impossible. In lafo, though
To twist plies over and under each other, as in braiding, is termed fili. Three-ply sennit braid has received the specific name of 'afa (cf. kaha and 'aha) but it may be further distinguished as 'afa fili-tolu (tolu, three). In braiding, some confusion in terms appears to exist as to cordage with more than three plies. There is no trouble as to the number, which is simply mentioned in the name. The lack of clarity is in the use of the terms tua (fold) and langa as in 'afa tua lima and 'afa langa lima. As langa seems to be derived from lalanga (to plait a basket or mat), it probably refers to a braid made with the technique of check plaiting, where each ply passes over one and under one, whereas tua would refer to the braids made like three ply braid, but where the outside ply crossing over to the middle position may cross over more than one ply. Thus the round four-ply braid is an 'afa langa fa (fa, four) whilst the five-ply braid seen in use is an 'afa tua lima (lima, five). All sennit cords and braids are called 'afa with a qualifying word. When 'afa is used alone, a three-ply braid is meant. The two-ply twist is also loosely called 'afa fili lua, which is not the correct usage of fili.
Threads, or fine cords are termed manoa, fine lines, ta'a and ordinary afo. A rope is maea and its strands fu'a. The word taura used in the taula. The term maea in punga which in Samoan is restricted to larger lumps of coral. punga and displaced the term taura (taula) to the rope which brought it up. The Samoan restricted punga to the coral, kept taula for the anchor and used maea for the rope which caused the anchor to come up to the surface.
The finer threads (manoa) seen in use were two-ply twists of fau songa, used for the finer lashings of fish hooks, headdresses, squid lines, and supporting red feathers.
Two-ply twisted cords were usually made of fau songa by the milo process on the bare thigh. The plies were lengthened by adding a fresh strand of material to the shortening ply with an overlap and then rolling them together on the bare thigh before rolling the two plies round each other. The specimen figured (Plate XVIII, C, 2) was made for a small fishing net. The cord has not been used in water and the material is between ecru-olive and buffyolive in color.
Two-ply twisted cords of breadfruit bast were seen in a special type of seine net at 'ulu manu'a were used. The bast was scraped on a board in the same way as in preparing paper mulberry bast for bark cloth. The cords were fairly thick but varied in different parts of the net.
A fine cord (Plate XX, B, 1) not quite 2 mm. in diameter was used for such purposes as tying stone sinkers to nets. The two plies could be rubbed together on the bare thigh by the milo process but owing to the rough nature of the sennit, it was usually made by twisting the plies with the hands and plaiting towards the body as in ropes, the commencing end being fixed to a post or stake. The fresh strands are usually joined to the shortening ply by simply overlapping the ends.
A coarser two-ply is made by strands rolled together with a binding fibre (fa'ata'a) such as are used in three-ply braid. Besides the simple join by direct overlapping, two other methods are used. (See figures 125, 126.) In
Two kinds of three-ply cords are made from the fau songa and from matiata: a, the ta'a cord made of fau songa is used for the fine lines with the trolling hooks known as pa ala, the thicker lines with the bonito hooks (pa'atu) and for rod fishing inside the reef (seuseu). (See Plate XVIII, C, 3, 4, 5) The three plies are rolled together by the milo process on the bare thigh. New strands are added by direct overlapping on the shortening ply and rolling on the thigh. The fau songa cords, after use in water have a typical pale neutral gray color. b, The name given to the three-ply cord made of matiata bast by a fatu a fau. The bast was split off from the epidermis, scraped, and dried in the sun.
Milo process. The bast was divided into appropriate thicknesses and rolled separately on the thigh into strands. Three strands were held between the left forefinger and thumb in such a way that they are slightly spaced apart. Still holding them firmly, they were laid transversely over the right thigh. The right palm towards the base of the fingers was laid over the three
The join in this thickness of three-ply was made by the doubling over method with the new strand added from below (fig. 125), but owing to the milo technique being used a slight difference exists as shown in figure 127.
The matiata three-ply twist (Plate XVIII, C, 7) is 3 mm. in diameter. The extra thickness gave the cord the strength needed for making shark nets which were anchored off the reef.
Sennit braid ('afa) is the most important single article in Samoan material culture. The quantity used is so great that even in the present period men use much of their spare waking hours in plaiting it. All the actual details of its preparation are thus readily available.
The coconut (niu) used is the special large variety known as niu 'afa (the sennit coconut). The fruit is about 13 inches long, the husk thick, and the contained nut comparatively small. (See Plate XVIII, B, 1.) The phrase applied to it, fete pulu (swollen husk), has come to be used for anything large with little in it. Other nuts used were as follows: niu ui, with a very pa'e pa'e); niu alava, with a medium colored fibre; niu malo, with a reddish fibre (mūmū). When niu 'afa can not be obtained, ordinary nuts, (niu sasa) are used. The niu 'afa is roughly triangular in cross section. The fibres within the husk run longitudinally with the long axis of the nut. The husk is, therefore, prized off in longitudinal sections. (See Plate XVIII, B, 2 and 3.) The sections corresponding with the convexities, or rounded angles of the cross section, contain longer and better fibre, and the sections are termed matai'a. The intervening sections between the three matai'a have shorter fibres and the sections are called lafalafa.
The husk of the green nut (mumu'a) or the mature nut (popo) were alike used. The outer skin of the husk is termed tua pulu, the fibres mm 'a'a, and the interfibrous material tae.
Treatment of husk. The husk is removed in even, longitudinal segments with a mele'i husking stake. The object is to separate the interfibrous material from the fibre. The husks of some kinds of nuts, such as the niu ui, may be beaten at once, but most of them, including the niu 'afa, require soaking in water to soften the interfibrous material. The green husks from mumu'a nuts require only 4 or 5 days soaking, but the mature husk of the popo requires much longer—three weeks to two or more months. Long soaking does not deteriorate the fibre but, according to the Samoans, rather improves it in strength and lasting qualities. The segments are collected in a coconut leaflet basket, submerged in a pool inside the reef, and weighted down with stones. The submerged basket of husk is termed taomanga. After the minimum time of soaking has elapsed, the number of segments required for use are extracted from the taomanga and the rest left in the water for future use.
Beating the husk (sasa pulu). The inter fibrous material is removed by beating the husk sections, also called pulu, on a wooden anvil with a wooden mallet.
The anvil is termed malaise, but in tu'itu'i. An ordinary log or a rock may be used as a makeshift anvil, but every family usually has a well made anvil cut out of a section of coconut wood. They are thus circular in shape with a flat upper surface and usually four but sometimes three, legs. (See Plate XVIII, B, 4.) The legs are cut out of the solid with the anvil and their outer surface is continuous with the outer surface of the anvil. The under surface is not usually horizontal, but slopes downwards and inwards from the circumference to a central point. The lower end of the legs may be square or come to a blunt point. The anvils vary in size with the taste of the maker, but the circumferential legs are a constant feature.
The mallet (sa'afa) is an ordinary straight piece of heavy wood, round in section and with one end trimmed down into a handle. No particular care B, 5.)
The outer skin (tua pulu) is readily pealed off and the inner short part of the soaked segment torn off and discarded. Holding one end of the segment with one hand, the other part is beaten on the anvil with the beater, the left hand every now and again turning the segment. The interfibrous material flies off under the blows of the mallet whilst that which does not is loosened and flicked off after every few blows. The ends are reversed and the sasa process of beating continued until only the cleaned fibres remain. The interfibrous material has a vile odor which is painfully evident whilst husk beating is going on. Each segment is beaten separately and tied at one end with one of the fibres to keep them distinct, and is called a mato fi. (See Plate XVIII, B, 6.) If water is available, the mato fi may be washed to assist in removing the smell as well as any particles of interfibrous material. They are then thrown up on the thatch of a roof and left exposed to sun and air to dry and also to complete the removal of the odor. Immediately after beating, the fibre has a beautiful, silky, yellow color, but this changes after exposure to the usual sennit brown. As already pointed out, some fibre has a natural deeper brown or reddish color. The mato fi may be left out on the roof for some days.
Rolling the strands (fa'ata'a). The first part in the braiding process consists of rolling the prepared fibre into suitable strands (fa'ata'a). The braider sits down crosslegged in the house with a matofi bundle beside him. From it he pulls out a number of fibres sufficient for one strand. The very short fibres (fungafunga) are separated from the good fibres, which retain the name of mui'a'a, and discarded. The little bunch of mui'a'a is held by the left thumb and forefinger whilst some of the fibres are pulled out slightly at each end, not only to lengthen the strand, but to thin the ends for joining purposes. A single fibre is separated, its middle placed against the strand, and one end twisted round it by the right hand. The other end of the single fibre is then doubled back and the strand twirled between the finger and thumb to finish the rolling of the binding fibre. The strand is next rolled on the bare thigh with the right palm. The rolling of two or three strands of other material to form a twisted cord has been described as milo. The rolling of the single strands of coconut fibre is termed fa'ata'a and the resulting strand is fa'ata'a. Here we have the causative fa'a combined with ta'a (a line), meaning to make as a line. The rolled fa'ata'a is laid down beside the worker and the process goes on until a sufficient quantity has been made. When work ceases, the fa'ata'a are bundled together and tied round the middle with a single fibre. (See Plate XIX, A, 2.) As the heap of rolled strands mounts up, so do the discarded short pieces (fungafunga) collect on the mat. Hence
Plaiting. A sufficient number of fa'ata'a having been rolled, plaiting (fili) commences. In plaiting 'afa braid, which is not very thick, the plies are held between the left thumb and forefinger with the thumb uppermost and plaiting is directed away from the body. The technique thus consists of pulling whatever strand is in the middle position outwards under a side ply; first on one side, and then on the other. Whilst the right hand pulls the middle ply outwards under, the left thumb rolls the side ply over into the middle position. The left thumb also, by pressure down on the left forefinger, keeps the plies in their relative positions after each twist is made. The plaited part, therefore, passes backward under the thumb towards the body. It is just the free edge of the braiding that protrudes beyond the thumb, but in this and following figures, the thumb is shown well back so as not to obscure the technique. (See figure 128.)
When the fa'ata'a rolled strand is added and becomes an actual ply of the plaited braid, it is called an anga. Hence anyone asking the number of plies in a braid, says, "Pe fia le malosi o lau 'afa?" (What is the number of the strength of your braid?). The reply comes, "E tolu anga" (Three plies) ! Another name for the ply is tafua. Some braids have more than three plies. "Ply" is therefore used here to correspond with anga, which leaves "strand" clear for material or fresh additions and corresponds to fa'ata'a. Thus, a new strand is added to a shortening ply and once incorporated with it after the join, is included in, and becomes, the ply.
The join (so'onga). A fresh fa'ata'a strand is added to a shortening ply in much the same way as in a twisted cord. The rule is to bring the short ply into the middle position and add the new strand to it with its short end projecting back on the completed work where it is held under the left thumb. figure 129 shows the process, but in actual practice the thumb keeps just behind the working edge and the completed braid works backward under it.
As the braiding proceeds, the fingers naturally feel the thickness of the plies. If one is felt to be too thin and thus likely to spoil the evenness of the braid, it is reinforced with a fresh strand added in a manner opposite to that of the join above. (See figure 130.)
The principle of reinforcing a thin ply is to add a new strand from below with its short end on a long ply in the middle position. A couple of turns are made to bring the thin ply into the middle position, when the long end of the new strand is doubled forward to join it.
The braiding goes on until the required length is reached. When the supply of rolled strands is used up, the end of the braid is stoppered by rolling the three plies together on the bare thigh with the right hand or by tying the two outer piles together in the first part of a reef knot. A fresh supply of strands is then rolled from the fibers of the mato fi hank and the braiding continued from where it ended.
At the end of a sitting, the worker measures the braid by holding one end with the left hand and running it through the right as he stretches the arms to full length. The full arm span is called a ngafa. The right hand holds the farthest point of the first span and draws it into the left hand which seizes the point. The second span is run through and so on until the number of spans or ngafa are counted. To lay by as reserve stock, lengths of one hundred ngafa are plaited.
The ordinary three-ply braid described above is used for lashing houses, canoes, and for general purposes. The braid looks loosely made and close inspection reveals two technical details which prevent a neat appearance: 1, the transverse turns of the fibres fixing the individual strands (fa'ata'a) show up on the plies of the braid; 2, the overlap of the strand joins (so'onga) can be detected. Quickness in manufacture and efficiency in use are, however, the guiding principles in the braiding of this type. The first detail keeps the plies of the strands together and assists quick work in braiding, and the second strengthens the braid which is subjected to firm pulling as it is applied. The average three-ply braid is shown in Plate XX, B, 3; a thicker braid in Plate XX, B, 4.
A neater braid, though smaller (Pl. XX, B, 2) half the size of the average type, is made by rolling the strands on the bare thigh without the addition of a transverse binding fibre, and omitting the doubled-over join. (See fig. 131.)
The join. The simple overlap and rolling together on the bare thigh with fau songa is not suitable with sennit fibre owing to the stiffness of the material.
The method of joining in fine braids consists, therefore, of bringing the short ply to the middle position and adding the new strand from above with the slightest projection backwards of the short end. The braiding is continued and the short ends in the middle line subsequently trimmed off short. There is thus no overlap to be seen and only the closest scrutiny reveals the cut off ends of the fibres. The same applies to the ends of the short plies. If some of these fibres stick out, they are also trimmed off. It can now be understood that
When the work has not proceeded far, small coils are made on the palm by winding the turns round the thumb and little finger. The coil is removed, pressed together at the middle and bound by passing the working end round it in a couple of half hitches. When the length of braid is too great for this method, it may be coiled round both thighs with the knees diverged as the man sits cross legged on the ground. A third way is to collect the braid in long loops of one span each as the plaiter measures his work. The end of the braid is tied around one part of the circumference of the coil. On recommencing plaiting, the working end is unloosed but the coil kept intact by tying some fibres around it. If a man forgets the length he simply counts the number of the complete turns in the coil which are each a fathom in length. (See Plate XIX, B, 2.)
The working material consists of the matofi hanks of fibre, a bundle of rolled strands (fa'ata'a), and the coil of sennit braid ('afa) that is being plaited. Combined, matofi, fa'ata'a, and 'afa constitute what is termed a to'oto'o ali'i (Plate XIX, A, 3) figuratively the staff (to'oto'o) of a chief (ali'i). As they are combined together for one purpose, they are used figuratively to denote unity of purpose in the saying, "E pei o le to'oto'o ali'i lou finangalo" (May your will be as the working material of sennit braid).
The rolled strands not used up at the end of the sitting, together with the matofi hanks, are tied with fibre to the coil or to the working end of the braid so that they will be together and not dropped as the chiefs pay visits and carry their sennit work with them.
Working coil. Carpenters working on a house or a canoe use small coils, (i'o fanga), as they are more convenient to work with than large coils or shorter lengths of loose sennit. The coils are made in two ways: a, by the simple method when the coil is wound on the palm round the thumb and little finger, pressed together to form an end, and, as close as possible, to the end, a few transverse turns are made with one end of the braid and finished off with one or more half hitches (Plate XIX, B, 1); b, by the crossed method when the coil is wound on the palm by diagonal turns round the thumb and little finger, carefully removed, and the first end of the braid located lest it be hidden away by subsequent lashing. The coil which forms a figure-of-eight is doubled at the crossing. The crossed end is closely bound with the outer end of the braid and finished off with a half hitch.
In using the coil, the first, or inner, end is pulled out and tied to the object with the running noose commencement or buried under overlapping turns. As the turns of the binding are made, the braid pulls out of the interior of the coil. The coil is easily handled and may be thrown over horizontals when
The making of these working coils is part of the duty of the house owner whilst the building is going on. I saw Misa at
The crossed method gives a shorter more compact hank. It is used nowadays with the very long imported lines used in deep sea fishing. In forming the coil, the line is wound diagonally over the two knees for its full length. The inner end is pulled out and the hooks and sinker attached to it. The line is then payed out until it touches bottom and the extra line remains coiled. The coil is called i'o fanga (from i'o, coil, and fanga, to pull out).
Permanent coils are those made to be stored away for future use. Each householder plaits continually in his spare time to lay up a reserve stock. The coils are then wrapped up in mats or baskets as a ta'ui bundle and stored on the fata shelves of the house. No one would think of plaiting sennit only when immediately needed. The permanent coils form part of the household property and wealth. The different forms of coils are as follows:
Loose coils (fanganga). The ordinary coil in fathom loops is used for shorter lengths in small tasks. (See Plate XIX, B, 2.) They save the necessity of breaking into the longer coils and thus spoiling them. They are also useful for satisfying people who come to borrow, and thus saving the larger coils which are kept concealed in their wrapping.
Ball ('afa tangai). The sennit is wound around a stick or some folded lengths of bark cloth into a round ball. Synonymous with tangai is ta'ai, to wind round.
Ornamented ball ('afa fa'aulu po'o, or afa manu lapotopoto). A ball is coiled to form a regular geometrical design on its surface. Some longitudinal lengths are first folded and the braid wound round the middle part to commence the ball. In the second name, manu means the sennit design and lapotopoto, globular. One of the longitudinal turns is usually prolonged to form a loop by which the ball may be hung up.
Cylindrical coil ('afa mamanu, or 'afa manu fa'aso'a). The cylindrical coil is the best and commonest form for a large quantity of sennit. (See figure 132.) In the name 'afa mamanu, we again have the geometrical design (manu), whilst in manu fa'aso'a, besides the manu design, attention is drawn to the cylindrical shape which resembles the collar beams (so'a) of a house.
The coil figured in Plate XIX, B, 3, shows the appearance, but some coils are much larger than this. It can thus be readily realized why the 'afa tanganga type to give away to borrowers to save his more elaborate 'afa many fa'aso'a.
The method of coiling (fig. 132) has been adopted by higher cultures for commercial purposes in coiling string. A piece of cardboard curved into a cylinder or truncated cone, takes the place of the longitudinal folds of sennit and in the example examined the ascending spiral turns were laid along the upper side of the preceding turns and the descending turns followed on the lower side of the preceding turns.
In addition to the types of coils, sennit braid may be simply wrapped up in a mat without elaborate coiling. Old sennit removed from dismantled houses was so treated and stored on the fata shelves. It was then called a ta'ui 'afa from being bundled in a mat to form a ta'ui. If merely put in a basket, it is termed 'ato 'afa (basket of sennit braid).
A four-ply round plait seen at Asau was used in a fishing line and besides being called by the descriptive name of 'afa langa fa (sennit of four plaits) was locally referred to as fili anufe from its roundness resembling a worm or caterpillar (anufe). (See Plate XX, B, 5.) Many Samoans seem to have forgotten the four-ply plait.
The technique is really a check plait on the round. The rolling of fibres into strands and the joins are the same as in the neater three-ply braids. (See figure 133.)
Another four-ply braid ('afa tua fa) is made by fixing one end and working towards the body, as it is easier with this technique to take the outside plies and work them over to the middle. A four-ply resembles a three-ply braid except that on one side, the outer ply passes over one ply to the middle whilst on the other side, the outer ply passes over two plies. (See figure 134.)
Five-ply sennit braid ('afa tua lima) is made for hand trolling lines and other purposes where fine braid is needed. Some braids are thin and finely made (Plate XX, B, 6), others thicker and stronger (Plate XX, B, 7). The technique is again similar to that used in making three-ply, except that the outer ply from either side crosses two plies, instead of one, in passing over to the middle position. (See figure 135.)
A five-ply braid made from fau songa is shown in Plate XVIII, C, 6. The specimen figured is reported from Samoa but I did not see any in use. fa tangi) were used and it is presumed that the fau songa braid was used with them. Five-ply braids are thicker at the sides than in the middle.
Two-ply twist (maea fu'a lua). Rough ropes of fau bast are quickly made. Little attention is paid to straggly ends. (See Plate XX, A, 1.) New strands are added by doubling down the short end on the other ply as in the sennit two-ply cord. The joins are easily detected from the wide strip of bast showing the change of direction where they cross to the other ply. No extra scraping is devoted to the material. The individual plies are twisted to the right and then crossed over the other ply from above downwards and to the left.
A better rope is made by scraping the bast more thoroughly and plaiting with a tighter twist. (See Plate XX, A, 2.) The new strands are added by
The better two-ply ropes were used in the house as lines between the walls on which to suspend dividing curtains or mosquito curtains. In lafo to distinguish it from the milo and fili processes.
Three-ply braid (maea fili tolu). A fairly thick three-ply braid rope is made from fibre in the same way as three-ply 'af a except that the end is tied to a stake or coiled round the big toe. The plaiting is therefore towards the body and the outside plies are brought over the middle ply to take its position. Thicker strands of fa'ata'a are rolled to the size required for a ply. The simple overlapping join is used in adding new strands as well as the doubled over join. A rope seen in the plaiting process in
Three-ply twisted ropes. Three-ply twisted ropes were stated not to have been made of fau bast. In the fau) were regarded as the strongest ropes and, in their mythology, was the type of rope used by Maui in snaring the sun. In Samoa, when extra strength was required, the material used was coconut fibre.
A three-ply twisted rope now in common use in Samoa (Plate XX, A, 3) is, however, of foreign make and material but is shown as it is now used in making the shark nooses still in common use.
Shark rope. The proper Samoan shark rope (maea noa malie) is a three-ply twisted rope in which each ply is formed of a number of strands of the common three-ply braid ('afa). popo) with three longitudinal grooves cut to correspond with the plies. This is inserted under the plies close to the upper binding around the strands. The three assistants then twist their sticks in the same direction so as to twist the strands of their respective plies. As the plies become closely twisted, they are allowed to twist around each other to form the rope. The chief rope maker manipulates the coconut husk gage by moving it downwards as the assistants walk around in
Three-ply composite braid ('afa'afalua). A composite three-ply braid rope is made of a number of strands of ordinary three-ply braid. One rope in A, 4, started with three, four, and five strands, but finished with four strands in each ply.
In these ropes, the old braid from dismantled houses was used over again. Short lengths could thus be economically used and as a particular strand in a ply shortened, a fresh piece was added by doubling over the short end in another ply. The rope, therefore, looked somewhat untidy, but was very strong. These ropes are used in lashing houses during heavy storms of wind.
Five-ply braid rope (maea tua lima). The five-ply braid rope (Plate XX, A, 5) is made in the same way as the five-ply braid illustrated in figure 135. The strands of coconut fibre are not rolled with individual binding fibres and the joins are by the simple overlapping of ends on the one ply. The rope besides being very strong is neat in appearance and finish. These ropes were seen in use as the bottom rope of large nets and the line for shark bait floats.
Plaiting is essentially a woman's craft and women plaited everything descibed except—according to my Fitiuta informant—the breadfruit cover (pulou 'ulu). Males, however, plaited certain articles without lowering their sex status. The articles plaited by men, in addition to the special breadfruit cover, were the two rough forms of coconut leaf basket ('ato) the coconut leaf thatch sheet, and the ridge sheet, the reason being that it was convenient for them to do so. The 'ato fil tasi and its later variant the 'ato fili tolu were both used by men to carry food from the plantations on a carrying pole. When no women were about, it was easier for them to cut a leaf from a neighboring palm and make a basket for themselves in a few minutes than to seek out a woman and so waste time. The basket and the pole belonged to men whilst the women more commonly used the carrying sheet strapped to the back. The young men did the cooking and it was convenient for them to plait the basket now termed 'ato fu'e umu to carry the cooked food from the oven to the dwelling or guest house.
House building was a male craft and it was convenient for men to plait the coconut leaf thatch sheet and the ridging sheet. They were often called upon to make them for emergency shelters. The making of the more tedious sugar cane leaf sheets they left, however, entirely to women. Thus articles likely to crop up in emergency were also plaited by men.
The fashioning of wall screens, food platters, and fish baskets required more skill and was left to women. The plaiting of the coconut leaf floor mat and all work in pandanus material belonged to women. The line was created by convenience and usage. There were no tapus to prevent men from plaiting other articles besides those mentioned. If they did, however, their own sex would regard them as effeminate or ask if he had no wife or female relatives. The plaiting of all forms of cordage was essentially the task of men.
Plaiting parties. Floor mats to furnish a guest house were sometimes made by a working bee of the unmarried women (aualuma) who congregated in a house which, for the period of working, was termed a fale lalanga (house for plaiting) from its use. The chief for whom the mats were made fed the laborers.
Begging parties. The alternative to the fale lalanga custom of making the mats locally is the tu'u papa, or tu'u fala custom of a chief accompanied by his talking chief and taupou visiting another village and obtaining the required quota by a levy on the various families. (See page 75.)
Braiding sennit. Attention is again drawn to the remarkable persistence of sennit braiding through its having been elevated to a chiefly custom for filling in time whilst making ordinary social visits, or attending meetings and even important fonos held in guest houses. The status of the occupation was recognized by terming the braiding material a chief's staff (to'oto'o ali'i) and incorporating it in the proverb already quoted. In marked contrast to the persistence of three-ply braiding is the rapid disappearance of the technique of other forms of braiding, such as the four-ply round and the five-ply flat. The person from whom the five-ply braid rope was obtained could not tell how it was made. The rope was carried round
The term used for a garment is 'ofu. 'ofu, the wrapping of food in leaves. The term 'ofu is now used to denote garments of foreign material as is also 'ie which originally signified a garment commenced with a plaiting technique.
The garment for everyday wear was formerly the titi kilt made of ti (Cordyline) leaves. Early voyagers and missionaries confirm the statement of present day Samoans that it was the only garment worn during the day.
On ceremonial occasions, a greater variety of garments were used which accorded with the rank and status of the wearers. For dances, besides the use of colored ti leaves, simple kilts were made of pandanus leaves and strips of fau and fanga i'o bast. With fau bast, the strips suspended from waist cords were also plaited.
A higher development with fau bast, were the short kilts ('ie) with hanging fringes that were commenced by plaiting. These again developed into the roughly rectangular garments ('ie fau) with a fine plait and an outer covering of tags. The same technique with the bast of the fau pata produced the white shaggy garments ('ie sina). The fine mats made of narrow widths of pandanus and most valued of all, were the 'ie tonga. All the types of 'ie were worn as kilts or skirts and were the special garments of the higher classes on ceremonial occasions.
Bark cloth (siapo) provided change garments for the evenings or was worn by women of higher rank. It was also used for wraps and other purposes besides clothing, but it never took the place of ti leaf kilts as an ordinary garment.
The term titi as applied to kilts is a reduplication of ti (Cordyline) from the leaves of which the kilt was made. The meaning has broadened out from actual material to the idea of a number of vertical elements suspended from cords tied round the waist. With other material than ti, qualifying words are used. Kilts made of fau and fangai'o bast become titi fau and titi fangai'o. The term titi, by itself, means a kilt of ti, but to make it perfectly clear when speaking of different kinds, it is a titi lauti. Here lau (leaf) is introduced before ti to prevent such an awkward combination as titi ti. The more modern dance ornament of strings of feathers hanging from a waist cord receives the name of titi 'ula, from the red feathers ('ula) which formed the original material.
Kilts may be conveniently grouped into the two classes of titi and 'ie. In all titi, the strips are first attached to a braid or cords which form the waist attachment. Some types may be plaited afterwards. The 'ie commence with the plaiting technique and the waist attachment comes last.
Kilts of Cordyline leaf (titi lauti). Different varieties of ti are recognized, such as:
The ti fonua, from the size of its leaves, provided the kilt for everyday wear. It was also used as a wrapping for food to form 'ofu. Hence the connection of 'ofu with the commonest garments as pointed out by ti tongotongo and ti fangasa were used for dance kilts owing to their color. The ti'ula was better still and was worn by the village maid and the young chiefs at village festivals. The ti was planted near the village and about the houses. Hence the saying:
The ti grows in clumps and there are always some leaf heads that can be used. Hence another saying: "E fa'apupu a ti e le ngase" (A clump of ti will not die).
The green, growing leaves are ti usi and those that are turning yellow and have fallen to the ground are ti pala'au. The golden-colored leaves are used to form anklets and armlets or a garland for the head (ti palea).
The base ends of suitable green leaves are pinched through with the thumb nail or cut and this portion with the stalk ('au) removed. The leaves may be plaited whole or split (tosi) with the thumb nail in such a way as to leave them connected together at the base with an unsplit portion.
The commencement of the usual kilt is with a three-ply braid (fili). The braiding may be done with the butt ends of the leaves alone or three strips of fau bast may be knotted and Commence the three-ply braid as in figure 136. The latter is the neater commencement as it provides a cord at the commencement end for tying, though a piece of bast can readily be tied to this end afterwards. The braid connecting the leaves together, serves as the waist band of the kilt. The length to which it is plaited depends on the type of kilt required. The titi lauti fall into two types, the narrow and the wide: a, the narrow kilt (titi fa'ale'a'u) was designed to provide the cover demanded by modesty without unnecessarily concealing the figure. It was practically a narrow apron, used in various dances but especially in the almost naked poula dance at night. By using this form, the village maid revealed the perfections of her well made figure and the young chief (manaia) enabled the designs of his tattooing to be fully admired. Before important functions, portions of the young chief's tattooing were gone over again with a tattooing instrument and pigment to darken it and thus show it up more. The term 'a'u means to surround or to meet as with the ends of a wide kilt. By using the negative le, titi fa'ale'a'u means a kilt made not to meet. Early writers state that the narrow kilt was also worn by men as an ordinary garment. b, In the wide kilt (titi fa'atuso'o) more leaf strips were added to the plaited band until the hanging strips completely surrounded the waist when the ends were brought together. From so'o, to join, the name titi fa'atuso'o means a kilt made to join at the ends. The garment was always worn by women on ordinary occasions and by men as an alternative to the narrow kilt. The term sulu is to fasten on a kilt and sulunga titi is the part which is fastened. Women fastened it over the right hip, and men fastened it behind. Women preferred longer strips of material for their kilts so that they resembled skirts rather taunga loloa. Hence a woman's long deep kilt received the name of titi lauti, from the kind of material, titi fa'atuso'o, from the technique, and taunga loloa, from the length of the hanging strips.
Some kilts are made very thick by using two layers of leaves attached in opposite directions at the band in order to balance the oblique direction as they enter the braid. (See Plate XXI, A.)
The foundation of the waist band is four strips of bast plaited into a braid as in figure 137. Here the back-ply from the left crosses two plies and the ti leaves are added to the plies crossing from the left.
Kilts were also named after the variety of ti used; as, titi fonua, titi tongotongo, and titi 'ula. The titi 'ula of the red variety of ti was also called lauti 'ula (red ti leaves) to distinguish it from the feather titi 'ula. They were fringed for dancing, hence the saying as applied to the equipment for the dance: "Le fonga tele ma le lauti 'ula" (The big top knot of hair and the red ti leaf kilt).
The ordinary wearing ti leaf kilt lasted only a day or two. It was part of the women's duties to plait a new kilt for the menfolk, ready to put on in the morning. It is the kilt of green leaves that is meant by the name titi lauti. When dried and withered, it was called a titi pa'upa'u or titi mangumangu, both adjectives meaning dry. The discarded dry titi were often used by old men whilst weeding amongst the stones of the platform or loosely paved areas surrounding the houses. They sat on the ground and moved forward by sliding in the sitting position instead of rising. The sliding movement is called se'ese'e and the cast off kilts used in weeding thus received the additional name of titi se'ese'e. Hence when anyone attempts to make use of a person to do unpleasant work, the following saying is quoted: "A fai ea a'u mou titi se'ese'e" (Do you wish to use me as an old dry weeding kilt).
The dance kilts are sometimes dyed black when they are termed titi pala, tualua, or pa'anga.
The chief's name for a ti leaf kilt is salinga, savalinga, or nau. noa. He also gives fusiua as a general name.
The kilts of fau bast (titi fau) differ from the ti leaf kilt in not having the braid commencement (fili). The long strips of bast are attached to one or two horizontal cords, the ends of which are used to tie the garment fau strips to the cord is termed fatu. In technique, the fau kilt is termed titi fatu to distinguish it from the braided kilt (titi fili). The method of attaching the strips differs in the single and the paired waist cords. The waist cords may be two-ply twisted strips of bast or a single thick strip twisted on itself to form a round cord.
With the single cord attachment, the method is exactly the same as in making a kava strainer. (See fig. 138.) The two-cord attachment is shown in figure 139.
The attachment of strips is continued toward the right until the kilt is long enough to pass round the waist. The completed kilts thus consist merely of strips of bast hanging from a single or double cord and constitute the true titi fau.
The kilt figured in Plate XXI, B, is unusually long and has been colored yellow in a solution of turmeric root (ango).
Kilts of fangai'o bast. The bast of the fangai'o splits readily into very thin layers which have a fine, lacelike appearance from the open texture of the fibres. A particular looped ornamentation is made with this material to form hanging strips for dancing kilts. (See Plate XXI, C.)
The hanging strips end below in a tuft of long strips. The tufts are fastened to a vertical strand either by tying them at one end or looping the strand round the middle of the strip forming the tuft. The vertical strand serves as the basis for attaching the loop. The loop technique is termed fatu fa'afeti'i. (See fig. 140.)
The waistband commences with three fairly thick strands of fangai'o bast plaited into a braid. After plaiting for a few inches, a loop ornamentation is introduced into the braid. (See figure 141.)
The kilt is used in dancing but owing to the spaced nature of the hanging strips, it is put on over some other garment as additional ornamentation. This type of kilt is now usually colored with various foreign dyes and made largely titi fili
Pandanus leaf kilts. Dancing kilts (titi fala) are made of strips of pandanus leaf and though the material may be lau'ie they are referred to as titi fala. The kilt figured in Plate XXII, C, is made from the discarded under layer of lau'ie provided by the preparation of material for fine mats. It is thus really a by product. The strips are about 0.5 inches wide.
Two suspensory cords of two-ply sennit twist are used as a waistband but the pandanus strips are attached to them in a different way to that in fau kilts. (See fig. 142.)
The ornamental wide strips in the kilt described were dyed with 'o'a native dye and also with some foreign trade dye. Some strips consisted of the full leaf unsplit into layers and ornamented with narrow, dyed strips from the upper layer of the lau'ie which were sewn across in various patterns. Though the idea of sewing patterns is foreign, the foundation technique of the kilt is native. Such kilts have 110 high status, but in the making used waste material made available by some more important activity, thus saving labor.
The feather kilt (titi 'ula) is shown in Plate XXII, A. The feathers are attached to fine two-ply twisted cords in small bunches. (See figure 143.)
The titi 'ula is worn outside of other garments for dances and festivals. As the name 'ula implies, the feathers consisted of the red feathers of the parrakeet, usually obtained from titi 'ula is now applied to any feather kilt irrespective of color. The feather kilt is kept in a bamboo cage (Pl. XXII, B.) wrapped completely in a thin sheet of bark cloth and suspended from the roof with a long cord by which the cage may be lowered when required.
Another class of kilts, made from strips of fau or fangai'o bast, resembles the previous ones of that material in the form of attaching strips to one or two suspensory cords and are hence classified as titi fatu. They differ in the
The check plait type (Pl. XXIII, A) is commenced with the single cord attachment after which the cord is turned towards the worker and the strips plaited in check in the same technique as plaiting a mat. (See fig. 144.)
In the type kilt, the plaiting was continued for a depth of 6 inches and a width of 2 feet 3 inches. From the last plaiting edge, the weft ends were simply left free without any special technique. When worn, the kilt had a plaited waistband with the ends of the wefts hanging down as a fringe for another 1 foot 10 inches, making a total depth of 2 feet 4 inches. The free ends were further split and combed out into their individual fibres. The upper suspensory cords formed the ties.
The twilled plait type of kilt (Pl. XXIII, B) has come into common use for dances and trade. Though showing modern influences, the technique is native. In the type figured, two suspensory cords are used but with a different method of attaching the strips to that described in figure 139. The full technique is shown in figure 145.
The exterior surface of the plaited band is decorated in various ways which show foreign influence as regards motives and the use of sewing. The type kilt has a border formed by a wide strip of fangai'o bast dyed navy blue and folded in a zigzag as it was stitched to a strip of pandanus leaf with a sewing machine. This native ribband was stitched at intervals to the bast band with dyed bast. Rosettes and artificial flowers of dyed bast are also stitched on with bast threads.
Much ingenuity has been displayed in forms of ornamentation. Though stimulated by trade competition, the adaptation of foreign motives to suit native material and technique are worthy of note. By the use of foreign dyes, the native craftswoman hopes to attract buyers from the people to whose culture the dyes belong. She thus commits atrocities which the foreigner, for whom they were committed, mistakenly attributes to inherent error in the native sense of the artistic. The fact that such objects command a ready sale at
The 'ie class of kilt is distinguished by an important change in technique. The braid commencement (fili) and the suspensory cord attachment (fatu) are both discarded. Plaiting (lalanga) now assumes primary importance for 'ie). The garments are thus called 'ie to distinguish them from the suspensory class of kilt (titi) in which the plaited band of the last two types is secondary as regards technique. To distinguish the 'ie kilt from the full sized 'ie garment, the qualifying word pupu'u (short) is added to 'ie while tutu refers to the characteristic braided tails which stand out from all the edges, hence the full name of the garment, 'ie tutu pupu'u.
Owing to the plaiting commencement and all the weft ends being braided into tails, the garment has to be attached around the waist by a separate cord which has no structural connection with it.
The 'ie tutu pupu'u is the highest development of the kilt. They were made for young men and women of high rank to use in the ceremonial dances on important occasions.
The 'ie kilts are made of fau bast. The wefts usually consist of double strips which make it possible for one weft element to be turned in to define the edge of the garment while the other is left projecting beyond the edge to
'ie garments including kilts, shaggy cloaks and fine mats. (See fig. 146.)
The completion of the commencing corner provides an oblique working edge formed by a number of working dextrals. From the working section thus provided, a section of plaiting is carried towards the right for the full width of the garment. As the plaiting proceeds, the lower edge is continued and fringe elements provided. On reaching the full width, the right lower corner is formed as well as the right edge of the garment for the depth of the section. In forming the right edge, fringe elements are also provided. (See fig. 147.)
On completion of the first working section, another section is commenced on the left. The left edge has to be built up above the commencing corner, and in doing so, fringe elements have to be provided for. In the lower working sections, the fringe elements of the left edge are provided by the one element of the double sinistrals as they reach the edge. Higher up the left edge the fringe is formed by adding new elements. The right edge with the addition of fringe elements remains the same throughout. The body of the garment continues in check throughout, but when the garment reaches a certain depth, a row of fringe elements is run across the body from left to right. The methods of providing the fringe elements on the left and on the body are shown in figure 148.
When the garment reaches a depth of about 11 inches, the upper finishing edge is defined as the last working section advances from left to right. The finishing edge must be formed as in the other three edges by turning in one set of elements to form a turned edge, and leaving another set of elements to provide the fringe. The usual method is shown in figure 149.
An alternate finish to the upper edge, not so neat, is shown in figure 150. There the last working section was left along the upper edge in the ordinary condition of plaiting edges. What would have been an insecure finish is remedied by the braided tail technique.
With the ending of the last working section, the body of the garment is completed with all four edges clearly defined, and fringe elements projecting from all. The fringe elements are then finished off by plaiting them into three-ply braid tails.
The fringe elements are braided for a short distance, tied, and the ends left free as a fringe. (See fig. 149.) A different technique is used in the garment with the plaiting edge finish. (See fig. 150.) All the fringe elements
Three varieties of textile kilts are figured in the Plates, each with some variation in technique, dyeing or material.
The garment in Plate XXIV, A, was dyed in a mixture of black candlenut, lama, and brown 'o'a and afterwards smoked to darken it further. All the edges were made with the usual half turns and the braided tails followed the technique in fig. 149.
The second variety (Pl. XXIV, B) is a larger and better looking garment, but three edges follow the ordinary plaiting edge technique. (See fig. 150.) The lower edge follows the orthodox technique of turning up one element of each double sinistral (fig. 147), but on the left edge, the whole sinistral projects to form the fringe while new double dextrals are added from the edge with their ends also projecting to form a double set of fringe elements. On the right edge, the dextrals as they reach the edge are not turned but the whole wefts are left projecting as fringe elements. Fresh sinistrals are added from the right edge also with their ends projecting to form a double set of fringe elements. At the upper edge, the ordinary plaiting
edge is maintained. (See fig. 150.) The upper, left, and right edges had each a double set of fringe elements which make the braid tails much heavier in this garment than in the first variety. The garment was dyed with 'o'a to a reddish-brown color.
The third variety (Pl. XXV, B), made of fau pata bast, shows a combination of the two other varieties in the formation of the edges and the provision of fringe elements. The lower edge is orthodox in following the other two. On the left, a combination is made of the two methods shown in figure 148, a and b. Thus one element of the double sinistrals is left out as a fringe but when the other element is turned in to define the edge, a dextral fringe element is added to it as in figure 148 b. On the right, one element of the dextrals is left out as a fringe element in addition to the elements added in the normal way. (See fig. 147 c.) At the upper edge, only one element of
the dextrals is turned down on the working sinistrals instead of both (fig. 149) while the other is left out to form an extra set of fringe elements. The technique is superior to that in the other two varieties, in that it forms a turned edge on all four borders and provides a double set of crossing fringe elements at the right, left and upper edges. The outer surface is covered with tags arranged in transverse and oblique rows. (See fig. 151.) Besides the braided tails on all borders, the highest fringe row on the body of the garment is braided at either end. On completion, the garment is washed and bleached repeatedly until it is white. The tags are combed out to separate the fibres, which gives the garment a fleecy appearance.
The garments, on completion, form a plaited textile wide enough to pass completely round the waist and averaging about 18 inches in depth. Owing to the thicker fringe at the upper technical border with the resulting greater number of braid tails, and also the transverse body fringe near it, this border is turned downwards when the kilt is worn. The garment is tied around the waist with a separate cord and the fringed edge above turned forward over it so that all the braid tails hang down over the cord. (See Pl. XXV, A.)
Fine mat kilt ('ie lavalava). A dance kilt of the same material and plaiting as the fine mats is called 'ie lavalava. (See Pl. XXVI, B.) The sinistrals on the left and the dextrals on the right are turned in to define plain side edges. At the upper border, the wefts are simply left free along the plaiting edge to form a fringe of natural weft ends. At the lower border, a similar plaiting edge was formed. Here both dextrals and sinistrals are taken in small groups, divided into three, and plaited as free braids for 3 inches. The two outer plies are then knotted and the free ends form a 12-inch fringe. To the lower ends of the braided tails, red and green feathers are tied. Other ornamentation consisting of a marginal strip of bast and strips of lau 'ie to form side fringes have been sewn on with a sewing machine and belong to the period of over-ornamentation created by foreign contact.
The larger garments, generally termed shaggy mats ('ie fau and 'ie sina), are a further development in technique to the short textile kilts. Though about the same width as the larger 'ie kilts, they are much deeper and may thus be termed skirts rather than kilts. All the edges have fringes of the
The material consists of the bast of the fau tu (Hibiscus tiliaceus) and the fau pata (Cypholopus macrocephalus). The fau tangaloa (Hibiscus tetraphyllus) is also used. The wefts are narrower but even in one garment there is considerable range in width. The plaiting stroke used is the check and all garments commence at the lower left hand corner with the typical corner commencement free of fringes seen in the 'ie kilts. (See fig. 146.)
The garments are divided mainly into 'ie fau and 'ie sina according to material, and varieties are also named from treatment to obtain change of color.
The brown shaggy garment ('ie fau) is made of strips of fau tu with fairly coarse wefts. Though in parts, not much finer than the 'ie kilts of the same materials, more care is exercised in the plaiting. The garment figured in Plate XXVII, A, is a good specimen. The shaggy outer appearance is produced by the close attachment of tags, but the turned-down corner shows the check plaiting technique. The garment is roughly rectangular, but the edges are never of even width owing to the difficulty of controlling exactly the width and thickness of the wefts, which also stretch. The wefts are double but alter
Plaiting commences with the clear lower left corner made in exactly the same way as in 'ie kilts. The plaiting of the first working section follows the same technique except that on the lower edge, in addition to the fringe elements provided by the double sinistral wefts, a second series of crossing fringe elements is added to the dextral wefts. (See fig. 152.)
The full width of the garment having been made, the right corner is turned and the right edge continued upward to complete the first working section. In both processes, the technique is the same as in 'ie kilts, except that a second set of fringe elements is provided on the right. (See fig. 153.)
In commencing the next working section, the left edge has to be carried upwards from that part already formed by the left commencing corner. The technique is again identical with that in textile kilts except that a fresh set of fausa) to prevent their becoming entangled. The fausa are formed by twisting a group together loosely, folding them backwards and forwards from the top ends and catching the folds in a half hitch from the slack on the end towards the plaiting edge. All wefts are thus treated. The two fausa on the right of the working edge are released and folded again on the higher level on the left when the working edge reaches the next two fausa on the right. (See Pl. XXVIII, A.) The working edge is fairly long, but a short one is shown in figure 155 to emphasize the identity of the plaiting technique with the ordinary technique used with coconut leaflets and coarser pandanus wefts.
The feature of the shaggy garments is the close covering of tags over the whole of the outer surface. The tags consist of fibrous strips of the same material as the wefts and they are placed in position at the working edge during the process of plaiting. The attachment of tags thus commences with the first working section and is continued throughout with each successive working section. Three methods of adding tags are in use:
(1) The first method consists of leaving out a length of new sinistrals as they are placed in the shed at the working edge. The method has been described with textile kilts in figure 148, c. Though used as a fringe in textile kilts the method is also used to augment the tags in shaggy cloaks. (See fig. 155) The upper ends of shortening sinistrals may be left out also on the upper surface of the garment and the same applies to both ends of dextral wefts. Thus, where in ordinary plaiting the ends of wefts are cut off close to the edge of the last crossing weft, in shaggy garments they are purposely left long to assist in ornamentation.
(2) The loop method of attaching distinct strips seen in kava strainers and the single cord attachment of kilts (fig. 138) is also used in shaggy garments, in which the looped tags are attached to sinistral wefts at the working edge. A horizontal line of tags so attached is shown in Plate XXVII, B, and the detail in figure 156.
(3) The third and most-used method is attaching the tag by two turns around dextral wefts at the working edge as in figure 157.
When the plaiting with double fringed edges and tag covered upper surface has reached within one section of the end, special technique has to be called
The knotted braid tail, which marks the ending of all the technical processes used can always be found at one of the corners of the garment placed in the proper position for study. (See Pl. XXIX, A.) The fringe elements at all four edges are left as tag fringes without braiding.
The white shaggy garment ('ie sina), made of fau pata bast, is a finer garment than the 'ie fau type made of fau tu. The material is scraped with a shell, dried, and split into wefts narrower than the fau tu. The technique is exactly the same from commencing corner to braided tail finish as in the previous garments. The tags are put on more thickly with the two-turn method. The working edge in an unfinished garment is shown in Plate
A. In some long, narrow garments, instead of finishing the long lower edge, a short working length is made, the complete left edge is formed with its upper corner and the garment turned with what will ultimately function as the left edge towards the worker. The plaiter then works across the narrower depth of the garment to prevent the use of a greater number of fausa, which leads to confusion. The technique and finish is the same as described, but, owing to the position while plaiting, the braided tail finish will be formed on the right lower corner instead of the right upper.
The completed garment is bleached by soaking in sea water and drying in the sun. Repeated washings in fresh water with the leaf of the fisoa (Columbrina asiatica) as soap results in a very white color. The white textile kilt (Pl. XXV, B) was treated in the same way. The fringes and tags are combed out with dry coconut leaflet midribs. The completed garment is bleached and not the unplaited material as
The garments in B, the thick outer coating of tags has not been combed out well and it lacks the whiteness resulting from repeated washings. The garments are heavy and clumsy in appearance but are well made, considering the material.
Attention has been drawn to the tags being attached during the plaiting of the garment as they are more easily wound round wefts which have one end free. The subsequent plaiting fixes them in position. 'ie sina in the
The 'ie fau retains its natural color, which deepens to a yellowish brown ('ena'ena). As no artificial methods were used to change its color, the garment retains the name of the material in 'ie fau. The 'ie sina is made of fau pata but as the garment is bleached white by a special process its gets the name of 'ie sina from its white color (sina, white). Two other varieties of shaggy cloak are distinguished by a red and black color respectively but the names given to them are derived from the coloring process.
The red shaggy garment gets the name of 'ie ta'ele (ta, to treat; 'ele, red earth) from being stained with a mixture of red earth in a wooden bowl.
The material appears to be fau tu. The technique is identical with the preceding shaggy cloaks and the dark red color gives it a fine appearance. (See Pl. XXIX, A.)
The black shaggy garment ('ie fuipani) according to uliuli) by being pressed down in the mud of a swamp. It is hung up to dry but away from the sun. Before a dance, the garments are sprinkled with coconut cream to make them shine. Samoans distinguish between drying an article in the sun (fa'ala) and drying in the air or a breeze away from the sun (fa'asavili). Direct heat from the sun would bake the mud on the stained cloak and cause it to flake off instead of soaking into the material. The term savili denotes a breeze as against a wind (matangi). No example was seen but the technique was probably identical with that of the three preceding varieties.
The shaggy garments are worn around the waist as skirts and kept in place with a cord or bark cloth belt. They are heavy and uncomfortable to wear for any length of time in a tropical climate, but are used only during ceremonial to denote rank and status. The 'ie sina was essentially the garment of the village maid and women of high rank. It was also used in the proof of virginity custom. In social value, the 'ie sina is almost, but not quite, in the same class as fine mats. Owing to their being less in number than fine mats, they are sometimes preferred during ceremonial to demonstrate that the family possesses them.
Garments occur which contain elements from different types. They may be important as indicating the natural progression in technique that took place as craftswomen sought to evolve better types of garment as a means of expressing social distinctions.
A kilt (L. 1571) in the 'o'a dye. (See figure 159.)
Another kilt (Pl. XXIX, B) shows the bottom left-hand corner commencement and body fringes characteristic of garment kilts but the body wefts at each upper corner are plaited into braids for tying around the waist.
A third kilt (lau'ie pandanus strips, follows the textile kilt technique of providing tags on the outer surface and the fine mat technique in the left lower corner, lower and side edges. The treatment of the upper edge is unique in that from a plain plaiting edge, the dextral wefts are plaited at the back into a three-ply braid running from left to right and continued beyond the right edge to form a free braid tail to be used as a tying cord. The sinistral wefts are braided from right to left
The term "fine mats" ('ie tonga) applied to articles plaited with thin narrow wefts in check and usually ornamented with fringes and red feathers; has been in use so long that it cannot be abandoned. The mats, however, are worn as skirts and thus come under the category of clothing.
The fine mats are made of lau'ie, the leaves of the cultivated 'ie, a pandanus distinct from fala and paongo. In lau'ie is said to have been first brought to Falealupo by Nafanua, who subsequently became an important war goddess.
The serrated edges of the leaves are removed and the spines on the back of the midrib trimmed off. The upper shiny surface of the leaf is the alo and the duller under surface the tua. Each leaf is cut transversely across on the under surface about two inches from the butt end. The cut passes through the under layer of the leaf and this tua, under layer, is peeled off between the cut and the butt end, leaving the thin upper, alo, surface intact. After a short exposure to the sun the leaf is folded in two-foot lengths and a convenient bundle made by folding other leaves longitudinally around it. The bundle is tied around the middle with a strip of fau bast.
The bundles, protected by a layer of green leaves from direct contact with the hot stones, are cooked in an oven for about half an hour, The leaves from the cooked bundles are folded around the hand, reverse to the previous folding, to straighten them out. The thin, shiny alo layer is peeled off (fofo'e) by grasping with one hand the butt part, already peeled (Pl. XXX, A, 2), and separating it sufficiently to allow the other hand to grasp the lower layer. Owing to the cooking, the two layers readily separate by pulling, but care must be exercised as the upper layer is liable to tear at parts along the midrib that form the sites of spines. The parts that stick are separated now with a metal knife, but formerly with the edge of a bamboo strip. The coarser under layers are discarded while the thinner alo layers, which form the plaiting material, are strung together on a three-ply braid. (See fig. 160.)
The braid with the leaves is termed filinga lau'ie from fili (to braid) and lau'ie (the 'ie leaves). During the plaiting, it was interesting to observe how the foot with its grasp of the braid end between the first and second toes stretched the braid taut to facilitate the plaiting. When the braid became long, the foot was shifted up on the braid to get a shorter grip. Civilized man deprived himself of extra hands when he confined his toes in boots, whereas the barefooted Samoan uses his toes to assist his fingers in many of his crafts.
Soaking in sea water. The braid is tied to a stake in the sea or weighted down at the ends with rocks. After soaking for a fortnight to bleach them to a lighter color (pa'epa'e), the leaves are cut off near the braid and dried in the sun. The braid with the butt ends is discarded. (See Pl. XXX, A, 1.)
Rolls. The leaves are split longitudinally down the midrib and the halves wound around the fingers with the inner split surface outwards. Conveniently sized rolls, much smaller than fala, are compressed to an ellipse and tied with strips of the same material. In this form, they are stored for use. (See Pl. XXX, A, 3.)
Preparing the wefts. The implements required are a scraping board and a shell scraper.
The ordinary scraping board (papa valu lau'ie) is a slab of wood about 2 feet long, 0.5 inches thick, 4 inches wide at one end and 3 inches at the other. The working side is worn down perfectly smooth.
The shells used are 'u'u (Mytilus sp.). The unsplit leaf is scraped with large shells, whole, or broken to provide an edge, while individual wefts are scraped with very small shells.
The half leaves are unwound from the roll and straightened out by reverse rolling around the fingers. They are laid on the board with the shiny outer surface downwards while the other surface is scraped to remove extra material. Any creases in the leaf are also removed by the scraping. The leaves are split into weft widths with a sharp point, such as the outer spines of the porcupine fish. The splitting always commences a short distance from the butt end and is run out to the tip. The unsplit butt end thus keeps a number of wefts together as in splitting fala leaf. The width of the wefts varies with the skill or desire of the plaiter. In a number of fine mats examined the wefts ranged from 12 to 14 to the inch, but some are finer. The pick of the material is used for the best mats while that not quite good enough (auaunga) is used for second-class mats.
Plaiting. The wefts are always double. The true outer surface of the material is smooth and shiny but the other scraped surface is duller and shows the longitudinal striations of the fibres. The double wefts are placed with the dull surfaces together so that the bright even surfaces will be exposed on both sides of the completed mat. Some individual wefts are formed by folding longitudinally instead of splitting. When plaiting is established, the long narrow wefts are folded into fausa groups as in shaggy mats. The small 'u'u shell scrapers are used to straighten out the working wefts and a porcupine fish spine for picking up the raised sets at the working edge.
The left commencing corner (Pl. XXX, B) is usually formed by the same technique as in shaggy garments but an alternate method of turning the corner with one weft is also used. The side edges are turned without providing fringes and no ornamental tags are used on the body. Both the lower and upper edges have two sets of crossing fringes but, besides defining the edges by the half-turns of appropriate wefts, an additional technique is introduced of using a fixation braid at both edges. (See fig. 161.)
The fixation technique, giving a better finish to the lower edge, is really a three-ply braid. A wider technique forming a five-ply braid and a method of forming a turned-back fringe are shown in figure 162.
The three methods of lower edge finish by thickening the lower edge also strengthen it. The right corner is turned by either of the methods used with the left corner. The right edge is defined and continued upwards for the depth of the working sections in the usual way of turning in the dextrals as they reach the edge. The body is completed by a series of working sections.
In a mat observed during plaiting, the working edge consisted of 64 dextrals divided into two sets. New wefts are formed individually by overlaying and dropping the end of the shortened weft on the under surface after the new weft has been fixed by a number of plaiting movements. The lower ends of the new wefts are on the upper surface of the mat. The loose ends are cut off close under the edge of a crossing weft for concealment, and trimming may take place at any stage or after completion of the mat.
The left edge of the last working section is built up, the left upper corner formed in the same way as the other corners and the upper finishing edge dealt with as in figure 163.
The upper edge is carried along with the three-weft braid finish and the right corner finished off with the three-ply braid tail as in the shaggy cloak garments (fig. 164). (See Pl. XXX, C.)
The knotting of the braid tail marks the completion of the plaiting technique, which may have taken months and even years of careful work, sometimes involving strained eyesight and damaged vision. It is little wonder
Owing to unavoidable slight variations in weft widths, the side edges are never quite straight and the completed garment never exactly rectangular.
The fringes at the upper and lower borders range from 6 to 8 inches in length. (See Pl. XXX, B, C, D.) Unsplit butt strips are sometimes found ill the lower fringe, due to the split individual wefts having been added as the dextrals and the united part not being subsequently split.
Ornamentation. The technical upper border is further embellished with two forms of ornamentation; triangles or pointed strips of the prepared weft material and rows of feathers. 1. The forms and technique of the weft material ornamentation are shown in figure 165. In one mat, a row of wide strips was sewn on about 0.5 inch from the edge and the lower ends split to augment the fringe. The lower technical border is left plain, but one mat was seen with a row of ornamental triangles at both fringe borders. 2. A fine mat is not considered perfect without a row of red feathers along the technical upper border. (See Pl. XXX, D.) The red feathers of the Fijian parrakeet (Lorius solitarius) were those used. They were probably obtained originally by barter and as reciprocal presents, but
The feathers are knotted along a thread in exactly the same manner as in preparing feather kilts. (See fig. 143.) The thread carrying the feathers is laid along the base of the triangular ornamentation and attached by stitches which loop over the thread and pass through the mat. In most mats the feathers form a continuous line, but, sometimes, owing to a scarcity of feathers, short lengths are used with gaps between. (See Pl. XXX, D.)
The fine mats are worn around the waist as a skirt, doubled so that the technical finishing edge hangs down with the red feathers in front. The two lower edges are adjusted so that the fringe from the edge at the back reinforces those of the front. In a mat made with the triangular ornamentation on the opposite edge and opposite surface to the red feathers, the edge at the back is allowed to hang down a little lower and show the triangular ornamentation below the interior fringe. The mats are kept in position usually by a fusi girdle of bark cloth. (See Pl. LIV, C.)
The Samoan name for bark cloth is siapo. The term tapa used in kapa), is not used as a general name for the material. tapa as "One of the white borders of a siapo.? In tapa means the border of a garment. The Samoan use of the word is applied to the border rather than to the material.
Before the contact with Europeans, and indeed for some time after, the use of
siapoas an article of dress was confined to a few unmarried females of the highest rank,O , titled ladies; all others being prohibited from wearing it upon pain of heavy chastisement. The privileged few only wore it in the house. For a long time past the rule has been broken through, andTausalasiapois now worn by all persons of either sex.
siapo is manufactured merely to supply curio dealers. No doubt the outside demand by tourists and dealers leads to extra material being made. The continuance of manufacture is also due to the persistence of certain social customs and needs. During the siapo. The same applied to women during the ta'alolo food ceremonies. At a wedding ceremony in siapo figured among the wedding presents. For ordinary presents to visitors, siapo has taken the place of fine mats. In siapo pieces at feasts, weddings and funerals. The giving of siapo with ceremonial kava drinking and the high chief's sua portion of food still persist in custom. A talking chief is not properly dressed unless he has a kilt of siapo as well as his orator's fly switch. Sheets are still used as partitions in guest houses and as bedding, while foreign influence is to be seen in their use as table covers in the various Samoan homes.
Though discontinued in most parts of Polynesia, bark cloth making is still an active craft in Samoa. Though there is evidence that some of the dyes have been forgotten and the wooden tablet (upeti) is displacing the original article made of pandanus leaves, the technical process remains the same. There are no lost secrets of the craft and the lack of exact detail in the Report of the Wilkes Expedition, that was deplored by upeti tablet for rubbing instead of printing.
The plant universally used is the paper mulberry (Broussonetia papyrifera). The name in common use is u'a, while the other name of tutunga is unknown to many of the younger people. The name u'a is also applied to a young breadfruit tree and to a net prepared from its bark. Tradition states that the plant was brought from tutunga to Fonga-savaii and Faiaai in the Salenga district of
The use of any plant except u'a is now denied, and breadfruit bast is restricted to making cords for a type of seine net. The paper mulberry,maunu), is cultivated in the food plantations. Saplings are used while a second growth (tuapipi) is also utilized.
Peeling the bark. The bark was peeled (sasae) from the wood (aumafuti) and the process termed sae u'a. The name for any bark is pa'u but that of the paper mulberry was also referred to as u'a. The saplings now generally used are about as thick as the thumb, though smaller sticks ('auli'i) may be used to provide material for patching holes. Large saplings are unsuitable as the bark is too tough to beat out into sheets.
The bark is separated at the butt end, usually with the teeth, and a short longitudinal slit made down one side. It is said that the longitudinal slit must be on the convex side of any curve in the stake. The reason given is that the texture of the bark is more complex on the concave side. The bark as it is pulled off, will split cleaner on the convex side, whereas on the concave side the edges are left ragged. The bark is worked down from the butt until the left hand can grasp the freed bark. Holding the wood with the right, the bark is pulled away from the wood in a long tube which splits on the convex side from the initial cut. Towards the tip end, care has to be exercised as the projecing buds of leaves adhering to the wood tend to split the bark on either side of them and leave portions attached. Any such tendency is watched and the base of the knob pushed off with the thumb nail to free it with the bark. When the strip becomes too small, the bark is cut off.
Peeling off the bast (fofo'e). The bast is usually separated from each strip, immediately after stripping it from the wood. The inner bark or bast has no specific name but is referred to as pa'u pito i totonu (the part of the bark on the inside), as against pa'u pito i tua (the part outside). The strip of bark is wound around the left hand, commencing with the butt end with the inner surface outwards. The somewhat tubular strip of bark is thereby opened out and flattened.
The strip is opened out in its length and a transverse cut made through the outer bark near the butt end. The outer bark is lifted up, with the edge of a knife or a piece of bamboo, until it separates across the width of the strip and sufficiently down its length to allow the left hand to grasp it. The bast is held down on the right thigh by the right palm while the left hand pulls the outer bark away from it. The two separate quite readily along a natural plane of cleavage but care is exercised in separating with the thumb nail any portions of outer bark that tend to stick to the bast. This applies especially towards the tip end where the leaf buds are. When the tip end has run out, the strip is reversed and the portion of the butt end on the butt side of the tranverse cut dealt with similarly. (See Plate XXXII, B, 1.) During the separation, the strip is kept taut between the two knees, or legs. The outer bark is discarded and the bast thrown into a bowl of water. Each strip of fofo'e (to peel).
Scraping the bast. The scraping of the bast (fai u'a or fafai u'a) requires a scraping board, a number of shell scrapers, and a strip of bamboo for a knife.
The special scraping board (papa fei u'a) is dubbed out of iliili or other wood and was formerly smoothed clown with coral. An average sized board is 37 inches long, 18.5 inches wide, and about 0.75 inches thick at the edges. Sometimes a piece from the side of a canoe is used.
Shell scrapers (Pl. XXXI, B) are referred to generally as 'asi though 'asi is also the name of a specific shell. Three kinds are used: Asaphis violacea, Antigone reticulata, and a species of Area. The names applied by the Samoans vary both in districts and with the particular use of the shell.
The 'ofe knife is merely a strip of dry bamboo about 0.25 to 0.5 inches wide and of any convenient length.
Scraping the bast is clone in a stream when such is available. Otherwise a wooden bowl containing water forms part of the equipment. The board is propped up at a convenient angle against a stone or other support. The lower end rests on the ground where the worker seats herself. Though the whole process is termed fai u'a, the actual scraping is valu u'a. There are usually four stages of scraping and rubbing and a different shell is used with each stage.
Scraping (valu u'a). A strip of bast is placed longitudinally on the middle of the board with the butt end towards the top and the surface that was next to the outer bark uppermost. On this surface, streaks of green coloring matter adhere and the first scraping is to remove them. In Asaphis violacea (pipi) is used (Plate XXXI, B, 1.) In Arca is in vogue (Pl. XXXI, B, 2) except at Safune where the pipi is common. The Arca shell so used is termed 'asi valu. A shell full of water is dipped out of the bowl and dropped on the upper part of the bast strip so that it runs down over it. The scraping is made with backhanded movements upwards and away from the body, the outer surface of the shell being towards the worker. Every now and again a shell full of water is dropped on the bast. When the green coloring matter and coarser particles of outer bark have been scraped off, the other stages of the process are completed before another section of the same strip is moved up. (See Plate XXXI, C.)
Rasping (mangeo or pae). The next stage consists in rasping the surface of the bast with the rough outer surface of an Antigone reticulata shell. (See Plate XXXI, B, 3.) In mangeo and in pae. In each area, the process receives the local name
Smoothing (pae or fa'amalu). This process consists of smoothing down the rough surface left by the rasping and at the same time removing any loose particles or ends. Here there is some confusion between the methods of western and pae is an Arca shell that has had its slightly irregular natural edge ground to an even sharp cutting edge. (Plate XXXI, B, 4.) This edge is used as in scraping. It removes the loose pieces and irregular ends; The movements are up and out and water continues to be used. The process is called pae which in Antigone reticulata shell and the process of rasping.
In Arca but the back is used as in rasping. The back or outer surface is grooved but the ridges are not serrated as in the Antigone reticulata. It thus acts quite effectively in removing loose particles and smoothing down the bast surface. In fa'amalu as was also the shell used for the process, though the general shell name was 'asi.
Drying (to). In ta. The shell used was an Arca that had been ground from both sides to form an even, obtuse-angled blunt edge. (See Plate XXXI, B, 5.) Water was dashed on and scraped off to clean the bast. The last movements were applied with firm pressure to remove as much of the moisture as possible. The shell was held as in scraping.
In fa'amalu. In some villages, however, an Arca shell corresponding to the ta was used in the same way but was called langalanga.
After the section of bast has been subjected to the four processes, the strip is moved up so as to allow the next section to be done. In this manner, the strip is finished in sections. The last section with the end of the strip is reversed in direction so that the end is efficiently dealt with by the upward
In pipi (Asaphis violacea) and an 'asi (Arca) for the first scraping process and a langalanga as well as a fa'amalu for the two last processes. She also used an Antigone reticulata rasp. Thus she used five different shells as against three in some villages and four in fai u'a process is completed as far as the individual strip is concerned.
Each scraped strip is folded separately into quarter lengths with the internal surface of the first quarter uppermost on the board. The ta shell is used on the folded strip to express as much moisture as possible. From the manner of folding, the external surfaces will be to the outside both below and above. Other folded strips after scraping are placed above the first and the whole bundle again scraped with the blunt ta shell after each addition. The butt end is reversed with each alternate strip so as to make the bundle as even as possible. The bundle may contain from two to seven folded pieces. The topmost one is left as folded so that the outer surface of the bast is to the outside of the bundle both above and below. The Samoans attach importance to the method of folding. The bundle is called tapapanga u'a. ulututunga.
The pressure with the ta shell, besides removing any extra moisture, presses the strips closer together so that they appear as one piece. The tapapanga bundles are put under cover of an old strip of cloth to prevent their drying before the next process, which must wait until all the bast strips have been dealt with.
The term ta means both washing and expressing as much moisture as possible. It is nowadays applied to the washing and wringing of textile clothing.
For the important process of beating out the bast (tutu) an anvil and beaters are required.
The anvil (tutua). The anvil was usually made of toi wood as it gave a more musical sound when beaten. An example seen in use was 5 feet 7 inches long, rectangular in cross section, the upper and lower surfaces 8.5 inches wide and the sides 7 inches. The upper longitudinal edges were rounded off (See Plate XXXI, D.) Others seen in use were merely sections of a tree trunk.
A well made anvil in
The anvils were hollowed out to give a better sound for in olden days, women amused themselves at work by beating out various rhythms. They had many signals by which they could warn one another of the approach of strangers, and conduct a limited conversation. Anvils were also made of toa and ala'a wood.
Beaters (i'e) were usually made of pau wood but some were made of toa (Casuarina sp.). The Samoans classified them into smooth beaters (i'e mole) and grooved beaters (i'e tosi). In shape, they classify readily into two types, round and foursided. Of sixteen beaters in i'e teuteu. (See Plate XXXI, A.)
The round beaters are all smooth without any longitudinal grooves and are thus all i'e mole. They are longer than the other type, ranging from 15 inches to 17.75 inches with an average of 16 inches. In cross section, the beating part is usually elliptical though some may be circular. The beating part slopes gradually back into the handle, usually with no shoulder or abrupt change to mark any junction. The handle is much less in diameter than the blade and is usually flared out at its proximal end. (For the type beater see Plate XXXI, A, 1.)
The main variations may be summed up: The cross section is elliptical in four beaters, with the greatest diameter at the distal end ranging between 2.5 and 2.7 inches, the difference to the lesser diameter ranges from 0.2 to 0.4 inches. One was circular.
Between the beating part and handle, one beater out of five had a distinct shoulder 7.5 inches from the distal end, the beater being 16 inches long. The others had no trace as the slope was gradual.
The cross section of the handle was circular in three being 1.3 to 1.4 inches diameter. The largest beater had a difference in two diameters of 0.1 inch with the longest diameter 1.6 inches. The smallest had a difference of 0.3 inches.
Flaring of proximal end of the handle was present in two out of five.
The four-sided beaters are shorter ranging in a series of nine from 11.5 inches to 15.25 inches. They are practically square in section, with the four surfaces narrowing towards the handle. The slope from the beating part to mole), or grooved longitudinally on from one to three surfaces. In none of the beaters were all four surfaces grooved.
Smooth beaters (i'e mole). A typical beater (c. 354) not figured is 12 inches long. The cross diameters at the distal end are 2.3 and 2.2 inches. The end is cut off square. All four surfaces narrow gradually towards the handle and at the indistinct shoulder, both diameters are 1.7 inches. This point is 7.25 inches from the distal end. The handle narrows down to cross diameters of 1.5 and 1.35 inches. The proximal end of the handle flares out to cross diameters of 1.9 and 1.8 inches. The beater is smooth throughout and is 27 ounces in weight.
Of two other beaters in this sub class, one conforms closely to the type beater, but the other is very long. (See Plate XXXI, A, 4.)
Grooved beaters (i'e tosi). Of seven grooved beaters, one is grooved on one surface, five on two surfaces and one on three surfaces. The type beater is shown in Plate XXXI, A, 5.
The cross section of the beating part is practically square, there being a difference in the two diameters at most of 0.1 inch and in one case of 0.15 inches. Leaving out the special small beater, the distal diameter ranges from 2.6 to 2.9 inches with an average of 2.7 except for an abnormally large beater which has distal cross diameters of 3.2 and 3.1 inches.
All surfaces narrow towards the proximal end; none have parallel sides. By dividing the length of the beating part by the difference between the width of the distal and proximal ends of the beating surface a slope index is secured. Thus in the type beater (c. 760), the length of the beating surface is 7 inches and the difference between the two ends 0.7 inches. The index of slope is thus 1 in 10, which is a fair indication of the slope in these beaters. The beating surfaces average about 7 inches in length. At the distal end they range in width from 2.5 to 3.1 inches. At the proximal end, they average slightly over 2 inches ranging from 1.8 to 2.5 inches. The beating surfaces are thus marked by comparative short length, extra width, and by narrowing towards the proximal end.
The grooves are deep and wide but badly cut as they are often irregular at the edges. In one beater, the proximal ends of the grooves are cut off square in line with each other. In the same beater, the ridges formed by the grooves are rounded off. (See Plate XXXI, A, 7.) The beater with one grooved surface has eight grooves. In the five beaters with two grooved surfaces, the grooves are distributed as follows: 5+5, 5 + 5, 5+6, 5+8 while
The handles are approximately circular in the narrowest part, there being a difference of 0.1 inches at most in cross diameters. The diameter ranges between 1.3 and 1.6 inches except in one rather thin long handle where it is 1.1 and 1.2 inches in cross diameter. At the junction with the beating part, the slope is gradual and in some form a rather indistinct shoulder. In one abnormal case (Pl. XXXI, A, 7) there is first a distinct bevelled shoulder with the beating surfaces and then a perpendicular cut down to the handle, which is thinned down to the same thickness throughout. The proximal end is flared in five out of seven beaters, the increase being 0.2 to 0.3 inches more than the narrowest diameter of the handle but in the most marked flaring, the proximal end has cross diameters of 2.3 and 2.1 inches as against the circular diameter of the handle of 1.6 inches. The slope of the flare runs evenly up at the end making a sharp acute angle with the rim but in two beaters the rim was found to be trimmed off.
Large grooved beater with shoulder. The beater (Pl. XXXI, A, 7) with wide surfaces, the cut away shoulder and long shaped handle is exceptionally large and heavy. The distal end surface instead of being flat or slightly convex, is cut in pyramidal form by carrying four surfaces down from the beating surfaces to meet in a middle point with four distinct edges running from each corner to the middle point. The projection of the middle point is nearly one inch. This feature, combined with the cutting of the shoulder and shaping of the handle, shows a marked departure from normal due to the excessive use of a steel tool.
Smooth beater of triangular section. The beater in Plate XXXI, A, 3 is 14 inches long and 30 ounces in weight. The beating part is triangular in cross section with the longitudinal edges between surfaces rounded off. At the distal end, the narrowest surface is 2.5 inches and each of the others 2.75 inches.
The beating surfaces shape gradually into the handle which is flared at its proximal end. It is smooth throughout. The beater is a variation of the round type of i'e mole, the triangular section being most probably influenced by a natural tendency of the wood in that direction.
Small grooved beater (Pl. XXXI, A, 8) is a departure from normal size. The owners gave i'e tusitusi for the beater but could give no other information as to any special use. As tusi means to mark siapo, this specially light beater may have had something to do with the beating down of the cloth on the dyeing tablet as mentioned by
Foreign type. The beater figured in Plate XXXI, A, 9 was obtained on the island of
The tapapanga bundle, still damp, is placed lengthwise across the anvil and beaten evenly with the i'e beater. The margins and then the middle were beaten so that the thinning and spreading went on evenly by gradual stages. When the package was fairly thin, it was doubled and again beaten thinner. It was then opened out and the bands or creases at the folds which were slightly thicker were beaten to make the material of an even consistency. They were again folded and beaten. The grooved sides of the mallet were used first and the smooth sides as the material got thinner. The beating continued until the expert judged that the material was thin enough. (See Plate XXXII, B, 3.) The various packages prepared were all beaten separately before going on to the next process. The beaten material was meanwhile kept under cover of a sheet of bark cloth to prevent it drying too soon.
From the description of the scraping and folding, it will be understood that the strips of bast though damp were, speaking comparatively as regards the proceedure in parts of Polynesia to the east, fairly dry. Hence when the packages were beaten there was no felting of the strips together. On opening out the beaten bundles, the material formed from each strip came out as a separate and distinct sheet of thin cloth. They were long and narrow, wider at the butt end and thus conforming in shape with the nature of the strip. Towards the tip end, there were usually holes that coincided in position with the parts formed by flaws in the bark made by the leaf buds. Thus at the end of the beating process, the number of sheets of thin cloth corresponded to the number of strips of bark contained in the packages.
Before the material is dried, the individual sheets have to be stretched while still damp. This distinct process is termed lelenga. The tapapanga beaten bundle is removed from cover and each individual sheet unfolded (tala, to unfold). The first one is laid out to its full extent on the ground and is called the lafi lalo (the lower cover). Succesive sheets are opened out and laid out above one another with the corresponding wide parts at the same end. Each sheet is thrown out (lafo, to spread out as in spreading ordinary sheets). This is continued until from 14 to 20 sheets are in one bundle. The ulu u'a and the uppermost sheet, lafi lunga (the upper cover). The term lafi means to hide so that the lafi lunga and the lafi lalo sheets hide the pile between them.
A row of stones to act as weights is placed on the upper cover along the middle longitudinal line. All the sheets on one side of the stones are doubled over the row of stones so that half of the lowest sheet is on top with the others in consecutive order. The top half sheet is turned back into its original position and pulled and stretched to remove all folds and wrinkles. Each half sheet is dealt with similarly by taking them in turn from the top of the pile and replacing them in position ere stretching them. The sheets being still damp stretch easily and retain their shape when released. When all the half sheets are so dealt with, the other side of the bundle is doubled over the row of stones. Each half sheet is similarly dealt with so that on completing the lelenga process, each sheet has been fully stretched in a simple and effective manner.
The term fa'ala is to dry in the sun (la). The ulu u'a bundle is taken out into the sun after removing the row of stone weights. They may be left in one pile or separated into smaller piles.
Stones are placed on the ends to prevent their being blown about. When dry, the sheets are folded up, wrapped in a sheet of bark cloth and stored to await the other processes of technique.
The sheets are thin and white and are called lau u'a, now generally pronounced lau'a. lau u'a is synonymous with tutunga for which word it has been substituted on account of superstitions in connection with fishing.
Attention has been drawn to holes unavoidably present in the lau'a form of beaten cloth. The process of closing them by means of patches is called puni u'a for short but the full name is puni mata o le u'a. The term puni means to close, and mata is really the mesh of a net. In the thin cloth, the crossing fibres of the material are plainly perceived. The holes show as openings between the fibres, hence the naming after the meshes of a net. The expression puni mata o le u'a means closing the meshes of the cloth.
The pieces of lau'a may be dealt with after they are dried or at some other convenient time. Sometimes they are closed immediately before the dye is applied or while the actual pieces are being stuck together to form the thicker cloth. The closing of the holes is done with patches of lau'a which are stuck
The board (papa) may consist of the papa valu upon which the bast is scraped or the board upon which the cloth is dyed. A board must be used, but a special one is not necessary for patching alone.
The glutinous material seen in use consisted of three kinds but
- Arrowroot (
masoa). The tuber of the arrowroot is washed and cooked in an oven. It then forms a ball of paste which may be dipped every now and again in water to moisten it. This is the usual form but a woman was seen using a cooked ball of the prepared arrowroot which she dipped in water before applying it to the cloth.- Breadfruit (
'ulu). The over ripe breadfruit is very sticky and tenacious. The top of the fruit is removed and the rind acts as a natural glue pot containing the softened over ripe fleshy material. To apply it, a longitudinal section of coconut husk is used as a brush. The outer skin of the husk envelope is left attached, and the section is about 0.5 inches to 1 inch wide, 0.25 inches thick, and a few inches long. One end forms the handle and the other end is cut off square. The sticky material is thick, very white, and has a strong odor. On asking a woman who was using another material why she did not use breadfruit, she said that the unpleasant odor attracted too many flies. The'ulu ueais the best kind of breadfruit for paste. Some varieties are not suitable.Pipturus propinquus(fau songa). Thefan songais the plant whose bark furnishes the best material for lines and cords. The bark contains a copious, clear gum which exudes freely when the bark is cut. The woman seen using it, had a number of narrow strips of bark a few inches long which had been sliced from the tree. They were arranged in a wooden bowl with a little water in it and with the inner or bast surface upwards. On the bast surface, the exuding sticky gum had formed quite an appreciable layer. In using, she took up a piece of bark, applied its inner sticky surface to the material and wiped the gum off on it.Cordia aspera(tou).Pratt (23, p. 324) states that the berries of thetouwere used as a paste in makingsiapo. No information was obtained from native sources. It was formerly used as a dye, butPratt 's use of the word "paste" would seem to indicate that he meant the sticking together of material in the making ofsiapo.
The bamboo knife was seen in use at Vaitongi, 'auli'i saplings are beaten for patching material.
The technique of patching is simple. The woman examines each lau'a sheet and draws the hole over the middle of the board. She smooths out and flattens the sheet. Judging the size of the hole she cuts out a piece of lau'a from the patching strip, taking care to allow for an overlap. She lays the piece over the hole to verify the size. If too large, she trims it down to suit. She then applies the glutinous material to the upper surface of the lau'a sheet around the margin of the hole. The patch is simply dabbed down on the material and pressed with the open palm to close the mata. The glutinous material is never applied primarily to the patch. The edges of the hole, if ragged are trimmed with the bamboo knife.
A completed siapo cloth is an assembling of sheets of lau'a joined together into various sizes and various thicknesses by sticking them together with one or other of the glutinous materials already mentioned in the patching of holes. The various sizes and thicknesses receive particular names but they are all siapo,
A single layer of lau'a termed lau'a tasi, is worn by young men. It is generally smeared or painted with the red-brown 'o'a dye.
The finished siapo before it leaves the hands of the craftswomen is always colored with various dyes. There are two methods of applying the dyes; one to each lau'a sheet that forms the thickness of the siapo, and the other to the last sheet that is added. In the latter process, the full size of the sheet is made in plain material before the painting of the upper outer surface is commenced. The material required is a board and the glutinous material.
The board (papa) used is that on which the painting of the siapo is usually done. It is generally formed from a portion of the hull of an old canoe. One examined was 67 inches long and 18 inches wide over its transversely convex surface.
The glutinous material is any one of those mentioned, but for plain cloth breadfruit is commonly used. It makes very white siapo which is fairly stiff ('otu 'otu).
The first lauu'a sheet is laid over the board with the widest part from the butt end of the bast towards the worker and its length transversely across the board. The section on the board is then painted over with the breadfruit paste leaving a fair margin on the right border of the sheet. The second sheet is placed carefully upon the first and the part on the board rubbed evenly to make them stick together. The upper surface of the second sheet is next pasted with the paste over an area corresponding with that pasted on the first. A third sheet of lau'a is now carefully applied and rubbed evenly to make them stick together. Three thicknesses are usually enough but a fourth and then a fifth sheet may be similarly applied according to the thickness required. In this description, three will suffice. The worker may now move the stuck sheets to her left or move along the board. It will be noted that the ends of the sheets beyond the board are separate as are also the side margins on the right. On the right, the free margins of the two upper sheets are folded back over the united part. The margin of the lower sheet is painted with paste for the width of the board. A sheet of lau'a is then reversed so that its narrow end is towards the worker. Its left margin is carefully overlapped over the pasted part of the first sheet and carefully rubbed to join them together. The section of the new sheet on the board is then pasted, care lau'a, two or three sheets may be dealt with at once to make a convenient working width. The principle of marginal overlapping is, however, the same.
When the depth has been obtained, the stuck portion is rolled or folded longitudinally with the length of the board so as to bring another section of the free parts on the other side of the board into position on the board. Commencing at the left margin, the far ends of the sheets are folded back towards the worker. The lowest sheet is left on the board. Its upper surface is painted with paste for the section on the board. The line of adherence is now transversely on the near edge of the board. The worker sees to it that this transverse line of adherence is actually on the board. The second sheet is then turned forward over it and rubbed down. The right margins as before are left free of paste. The second sheet is smeared with paste except for the right margin and the third or upper sheet turned forward and stuck down. The next set of three on the right are dealt with successively as before. Here, however, there is now a line of adherence on the near side as well as the left. If the sheets of lau'a fit well through the alternate reversing of the sheets, when first stuck together, the margins will overlap quite well. Should there be any gap, a piece cut to the appropriate size is fitted in. When the second segment of pasting has been completed, the pasted part is further rolled to draw the next unpasted segment on to the board. Commencing on the left, the process is repeated until the required length is secured. If the length of the sheets of lau'a is not sufficient more material is joined on end to end in exactly the same way as side to side joining.
The fa'apa'o'o process. In the method of preparing the bark that has been described, it will be noted that the bast is separated from the outer bark without previous soaking in water. The bast separates quite readily as I have personally observed on several occasions.
The tree is cut clown, the bark peeled off, and soaked for forty-eight hours in water. The outer or brown bark is then separated from the inner or white, and the woody parts of the latter removed by scraping with a particular kind of shell.
The present day Samoan, however, denies that the bark was soaked beforehand as a necessary introduction to the scraping and beating of the bast. It is only done according to them when the bark cannot be dealt with immediately after it is brought in from the plantation. This occurred in past times when the paper mulberry was grown more extensively than now. When a large quantity of saplings matured at the same time, they were all cut to save the bark from becoming too old and thus useless for cloth. The bark was stripped and the bast separated (fofo'e). When the quantity of bast was too great to be scraped and beaten at once, it was stored away to await a convenient time. Before dealing with such stored bast which had become dry, it was soaked overnight in water if the fafai process of scraping was to be carried out on the following day. The scraping and beating were then exactly the same as that already described for the fresh bark. The material produced was termed u'a fa'apa'o'o to show that the u'a had been kept till it was dry instead of being beaten when fresh. The term pa'o'o means the dried gills of a fish such as the bonito which were used as a shark bait.
The
Joining technique. In the now prevailing technique, the individual sheets obtained from each strip of bast are stretched by the lelenga process before they are dried. They are joined together with arrowroot or other material after having been dried.
The bark is procured from the plant in strips of three, four or five inches wide, but by scraping and beating it is spread out to some ten inches, and made so thin that it is quite transparent. Several pieces are then put together, over each other, according to the thickness of the cloth required, arrowroot being used to make them stick together. The strips are then put together in widths to suit the purpose and beaten again, until they are made into one. The whole is then dried in the sun.
Again one cannot help thinking that confusion in the detail of Samoan technique has occurred through unconsciously transferring ideas formed in another area, or general deductions made through not carefully observing each detail in its proper sequence. The latter contingency would arise in regard to cloth making with any general observer who was not forced by special circumstances to carefully record the technical details of each stage in the manufacture.
Before describing the other process of making siapo, it is necessary to deal with the various dyes. The plant dyes in common use at the present time are a reddish-brown ('o?a), red (loa), yellow (ango and lenga), and black (lama). A red earth (ele) is extensively used. A number of others that have gone out of active use are mentioned by various writers.
The 'o'adye (Bischoffia javanica). The 'o'a dye procured from the bark of the Bischoffia javanica gives a reddish-brown color and is the one in common use. The bark is scraped with a shell by women from the growing tree and falls into a coconut leaf basket placed in position at the foot of the tree. Many 'o'a trees may be seen growing with the trunks showing the scarifications made by dye gatherers. The baskets of bark shreds are brought home where the wringing takes place.
The 'o'a wringer (to tau'o'a). The to tau'o'a acts as both strainer and wringer, hence to, a strainer, and tau, to wring. Another name is unu. It is a long plaited band made of strips of fau bast. A wringer presented to
The plaited band as shown in Plate XXXII, C is square at one end and finished off with a number of three-ply braided tying cords at the other. The plaiting is in check and commences at the left corner of the squared end in much the same way as the commencement of the 'ie fau garment in figure 153. In the wringer, however, there are no complications with adding fringe elements. (See figure 166.)
The dye is prepared by women in the vicinity of the cooking house. The basket of bark shavings, the wringer, some banana leaves, a wooden bowl, a stout stake, and a cross beam of the cooking house are all that is required.
The banana leaves are spread on the ground for the length of the wringer. The wringer is stretched on the banana leaves and opened out. The bark shavings are distributed over its length as shown in Plate XXXIII, A. Less material is placed at the ends. The side edges of the plaited band are then folded over the material and made to overlap. To the middle of one side edge a couple of lengths of sennit braid are attached by one end. These are run spirally round the folded wringer towards each end and tied. The wringer with the contained material now looks like a huge sausage. Two women, holding the ends carry the wringer beneath one of the cross beams (utupoto) of the cooking house. The square end is placed on the beam with the slightest overlap. Holding it in position, the other end is raised and placed over it from the opposite side. The braided ends are used to tie the two parts unualunga and the lower stake used as a handle, unualalo.
As the pressure takes effect, the liquid simply pours out of the wringer and runs into the bowl. When the liquid ceases to run, the women reverse and walk round in the opposite direction. Again when the pressure is felt, another lot of liquid runs down. The pressure is kept up steadily until no more liquid can be obtained. The wringer is taken down and the drained bark refuse discarded. It burns well when dry. The rest of the material is dealt with similarly.
It is surprising the amount of liquid that is obtained from the bark. The bark has to be dealt with the same day as procured. No water is added to the bark as it weakens the dye. The liquid is stored in large coconut shells; similar to those used as water bottles. (See Plate XXXII, D, 1.) It is'
Besides being used by itself it is used with the black dye from the soot of the candlenut and also with the yellow dye.
The loa dye (Bixa orellana). The loa is an introduced tree but the bright red five obtained from the seeds is simply squeezed with the fingers in a wooden bowl or a half coconut shell. It has to be used when procurable as there is no method of keeping it for any time.
The ango and lenga (Cucuma longa). The turmeric plant (Cucuma longa) is called ango as is also the root and the dye obtained from it directly after grating. The turmeric powder obtained by a complicated process from the root is called lenga.
The terms ango and lenga are quite distinct and must not be confused: a, in making ango, the root is grated and used directly by mixing with water when it give a dull yellow color, or with 'o'a when the yellow is much brighter. This is the usual form used for dyeing cloth in, both the rubbing and painting processes. b, The making of turmeric (lenga) is a complicated process that is carried out by a number of women. It is never an individual effort. When carried out by a family, it is termed fainga; by the village, nuanga. Like other communal activities it was marked by a set routine accompanied by prohibitions as well as feasting and enjoyment.
A house is set apart as the fale nuanga and an enclosure of pola mats made round it. A small paopao canoe is carried up into the house and the apparatus set up for grating the root. A skilled woman is appointed sisili (director of the work). Meanwhile preparations for a feast are made by the menfolk.
Gathering the root. Parties of women go out into the bush and dig up the roots with pointed sticks. The baskets of root are carried down to the seaside and washed. The roots are then scraped within the enclosure with pipi shells. The material being ready, the process is as follows:
Grating (
ngungu). Special graters are prepared from poles about 1.5 inches in diameter. Five or six of these are attached at intervals to one gunwale of thepaopaocanoe placed within the house orfale nuanga. The poles slant obliquely across the hold of the canoe and the upper ends are attached to a convenient part of the frame of the house. The lower parts that are above the canoe hold are closely wrapped for about 18 inches of their length with sennit braid, which supplies the grating surface. Above the sennit surface, a banana leaf is tied so that it hangs vertically down against the inner surface of the far side of the canoe hold. The leaves prevent the grated material from falling outside the canoe receptacle. A woman sits down beside the lower end of each pole. Using both hands, theangoroots are rubbed against the under surface of the sennit grater and the gratings fall down into the canoe. The usual term for grating isolobut the grating of turmeric receives the special name ofngungu. When a sufficient quantity has been grated, it is strained.Straining (
'po). The strainer (po) is formed of three or four stakes set upright in the ground to form a triangle or square. Coconut leaves split longitudinally down the midrib are used to form an upper continuous rim. The hanging leaflets are gatheredtogether below in the middle and their tip ends tied together to close the bottom. Selected large strips of the textile-looking material ( lau'a'a) from the base of coconut leaves are used for lining the inner surface of the strainer. The edges of the pieces are made to overlap and the edges are sewn together with strips offaubast. A pointed stick is used for piercing holes through the material. The strainer is called apoas is also the process of straining. Another small canoe or a large wooden bowl is placed below the strainer.The grated
angofrom the first canoe is placed in the strainer. Water is added and the mass worked with the hands (lomi). The liquid part strains through the fibrous lining oflau'a'ainto the vessel below. When the grated material has been treated, the liquid in the vessel or vessels is allowed to settle. The refuse within the strainer is cleaned out and discarded.Decanting the water. As the liquid settles in the vessels, the heavier extract from the
angosinks and the water used in washing through the strainer rises to the top. The water is then decanted by tilting the vessel.Decanting the
lenga, fa'amămā. Theangoextract left in the bowl divides into a thicker lower layer (malasina), and a thinner, lighter layer (lenga) which floats to the top. The process of separating these two layers is termedfa'amama. Thesisilior director here assumes direct personal control. She uses a coconut cup with a hole in the bottom as a dipper. An empty bowl is placed beside that containing the liquid. The latter is carefully tilted and the upper layer oflengaslowly decanted into the second bowl. Some bowls used to containangoare made with a short spout and some have merely a groove cut at the narrower end. These forms facilitate the pouring off of thelengabut ordinary bowls are also used. Thesisilidips up the liquid in her cup and allows it to run back into the bowl to keep thelengain motion and assist in running it off. The expert judgment of thesisiliis exercised in deciding when all thelengahas been poured off. On completion, thelengais in the second bowl and what remains in the first is themalasina. The above process is referred to in the saying, "O le atete'a nei lenga ma le malasina" (Thelengais being separated from themalasina), which is quoted metaphorically when selections and divisions are being made.
Use of malasina. The thicker malasina is not used as a dye but it is cooked for food, by dropping-heated stones into it.
Cooking the turmeric. The yellow turmeric is prepared by cooking the liquid lenga that has been decanted. Stones are heated in an oven. The firewood used is always breadfruit. As a fierce heat is not required, only the small branches (lala) of the tree are used.
A number of half coconut shells are cleaned and prepared as containers for the liquid lenga. The half shells used are always those which contain the mata depressions at the end. The single mata is pierced with a coconut leaflet midrib and cleaned out so that each cup has a hole through the bottom. The holes are plugged with strips of banana leaf to close them.
The heated stones are arranged in threes with the iofi tongs to form stands upon which the round cups may rest securely without spilling. The sisili fills the prepared cups with lenga and they are placed on the heated stones. The cups are lastly covered over with leaves such as those of the fau pata and left for an hour.
On removing the oven leaves, the lenga is found to have solidified within the cups. The plugs are then removed from the bottom of the cups. The cups are inverted on the palm of the hand and by blowing through the holes, the lenga comes out in perfect molds the shape of the shell cups. They are then dried (fa'ala) in the sun. The dry lenga is wrapped in thin lau'a cloth and stored. Care must be taken both in wrapping and in storing as the material easily crumbles into powder.
When thus prepared lenga is a rich, deep yellow in color. When crumbled, it forms a fine powder, which was formerly much in demand.
Uses. Cloth dye. For dyeing siapo, the grated material in the form of anno is that in general use. It satisfies the purpose and can be readily prepared in the quantity required by the siapo maker who also avoids the complicated manufacture of real lenga. When made lenga serves other important purposes not served by ango and it is usually kept for those purposes. When I asked some women if they used lenga for dyeing siapo they returned a negative answer. When I triumphantly pointed to the yellow color used on some of their siapo, they said it wasn't lenga but ango. It was not until the whole process of the preparation of lenga had been explained that I could understand that the yellow dye was not lenga. Although it is not usual, lenga may be used with 'o'a to produce a fine yellow dye.
Ornamentation. The yellow powder mixed with coconut oil is used for rubbing on the body for ornamentation.
Medicine. Mixed with oil, it is used for rubbing on inflamed parts. It is especially used to rub over recent tattooing to sooth the painful parts.
A usage formerly in vogue is lenga o le matua (the lenga of the father). When making lenga, a family provides one very large shell cup to form a mold. The lenga prepared in it is carefully stored away against the time of death of their father.
When that time arrives and the mourners have taken their place beside the corpse, the family produce the special piece of lenga and hand it to the mourners with the announcement, "O le lenga o le matua" (The turmeric of the father). The mourners then smear themselves with it as a sign of mourning. Such provision beforehand is regarded as fa'aaloalo (high respect for the deceased) and the dutiful offspring gets much credit from the public. The beneficiary, while alive, also derives a certain satisfaction from his present status being respected as well as the more melancholy one that part of his future has been provided for. The lenga o le matua is never given away or used for any other purpose but distribution to the mourners.
Babies. Turmeric is still used as a dusting powder for babies.
The nuanga process of making turmeric is regarded as an important event by the community. The womenfolk enjoy it as it is one of the few occasions upon which they rule the proceedings. Singing and dancing goes on. The young men come to assist those to whom they are paying court. They do the menial tasks in order to have the opportunity of penetrating behind the walls pola enclosure. A feast is prepared and pigs are killed for the sisili who is head of the organization.
A further indication of the importance of the occasion is that a number of things are sa or prohibited.
- In preparing the oven for the feast, any stones that are scattered about the edge must not be replaced. The
umuis not trimmed.- In grating coconuts for the coconut cream which accompanies feasts, the parts known as
tuatuathat stick to the shell, must on no account be eaten. The grated nut, after straining off the cream, must not be thrown on the fire or thelengawill become dirty.- Grimacing, even in fun, must not be indulged in or the
lengawill not solidify.- Crying is repressed. If anyone cries, the
lengawill cry also (tangi le lenga) and consequently will not solidify. The phrase,tangi tangi le lenga(thelengais crying), is used for crying babies and originates from the above idea.- The person for whom the
lengais being made must not eat by himself any delicacies he may have, such as fish. If he does, flaws will appear on thelengaafter removal from the molds. The flaws take the shape of fish or the articles that have been consumed secretly.
In connection with the last prohibition, it should be explained that a chief may ask that lenga be prepared for him. He and his family are then responsible for the entertainment of the workers including the supplying of food for the feast. When the idea originates in the family, the prohibition applies to the head of the family.
The above prohibitions are good examples of sympathetic magic even to the contortions of the face causing movements of the lenga within the molds and thus preventing it from solidifying. They form good excuses for the accidents likely to occur through bad work or ill judgment, dirt, failure to solidify, and flaws in the solidified shapes. The prohibition against secret eating was also to prevent such an occurrence depriving the workers of the choice articles of food. The same idea was carried out by the guild of carpenters when they prohibited the house owner from giving away presents, mats, or food during the construction of the house. The builders retaliated by abandoning the work and leaving a flaw in the building. The turmeric makers threatened a flaw in the shape of the turmeric they built up in the shell cup.
Some districts were famed for their turmeric. Such a one is
The lama (Aleuritea molucanna) is a black dye obtained from the soot of the seed kernels of the candle nut tree also called lama. The material is generally prepared in quantity by a number of women banded together into a working party. The hard-shelled nuts are gathered in baskets from beneath the trees and cooked in an oven where it is left for one or two days. The nuts are then cracked between ordinary stones and the kernels threaded on dry coconut leaflet midribs (tuaniu) in the manner of preparing the lighting sometimes used in houses.
A building of the cooking-house type is set apart for the preparing of the lama. The sides must be closed so as to exclude any wind. Within the house a special fireplace is built of stone. The sides and end are built-up and flat stones with a rough surface are laid across. Within this fireplace the candlenut kernels are lighted and the fire kept going by adding the prepared skewers of nut kernels. The nuts are very oily and burn readily, emitting a black, oily smoke. There may be more than one fireplace, according to the number of the party. Women tend the fire in relays, for it must be kept constantly going. The fine black soot adheres to the surface of the stone. When enough soot has accumulated, the cross stones may be removed and the soot scraped off onto a banana leaf. The stones are replaced and the fire kept going again. The work often lasts through the night.
The soot is stored in large-sized coconut shells which are then called pupu lama. Enough is prepared at one time to supply the family group.
The dry powder, when used, is mixed with 'o'a dye and not with water. The lama is poured into a half-coconut shell and usually rubbed with a stick as a pestle to break up any lumps that may have stuck together. The 'o'a is poured in and a mixture made. The dye is perfectly black and the 'o'a gives it a shiny appearance. It is used for painting the siapo.
Red earth ('ele) is obtained from some parts of 'ele. The material was dug up in lumps with a pointed stick in olden days and it was quite an expedition to obtain it in some of the rougher country. Even now 'ele passes round in the way of trade to the districts that have none. To a siapo maker, her lump of 'ele is an important adjunct.
The 'ele is rubbed on the back of a mangeo shell (Antigone reticulata) and the powder dropped on the surface of the siapo as it is being rubbed. The 'lump is moved about so that the powder is evenly distributed. It is then rubbed over the surface with a piece of bark cloth dipped in 'o'a. The 'ele is generally used with 'o'a and brightens up the color into a more reddish or brick-colored tint. It is never mixed with water or other liquid beforehand as a pigment.
The shells seen with the red 'ele on them are graters and not containers of the powder. The 'ele is kept in lumps and only grated as required onto the surface of the siapo.
Soa'a (Musa sp., plantain) gives a purple dye. The trunk of the plant is cut through and the sap allowed to drip into a coconut-shell cup. It is used for painting and not rubbing.
Other dyes. taro patch for two types of cloth. (See 39, pp. 82, 83.)
Dyes mentioned by early writers but not now in use are as follows:
'o'a. On the other hand, the vavai hiri) in dyeing their cloth. One cloth termed pareu was soaked in the mixture and on drying assumed a reddish-brown color. It is curious, however, that the present Samoans, though asked about it, know nothing of its use in the past.
nonu fi'afi'a, Malay apple ( Eugenia Malaccensis) with sea water and lime."
nonu, but ango and lenga seem to have completely displaced it. In the Morinda citrifolia, under the name of nono, was used as one of the sources of a yellow dye, and sea water was formerly used to brighten up the color.
pani with sea-water." pani as the name of a tree but does not give the scientific name. He also gives pani as a verb, to dye the hair with the juice of the pani.
With the exception of red earth ('ele), the dyes described are applied to cloth in a state of solution, and penetrate the pores of the object so treated. This general method, universal throughout Polynesia, comes under the term "dyeing." The methods of dyeing vary. To clearly appreciate one of the Samoan methods, not hitherto described in detail, it is necessary to define the described methods used by the Polynesians. While recognizing that they are all forms of dyeing, it seems permissible to use the term in the more restricted sense that the word conveys to the average reader.
The material is immersed and soaked in the dye solution. In the
Painting. The color designs were painted freehand on one surface of the
Ruling. Single, double, or multiple parallel lines were drawn across one surface with special implements devised for the purpose. This method was common in
Stamping or printing. Special patterns carved on wooden stamps were stamped or printed on the surface of the completed cloth. In ohe kapala. In the rakau takiri pahoa. For both, the dye was dabbed on the stamp or frame and then pressed down on the material. The process was repeated over the surface to complete the design.
It has been erroneously stated that the Samoans used the printing method. The error evidently dates from Wilkes, who was in Samoa in 1839, and who states (42, vol. 2, p. 150):
The tapa is often printed in colors in patterns. This is performed in a mode similar to that practised in Europe before the introduction of copper rollers. Instead of engraved blocks, they form tablets, about as thick as binder's boards, of pieces of large cocoa-nut leaves, by sewing them together. One side of the tablet is kept smooth and even, and upon this coconut fibres are sewed so as to form the required pattern, which is of course raised upon the surface of the tablet. These tablets are wet with a piece of cloth well soaked in the dye, after which the tapa, which for this purpose is well bleached and beautifully white, is laid upon them and pressed into close contact. The dye is made from herbs and roots, and is of various colors.
upeti frame as a board on which "cords of sennit or hisbiscus bark were fastened across and across," says, "These cords were smeared with dye, by a cane brush or ball of siapo and then the native cloth was carefully adjusted upon the board and rubbed by hand or beaten with the betel to take the impression."
The statement that the tablet was first moistened with the dye and the cloth afterwards placed upon it was evidently accepted by upete, and the similar wholesale stamp of the other southern groups." The inference is that the method of applying the dye was the same.
The correct technique had been mentioned by
The printing of the patterns is done by spreading the cloth over a large board, on which are fastened (by a particular process) the ribs of the coconut leaf, and while stretched out it is rubbed over with a reddish-brown juice, obtained from the candlenut tree, which grows in all the islands of the
Pacific . This juice marks the cloth only where the ribs of the cocoa-nut leaf raise it.
The statement that the cloth was rubbed and not the ribs of the coconut leaf is correct. Though it may appear a detail it makes a totally different technique and constitutes the vital difference between rubbing and printing.
The Samoan tablet, which is called upeti (not upete), was not used as an engraving block or a stamp with the printing process. Wilkes commenced wrongly by stating that the tablet was made of large coconut leaves, whereas large pandanus leaves (preferably paongo) are used. In rubbing the cloth, the dye soaks through and stains the tablet as if it had been directly applied. Wilkes saw the stained tablets and the cloth stained with similar designs. He, therefore associated them together in what seemed the only obvious technique. I was six weeks in Samoa before I saw the upeti actually used. Not having read
Rubbing. Dye is applied to the upper surface of the cloth by rubbing to bring out the design of a tablet placed beneath the cloth.
The Samoans divide their present methods into two: tutusi or mamanu (painting) and elei (rubbing). The expression tutusi is the act of printing or drawing, and mamanu, the design obtained by painting. o'ai (to mark or paint siapo) and o'ainga (a marking of native cloth). He also gives o'ainga as a synonym of elci, but it is obvious that he was not clear in either the technique of the exact shade of meaning of the words. Dyeing by immersion was formerly practiced.
Dyeing by Immersion (fui). Cloth was formerly dyed red in the nonu dye and the process termed fuinonu. So, also, the cloth dyed black in pani was called fuipani, as was also the process. The method of immersing in the black mud of a taro swamp also comes under this heading. lufa as a large, black siapo and sema (23, p. 262) as a red one.
Painting (tutusi). Cloth painted freehand with a brush is called siapo tutusi or siapo mamanu. (See Pl. XXXIII, C, 1, 2.) The plain white cloth is pasted together to the required thickness and size. When ready for the brush, it is called tasinga. The materials required are a board, dyes, dye cups, and brushes.
The board (papa) is the same one used in pasting the sheets together.
The dyes, such as ango (yellow) and loa (red), are usually freshly made. The 'o'a and lama are kept in large, stoppered coconut shells. (See pupu 'o'a, Pl. XXXII, D, 1.)
Dye cups to contain the dyes for use are formed of large, half coconut shells, that are simply cleaned out and kept after use. (See Pl. XXXII, D, 2-5.) The brushes consist of four types:
The cloth is laid on the board, which gives it an even, hard surface to be stretched on. The painting is done in sections. The artist dips her tusi brush into the appropriate dye cup and deftly draws the outlines of the section design. The design is filled in with the various colors, each color having its own cup and brush. Some colors are given a second coating, to show them up. One coat of 'o'a gives a dull, red-brown color but a second coat gives it a varnished appearance. The second coat is called 'āle, or 'ale'o'a. The lama mixed with 'o'a gives a shiny black. A second coat is called talama. The (ordinary repainting of black cloth is called amoamo.
As one section is completed, another part is moved into position on the board. Large sheets of cloth are often spread out on the floor upon the mats. The part done is usually rolled or folded so that the artist may have her work before her as she sits cross-legged on the ground. The part spread out is weighted down with stones at the edges to prevent it being blown up by any breeze.
Some artists use a shell for scraping lightly over the painted design to trim it off.
The varieties of painted cloth are few as compared with those dyed by rubbing. They are easier to do by painting and there are probably more of them in circulation through trade than of the others. The varieties enumerated in lau'a tasi, the single sheet kilts worn by young men were usually painted or smeared over with 'o'a; taloa, the corresponding garment worn by young women of rank, but consists of two thicknesses of cloth; siapo mamanu, which class includes any cloth of three or more thicknesses, C, 1,2.)
Rubbing. In dyeing cloth by the elei method of rubbing the process must commence with the thin lau'a sheets. The rubbing and the pasting proceed together until the siapo cloth is completed. In addition to the requisite quantity of lau'a, the materials required are a board, design tablets, dyes, paste, dye daub, and paste brush.
Board (papa elei). The board is the same large slab used in painting. It must be wider than the design tablet and is now called papa elei from its function. Being usually formed from the side of an old dugout canoe, it has an outward convexity transverse to its long axis. (See Pl. XXXIV, B.)
Design tablets (upeti). The original tablets were constructed on a foundation of pandanus leaves. Some of them are still in use but they are being supplanted by the more modern article made of wood.
The tablet of pandanus leaves is called upeti fala, though the kind of pandanus preferred is the paongo, which has larger leaves than the fala.
A typical tablet (Pl. XXXIV,
A) is made of two layers of leaves. The lower layer consists of 36-inch lengths of 4 leaves and a narrower strip, laid together with overlapping edges stitched together withfanbast. The upper layer consists of 15.75 inch lengths laid transversely to the long axis of the lower layer. The design is in 4 distinct sections, each of which is dealt with separately. The design material consists of pandanus strips and two-ply twisted cords of sennit fibre. The pandanus strips, one leaf thick and 3 to 6 mm. wide, are used in the design to form vertical panels separating the sections and are also placed horizontally and obliquely to divide the sections into smaller areas. The pandanus strips are laid in position on the leaves of the upper layer of one section and stitched to it with loops which pass through that layer alone. The section is then laid in position on the lower layer and the sennit cord elements of the design are stitched in position, through both layers of the tablet. Each section is treated in turn from the left. The stitches of the pandanus strip therefore keep the leaves of the upper layer together while those of the cords bind both layers together. The pandanus elements in straight or zigzag strips map out the design into smaller areas which are filled in by parallel lines of cord arranged in horizontal or oblique sets. The pandanus strips are stitched to the upper layer with single thick sennit fibres or with fine twists of two fibres which look like fine copper wire. The stitches over the cord elements are offaubast, while here and there, a strip of the same material is passed over the pandanus strip and through both layers of the tablet. The tablet is finished off with a piece of sennit braid stitched around the upper surface close to the edges, forming loops at both long edges and ending at a corner with a coiled length.
Some upeti are smaller than the one described and the designs on them vary but all are rectilinear. The tablet, being soft and yielding, is of no use unless it can be fixed firmly to some solid material that will give it support.
Setting up the design tablet. The tablet is laid longitudinally on the long axis of the board on its convex outer surface. The sennit coil at the rights lower corner (Pl. XXXIV, A) is untied and stretched across the under surface of the board to pass through the loop opposite. The loop is drawn taut on the under surface of the board and the braid brought back to the next loop on the near edge. In this manner, the braid zigzags from loop to B.) The design is fixed to the wood as if it had been carved on it.
The pandanus tablet is the invention of a people who did not have the tools or the skill to cut out elaborate designs in wood. They added the design to the leaves in the technique possible to them. After the advent of steel tools, however, the designs were cut on the wood (papa) with the superior tools available to form wooden tablets (upeti papa). The wood used is usually part of the side of an old dugout so that they are usually convex over the short axis. (See Pl. XXXIV, C.) The same rectilinear designs were used in the older forms but with the more modern make on ordinary straight boards, curves have been introduced. Through the comparative ease with which wooden tablets may be made and their longer life, the pandanus tablet is being gradually abandoned in favor of the modern article.
Technique of rubbing. The upeti is placed on the ground with its length transverse. Beside it are placed a pile of lauu'a sheets, a dye cup with the 'o'a dye, a tata wiper of bark cloth, a lump of red earth, a shell grater, a tuber of cooked arrowroot, and some sharp-edged strips of bamboo.
A single sheet of lauu'a is spread over the tablet with its wider butt end towards the worker. The left side is adjusted to leave a clear margin beyond the left edge of the tablet. The near end edge of the sheet is similarly adjusted to the near edge of the tablet. The sheet is not as wide as the transverse length of the tablet so another sheet must be added, but with its narrow end reversed towards the worker. The right edge of the first sheet is rubbed with the cooked arrowroot paste for about an inch deep along the part of the edge that rests on the tablet. The left edge of the second sheet is placed over the pasted part and rubbed smoothly with the tata wiper to stick them together. A third sheet is usually necessary to cover the remaining part of the tablet and leave some spare material to the right. It is reversed with the broad end proximal and stuck by pasting the right edge of the second sheet. (See fig. 167, also Pl. XXXIII, B.)
The red 'ele earth is used to brighten up the 'o'a color. The mangeo shell grater is held above the cloth and the lump of red earth rubbed against its rough outer surface. The grater is moved about so as to sprinkle the powder evenly over the surface of the design. The tata wiper is rubbed over the surface and the brick-red powder sticks to the wet surface and brightens up the color. When the design is sufficiently clear on the first layer of sheets, another layer is added above it.
The area of the design is rubbed over with the moistened arrowroot, care being taken to restrict the pasted area to the actual area of the design. A sheet of white lauu'a is laid on the left side above the first sheet. It is tata wiper to make them adhere to the first layer. The wiper is dipped in the
'o'a dye and the design brought out on the second layer by rubbing. If thin cloth is desired, the process may stop with the addition of the second layer of sheets. For thicker cloth, a third layer is added in exactly the same way as the other two. The last layer usually gets a more careful rubbing and a more liberal sprinkling of red earth. The rubbing of each layer is, of course, to make the sheets lie down close to conform with the ridges and depressions of the design. The dye which damps the cloth enables this to be done. If each layer were not done, the design would not come out clearly on the surface of the upper layer. The dye also soaks through the cloth and not only shows up the design on the under surface of the first layer but also stains the tablet. As pointed out before, it is this staining which filters through the layers of cloth that gave the idea that the dye had been directly applied to the tablet.
The completed first section is rolled back to the left and after manipulating the right edge of the design on the cloth to coincide with the left edge of the frame, the rolled part is anchored in position by placing stones on it. The extra material beyond the right edge of the dyed part is now on the tablet. The sheets of the middle and upper layers are folded back to the left and kept there with a stone leaving the lower layer alone on the tablet. Remember that the three layers are only stuck together as far as the right edge of the design. The lowest layer sheet is smoothed out evenly on the tablet and its right edge rubbed with arrowroot for the depth of the tablet. Another sheet of lauu'a is joined to it and another to the right of that until the o'a and red earth, the second section is complete. In this manner by successive sections added to the right, the desired length is obtained. Some women may start from the right and work to the left. In large sheets, the measure is always the number of upeti or tablet lengths.
Depth is obtained by adding successive rows of rubbing to the first row. It is now better to unroll the completed part which has been rolled parallel with the left edge. It is rolled again from the proximal edge. On the left, this brings in the first section rubbed until its far transverse edge coincides with the near edge of the tablet. The free parts of the two upper layers are folded back proximally and the sheets of the lowest layer are spread out on the tablet. The overlap of the individual sheets are pasted together at their side edges for the depth of the design. If there is any gap between individual sheets in the same layer, a fresh piece is cut to size with the bamboo knife and pasted in to fill the gap. So also any trimming of frayed edges is done with the bamboo knife. The first layer being satisfactorily arranged it is rubbed with dye and then pasted. The sheets of the second layer are then turned forward individually and stuck to the first layer. Any holes or gaps are remedied so as to maintain an even thickness in the cloth. The second layer is rubbed and pasted, the third layer is turned forward and carefully rubbed. This completes the first section of the second row. The material is pulled to the left and successive sections added on the right until the full length is secured. The only difference to the first row is that the near edge is stuck together as well as the left. Care must therefore be exercised in getting paste down into the left-hand corner of each new section so as to stick the sheets evenly together. Successive rows are added until the depth is secured. In arranging the lower layer of each new section, the edges must be carefully adjusted to make the design continuous without any breaks between the tablet areas.
The technique of increasing length and depth is the same as in preparing plain cloth for painting, but the layers are adjusted to the smaller area of the tablet and each layer is rubbed with the dye before paste is added. The above detail should dispel the error that the dye tablets were used as printing frames with the dye directly applied to the design and thence stamped on the cloth.
The tata wiper served as both paint brush and brush for smoothing out the sheets laid on the paste. upeti was rubbed directly with the dye, states that the cloth was pressed and beaten down with the hands and with a betel (beater). On i'e tusi tusi as already mentioned may, however, have had such a use.
The upper surface of the completed cloth is in one color. The lines are less sharply defined at the edges than in painting, and a certain amount of smudging may be seen in the lighter areas between the colored lines. The design may be enhanced by painting over some of the wider lines when they show up with the brighter varnished color that a second coat of 'o'a gives. Parts may also be gone over in black, which gives a brighter varnished appearance because the black lama is mixed with 'o'a as a vehicle for the color. Apart from the close, red-brown background that rubbing with 'o'a gives, rubbed siapo (elei) can always be recognized by the same, though lighter, design appearing through on the under surface. (See Pl. XXXIII, C, 3, 4.)
Cloth prepared by the rubbing process receives names according to the size and the uses to which it is put: siapo, though a general name, also denotes the shorter-sized pieces suitable for use as a kilt—the reward given to a talking chief for calling the kava, is referred to as one or more siapo; potu, a sheet larger than a siapo, often used as a small screen. pupuni, a large sheet used as a screen to shut off an end of a house; ta'i namu, a sheet used to make mosquito curtains; ululima, a large sheet measuring fifty lengths of the upeti tablet; uluselau, a very large sheet measuring 100 lengths of the upeii.
Loin cloth (malo). maro (malo) made of siapo was a narrow strip of a foot wide and six feet or so in length. The maro of other Polynesian islands was wound around the waist and an end passed back between the legs. maro (girdle), tied loosely around the loins, was worn by thieves in their night operations.
Kilt or skirt (lavalava). The single sheet (lau'a tasi) worn by men and the two thicknesses (taloa) worn by women, when used around the waist as kilt or skirt are called lavalava from their use.
Belt or girdle (fusi siapo). Pieces of varying length and about a foot wide were worn around the waist as girdles after the fine mats had been put-on. The girdles were often for keeping the mats in position. Cloth rolled around the waist was also called fanga'au or fangau.
Wrappers ('afu). A sheet of cloth was used as a wrapper in the evenings or early mornings.
Bed clothes ('afu loto). Sheets were used as bed covers at night after discarding the kilt. It was usually the same as the 'afu mentioned above.
Poncho (tiputa). According to tiputa was introduced to Samoa from
Bodice without sleeves (pi). This was also a missionary introduction contemporaneous with the teaching of foreign ideas of modesty.
Plain cloth was occasionally colored by subjecting it to the action of smoke, whereupon it became smoked cloth (siapo fa'aasu). Such cloth was worn only by the village maids or chiefs' sons.
A smoke house (fale fa'asu) was formed of a number of light poles arranged to enclose a fireplace at their lower ends. The upper ends were tied together at a point.
The cloth after being pasted together was rubbed by hand with coconut cream (pe'epe'e). A fire was lighted in the enclosure and after it had burnt down to glowing coals, the cloth was wrapped around the frame to form a tent-like structure. Outside of that, old, used mats were wrapped around to prevent the escape of the smoke about to be generated.
Material, to give off smoke, was placed on the coals by raising the side of the tent. The materials used were:
The Cordyline and sugar cane smoke produce a yellowish brown; ngatae gives a much deeper brown. The soaking with coconut cream enables the smoke to penetrate the texture and thus give a lasting color. The cloth is moved from time to time in order that the whole surface may be evenly smoked.
The soles of the feet of the Samoans became so thickened from childhood that no protection is needed for ordinary locomotion. On occasion, however, sandals (se'e vae) are used to protect the soles from particularly sharp coral while wading inside the reef. The method of attaching by cords passed between the toes causes considerable chafing to the more tender skin between the toes and on the dorsum of the foot, especially when coral sand gets under the cords. For this reason, sandals were not much in favor. I have seen
Sandals are thus said to be used principally to protect the soles when
Three kinds are made; from fau, lau'a'a, and pulu respectively.
Sandals of fau bast (se'e vae fau) are made with a particular technique, but the other two are makeshift articles. The material required is a quantity of fau bast strips about an inch wide and a stick a little thicker than a lead pencil sharpened at one end. (See fig. 168.) The sandals are always made short, coming to just beyond the ball of the foot, with the toes projecting unsupported in front. (See Pl. XXXV, A, 2, 3.)
A temporary form of sandal (se'e vae lau'a'a) is made from a couple of sheets of the fabric-like material at the base of coconut leaves (lau'a'a). (See Pl. XXXV, A, 4.) The material is soft and affords protection to a cracked sole. They do not last long but are easily made (fig. 169).
Another temporary form (se'e vae pulu) is made of coconut husk pulu (Pl. XXXV, A, 1). A longitudinal slice of husk, about 7 inches long by 4 inches wide, is cut off level to fit the foot. The section is naturally oval with thin edges and deepening to a little over an inch in the middle. The method of attaching the fau bast tying strip is shown in figure 170.
The articles grouped under clothing were made by braiding, looping, plaiting, and beating out bast. In none of the technical processes was twining or weaving, either downward or upward, used. The primary use of clothing is protection from the weather and concealment of the person. The secondary use, which may become more important, is decorative. On distinguishing individuals decoration leads on to distinguishing social classes. From use as indicating rank and status, they became valuable property and certain articles have intertwined with social customs, finally becoming a medium of exchange and an economic standard of value in themselves.
Clothing. The primary need for clothing dwindled down in the tropical climate of Samoa to one causative factor, concealment. Modesty, however, demanded little. The exposure of the upper body in both sexes was not considered immodest, or rather, was not considered at all. Hence the titi kilt of Cordyline leaves fulfilled all the actual requirements of clothing. A few titled women and sons of chiefs used bark cloth lavalava as a change garment. All other articles, therefore, listed under clothing were created by other needs.
Decoration. Festivals with accompanying dances demanded something beyond the ordinary ti leaf kilt. The demand was easily met by using the brighter colored ti leaves and change was also secured by making kilts of bast, first by simple looping to waist cords and later by plaiting.
Rank and property. The desire to further decorate and distinguish the daughters and sons of high chiefs, as in the persons of the village taupou and manaia, gave a stimulus to textile development that the need for ordinary apparel would never have supplied owing to that need being of so little importance. The short 'ie tutu pupu'u with wide wefts but highly decorative effect was the natural outcome. The present form of these kilts with their many free tails may be later chronologically, but coarse plaiting in bast certainly preceded the finer forms with their more elaborate tag attachments.
With decoration heading in the direction of finer plaiting, the 'ie fau and 'ie sina garments were evolved and became a still more material expression of rank. They were not needed and could not be worn as everyday clothing. They were purely an expression of rank to be worn during ceremonial and discarded as soon as possible afterwards for the sake of comfort. Fine plaiting also proceeded with pandanus material and reached its culmination in the ie fau, 'ie sina, and fine mat discharged that duty.
As these mats distinguished rank, they became valuable property. Rank besides its social significance was distinguished by the possession of material property or, in other words, wealth. The wealth of a family was demonstrated by the numbers of ta'ui bundles of fine mats stored on the cross beams of the guest house.
Customs. The 'ie sina and fine mat became the necessary equipment of the village maid. Through association with the village maid, the 'ie sina also became the garment on which her virginity was officially proved before marriage. The 'ie sina thus became a highly specialized garment necessary to the particular custom of testing the virginity of the taupou.
The hue mat, however, gained a wider sphere of influence. It supplied the dowry of the village maid and chieftainesses of rank. The husband's family supplied the food, including pigs for the wedding feast, which, with other presents such as weapons, noraments, canoes, or houses, were grouped together under the name of oloa. The bride's family supplied an equivalent value in line mats distinguished by the term tonga. Hence the derivation of the name of the fine mat, 'ie tonga, and we get the fine mat constituting tonga (valuable property).
From association with marriage, the fine mat became associated with birth. It formed such an absolute necessity at marriage that after the birth of a daughter, mothers commenced plaiting a special fine mat termed an 'ie fa'atupu to form the nucleus of her daughter's dowry. These were carefully plaited in spare moments between the completion of other fine mats and were often finished just before the marriage of the grownup daughter took place. Some of them thus took over sixteen years to complete. Appropriate presents on the birth of a chief's daughter were fine mats to mark the occasion.
At death, fine mats again figured importantly. The ceremonial connected with the death of a high chief was termed langi. The blood relations of the deceased contributed fine mats for the langi. Amongst them had to be one of special note, either from the extra fineness of the plait or from its historical associations. The mat was called 'ie e fai a'i tonga (the mat which gives distinction to the property), or 'ie c fai a'i le langi (the mat which gives auala on account of their ala or blood relationship with the deceased, recited the langi ritual as they sat before the house of death with palm leaves laid on the ground before them. If their ala was proved or well known they were allowed to proceed. If they acquitted themselves with scrupulous correctness they received fine mats at the distribution in accordance with their status. Apart from the special fine mat of note, the tama sa or son of the sister of the deceased had the first pick of fine mats.
Thus in birth, death, and marriage customs, fine mats played the most important part from both an economic and ceremonial point of view. The Samoans attached the greatest value to them and neglected no chance of acquiring them. They themselves admit that a talking chief will enter into negotiations for the marriage of his chief influenced solely by the bundles of fine mats he has seen hanging up in the guest house of a girl's father. He knows that in the subsequent distribution of the marriage dowry among the husband's family, he, by virtue of his official position, will get the most important share. Similarly, at funerals a high chief will often refrain from attending because his talking chief has to recite the langi and will thus get the fine mat that is given to his party. Talking chiefs on the other hand will never neglect attending a funeral if he can prove an ala pathway to the genealogical tree of the deceased chief.
An unscrupulous yet humorous talking chief attended a funeral near
Apia when he had no truealabut hoped to bluff his way through to a fine mat. Seated with his palm leaf before him, he began to recite thelangi. The astute talking chief of the family of the deceased interrupted calling, "Never mind thelangi, tell us thealaby which you find yourself here." Having no genealogical pathway, the impostor kept on reciting thelangi. When he had recited five out of the customary tenlangi, the interruption became so marked that he realized his scheme had failed. Now, the termala, besides being a genealogical pathway between blood kinsmen, is also the ordinary material pathway between villages and houses. Before vacating an untenable position he replied to the aggressive defender of the family fine mats. "Thealaby which I came here?" he cried pointing at the road, "There it is, thealawhich leads fromApia to this village." "Kill him," yelled the infuriated official to the family henchmen. The visitor sprang to his feet. "Quick," he cried as a parting shot, "Show me thealato the missionary's house." At the missionary's house, he found refuge until the storm subsided. Then without a mat, he returned along the emptyalatoApia .
The above incident shows that while the desire to acquire fine mats was great, the desire to keep them was equally strong.
It may be truly said that in their present elaborate form, marriage and death customs could not be correctly observed without fine mats.
Standard of value. Through their value as property, the fine mat became the Samoan standard of value. We have seen that in the building of a house, a fine mat ratified the agreement between the prospective owner and the head builder. It paid a fine and pacified the builder's guild into overlooking a serious breach of etiquette. It formed the principal medium in paying for the building of a house at the umu sa ceremony. Canoes, tatooing, and various services were paid for in fine mats. High chiefs rewarded their talking chiefs with fine mats. No important function or activity in olden times could take place without the passing of fine mats. In Samoa, fine mats became the equivalent of the coin of higher cultures and the shell money of the
Fine mats with a history assumed a sentimental value out of all proportion to their intrinsic value. Songs have been composed giving the history and changes of name of some of them. The historical value exceeded that of the fineness of plait, though where both existed the value was naturally enhanced. Some of them, worn by age, have been patched again and again and the number of patches adds to the value. A mat seen at
Interaction of custom and technique. Though tradition holds that the fine mats are very old, any fine technique must not only have time but a strong incentive to develop it. It is natural to expect that there has been an evolution in the technique of the fine mat. The gradual improvement in technique led to an increase in the material value of the mat. With increased material value it became tonga (valuable property). As such it became the material medium for expressing certain customs. This gave it a still greater value which must have reacted on technique by stimulating the women to greater effort. The finer the plaiting, the greater the value. The greater the value, the greater the incentive to fine plaiting. Women became famous for their fine plaiting and the fine and arduous work rendered the mat worthy of the high recognition it ultimately received.
Plaiting houses. The ordinary floor mats for furnishing a new house were made by the aualuma (assembly of unmarried women) who might meet together in a house set apart for plaiting. The real plaiting house (fale lalanga) was, however, occupied by an assembly of expert women, mostly married, who met to plait fine mats on the invitation of a high chief who kept them in food and made appropriate presents. Apart from such a working bee, women were constantly at work plaiting fine mats to supply their own family needs.
Bark cloth. Bark cloth had little value in old Samoa as everyday clothing material. Its other uses in the house have been mentioned. It formed, however, a useful if less valuable medium of reward and exchange. Talking chiefs were rewarded with siapo at the chief's sua meal and in ceremonial kava drinking when he called the cups. A high chief desirous of adding to his prestige would perhaps pay with fine mats but for many occasions, siapo was correct. At lesser weddings, bark cloth enters into the presents. In more recent times as fine mats are becoming scarcer, bark cloth is taking their place in ceremonial. Talking chiefs and others attend ceremonial gatherings wearing siapo. Bark cloth now figures largely in presents to visitors. In ie fau and 'ie sina, and partly to the difference of technique in manufacture.
On first passing through a Samoan village, the impression is formed that the Samoans went in for a good deal of stonework. This is especially true where the villages are situated on rocky ground. On coming to details, however, it is seen that the impression is largely due to stones being extensively used as house platforms. On rocky sites, nature has provided the abundant supply of stones, but the Samoans for their own convenience have been forced to arrange them in some sort of order about their houses. Pig walls and paved roads also add to the false impression that the Samoans were stone workers of considerable ability. The quantity of stone that influences the observer is unworked stone. When it conies to worked stone, as exemplified by stone adzes, the Samoans lag far behind their fellow Polynesians to the east, in stone technique.
House platforms. The use of unworked stone in house platforms has already been described (p. 66) together with the loosely paved terraces common in sites where stone was plentiful. The extra terraces, built by a chief in honor of successive wives, must also be remembered. Some of these are built a little out from the completed terraces of the house platform and have no structural connection with it. This lack of continuity was puzzling to me until the reason was explained. A heap of stones, a cairn, or a terrace may thus be used to celebrate some event that has been forgotten or lost with the dead historians. The tendency to ignore the psychological and seek a material use may lead to an explanation far from the real cause.
Stone seats. The stone seats made of a large slab with another set upright or slanting at its back, characteristic of some of the house platforms in the
Pigeon netting platforms (tia seu lupe). The cleared platforms associated with the chiefly sport of netting wild pigeons were located in the forest usually on some commanding ridge. The ground was levelled off to provide concealment shelters for the fowlers. To obtain the required space, the sides and lower end were built up and faced with stone. One, named Muli-maunga, situated on a ridge at the back of tia vary, some being raised completely above the ground on all sides. The stones were the available stones scattered about, and built up loosely merely to face the earth tia are famous in local history.
Monumental cairns (tia).
For higher chiefs, the cairn is built up in loose rectangular piles that may be stepped as in Plate XXXV, B. These again are quite common throughout the villages. Some cairns were fairly large and high but always of loose stones arranged to maintain the rectangular form.
Commemorative heaps. The people often visited some hilltop or other on a walking trip or malanga. To mark the occasion, stones were carried up or collected in the vicinity and piled in a heap. Some of these heaps may be seen in places where the stones did not occur in nature but were obviously carried by human agency. ali'i predecessors for the sake of the view. The guard (soatau) took stones up on each occasion and piled them in a heap. They served not only as a memento of the visits but could be used as sling stones and throwing stones if they were suddenly attacked while there. Where people of a different culture commemorate a visit by carving their names on rocks and cliffs or defacing historic objects, the Polynesian made a cairn of stones.
Walls of loosely built rock or stone may be divided into two kinds; for defensive purposes and for restricting the wanderings of pigs.
Defensive walls ('olo or pa taua). Many of the villages in olden days had stone walls built around them to assist in keeping out attacking war parties. For such purposes, the wall was called a pa taua or 'olo, though 'olo usually refers; to the village or place thus fortified. Such a wall in ancient days was built around
The village on pu'e
Pig wall (pa pua'a). Pig walls are usually built at the back of the village or a little distance away to enclose a space termed sauauli in which pigs may roam.
The village pig pen is a community institution, and from all appearances an ancient and honourable one. The walls are substantial and look almost like fort walls. The acreage enclosed is considerable. Pig pastures is a better term than pig pen. Because of the steep hill back of
Ofu , the inland boundary of that village pig pasture is cut into the bank. Pigs are fed coconut meat refuse from whichpe'epe'ehas been extracted.
The walls are built of loose stone and range from 4 feet to 4 feet, 6 inches in height and about 2 feet, 6 inches to 3 feet in thickness. Steps are sometimes made on the outside with stone, or styles of coconut tree trunks cut to make steps are formed by leaning the trunks obliquely against the top of the wall from either side.
Though the paving of the old tracks and roads have received much attention from various foreign administrations, largely as an occupation for law breakers, many of the roads had been roughly paved before the advent of white influence. Some of the roads in itu (Supernatural beings) the one from the Faasaleleanga district commenced at the Puapua end and owing to rough work and less care he moved more quickly. He got well past the midway mark before he met his rival who had been doing better work as regards quality. The result was that the bad worker opened up more territory for his descendants for the boundary was formed where they met and where the difference in the quality of the road can still be observed. Hence the saying applied to the Faasaleleanga people: "Faasaleleanga, mata e vave" (Faasaleleanga, the quick eyed).
The moral is to do things quickly, no matter how roughly, in order that you may win.
A very fine, high-built road extends between the two ends of the Fitiutan village and forms an important boundary between the inland and outer parts. Though built up and improved in modern times, its foundation was laid in olden days.
The stones are naturally selected to form a flat surface but there is no attempt at cutting them to fit, though, of course, the natural stones are fitted as far as possible along the sides.
The so-called stone house (fale-ma'a) known as
I visited
Side pillars. A number of side pillars stood around the circumference. They were not counted as it became evident that some of them had been recently propped up. Not one of them had been sunk into the ground firmly in the manner one would associate with permanent fixtures. They had been placed standing on end and other pieces of rock laid against them on the ground to keep them upright. Some were leaning over in a slanting position and one was leant against another to prop it up. The dimensions of those measured are given below.
The pillars consisted of basaltic prisms and columns which followed a natural cleavage. Most were four sided but some were three sided where the wider surface was curved. Only an occasional one was rectangular. In the others, the angles between the surfaces formed two acute angles and two obtuse angles. The height of the side pillars ranged from 2 feet, 2 inches to 3 feet, 4 inches. None of them was worked by man. Though some had broken off at an angle at the upper end there was absolutely no sign of the
Center pillars. The place was so overgrown and interfered with in parts by the roots and stem of a growing tree that it was impossible in the time available to clear the ground sufficiently to make a thorough examination of the floor of the site. The largest piece of stone found near the center, was 4 feet 5 inches in length and was lying on the ground. This was evidently
Rafters. Nowhere could anything be seen of the 12 feet and 6 feet lengths which formed
The quarry. A perpendicular basaltic cliff formed part of the hill at the back of the house. The exposed base of the cliff was a few yards up the hill. Some detached pillars stood out at a slight angle from the cliff face, others had fallen to its base and were scattered in short lengths over the sloping ground below. Other pieces were lying on the ground between the base of the hill and the house site. The material had fallen from the base of the cliff by natural causes and had been transported to the house site. The floor of the house in addition to being covered with short lengths of basalt had been filled in with waterworn stones from neighboring streams.
Both fate tele house. We have seen that the fair sized round house in Fitiuta was 31 feet 7 inches long by 27 feet 6 inches. The exceptionally large house in Iva was 54 feet by 45 feet. In the house at Fitiuta the center pillar was 20 feet high and in the house at Iva, 32 feet. Both
When
In assessing the value of
About 9 feet from the house site in line with the cliff, an arrangement of basaltic stone was called tia raised over a grave. (See Plate XXXV, C.)
Prehistoric remains discovered in the mountains of
A number of natural, large rocks are connected with myth and tradition. Some are connected with historical events and though pointed out, have no special significance beyond illustrating how the imagination is given full play in associating natural physical features with historical characters. Thus the number of stone seats, kava bowls, and other-impossible stone artifacts associated with the name of
The true tupua rock is the petrified body of an ancestor. Such are the two rocks, Ui and Luamaa, near the ancient village of
Some rocks seem to have been placed in particular positions by human agency and an ancestral name with supernatural qualities given to them. Such a one stands in the village of
Other natural rocks are also called tupua and are associated with prohibition. A tupua rock on the beach near Amanava in
Inquiries about quarries elicited no results except in tia) was passed. Before reaching the actual quarry, the hillside was found covered with chipped stones and large flakes. In spite of the large quantity of flakes very few discarded adzes were found. In many places there were heaps of finer chips showing where trimming had occurred. Further up the hill, on either side of the main ridge, large pits had been excavated. On the northerly side were a large, deep circular pit and a smaller four-sided one. On the southerly side, a deep, wide trench ran for a short distance in the line of the ridge and then turning at right angles ran out on the hill slope. In all the pits, the earth had worked down the sides and there was no clean-cut face of rock visible. The rock could, however, be made out on clearing away the earth and it was visible here and there in the side of the deep trench. The overgrowth and tumbled earth was too great to clear away in the time available.
On the middle of the ridge between the pits, there was a large accumulation of fine chips, much finer than those on the lower slope of the hill. The accumulation was in two distinct areas of several feet in extent and some depth.
Breaking the stone (foa). The stone or rock obtained from the quarry was broken (foa) by striking it with another rock. The pieces were then selected for chipping into shape. It would seem that the Samoan accepted the pieces that the rock broke into and over the shape of which he did not have much control. The adzes were shaped to suit the sizes of the pieces and the variations in thickness in the same general type of adz may have been largely influenced by the nature of the pieces into which the stone broke. Though a piece might be thinner than usual, it was a pity to waste a good piece of stone and a thin adz resulted.
Chipping (tanga). The verb ta means to strike, usually with something. The pieces of stone were struck with a hammer stone to remove chips and so shape the stone into a matau.
Grinding (olonga). The grinding of the chipped matau had to be undertaken where water was available. The term olo means to grind and olonga the grinding. It is only after the matau which was shaped by chipping, has had the cutting edge ground that it receives the name of to'i, the general terra for a completed adz. I say, "after the cutting edge," advisedly, because the large number of adzes in the matau from the chipping workshop above were ground. Large stones with a wide surface, standing in the stream, were used as grindstones. The chipped adzes were dipped every now and again in the stream and rubbed (olo) back and forth on the surface of the basaltic grindstones. In course of time circular hollows were formed which served the purpose better as water could be poured into the concavity and the adz and grinding surface kept wet. Some stones contain more than one grinding facet. Evidently, a deep concavity was not so good for grinding for some of the flattish stones have been turned over and grinding facets produced on them. Too sharp a concavity evidently interferes with the length of the rubbing stroke. Hence when a facet wore too deep, another was started. It would seem from the fine nature of the chips that the adz makers had shaped many of their adzes on the ground before taking them elsewhere for grinding. Judging from the quantity of chips an immense number of adzes must have been made on the spot, no doubt extending over a considerable period of time. The quarry was well known to the older men who stated that people came from all parts of matau, whilst ta is to chip. The name of the quarry is, therefore, the "chipping of stone adzes." Among, the fine chippings only one poor discarded adz was found. Diligent search discovered only two round stones that could have been used as hammer stones.
Working the stone. The attempt to obtain details as to how the stone was worked proved hopeless. Whether or not fire was used to split the stone was unknown. As stated, there was no distinct face of rock apparent in the quarry. In the larger circular pit, it seemed as if the stone had been in detached lumps that were dug out.
Samoan stone adzes show that two distinct processes were used in shaping the stone; chipping, and grinding. Before the stone was chipped, it had to be broken into suitably sized pieces. On a very large stone in the stream near Nua, tia of the Tuitele family at olonga refers to the grinding process foanga. A historical grinding stone near Fitiuta (about 37 inches long and 17.5 inches at its widest part) is called
Another type of grindstone is provided by the flat lava formation seen in some parts on the seashore. At
The above were the processes observable in Samoan adzes. Grinding as we shall see was limited to parts of the adz and rarely were all surfaces ground. The constant parts ground were the anterior and the bevel surfaces, and the parts of the lateral surfaces to trim off any sharp edges left by the chipping. In spite of the number of grindstones, the impression left by a study of large numbers of Samoan adzes is that the craftsmen did as little grinding as they could.
Besides foanga grinding stones, the Samoans referred to ma'a fa'amata to'i (stones for putting an edge on adzes). Whether this was merely a functional name for the foanga or another type of sharpener could not be ascertained. It is improbable that hones were used. Ground stones from
Native names. The name to'i holds for all types of adzes but qualifying names were given to the different types. The general terms to'i laitiiti and to'i tele were used for small and large adzes. A large adz used for commencing the work is a ta'i lau. Here lau probably refers to a wide surface and hence a wide cutting edge. The term to'i fau tonu is used in the second stage of cutting finer chips or shaping and to'i sila is the adz used for finishing off. It has presumably a sharper edge. An adz halfted sideways like an mele'i. It was also said to be used hafted end on like a chisel and used as a battering ram in felling a tree. An adz said to be bevelled on both sides was termed ololua (two grindings). to'i ololua as an adz ground on both sides but as some few adzes have the posterior surface ground as well as the anterior, the name ololua may possibly refer to an adz ground on both surf aces and not necessarily bevelled on both surfaces. Adzes bevelled on both sides are not present in the fairly large mele'i adz may be correct as regards use, the name mele'i sounds suspiciously like meleki the modern term coined to represent the word American.
The sele ta was described as a war adz with a cutting edge about 8 inches wide, hafted like a chisel and used in fighting with a forward thrust. It was also used to cut off human heads. Another name for a war adz is leu tasi. The term to'i vete was given to to'u is another name for adz.
An adz worn down so much by repeated sharpening on the grindstone that it is too short for use is a to'i fatu. A man who did not want to lend his adz to a neighbor importuning a loan would put him off by saying that his adz was no good as it had become a to'i fatu. These to'i fatu have naturally found their way into collections and should always be considered when classifying adzes into types and sub-types.
In Samoan ceremonial language, the terms used for adzes naturally differ from the universally used to'i. The terms ulaone, fa'alafanua, anga'ese, and 'ausulu are the names used by or to chiefs. It is probable that they indicated different types. Other terms were applied to adzes to indicate the type of lashing.
The terms used follow as far as possible a compromise arrived at by
An adz is a cutting implement of stone, shell, or other resistant material with the cutting edge running transversely to the long axis of the haft. It is described as if in the following position: the long axis vertical, the cutting edge inferior, and the bevel which forms the cutting edge, posterior.
The adz is regionally divided into two parts; the butt and the blade. The butt is the upper portion which is engaged by the lashings when hafted. The blade is the lower remaining portion. In Samoan adzes the exact line of division between butt and blade is indistinguishable unless the adz is hafted. Individual craftsmen vary as to the length of the adz included in the lashing. (See fig. 171.)
In triangular adzes with three main surfaces, the two sides meet to form a median ridge which may be in front or at the back. (See fig. 172.) Median ridge is adopted in preference to longitudinal edge, previously used (39, p. 221).
The width of an adz or a surface is the transverse measurement. The thickness is the antero-posterior measurement. "Chipping" is the process of removal by blows of flakes, large or small. (The practice of American ethnologists is to reserve the term "flaking" for removal of pieces of stone by pressure, a process believed not to have been used in Polynesia.)
The adzes available for study in
Type I. Adzes (Pl. XXXVI,
The front is convex longitudinally and usually slightly so transversely. The actual front surface, owing to coarse chipping at the sides of the butt, may be narrowed to a point, thus making the surface triangular in shape. The chipping of the back is fairly level but it is convex longitudinally and slightly so transversely. The edge is straight but may be slightly raised at the ends. In most adzes the edge is the widest part of the adz, but in a few it is slightly narrowed by grinding the lower end of the sides. (See fig. 173.)
Variations. Thick adzes. An adz from A, 2, and in figure 174.
Long, narrow adzes, showing greater length in proportion to width, are illustrated by a well-made adz from.
A small adz from
Wide, thin adzes, comparatively short, are illustrated by an adz from
Another wide, thin adz from
Type II. Adzes (Pl. XXXVI, B). Roughly-made adzes of Type II look: like unfinished forms of Type I but, from well-finished examples, the employment of a different technique to create a distinct type is obvious.
On the front, the anterior surface resembles Type I in being narrower than the back, well ground, and narrowing towards the poll. Owing to the coarser chipping of the sides, the front edges are more irregular and in poorly made adzes this feature is marked at the butt end.
The sides are formed by coarser chipping with more of an outward slant so that the edge between the front and side surfaces forms a more obtuse angle than in Type I. The side surfaces are narrower and in a side view they may occupy less space than the backward projection of the posterior surface. The back edges are well defined but irregular owing to the coarse chipping of the adjoining surfaces.
The back of the adz, instead of forming a fairly level surface, projects considerably in the middle line. In chipping the back inward from the sides, a marked longitudinal median ridge may be formed, or the ridge may be rounded off, yet leaving the back with a pronounced transverse convexity.
Though much variation occurs, the method of treating the back by leaving a median ridge or rounding it off justifies the adzes of Type II being divided into two subtypes.
Type II, subtype a. The chipping of the back results in a median longitudinal edge which divides the back surface into two, giving the appearance in cross section of five surfaces instead of four. The median ridge may be irregular, diverge to either side, but is yet distinct. The bevel is triangular though the two sides formed by the shoulder may not be quite straight owing to the depressions and edges encountered by the ground plane of the bevel when it meets the chipped surface at the back. A well-made adz from
A chipped unfinished adz from
Type II, subtype b. The chipping of the back, instead of forming two inclined planes which meet at a sharp ridge, forms one rounded surface with a marked lateral convexity which results in a curved chin. An adz from
A well-finished adz, also from
Adzes occur which combine features of Types I and II, with regard to their sides; one side may have the more abrupt slope and greater thickness of
Progressive grinding. The outstanding feature of the two types described is the reduction of grinding to the minimum required in providing a cutting edge. In adzes the cutting edge is transverse and is formed by an approximately straight anterior surface and a posterior bevel surface. The great majority of Samoan adzes range under Types I and II in the working of which the craftsmen adhered to the principle of minimum grinding. In grinding the front of the adz, practically the whole surface came in contact with the grindstone and shared in the removal of the chipped hollows. In grinding the bevel at an angle with the front, the plane of the back surface was never in contact with the grindstone and hence remained in the chipped condition, except for small trimmings at times near the chin. To define the all-important cutting edge exactly the neighboring parts of the sides were trimmed up by grinding. Except for the lower part and some grinding down of sharp chipped edges, the sides also remained in the chipped condition.
The slight grinding of the sides, however, introduced a slight departure from the principle of minimum grinding, which may progress to the extent of grinding the sides over the full extent of the lateral surfaces as in the
A further progression in grinding includes the back as well as the sides which occurs in an
The two preceding adzes though typical in shape to Type I have introduced new departures in the extra employment of grinding. The application of grinding to all surfaces leads the way to Type III.
Type III. Adzes (Pl. XXXVI, C). Adzes of Type III are quadrangular and well ground on all surfaces except the poll. The back is still wider than the front but the difference is not so marked as in Types I and II, owing to the steeper transverse slope of the sides. The narrowing of the adz towards the poll is not so marked as in the previous types. All longitudinal edges are straight and the chin is marked and straight horizontally. The front, back, and sides are convex transversely and the front and back slightly convex longitudinally. A good example from
The lower end of a broken adz from
The
In the three types of adzes described, the common feature is that the back is wider than the front. A reversal in the width of the two surfaces marks a new type.
Type IV. Adzes (Pl. XXXVI, D). In adzes of Type IV, the sides instead of sloping outwards from the front, slope inwards and backwards. The inward slope is not a mere accident due to overgrinding past a right angle, but is due to deliberate technique to make the back narrower than the front. It is thus a reverse technique to that of the three preceding types. The inward slope is marked so that the sides form acute angles with the front at the front edges. The slope of the sides may end in longitudinal edges which define a narrow back surface or the slope may be continued to form a rounded back without edges. The treatment of the back results in two subtypes.
Type IV, subtype a. The subtype may be termed the reverse quadrilateral as the front is wider than the back. The sides slope inwards and meet the narrower back surface in distinct back edges. An adz from
A small
Type IV, subtype b. In the second subtype there are no posterior longitudinal edges and the sides and back form a continuous rounded surface. A small
A small adz from
A fragment from
Four adzes of Type IV were picked up on Manase,
Type V. Adzes (Pl. XXXVII, A). Two adzes, showing the reverse form to Type IV, have been grouped in Type V. One of them from
A longer and more slender adz from
Both the preceding adzes have a curved cutting edge. A much thinner adz from
Adzes which are triangular in a typical cross section taken through the blade above the bevel, are usually termed triangular adzes. The triangular cross section must be obtained by purposive chipping to obtain three surfaces. Sections through the butt are not always typical of construction. In some quadrilateral adzes of Type I, sections of the butt near the poll may be triangular owing to the rough chipping methods employed, or to the hammering process of breaking the rock into the pieces required. Of the three surfaces in triangular adzes, two always form the sides while the third may form either the front or the back. In adzes with a front surface, the median ridge
Two problems arise with the two classes of triangular adzes; hafting, and the formation of the cutting edge. In adzes with a wide back surface, hafting presents no difficulty as the flat back of the butt will rest naturally against the foot of the haft. The problem of the cutting edge has to be met by chipping and grinding the lower end of the front median ridge into a surface wide enough to form a transverse cutting edge with the bevel from the back. In adzes with a wide surface in front, the cutting edge is provided for by the technique used in quadrangular adzes, but the presence of the median ridge on the back requires special technique in hafting.
Of triangular adzes with the wide surface at the back, two well-established types occur in Samoa. Of the class with the wide surface in front, one solitary example occurs in the
Type VI. Adzes (Pl. XXXVII, B, 1). The triangular adzes grouped under Type VI have a wide surface at the back and a front median ridge chipped and ground at the lower end into a fairly wide triangular front surface to provide a fairly wide cutting edge. The width of the adz is greater than the thickness. A chipped, unfinished adz from B, 1, and fig. 195.)
Of five large triangular adzes of this type in
A study of the large adzes shows that if they were cut off below the apex of the triangular front surface, they would resemble the shorter adzes of Type I, wherein the chipping narrows the front surface towards the poll. In some adzes of Type I, it has been noted that the sides actually meet in front and make the poll end sometimes triangular in cross section. With longer pieces of stone, the tendency in Type I has the opportunity to develop and the two sides meet above the apex of the front surface to be continued upwards as a median ridge in the large adzes. Large adzes in which the chipping of the two sides does not meet, constitute Type I, but those in which the chipping of the sides does meet in front, belong to Type VI. The large adzes of Type VI may therefore be regarded as being derived from the smaller adzes of Type I.
Type VII. Adzes (Pl. XXXVII, B, 2). Adzes grouped under Type VII resemble Type VI in having the wide surface at the back and in the treatment of the lower end of the front median ridge to provide a triangular front surface to oppose the bevel in forming the cutting edge. They differ in being relatively much thicker and narrower. As a consequence of the general decrease in width, the front triangular surface and the cutting edge are also relatively narrower. The type is well established with 2 examples from
A broken adz from
Type VIII. Adzes (Pl. XXXVII, B, 3). Adzes forming a reverse triangle with the wide surface in front are not only the last type but also the rarest form in the collection. The one example was secured in
Large adzes. The collection contains 9 adzes, of which three are quadrangular, belonging to Type I, five are triangular adzes of Type VI, and one triangular of Type VII.
Very small adzes, which are plentiful, are usually well ground on all surfaces except the poll, though occasionally a hollow too deep to remove may be left, especially on the back. In shape, they group under Types I, III, and IV. An example of Type I is shown in figure 200.
Owing to the more extensive use of grinding, adzes of Type I shade off into Type III, of which a fair example is shown in figure 201.
An example of Type IV, subtype a, is illustrated in figure 202.
Shell adzes. Two shell adzes are in the collection of
Diagramatic cross sections of the types and subtypes described, illustrating the fundamental differences between them, is shown in figure 203.
Intermediate types. Adzes intermediate between Types I and II are quite common and are illustrated by cross sections in figure 204.
A number of small adzes from
The outstanding feature of Samoan technique is the rough chipping, the reduction of grinding in the common types to the minimum, and the absence of bruising and pecking. The chipping is especially coarse on the butt part of the adz. Even in well-ground adzes it is rarely that all depressions are ground out. The order of grinding, as proved by numbers of adzes in various stages of finish, was the anterior surface first, the trimming of the sides next, and the bevel surface to form the cutting edge last.
Samoan adzes come under the class of tangless adzes.
Another feature is that by far the greatest number of adzes are quadrilateral and deliberately chipped to make the front narrower than the back.
That the Samoan was on the way to obtaining the broad cutting edge of other areas is shown by the occurrence of the reversed quadrangle in Type V, and the reversed triangle in Type VIII, though only one example of the latter has come to hand.
The haft ('au) is usually made of toi wood. A branch of suitable size for grasping is selected, after estimating the suitability of the upper angle that it forms with the tree trunk. The trunk is cut transversely 2 or 3 inches above the branch junction and about 1.5 to 2 inches below it. The transverse cuts or scarfs are made about 2 inches deep or more and a piece of the trunk about 6 to 7 inches long, 2 inches thick and 2.5 inches wide is removed with the
Terminology. The haft is described in the position for use, with the shaft to the back and the acute angle formed by the shaft and the foot, below. The terms used may be followed out in figure 205.
The shaft is the part of the haft grasped by hand to swing the adz. It is formed from the selected branch.
The foot is the part of the haft set at angles to the shaft. It supports the adz and the lashing and is formed from the wood cut out of the tree trunk.
The heel is the upper part of the foot. It forms an obtuse angle with the shaft which may be trimmed off into a curve or the upper line of the handle may be continued to meet the front of the foot and so remove the projecting heel. The heel has a flat front surface which narrows towards the top. The sides of the heel are convex transversely and usually meet in a posterior median edge which runs up from the top of the shaft junction. The cross section is thus triangular. The sides also narrow but instead of reaching a point, the end of the heel is cut off with a forward upward slant to form an upper triangular surface. In some hafts, the back of the heel is rounded off. The heel gives support to the upper turns of the lashing and was devised
The toe is the lower end of the foot and forms the part projecting below the shaft with which it makes an acute angle termed "the toe angle." It supports the adz and must be a little wider than the butt of the adz. The front of the toe forms a surface which is usually flat or may be hollowed to fit the back of the butt. The sides of the toe are convex transversely and may meet in a median edge at the back or form a general rounded surface. The toe is shorter than the heel and supports the lower turns of the lashing. In adzes without a projecting heel, the toe is longer and may support the whole of the adz butt.
The front of the foot, including the heel and toe parts, is dubbed out in one plane. The adz butt is laid upon it and the craftsman decides how much of the adz will be included under the lashing. A mark is made where the poll ends. A transverse cut is made and the part of the surface below it is removed to form a deeper plane, the depth of which depends on the thickness of the butt. The right angle formed with the upper plane is termed the shoulder (le'i) and it prevents the poll from working upwards under the lashing. The site of the shoulder varies as indicated above with the amount of the adz that the craftsman intends to include under the lashing. In some adzes, the site is in line with the lower line of the shaft (fig. 205, a); in others, it is above the upper line of the shaft. (See fig. 205, b.) In hafts without a projecting heel, the shoulder site may be below the level of the lower angle. (See fig. 205, c.) In the hafted adzes in
Hafting wedges of wood termed matalafi (mata, an edge; lafi, to hide) were used when a butt such as those of Type II (fig. 205, f) would not lie evenly against a level toe surface. While a level toe front is suitable for adzes with flat backs (fig. 205, e) no amount of lashing alone will keep an adz with a median back ridge or a rounded back in an immovable position on a flat surface. Even in a haft hollowed out with a steel tool by a master carpenter in g), and that the old method of fitting was by using wedges and a flat toe surface. It was easier to build up a bed for an uneven lafi) and for all practical purposes were as good as if they formed a structural part of the toe.
The lashing. To lash is fafau, but an adz lashed to a haft is termed to'i fafao. The lashing material was always sennit braid. With the adz fitted to the foot and the poll resting against the shoulder, a sheet of the coconut fabric-like material (lau'a'a) was laid over the butt with a fair margin projecting down beyond the toe. The material was bound on by the lashing and when the adz was used, the lower projecting part of the sheet was turned back over the lashing to protect it from wearing against the wood being worked.
The lashing turns are arranged to bring out some design and so combine decoration with utility. The commonest hafting design of repeated chevrons consists of a mesial line of crossings which is also used on wall posts. As applied to the hafting of adzes, it is described under figure 206. (See Plate XXXVII, D, 2.)
Another common form of lashing, also used on wall posts, is the single lozenge design in figure 207. (See Plate XXXVII, D, 1.)
A more elaborate lashing brings out a double lozenge design. (See fig. 208, and Pl. XXXVII, D, 3.)
In the above three forms of lashing, the use of the heel in making possible the upper turns is obvious. In handles without a projecting heel, the upper turns have to be made around the shaft as in the first stage of the chevron design. (See fig. 206, d.) The technique is shown in figure 209.
The Samoan guild of carpenters now use steel plane blades lashed to the old time short handles with sennit braid. The le'i shoulder is always used. The carpenters maintain that the lashing designs now used are the old patterns originally used with stone adzes and that they have been transmitted
Another type of wedge termed olaolatina was sometimes driven in under the completed lashing to tighten it up if there was any tendency to slackness.
Straight hafts. Though no adzes of the axe type, bevelled from both sides, were seen,
The name of to'i laitiiti given to the slot lashing refers to the type of adz (laitiiti, small) and not to the style of lashing. The adz in the split handle was termed to'i pito tele (pito, navel; tele, large) and evidently likens the poll of the adz, as viewed from above, to the anatomical region mentioned.
Reverse hafting. An old man in
The contention is interesting in view of similar opinions held by some of the
"A chisel is a cutting implement which is hafted with its long axis continuous with the long axis of the haft. The motive power is supplied sometimes by pressure and sometimes by mallet blows." The above is the definition of a chisel accepted by the conference on terminology alluded to on p. 334. Long, narrow implements are termed tofi by the Sampans, but no accurate information could be obtained as to how they were hafted, or whether a mallet was used. As Samoan carpenters were seen hollowing out bowls with steel chisels and gouges hafted as adzes, it may be inferred that the larger stone implements with fairly parallel sides were also hafted as adzes. samala la'au but, as he also gives samala as the Samoan form of the English word "hammer," it would appear that the Samoans had no old word for mallet. Failing a true Samoan word for mallet, the inference is that the mallet was not used in Samoan woodcraft. The Samoan tofi were therefore probably hafted as adzes, and such as might have been hafted in the same axis as the haft were used with pressure and not with a mallet. Some of the longer ones were probably used with pressure without hafting.
The smaller implements regarded as chisels in appearance may be grouped into quadrangular, and triangular, and they follow some of the types of adzes. Quadrangular chisels may resemble adzes of Type I in being narrower in front and narrowing towards the poll. A short implement is shown in figure 211. (See Pl. XXXVII, C, 2.)
A longer quadrangular chisel resembles adzes of Type III, but the back has been ground into three planes towards the bevel. (See figure 212 and Plate XXXVII, C, 3.)
Triangular chisels. Many roughly chipped triangular implements were picked up on house platforms. They had no ground edge and resembled large drill points. One, however, that was partly ground, showed that the median ridge was in front. (See fig. 213, and Pl. XXXVII, C, 4.)
A broken triangular chisel with more grinding indicates a better-worked specimen of the preceding type. With a narrow triangular surface in front, it resembles adzes of Type VII. (See fig. 214, and Pl. XXXVII, C, 5.)
Another broken triangular chisel with a wider front surface at the lower end of a front median ridge, and a straight transverse chin, points to affinity with the adzes of Type VI. (See fig. 215, and Pl. XXXVII, C, 6.)
A well-ground but broken chisel shows the reverse triangular form with the back median ridge and a triangular bevel, thus forming affinity with adzes of Type VIII. (See fig. 216.)
"A gouge is a special form of chisel in which the edge is curved to such a degree that the bevel is hollow or grooved." (Terminology Conference.) No implements in the
A number of pieces of stone picked up on house platforms were found to be flat on one surface and chipped from the other rounded or irregular surface to form a curved edge. The shape was unsuitable for adzes and yet the curve had been deliberately worked. The problem was solved by the Samoans diagnosing them as coconut graters (tuai ma'a). Suitable pieces of stone were chipped to form a serrated curved edge which formed a better grater than an even ground edge. The grater was lashed to a wooden stand ('au sa'alo) with the flat surface upwards. A fairly thick implement is shown in figure 217.
Old stone adzes, retouched by chipping on the bevel side of the edge were also used as graters. A Tutuilan adz so treated is shown in figure 218.
In tuai to the arm of the wooden grater stand. (See Plate V, F.)
Basaltic flakes, standing out in contradistinction to the gravel, are found on the house platforms. Some of them were probably brought from the
Net sinkers were selected from suitably sized natural stones. They were firmly tied with a special cord and then tied to the bottom rope of the net. Even the small, rounded, waterworn stones seen on casting nets were held firmly by the tying technique. The stone sinkers of ordinary nets were not treated by grooving. Line sinkers do not seem to have been used in former times because the hooks in use were trolled and did not require sinkers.
Sinkers for shark bait and shark nets. Large waterworn stones, used with set nets with big meshes for catching shark outside the reef, had neat funnel shaped holes bored through from either side with a stone pointed drill to complete the perforation for a rope attachment. In a shark bait sinker (ma'afa'amalie) a chipped groove extended from either side of the perforation to meet over the end. (See Pl. XL, D, 1.) taula (anchor). (See Pl. XL, D, 2.)
Canoe anchors are termed taula. Though suitably shaped, natural stones were more commonly used, some such as the anchor of the famous Lauti war canoe that belonged to Faumuina of D, 3) were said to be tied together to form anchors for the small paopao canoes while fishing within the reef.
Stones shaped like ordinary wooden spinning tops were used in squid lures.
Discs (te'a) of coral stone, used in the game of tanga te'a, were usually selected from flat pieces which had been rounded by wave action, but some are said to have been trimmed to the right shape.
The stone nail figured by fao is unique and no information could be obtained regarding such objects.
Samoan stonework is characterized by negatives. In stone structures, cut stone was not used. The stone posts of the Fale-o-le-Fe'e, often quoted as having been cut by human agency, are natural basaltic prisms. The outstanding religious stone structures of marae for such structures exists in the Samoan form of malae but the term is applied to an open space in a village where public meetings are held. The Samoan malae has the same meaning as the marae but though both areas have retained the social significance of the term, the special stone structure associated with it in
Among the necessary implements, tanged adzes as a form of purposive technique are entirely absent. The most common type of adz are quadrangular with the widest surface at the back. The commonest types are marked by a minimum of grinding, but full grinding on all surfaces except the poll is present in many of the smaller adzes and the less common types. Triangular adzes are characterized by the widest surface forming the back while the reverse is exceedingly rare. In spite of the cruder appearance of the adzes, good work was accomplished with them as evidenced by the technique of the arches of the guest houses and the flanged plank canoes. The aesthetic sense of the Samoan craftsmen did not express itself in stone but sought some other medium. Similarly the pounding of food and other material was performed with natural objects and stone pounders and pestles did not enter into Samoan domestic economy.
The Samoans enumerated seven types of canoes differing in size and varying in construction. The different types may be grouped into two main divisions; the dugout, and the plank canoe. Both divisions have a single survivor that remains in active use. The differences between types are connected with size, involving the number of outrigger booms and whether the canoe was used for sailing in addition to paddling. Sailing involves the addition of supports for the sail, bow and stern covers, and a projecting balancing boom on the side opposite the outrigger.
All have top sides and bow and stern covers.
Outriggers are always on the left side. They consist of a float, cross booms, connecting pegs, and a connecting lashing. The float (ama) is a large spar of light wood set parallel with the canoe at a little distance and floats on the surface of the water to give the narrow canoe a wider support on the water. The outrigger booms ('iato) are lashed to both gunwales or upper edges of the sides of the canoe and project out to the left over the float. Being straight they are connected with the float by an indirect attachment of connecting pegs (tu'itu'i) and a lashing of braid (li) which extend between booms and float.
The 'alia voyaging canoe disappeared as foreign transport afforded an easier way of getting to the various islands in the group. Its disappearance was hurried by the coming in of the taumualua as far back as 1849, though the 'alia survived for some time after that. The taumualua was created from a foreign model and made wide enough with sennit-sewn planks to dispense with the outrigger float. It was much used in military operations to convey armed troops, and its sides could be barricaded to protect it from gun fire from the shore. It was paddled after the Samoan fashion with the paddlers facing the bow and propelling it like an ordinary canoe. As inter-district wars died down, the taumualua in turn gave way before the fautasi, a boat built purely for transport in the form of a large whaleboat with planks nailed together and rowed with oars resting in rowlocks. The fautasi in turn are rotting in their boat sheds, as the desire for speed and less labor has reached fautasi are also community boats which require large crews. They are unsuited to the needs of the few. Governments have also extinguished any flickering remains of the Polynesian voyaging spirit by prohibiting travelling in boats between distant islands owing to the danger. In this way, the descendants of one seafaring race is protecting the descendants of another from the element that made their ancestors famous.
The 'iatolima and amatasi are no longer seen but an odd soatau still survives as an interesting relic of the past. The Samoans have reduced the canoe building craft to a minimum. Only such craft as have a material advantage which cannot be otherwise supplied are made. The types still made have therefore dwindled down to two, the va'a alo (bonito boat) and the paopao (dugout). The bonito still swims in Samoan waters and no easier method of securing them has been devised than the pearl shell hook trolled in the wake of a fast-paddled plank canoe. Even now change is taking place for more and more bonito canoes are being made of dugouts. It only remains for a cheap kind of oil engine to be put on the Samoan native market and the lingering type of plank canoe may join its contemporaries. The paopao is a necessity which cannot be discarded. So long as the people obtain an important part of their food supply from the lagoon, so long must every family have some kind of vessel to assist in obtaining it. The paopao is light, easy to manage, and not expensive in building. The complex of the hollowed log, the outrigger float, and the direct paddle is so inground in the methods of the people that the paopao is assured of existence for many years to come.
The paopao dugout is in active general use throughout the group for fishing or transport inside the reef and is an indispensible part of every male adult's equipment in life. In fine weather, they are used in alafanga fishing with the paala hook trolled from a line outside the reef. (See Pl. XXXVIII, A.)
The timber for the hull is papaongo, pipi, fau, 'ulu, tamanu, and mosooi. A tree with a trunk of suitable size is selected in the forest, cut down and roughly shaped to lessen the weight. The stern end is invariably shaped to a knob with a constricted neck. Around the neck, a rope is tied with which to haul it to the village. While at tamanu was floated round from the village of
The hull was subsequently dubbed out at liu) was hollowed out. The process of hollowing out the solid trunk in one piece (fu'e fua) is in contradistinction to hollowing out the inner side of the planks of a plank canoe (fufu'e). The shape and the amount hollowed out is guided by experience. When hollowing the hold, the carpenter often taps the side of the hull with his knuckles to judge by the sound if more wood should be removed. He hollows down until by percussion he gets the normal note of the required thickness. The shape of the hull of a paopao is shown in figure 219.
In preparing the log, the removal of the wood to clear the upper edges of the canoe leaves a plane surface which is above the greatest diameter of the log. In further shaping the log, though the ends are narrowed to form the bow and stern, the middle part of the hull retains the original contour of the log. In excavating the hold, the dubbing out in the middle part thus follows to some extent the original curved circumference of the log with the result that the greatest width of the hold is below the upper edges. This is a characteristic feature of the Samoan paopao but naturally where the hull narrows towards the ends, the greatest width works upwards to the upper edges. The narrowing of the hold at the upper edges has been held by some writers to be so characteristic of Melanesian dugouts as to form a diagnostic point of
The solid parts of the bow and stern that project beyond the edge of the hold vary in length. In a Tutuilan canoe 15 feet long, the solid bow was 17 inches in length and the stern 7.5 inches. In some canoes, the upper surface of the solid bow and stern parts were in the same plane as the upper edges of the sides, while in others the edges were cut down at a slight slope fore and aft to place the canoe edges at a lower plane.
Many of the
The occasional knob ornamentation of paopao canoes follows the orthodox treatment of bonito canoes in which the knobs form stands for pule shells. As such shells were not easily obtained in Samoa, it is improbable that they were used in the paopao canoe.
Topsides. The typical dugout paopao has no added topsides (gunwale) or oa. In
As the canoe is for use within the calm lagoon, bow and stern covers to roof over parts of the hold and prevent waves breaking in are not necessary and consequently are not used.
The outrigger. The outrigger structure consists of two cross booms ('iato), a float (ama), and connecting pegs (tu'itu'i). (See Pl. XXXVIII, A.)
The booms. These are preferably of milo wood. In the 15-foot paopao from
In the typical paopao single fair-sized holes are bored through the sides of the canoe just below the inside upper flange, or about an inch below the edge and directly below the booms. The method of lashing is described on page 393.
The float. This is made of fau and is 10 feet 6.5 inches long, sharpened at the fore end and cut off square an inch behind the aft peg connections of the aft boom. The vertical distance between the upper surface of the float and the under surface of the fore boom was 12 inches and between float and aft boom 7 inches. The float was selected from a pole of appropriate size and usually receives no trimming except at the pointed fore end. In this particular canoe two flat upper side surfaces were trimmed to form a mesial upper ridge between them. The float was approximately parallel with the middle line of the canoe.
The above measurements give an approximate idea of the canoe for comparative purposes. As will be seen in describing the more elaborate bonito canoe, measurements were not made beforehand, but the set of the outrigger was judged by eye and after lashing the parts together the superfluous ends were cut off.
The connecting pegs. These are of two types; separate pegs, and the branched boom. Separate pegs consist of two pairs to each boom. They are made of ironwood rods a little over half an inch thick. The inner pair are connected with the float by pointed ends stuck into' holes on the upper surface of the float to the canoe side of the middle line of its long axis. The upper ends incline inwards, one on either side of the boom to which they are lashed. The outer pair are fixed similarly below to the outer side of the middle line of the float, and above they incline outwards to embrace the boom to which they are lashed together. Below, the two sets of holes are between 1 and 2 inches apart across the middle line. The individual members of each pair are
The booms are perfectly straight so the connecting pegs fill the distance between the booms and the level at which the float is placed. In most of the paopao examined for this feature, the outer pair of connecting pegs are longer than the inner pair on the same boom. This is not so noticeable with the aft boom, where the distance between the boom and the float is shorter. Thus in the paopao from paopao may be due to less careful work. As stated, the relative position between float and boom is decided without the pegs. The pegs are afterwards lashed on to maintain the relative position, and one craftsman may merely place them against the boom and lash them in position without considering the other pair; another craftsman, after fixing the inner pair, may measure the outer pair by laying them against the inner pair and so arrive at equal lengths and equal angles. The relation of the upper ends of the outer pair of pegs to the outer end of the boom does not come into consideration. There is no fixed outer end to the boom until the upper ends of the outer pair are lashed to the boom. The outer part of the boom is then cut off 4 inches to the outer side of the lashing. This, however, may be after technique due to sharper tools.
The complementary lashing. The peg type of connection keeps the float off at the proper distance but, owing to the insecure connection below, something is needed to keep boom and float from coming apart. This additional connection, by lashing with braid, is complementary to the peg connection. It is easy to pass turns of braid vertically round both boom and float but if the turns are taken under the float, they will be subjected to much wear and tear, as the canoe is dragged up on the reef or shallow parts within the lagoon. In the paopao canoe, therefore, the braid lashing must not pass under the float. To obviate this, a transverse hole is bored through under the median ridge on the upper surface of the float. This is the reason for shaping the upper surface of a paopao float. If it is not shaped, then a hole must be bored downwards and inwards on either side of the middle line to meet. Through the hole, a number of turns of braid are passed which loop over the boom, the hole being directly under the boom. After sufficient turns, transverse turns li, thus keeps the float attached to the boom and by passing through a hole through the upper surface of the float, is saved from wear on its under surface. The details of lashing the pegs to the boom are described on page 398.
The branched boom. This is of milo selected with a small branch of the right size to form a connecting peg. The lower end of the branch is sharpened and inserted firmly in a hole made in the upper surface of the float. An example was seen at
The seat. The paopao is a one-man canoe. A seat (nofoanga) about 6 inches wide is dubbed out of any convenient wood and the outer ends usually cut away on the under surface to fit over the gunwale of the canoe and so lock it into position. When a man hauls up his canoe, he usually picks up the seat and carries it into his house together with his paddle. The position of the seat is in front of the aft boom. The canoe will of course hold two in an emergency, but the gunwale edge comes down to the water line. For more than one the larger soatau dugout was made. A longitudinal pole is sometimes tied to both booms towards their outer ends.
The float is deeper forward in construction, but when the canoe is in the water with the weight of the occupant more astern, the aft part of the canoe sinks down and the float is then level in the water as the bow part of the hull is further out of the water.
The outline of the hull in a well made canoe is regular. Most paopao not so regular, may look very badly made as the side edges are often twisted to one side or the other. Much of this irregularity may depend on the timber, or on the individual craftsmen. The paopao canoes are made by the householders who are not expert carpenters. A master builder while enumerating the canoes made by the carpenters' guild omitted the paopao. On my mentioning it, he smiled and said, "The paopao is not a canoe." Neither is it from the expert point of view. In the eyes of the guild they rank with the cooking houses and are beneath their dignity to build. Hence, unskilled labor gets employment to its own content for no one would pay the price of skilled labor for either paopao or cooking house.
Rod rests. In
The soatau is hollowed out of a longer section of tree trunk than the paopao, and is provided with a longer float and an extra outrigger boom. Top-sides, bow and stern covers are not added. The general lines of the bow are similar to the paopao but the stern is wider and cut square across without the hauling knob characteristic of the paopao. A full sized paopao figured by
A model in B), shows a more sloping stern but the divergence of the lower ends of the connecting pegs is present. The connecting pegs are in two pairs but their lower ends are more widely apart in the long axis of the float than they are in the paopao. The upper edges of the sides are provided with inner flanges under which the lashings of the straight booms pass. The float is cut off square close behind the aft connecting pegs, while the fore end reaches the level of the bow. Longitudinal poles are attached to the outer ends of the booms close to the lashings of the connecting pegs. A complimentary lashing to the connecting pegs was also used. The only soatau seen was at
The 'iatolima ('iato, outrigger boom; lima, five) receives its distinctive name from having five outrigger booms. It is an advance on the soatau in size with additional booms and being sailed, it usually has additional depth provided by topsides and protection from an inrush of water by bow and stern covers. A suati balancing spar also projected from the right side of the canoe.
No 'iatolima was seen, but soatau which is rigged as a 'iatolima, including the sail. The hull is undoubtedly a soatau dugout with finer lines at the bow and stern. The upper edges have wide inner flanges and there are no topsides, or bow or stern covers which my informants stated were characteristic of the 'iatolima. On the other hand, there are five outrigger booms; a forward pair, a middle pair, and an aft single. Each boom has its own set of two pairs of connecting soatau. The float is cut off short behind the aft connecting peg. Two longitudinal poles are lashed to the booms above the connecting peg lashings. The suati balancing spar is present. The characteristic features of the rig of the 'iatolima may be followed out in figure 223, taken from
The significance of the five booms can now be followed. Three booms would have been quite enough if the canoe had been propelled by paddles alone, but the use of a sail necessitated special provision for a balancing spar and the support of the sail spars. The balancing spar was provided for by placing two booms close together in the middle with enough space for the spar between them. A cross piece was lashed above the booms over the left edge of the canoe and another similar cross piece was lashed above the right edge of the canoe. The inner end of the spar was placed under the left cross piece and as it projected out over the right side it rested on the second cross piece to which it was lashed. The forward pair of booms were placed closed together to give support to the lower ends of the spars supporting the sail.
The 'iatolima has the distinction of being the only dugout that was sailed and it received its name from the special technique of providing five booms to enable a soatau hull to be sailed. The provision of topsides, and bow and stern covers are improvements to a craft that sails out beyond the reef.
Before passing on to plank canoes, it may be stated that dugout bonito canoes are now being made quite extensively in
The bonito canoe (va'a alo) was built for speed so as to keep up with the schools of fish being pursued by the bonito. To obtain speed, the hull had to be made as light as possible. The size of the canoe was no problem as trees larger than the canoe were readily obtainable and were used in the soatau and 'iatolima types. To get the hull thin enough, it was easier to control the thinness of the material by dubbing out short sections of planks than by excavating the whole hull in one piece. Of later years, better control over a one-piece hull has been obtained with the sharper steel adzes, and has led to the manufacture of dugout bonito canoes. Before the advent of steel adzes, however, the technique of the plank bonito canoe had become established and many craftsmen despise the dugout bonito canoe as not being true to type. The manufacture of the plank canoe came within the field of the guild of expert carpenters. The canoe is made in the old style except for the use of steel adzes.
Besides timbers and adzes, a good supply of sennit braid is needed for the lashings. The braid is made up into working hanks which in canoe building are distinguished by the term tanganga instead of being called i'o fanga as in house building. A length of braid is used for measuring. For marking the wood, a piece of charcoal, or some chacoal mixed with water in a half coconut shell with a piece of coconut leaflet midrib, are needed. For fitting the planks, red earth is mixed in another half coconut shell and a section of coconut husk (pulu) is used as a brush. A coconut leaf to supply leaflet midribs to serve as needles, and some bits of wood to form wedges are also on hand. Holes are foafoa (Terebra) shell with its spiral whorls running to a point was used as a drill or gimlet.
The canoe hull was built in a canoe shed (afolau) or in an unoccupied dwelling house.
The hull. The plank hull consists of separate sections of which the keel, bow piece and stern piece are single elements sharing in the construction of both sides. The other separate pieces of plank (laufono or lauva'a) are arranged in two tiers (taloa) of which the lowest, consisting usually of five pieces, is termed the laulalo (bottom tier), and the upper the laulua (second tier). The vertical or oblique joins between individual pieces are termed tautu while the horizontal joins between tiers is distinguished as aufono. The arrangement of pieces in the hull is shown in figure 224.
The five sections of the lower tier have been given individual names by the carpenters as they have to be very carefully shaped and fitted to form the foundation lines of the canoe and are often referred to during work. The four sections of the upper tier have no individual names but are referred to as ola o le laulua (sections of the second tier).
A flaw in an otherwise good piece of timber was met by cutting out the flaw and putting in a patch rather than to waste material. Such patches (fa'asosolo) are often seen in perfectly good canoes and their presence does not depreciate the value of the canoe. Canoe builders were masters of shaping, fitting, and lashing, and though they had a general rule as to the number of sections to be used in each tier, they had no hesitation in altering details to make the available material suit their purpose.
The keel. After assembling the material, work commenced with shaping and setting up the keel (ta'ele). Bonito plank canoes were seen in various stages of construction throughout the group but the actual setting up of a keel
The keel is made in one piece of ifilele or talie wood and may be simple or compound in form.
The simple keel in the C) is 23 feet 5 inches in length and ranges in width from 1.3 inches near the pointed ends to 4.5 inches in the middle, while the greatest depth is 2.5 inches. The greatest width is that of the convex under surface which is exposed throughout its length. At either end, the keel is triangular in section, the two inclined sides meeting in a median ridge. As the keel widens out, a grooved upper surface is dubbed out in such a way as to provide flanges for the lashing technique. For details see figure 225.
In some canoes two raised longitudinal ridges are left on the upper surface close to the inner side of the lashing flanges. (See fig. 225, d.) They are situated in front of where the aft boom will cross and are called tali tata (tali to receive; tata, bailer). The ridges are long and protect the lashings as the bailer is scooped along the bottom.
The compound keel has the complete stern piece cut out of the solid with the keel, and also raised portions at the bow end (fig. 226). This form of fa'autouto. Keels vary in dimensions, one seen in
The individual sections are shaped and fitted to the keel, the order being the stern piece, bow piece, and the five sections of the lower tier.
The stern piece. The separate stern piece (taumuli) has two main features in construction; the narrow part aft is solid with a groove cut on the under surface to fit the solid triangular part of the keel, and the forward part is hollowed out to form thin sides. When the lower, and upper sides are freed by hollowing out, these edges have to be provided with inner raised flanges which are necessary for the lashing technique. The right edges have also to be provided with flanges for lashing to the next plank section. The compound keel requires no preparation for attaching to the keel but it has to be hollowed out and flanges provided at the free upper and front edges. In shape the compound stern piece follows more on the lines of the paopao dugout. For details regarding both forms of sternpiece, see figure 227.
The bow piece. The bow part is generally referred to as the taumua but the actual bow piece is termed pale in la'ei in taumua to the bow cover. Where the term taumua is applied to the bow piece, the bow cover is distinguished simply as tau.
The bow piece is dubbed out of the solid, the forward narrow part remaining solid while the aft part as it widens out, is hollowed out to form the two diverging sides. It runs the full depth of the deep bow and is hence attached to both sections in both tiers. See figure 228.
Keel fitting. It is most important that the lower edges and lower surfaces of all pieces and sections should fit exactly against the sides of the keel. Exact fittings are made by using the mixture of red volcanic earth (eleele) which must not be confounded with the special red earth ('ele) used in dyeing bark cloth. The mixture with water is termed sama, the husk brush becomes au sama, and the process of marking the boards is termed ango. The mixture is applied to the narrow side surfaces of the keel and the stern piece fitted over it to take the impression of the wet sama. The parts marked on the stern piece are carefully chipped off and by repeated trials the fit is made perfect. The same procedure applies to the bow piece and all plank sections.
Temporary lashing. In order that all end edges as well as lower edges may be exactly fitted, a commencement is made by temporarily lashing the stern piece to its place over the keel. The lashing through the solid grooved part is different to that through the flanges of the diverged lower edges. The temporary lashing is made through the flange part and the solid part is left until the permanent lashings are made. With the stern piece in position, the flanges in its fore part fit exactly with the flanges of the keel. Flanges to the inner side of all edges made for the purpose of lashing are termed fa'aopoopo. The first step is to bore holes through both flanges as in figure 229.
The lashings are made through the paired holes and round the flanges on the inner side so that they do not appear on the outer side of the canoe. Though the lashing is temporary, the special technique involved is also used
The completed lashings consist of a number of transverse turns over the flanges and neither the commencement nor the ending of the lashing can be seen. In temporary lashings, the turns are not made very tight but wooden wedges (tina) are driven in under the lashing to make them taut. When the permanent lashings are made later, the wedges are easily driven out and the slack temporary turns of the lashing are more easily removed than if they had been made taut originally. The piece is lashed to both sides of the keel and the fitting of the bottom tier of plank sections is proceeded with.
The lower tier. The shape, position, and names of the individual sections of the lower tier have been described under figure 224. They are all made of fau wood dubbed out to the required individual shapes with the appropriate outer curve. Each section is 1.5 inches thick. Fitting commences with the companion of the stern piece and the sama mixture is used to make an exact fitting with the keel and then with the forward edge of the stern piece. For exact fitting, the section must be temoporarily lashed to the keel. The temporary lashings differ from that of the stern piece in that the lower tier sections are fitted as regards their edges but the lashing flanges have not yet been provided. For the time being slots are cut on the inner surface to coincide with the holes through the keel flange. (See figure 231.)
The other sections of the lower tier are all fitted and lashed in the same manner and last of all the bow piece. When the carpenter is satisfied with the fittings, and that the right angle with the keel has been obtained, the temporary
Dubbing out the side pieces. The dubbing-out process (fufu'e) is to complete the flanges at all edges and reduce the remainder of the section to the permanent thickness which ranges from 0.25 to 0.5 inches. The flanges at the edges are kept at the original plank thickness of 1.5 inches or slightly less. The sections are carefully sloped in from the edges to get the inclination of the flange and the remaining material dubbed out. On the lower edges, the flange portions in the temporary slots are made continuous. In addition to the flanges, intermediate raised ribs (fa'aau or iviivi) are left at intervals to strengthen the plank. A hook projection (fa'alave) is also provided for. (See figure 232.)
After the temporary fitting of the lower tier, the expert carpenter knows exactly where the booms and seats will be. In dubbing out the pieces of the bottom tier, he leaves strengthening ribs at appropriate places to form foot tulanga vae (tulanga, standing place; vae, feet.)
Permanent lashings are made after the side pieces have been finished off with flanges, and ribs and made as light as possible. The lashings through the flanges are a repetition of the temporary one but each turn is carefully made and pulled as taut as possible as no wedges are permitted on the finished canoe. The attachment of the solid parts of the stern and bow pieces to the triangular section of the keel created a problem that was met as shown in figure 233.
The lashing of the solid parts of the bow and stern pieces to the simple keel involves a distinct technique. It is probable that the compound keel which avoided the above problem is the older technique.
Before the pieces and sections are permanently fixed, heated breadfruit gum is smeared over the joining surface of the keel and acts as caulking. The best gum is obtained from the varieties of breadfruit known as puou and 'ulu uvea.
The lashing of the end edges is exactly similar to the lashing with the keel flange.
The second tier. The four sections of the second tier (laulua) are shaped fitted, and lashed in the same way as those of the lower layer. The upper edges of the lower tier being fixed, the red earth mixture is painted over them and the upper sections trimmed accordingly. Care is exercised in getting the proper outward inclination to form the lines of the canoe. The planks of the second tier are wider than those of the lower and bring the canoe up to full depth except for the top sides and the bow and stern covers. When the second tier is temporarily lashed, the long upper line is marked from end to end with charcoal and the upper edge of each section trimmed accordingly when the temporary lashings are unfastened. After the permanent lashings of the second tier, the canoe hull is finished off with the top sides and covers at the bow and stern.
The gunwale. In the type canoe being described, the 12-foot gunwale (oa) covers the middle part of the upper edge while the remaining parts are filled in by the bow and stern covers. The gunwale adds another 2 inches to the depth of the canoe. The shape in section and method of attachment is shown in figure 234.
In pu fangota, in which the correct number to bring luck in fishing (fangota) is 15 on the right gunwale and 16 on the left and all must be on the inside. In the type
The bow cover. The bow cover (tau) is flat. It covers that part of the canoe between the bow ends of the gunwales and the small upper surface of the bow piece. In the
The lashing of the bow cover to the gunwales receives the special name of tuanga. (See fig. 236.)
The lashing of the bow cover to the side pieces was done by two methods. Those with a method of joining two pieces of the cover together are shown in figure 237.
The stern cover. The stern cover (velo) is shorter than the bow cove and covers in the space between the ends of the gunwales and the upper surface of the stern piece. In addition to the ornamental shell stands in the middle line, it carries the important fishing rod post (pou'ofe). (See figure 238.)
The shell stands are termed tulanga pule (tulanga, stand; pule, ovula shell) from their function but they also receive the specific name of salue.
The stern cover is lashed to the gunwales and the side pieces in exactly the same manner as the bow cover.
The outrigger booms. In bonito canoes, two booms ('iato) are connected with the float and a middle one is short. The booms receive the iato mua, 'iato muli and 'iato loto). The fore and aft booms were made of poumuli poles averaging 2.6 inches in diameter. The aft boom was 4 feet 10 inches long and the fore boom 2 inches longer. The middle boom was 3 feet 9 inches long and barely 2 inches in diameter. On the hull, the aft boom was 4 feet 5 inches from the fore edge of the stern cover. Between the aft boom and the middle boom, the distance was 4 feet 2 inches and between the middle and fore booms, 3 feet. The booms were laid across the canoe and the correct positions marked across the gunwales with charcoal. A single large hole to each boom was bored through the gunwales an inch below their upper edges and directly below the middle of the boom. The boom is replaced in position, with one end projecting but slightly over the right gunwale while the thinner end extends beyond the left gunwale. The under surface resting on the gunwales may be trimmed flat. The usual boom lashing with the single lozenge design and two alternative forms are shown in figure 239.
The float. The float (ama) is made of fau, a piece of the required thickness being selected and merely the bark removed. The length cannot be determined until it is fitted to the canoe. The rule is that the fore end comes level with the cutwater of the bow. When the position of the aft pair of connections with the boom has been determined on by temporarily trying in position, the aft end is cut off square about an inch behind the insertion of the pegs into the float. The short, square aft end of the float is typical Samoa. Never under any circumstances is the aft end of the float left or sharpened. The fore end is trimmed to a blunt point by cutting away upper surface and the sides, but very little of the bottom. The float of the
The float does not lie parallel with the keel of the canoe but is closer in towards the bow. At the aft boom, the nearest part of the float and the keel are 38 inches apart while at the foreboom they are 33 inches apart.
Attachment of booms to float. The booms being straight, the attachment to the float is always indirect by means of two pairs of pegs (tu'itu'i) to each boom. The middle boom, falling short of the float has no connection with it, but serves as a brace to the gunwales, a back support to the front seat, and an additional means of carrying the canoe if needed.
The connecting pegs are made of ironwood (toa) about 0.7 inches in diameter at the thicker lower ends. Their length is decided when fitting takes place. The lower ends are sharpened for insertion into holes in the float, and the bark is usually left on the pegs. The fitting and lashing of the booms takes place outside the shed, as after it is done the canoe is carried down to the lagoon and tested. The float is laid on the ground at its correct distance from the canoe and then raised by putting pieces of wood under it to get the right level with the keel of the canoe which rests on the ground. When the float is adjusted to suit, stakes are driven into the ground to the outer side of the float and the long booms tied to them to maintain their relative position with the float.
The aft boom is lashed first. The pegs consist of an inner and outer pair, each of which embraces the boom and is lashed together as a pair. The boom and the float being in correct relative position, the pegs are tried in position. The inner pair is tested and the length of the outer pair made the same with the result that the angles made by each pair with the boom and the float are the same. The points where the four lower ends touch the float are marked and four shallow holes not more than an inch deep are drilled in the upper surface to form a pair on either side of the middle line. Transversely between pairs, the holes are 1.75 inches apart and longitudinally between individuals of a pair, they are 2.25 inches apart. As the diameter of the boom at this part is 2.6 inches, the individuals of a pair diverge slightly at the upper ends. The sharpened lower ends of the inner pair are placed in the holes and the upper ends directed upwards and inwards towards the canoe to embrace the boom and get the right angle of inclination. Holding them together against the boom with the left hand, the upper ends were marked so that they come above the point on the outer circumference of the boom that they touch but barely up to the highest upper level of the boom. They are then cut off square at the marks. In this way the length of the pegs is decided. It happens in the
The outer pair of pegs is lashed in the same way as the inner pair. The distance between the peg pairs on the boom is 10 inches. The
The depth from the under surface of the aft boom to the under surface of the keel in the
The lashing of the fore boom differs only in the greater depth of the float from the boom. Thus where the nearest parts of the aft boom and the float were 8.5 inches apart, the distance of the fore boom was 12 inches. The pegs were correspondingly longer, each pair being 17 inches as against the 12.5 inches of the aft pegs. The lower ends were the same distance apart (1.75 inches) transversely, but the individual pegs of the same pair were 3 inches apart. As the boom in this part was 2.5 inches in horizontal diamteer, the upper ends of the pegs converged slightly as against the slight divergence of the aft pegs. Neither the slight divergence or convergence is of much moment except to show that the lower ends of each pair of pegs is approximately a similar distance apart on the float as they are on the boom, which is important as a type of local technique. The upper ends of the pairs were 15.5 inches apart. (See Plate XXXIX, B.)
The peg connection with the float is not made secure. The points are placed in the shallow holes of the float, not to hold the float to the boom but to keep the lower ends from slipping off. Strictly speaking the pegs in a bonito canoe are not really attachments. They act as rigid struts to keep the float at its proper distance from the boom. When the canoe is in the water, the force is mostly downwards against the float from the boom through the pegs. The float withstands this downward thrust and so balances the canoe. When the canoe rolls to the right, however, the booms lift and there is a strong tendency for the pegs to come out of the shallow holes in the float. Also in carrying a canoe, the float will drop off through its own weight if it has to depend on the connecting pegs alone for its attachment to the boom. A suspensory sennit is therefore used to supplement the peg attachment.
The fallacy of generalizations is well illustrated by the pegs of the Samoan bonito canoe. In Y-shaped connecting peg is used without any additional lashing. The forked limbs are attached to the boom and the straight stem driven well down into the float. In some cases, it shows through on the under side of the float. Canoes are sailed with the float at times well out of the water as the canoe heels over. Here the connecting peg serves as a suspensory attachment as well as strut and so differs materially from the Samoan connecting peg.
Suspensory float attachment. The suspensory attachment (li) between the boom and the float is of vertical form, the braid passing around the boom
In
The li lashing, therefore, keeps the float attached to the boom when any pull comes on the float or boom that would tend to separate them. It also keeps the float firmly up against the lower ends of the pegs and helps to keep the li lashing which pass round the bottom of the float are thus not worn by any friction. When the peg connections and li lashing have been made, any extra length of boom is cut off about 4 inches beyond the outside peg lashing.
The steering seat. In
The steering seat (nofoanga), however, is specially made as it takes part in supporting the lower end of the bamboo fishing rod for the use of which the canoe was made. (See fig. 243.)
With the seat in position a length of braid is passed through each of the outer holes of the seat and a loop taken around the backward projections cut on the piece in the bottom tier (tatao) already mentioned. This projection (fa'alave) is behind the steering seat under the stern cover. (See fig. 232, b.) The turns of braid (laoa) prevent the seat moving forward. Some seats have holes through the front flange of the seat and have laoa lashings passing fa'alave projections formed on the planks of the bottom tier in front to prevent the seat from sliding back.
Through the two inner holes a loop is formed of a few turns of sennit braid. The loop which projects backward is then closely seized with spiral turns of the braid so as to stiffen the loop and keep it open. The loop ends are on either side of the groove at the back of the seat. The two form an opening into which the small knobbed end of the fishing rod handle is set, while the rod is leaned back against the slanting surface of the rod post on the stern cover.
Much difference of opinion exists regarding the correct names of the various parts for holding the rod, but futia. The handhold (pu'enga) is shown in figure 243 a, 6.
The diagonally lashed stick forms a hold for the left hand of the steersman as he leans forward to swing in the bonito rod with his right. In pu'enga and in manu. The handhold should be grasped quietly without any sudden jerk and the swing of the rod made evenly and smoothly. This applies to the activities of life, hence the saying, "Ave malu i le pu'enga" (Grasp the hand support gently).
Outrigger rod rests. In action, the bonito rod stretches obliquely upwards over the stern with its lower end supported by the futia loop and the rod post. In going to or returning from the fishing grounds, the rod lies horizontally across the outrigger booms. There are two supports for the rod to prevent it from rolling off the booms: a, the fore support (lango'ofe) consists of a forked stick attached to the outer end of the fore boom and one of the connecting pegs (see fig. 244); b, the aft support is tied to the aft boom closer to the hull than the fore support, being on the hull side of the middle of the boom. It is also low, the fork or loop being level with, or just above, the boom. Three forms seen are shown in figure 245. In manga mate and in umele, which differs from the usual Savaiian meaning of that word. The narrow knob on the butt of the rod handle rests on or in the rear support. The rod thus lies obliquely between the rests with the handle part within easy reach of the steersman, and the outer end well raised.
Shell ornamentation. The properly set-up bonito canoe has a row of shells down the middle line of both bow and stern covers; one mounted on each knob (salue). The numbers vary, but in the canoe from pule (Ovulum) and are difficult to get in Samoa. They came in the way of presents (taulanga) to high chiefs. Some of the sets have been in the possession of some families for a considerable time. When a bonito canoe rotted and was abandoned, the shell ornamentation was transferred to the new canoe.
Each shell had a small hole cracked through the back not far from the rim to allow the sennit braid lashing to pass through. Commencing at the stern with the craftsman on the right side of the canoe, the lashing commences with the end support on the left. There is one shell to each support. The raised supports have a hole drilled through each from side to side
In B.)
The lashing of a shell to the end of the middle boom was a sign of distinction in
Owing primarily to the scarcity of the proper ornamental shells in Samoa and secondarily to the breaking down of old time values which accompany the establishment of a foreign culture, the shell ornamentation has been abandoned in most parts, with the exception of the
Timber used. It can now be appreciated more fully that various timbers were used for different parts of the same canoe. In the canoe from fau but the bow piece was of breadfruit wood and the stern piece of fu'afu'a. The right gunwale was of pata, and the left of fau, the bow cover of fu'afu'fa, and the stern cover of futu. The float and seats were of fau, the booms of poumuli, and the connecting pegs and handhold of ironwood. In the selection of timbers, expediency plays a large part. There are always alternative timbers and if one kind is too far away in the forest, the nearer one is used. Timbers that have been proved unsuitable were, of course, never used. The craftsmen knew the material of their particular
Repairs. When damage occurred to a plank resulting in loss of material, the whole plank had to be removed and a new one of the same size made, with flanges at all edges for joining. Patches could be put in before the sections were thinned down, as the thick material allowed flanges to be made at any shaped edge. Similarly, cracks could be repaired by the flange method, if the cracks occurred before the timber was thinned. Cracks in finished canoes could not be repaired by the flange method so alternative methods were adopted, as shown in figure 247.
Two-rod fishing. In matila) is sometimes carried in addition to the full length rod. The rod post is wider and has a lower groove on the right for the short rod. At the back of the steering seat there are also two sennit loops (futia), the right one again being for the end of the matila. The forward high-forked rest on the fore boom has two forks; the right one being for the short rod.
The amatasi was described to me by an old man in suati) projected on the right side. This he remembered and also the phrase "Tatao le suati" (Press down the suati), which was used when the canoe heeled over on the outrigger side.
The amatasi was very fast and was used in catching masimasi (dolphin) which were trolled for with a baited hook. amatasi model, the main features of which are shown in figure 248. Another of
Wilkes (42, vol. 2, p. 150) who did not see any double canoes, states that the longest canoes were 30 to 60 feet in length and built of planks with the typical flanged edges. They had a deck fore and aft and could carry 10 to 12 Ovula (marine shells) attached to a row of pegs. This part formed the seat of honor upon which the Samoan chief easily sat with safety, but one of Wilkes' gentlemen had a very precarious time in keeping on it. The song sung to him was "Lelei tusa lava le tau mua" (Good above all is the part before).
"Having both a prow and stern, these canoes cannot be manoevred without tacking; consequently the outrigger that constitutes their safety is, in using their sail alternately to leeward and windward, and does not, when to leeward, add much to the stability of the canoe. They carry less sail than the canoes of the other natives of Polynesia and to guard against the danger of upsetting, the natives rig a sprit or boom (
suati) projecting from the opposite side to that on which the outrigger is fitted. The boom is secured with guys to the top of the mast. When the wind blows fresh, some of the men go out upon it and thus balance or counteract the force of the wind. Those on the other side of the canoe are kept ready to go out on the outrigger when that becomes necessary. The sail is made of mat, of a triangular shape, with its apex below; some of these are ten feet high.
From the Wilkes' description, the canoes referred to were amatasi. As all canoes except the recent taumualua and the 'alia had one float (amatasi), it seems likely that the use of the name amatasi for this type was to distinguish a larger plank sailing canoe from the double canoe.
The taumualua was so called because it had two bow-shaped ends (taumua bow and lua, two). None are to be seen in Samoa at the present time. They were described to me as a long, wide canoe made of planks with raised flange 'alia and the bonito canoe. They had raised projections at the bow and stern which were ornamented with pule shells. The outstanding feature was that they had no outriggers but were sailed as boats. They were propelled with paddles, the crew facing the bow and using the paddles in the same way as in other types of canoe. They were much used in the Samoan fighting of the last century, when barricades were erected along the sides for protection against the firearms which had come into use.
The length was 60 to 70 feet and the width 7 to 8 feet. The planks were 7 or 8 feet long, 0.5 inch thick and sewn together like the
'aliadouble canoe. The depth of the hold was about 3 feet and there were ribs 4 feet apart. It was decked fore and aft for 8 feet. The seats for the paddlers were 3 inches below the gunwale. The bow and stern pieces were ofmaliliwood and the hull ofifi lele, orfatau. A triangular sail was used with the apex down. The mast rested on top of thwarts, where it was kept in position by stays. The steering paddle was 14 feet long with a 12 inch-wide blade.
taumualua in the water, and two models. In all, the upward projections at bow and stern are present but have no bearing on ancient technique, as they are only present in the taumualua. The curved ribs extending from gunwale to gunwale across the bottom are distinct pieces of wood lashed to projections or flanges on the inner surface of the side planks. A horizontal piece on either side is lashed to the ribs and in turn supports the thwarts. One of the models shows the mast stepped to a longitudinal board attached in the middle line to two cross booms which rest on the gunwales. The mast is stayed with ropes. The triangular sail has the apex forward over the bow cover, while one side runs parallel with the lines of the gunwales, and the base is directed astern.
The whole technique is apparently native and surprising as a departure from the use of the outrigger. My Samoan informants in various parts of the group always enumerated the taumualua among the types of Samoan canoe and stoutly maintained that the canoe was not due to foreign white influence. Their proof was the native technique in building and the use of paddles. They failed to see my difficulty in accepting their claims to having invented a vessel without an outrigger. Only one old man said that the taumualua was not very ancient but must have come in about the same time as the missionaries. However, he, too, maintained that it was native to Samoa in that the Samoans made it. The claims of the present-day Samoan to the invention of the taumualua is an excellent example of the intentional fallacy of human memory. It gave them great satisfaction to think that they had a distinct type of canoe that others did not have.
However, a, p. 247) states that the taumualua was introduced into Samoa in 1849 owing to an incident in the Samoa "war of 1848-1851." Owing to damage to the property of British subjects, taumualua, which is interesting in the subsequent adaptation of Samoan technique to supply the lack of nails and sawn timber in carrying out a foreign idea.
va'a tele (large canoe) was much larger than the more recently used one which has supplanted it. Two canoes, one much larger than the other, were lashed with crossbars amidship. A thatched shed or cabin was built on a stage projecting over the stern instead of amidship as in the Tongan type. They were much larger but more difficult to handle and could carry two bonito boats on deck if required.
The double canoe used in 'alia and evidently warrants as little description as the taumualua. The canoe hulls were made of planks fitted together as in the true Samoan plank canoes.
The last full-sized canoe rotted away recently at Mulinuu near A) serves to illustrate the general principles of construction and throws light on such important points as the technique of the deck, house, stepping of the mast, and arrangement of the sail. In the model both canoes are dugouts instead of being made of planks. The one on the right is larger and longer and has the deep Samoan bow and sloping stern. The smaller canoe on the left has the bow sloped in exactly the same lines as the stern and thus from size, shape, and position represents a large outrigger float. The fact that the canoe acting as a float is as high above the water as the main canoe does away with the need for intermediate connections between booms and float. Both canoes have front and stern covers. The raising of the woodwork to which the connecting cross booms ('iato) are attached and the provision of a deck are shown in figure 249.
In the model, the raised longitudinal planks and the raised transverse planks which well in the holds are dubbed out in one piece with the gunwale and the bow and stern covers respectively. In the full-sized canoe separate planks were used, the longitudinal planks (vaeoa) being lashed to the upper edges of ordinary gunwales (oa) and the transverse planks (faniu) being lashed to the bow and stern covers.
In the Samoan double canoe ('alia), the method of keeping the smaller canoe on the left in place of the float is similar to a similar procedure adopted
va'a tele), one canoe was smaller than the other.
The deck house and upper platform. The deck house (falemanu'a) is built over the middle deck with the ends about level with the companionways.
fa'asoata type is erected and thatched to provide protection as much from sun as from rain. Part of the framework is utilized to form an upper platform (fata tele) above the roof of the house. See figure 250.
The upper platform serves as a bridge and is occupied by the most august personage who is travelling.
The mast and sail. For the few details of Samoan names for the ropes and parts of the sail and the order with which they are dealt, I am indebted to the Samoan text in fana), and the upper top mast (tomotomo) with an expanded end. (See Pl. XL, A.) The expanded part of the tomotomo looks like the end of a netting needle and the end may be open in some and closed in others. The main mast and top mast are fitted together with a slanting surf ace broken in the middle by a short surface at right angles to the main ones and lashed together with transverse turns of sennit braid.
The lower end of the mast fits against a wooden block (tulafana) which is fixed to the deck above the middle cross boom opposite to the right side of the house. (See figure 251.)
The sail (la) is triangular in shape with a wooden spar along each of the long sides and it is made of sections of plaited pandanus leaf treated like floor mats with a check plait. (See figure 252.)
The staying of the mast. On the tomotomo a little below the expanded portion are a couple of T-shaped projections situated fore and aft with the cross upwards and the whole cut out of the solid with the tomotomo. This may be merely model technique but it prevents ropes tied above them from slipping down. The mast stays are tied to the tomotomo above the projections. The mast is stayed on the right side by a single rope which is tied below to the middle cross boom ('iato) end which is made long enough to project beyond the right edge of the deck. On the left side there is a curious support called the maile consisting of two uprights let into the planks of the left side deck and lashed to them. They are on either side of the middle transverse line of the deck on which the mast is stepped. A rounded crossbar is lashed to their upper ends. To the outer side of the maile are two pairs of holes bored through the deck plank towards the outer edge. Two side stay ropes pass from the mast tomotomo down through the inner hole of each pair, up through the outer, and are then wound around the crossbar of the maile.
To accommodate the fore and aft stays there are two projecting lugs, on the inner or left side of the bow and stern covers of the right canoe, and close towards their outer ends. Through these projecting lugs, holes big enough to take a rope are bored. In the model, one end of a long rope tied to the tomotomo passes forward through the fore lug, back along the deck, through the stern lug and back to tomotomo. It is a continuous rope and being taut holds quite firm owing to the oblique pull through both lugs. This rope is called the tu'u and sets the forward inclination at which the mast is set. The angle of the mast can be set in a fore and aft direction owing to the nature of the tongue at the lower end which fits into the tulafana groove. The rope for hoisting the sail is tied to the upper spar (tilatu) and passed through the large-upper opening in the tomotomo. Everything being ready, with the stays tied to the tomotomo and the sail rope termed maea sisi (hoisting rope) passed through the hole of the tomotomo, the mast is stepped and raised. The side stays are held and the tu'u fore and aft stay rope pulled or slacked until the mast has the right angle. The side stays are lashed to the middle cross boom and the maile. The sail rope is pulled to hoist the sail, and the lower end of the top spar placed in the bow hole. When the top spar reaches the tomotomo and the set of the sail is right, the sail rope is lashed to the projecting middle boom on the right.
The canoe is steered with a long paddle. The upper platform is reserved for the high chief of the travelling party, while the lesser chiefs, adherents, and crew occupy the deck. There are other details of decoration and construction in the model, such as handrails but they do not affect the principles of general construction.
Calking. In calking the joins between planks in the larger canoes, a strip of lau u'a bark cloth was smeared with breadfruit gum and laid on the upper
Paddles. The favorite wood for paddles is milo, but tava, ifi, and manaui are used. The typical shape is shown in Plate XL, B. The handle is straight and round, or bluntly elliptical in section. Its length ranges between 34 and 38 inches, about 1.5 inches in diameter at the upper end and thins down gradually to about 1.25 inches in transverse diameter at its junction with the blade, and less in the antero-posterior diameter.
The blade is a straight continuation of the handle though there may be a slight longitudinal concave curve on the front surface. The paddle is dipped into the water in front, is pulled towards the paddler and finishes the stroke behind him. The surface towards the paddler at the commencement of the stroke is thus the surface which does the active work and will be alluded to as the front while the other surface which faces the bow is the back. The blade meets the handle on either side at a clean-cut angle and then slopes outwards and with a fairly sharp curve until the blade reaches its widest part. The widest part is not far from the junction with the handle. The length of a fair-sized paddle blade is somewhere about 29 or 30 inches. In such a blade the widest part is from 5.5 to 6:5 inches from the top. From the widest part, the sides slope evenly down to a blunt point. The back of the blade continues straight down from the back of the handle. Longitudinally it is straight but may show a slight concavity owing to the slight hollowing out of the wide part. Transversely the back is slightly concave in the wide part due to deliberate working to thin the blade. The front of the blade commences with the full thickness of the handle. This thickness may be continued down the middle line as a distinct raised rib of the full width of the handle above and narrowing gradually until it merges in the front surface, or it may continue down as a median edge from which the sides slope evenly outwards. In either, the surface slopes backwards and outwards from the middle line towards the sides. This gives the front a transverse convexity which is continued to near the point, and according to the Samoans allows the water to run off. It is to comply with this front convexity that the back is hollowed out to obtain the lightness of the paddle.
The blade below the upper thicker part is dubbed down to an average thickness of 0.3 inch with 0.2 at the edges. The greatest width of the two paddles figured is 7.5 and 8.25 inches respectively. The surfaces and edges meet at the lower point evenly without any special curves, knobs, or projections.
A variation of type is seen in the paopao paddle (Pl. XL, B, 1). The blade of 21 inches length, reaches a maximum width of 8.75 inches only 2.75 inches from the handle junction. This results in a distinct rounded angle at the widest part.
Bailers. Bailers are called tata and to bail out a canoe is tata and asu. Wooden bailers are usually made of fau. The bailers seen consist of three kinds:
Anchors. Pieces of discarded iron have displaced the stone anchors (taula) used with paopao. Some of these had holes drilled through. A fue vine which lasts longer in the water was used as a rope.
Masts. The mast (fana) of the double canoe has been described. It is called fana while the word tila is used for the spars of sails. tila is now being used for mast. In tira is used for the mast.
There is nothing to be added to what has been said about the making of the smaller dugout canoes. They were common and did not enter into the ceremonial which surrounds expert craftsmanship. The plank canoes which involved the laying of keels and expert shaping came within the sphere of the Sa Tangaloa, the builders' guild. The builders built canoes as well as houses. In the mythical tales of the early meetings of the guild, with Tangaloa himself presiding, one of the subjects that came up for discussion was whether sennit braid ('afa) should be used first on a canoe or a house. It was decided in favor of the house and sennit braid was thus used on houses before the canoes. As the houses were the better type of house made by the Sa Tangaloa we may take it that the canoes were also the better types. The story would seem to indicate that the higher development of the house preceded the making of plank canoes.
The person desiring a better class of canoe had to approach a master builder with all the ceremonial preliminaries observed in house building. The canoe and the house were on the same level. The chief, therefore, mobilized his family and his resources beforehand. He planted food crops and collected fine mats for he had to feed and pay the master builder and his associates. Everything being ready, he approached the desired builder with a fine mat and over the ceremonial bowl of kava made his request and proffered his mat. The builder replied and if he accepted the mat, the contract was sealed. If he refused, the chief sought another expert with the rejected mat.
The builder on an appointed day arrived with his party, selected the timber and did the preliminary shaping whilst the chief's family did the rough work in transporting the timber from the forest to the village. The wood was allowed to season while the builders returned home. The wood sufficiently seasoned, the builders returned and dwelt on the hospitality of the chief and his family. All the general observances described in house building were carried out in canoe building. The builders had to be fed on the best of food with variety in delicacies or they abandoned the work which no one else would take up. A member of the chief's family had to be in constant attendance to show the respect evinced by an active interest in the work, as well as to anticipate the material wants of the builders. Interim payments had to be made and if they proved unsatisfactory to the builders, they left on the pretext that they had not been treated with sufficient respect.
According to
In house building, a similar enumeration took place at the final payment.
During the building of the canoe, however, the builders could adopt a rather mean way of venting their spite on a chief when it was net deemed advisable to go on strike. They could make the canoe, if it were a fishing canoe, unlucky. There were two ways of doing this. The lashings of the topsides or gunwale to the side pieces in a bonito canoe are called the pu fangota. The correct number of lashings as already stated are 15 on the right and 16 on the left. All the builders had to do was to change that number and the canoe would never catch more than ten bonito. This may apply only to tina or mata lafi, were driven under the lashings to tighten them. In permanent lashings, these wedges were of course removed. All the dissatisfied builder had to do was to leave a temporary lashing with the wedge under it and the finest bonito hook could not overcome the evil influence of that one wedge. The wedge was left in a lashing under the bow or stern narrow part. When the bow or stern cover was lashed, the keenest-sighted owner could not locate the wedge. He found out afterwards from results or rather from lack of results.
In Savaii, I watched a master builder solemnly strike an unhusked green coconut against the bow piece of a newly-built bonito canoe, walk sedately around the canoe twice and as I waited with a camera until he struck it again sufficiently hard enough to crack it to make the contents flow, he cast it into the sea. He then pushed the canoe out into the lagoon and critically watched the set of the float and how the canoe rode in the water. The owner then waded out, got aboard and showed her speed to the admiring family gathered on the beach. Whether the coconut was a substitute for the bottle of wine of another culture, I was unable to determine. In spite of the head builder's
Every person of any status had a bonito canoe. These went out in fleets during the season and sometimes in pursuing shoals got out a good distance.
Samoan historical narratives are singularly lacking in detailed stories of long sea voyages, and present a marked contrast to the wealth: of such material in marginal Polynesia. This may be attributed to the fact that the Samoans were early located in a group of islands sufficiently large to absorb their population. Many of the trips of Samoan legendary ancestors were accomplished by swimming, which shows how little pride they took in voyaging canoes, else they would have handed on more details of the form of actual transport. 'alia and the taumualua, were both adopted from foreign patterns. The Samoan evidently not only did not have the necessity for long sea voyages but he did not have a satisfactory type of large craft in which to make them. It is fair argument to suppose that he had not evolved a satisfactory type because, though he had an organized guild of canoe builders, he had not developed, or had lost, the voyaging instinct. There seems to be no material reason why he should have gone searching for land. Thus, both
The waters within the bounding reefs provided the main source of fish so all important in furnishing the chief flesh food supply of the Samoans. These waters were combed over and over again and day after day in every conceivable fashion, from simple groping between the rocks with bare hands to skilled devices with traps, nets, and hooks. The habits and movements of various kinds of fish had been practically studied by generations of fishermen and the knowledge influenced invention and method. Method varied from the efforts of individuals to the organised cooperation of the community. The sphere of women was restricted as they were denied the use of the fishing canoe. The canoe took men outside the reef to seek the deep sea fish that came within the possibilities of their attainment. The deep sea fish pursued were the bonito, dolphin, shark, flying fish, and some that frequented the outer side of the reef. Of the migrating fish which entered the lagoons, the most important were the mullett and the mackerel. In fresh waters, the eel and fresh-water crayfish were obtained and the migrating fry of the ingana in two rivers of
Groping. Both men and women are expert at groping (naonao) in the crevices between rocks for the rock-frequenting fish which rest there. The men often use the short spear but women use the bare hands, or nowadays a piece of cloth. The narrow clefts from which there is no escape are naturally productive of the best results. The constant search that commences in childhood naturally leads to the villagers coming to know every suitable cleft and crevice in the lagoon that adjoins the village. They submerge and with open eyes swim around the rocks peering and feeling in the crevices. When a fish is caught, it is brought up, the head bitten to kill it and the catch deposited in a basket of the ola malu type tied around the wrist or slung over the back. The groping method also accompanies the community fishing with nets as the people assisting in a drive simply cannot pass suitable looking rocks without diving down and groping amongst them.
Rock heaps. To furnish extra resting places to attract fish, rocks are piled up in heaps (ma'a) in suitable places within the lagoon. In nia'a (rock), or fatuati. Piles of branching coal ('amu) are also used and called fatu'amu. These form suitable places for simple groping. An improvement is the use by women of the ola tu fish basket. The basket is held open against a suitable part of the pile while women remove the stones gradually from the other side and by taking away the cover of the fish, drive them into the basket, The fanga (trap) from its use. At high tide men form a wide circle around the pile in their paopao canoes. They then paddle slowly in towards the pile, beating the sides of the canoes with their paddles or with the bailers to drive the fish into the pile of rocks. Each canoe has a light stone anchor which is thrown overboard as it nears the rocks. The men, armed with short spears, jump overboard and surround the rocks. The fish are speared as the rocks are removed. The method employs the largest number with the least results, as many of the fish escape. A better method is to surround the pile with a meshed net, throw the rocks outside, and leave the fish enclosed. After disarranging the pile of rocks in these operations, the stones are always heaped up again before leaving. The rocks used are the waterworn blocks of coral which are not very heavy. The manini fish which come in shoals in November are caught in piles of stones that are heaped up (ma'a fa'aputu). manini to Auala near Asau in lo are also chased into the stone heaps by throwing stones at them, which process is called tilo lo.
Sticks. Battering sticks are used in another form of fishing with the ma'a piles of rocks. People surround the rocks and work in towards them, poking and pounding the branched coral amongst which the smaller fish lurk. The process of pounding the rocks is termed tu'itu'i, and being associated with the ma'a piles is termed tu'itu'i ma'a. From the general process, the sticks are called la'au tu'itu'i ma'a, and for short the word ma'a is dropped. By this means the fish are driven into the ma'a piles and caught by groping, spearing, the ola tu basket, or the net. The method may also be used without the ma'a piles in parts where a plentiful growth of the small branched coral affords cover for fish. The ola tu basket is set across a convenient small channel and the coral pounded with the sticks by women usually, who work in a decreasing oval formation with the fixed small end of the oval at the basket. The branching coral is termed 'amu and the method tu'itu'i'amu. The water is never above waist high during these operations. Both sexes are expert in floating over the sharp coral by paddling with their hands while the face is under water watching for fish. They rest on the stick or put a foot down here and there and by keeping their full weight off the sharp coral, progress quickly without needing sandals. To a foreigner progress even in boots is slow and fraught, with danger from cuts about the ankles and legs which inflame and are slow to heal. Cuts from live coral take months to heal as they break down again and again, yet the Samoan constantly explores the sharp 'amu patches with impunity even after sustaining cuts. His culture trains him from infancy to cope with the watery element. He paddles over sharp coral groves, swims
The octopus stick (sao fai fe'e) about 3 to 4 feet long, and as thick as the finger, is used by women for drawing the octopus out of holes in the coral within the reef. At low tide, women may be seen prowling about the shallow parts of the lagoon looking into pools for likely holes. The sao stick for obtaining octopus (fai fe'e) is thrust down into the hole and twirled about. The irritation drives the octopus out, when it is quickly seized by the body and bitten between the eyes to kill it before it can get a grip with its tentacles. They are, too small, however, to do much damage. I heard of a woman getting into trouble with a larger octopus than usual. When she attempted to bite it the octopus got its tentacles around her head and neck. The woman's screams brought assistance. The rescuer said she presented a ludicrous sight with the writhing tentacles waving from her head like hair. Had the "Medusa's head" myth occurred in Polynesia, the moving tentacles of an octopus would offer a more rational explanation than writhing snakes. A second short slender stick about 18 inches long is often carried in addition to poke into awkward holes with a bend. The octopus, after being killed, is placed in the ola malu slung on the back. The octopus is a great delicacy much sought after. The sao method is used by women alone.
Snares and nooses, which come under the term sele, are used to catch the sea centipede, fresh-water crayfish, and sea eel. A more elaborate rope noose is used as the orthodox method for catching shark.
Sea centipede snare (sele valo). The sea centipede (valo) is caught at alava, the outer skin from the butt end of a coconut leaf. The sea centipede lives in holes in the beach which are burrowed out with a vertical entrance and then run horizontally under the surface. A hole is located and a long strip of flexible alava thrust down into it and worked along the burrow to indicate its direction. The sea centipede is a. As soon as its body is partly out of the hole, the man with the snare draws the two ends apart and thus catches the sea centipede in the knot.
Skipjack snare (sele malauli). At Asau in penu penu fish and thus attract the large fish called malauli by the Samoans and skipjack by the whites. (See figure 253 b.) The fishermen beat the sides of the canoes with their bailers and in this way, it is said, draw the malauli to the heaps of coral, where they pass into the crevices in search of the smaller penu penu.
One hand may assist in keeping the loop open while the right holds the end near the small loop. The malauli swim about freely and brush against the men. When one passes into the loop, it is drawn taut. Holding the end and keeping the line taut, the fisherman jumps into his canoe and plays the fish. It was looked upon as great sport to let the fish tow the canoe. When tired, the fish was hauled into the canoe. The snaring of malauli thus partook of the nature of a game as well as adding to the food supplies.
Crayfish snare (sele ula). The fresh-water crayfish (ula vai) frequents pools in the various streams and is excellent eating. The snaring method is much used by boys and even adults as a makeshift method readily available. (See figure 253 c.)
Sea eel snare (sele pusi). Sea eels (Muraena), called pusi, are usually caught in a set trap but they are also caught inside the lagoon with a snare set on a wooden handle, while a separate bait is used in conjunction with it. (See figure 253 d, e.)
When a sea eel is located amongst the rocks in pools inside the reef, the noose is adjusted to suit the size of the fish. It must be larger than the fish, but not too large. The free end of the cord is then twisted around the fingers of the right hand; the handle with the noose is held in the left. Owing to the stiffness of the cord, the noose is easily adjusted to stand open at right angles to the stick. The bait stick, held in the right hand, is advanced through the loop to attract the eel. As the eel follows it up, the bait is withdrawn through the loop. The fish naturally follows and as soon as its head has passed through the loop, the cord is drawn taut by the right hand and the eel secured against the stick handle. (See Plate XLI, A, 1.)
Shark noose (maea noa maliei). The shark noose is made of full-sized sennit rope and, the noose being rather large for the term sele, the name is applied to the rope (maea) by which the shark (malie) is strangled (noa), hence maea noa malie. The older name for a shark was the universal Polynesian word mango but, according to tradition, it carried malie, to thank) and rewarded it with the agreeable name of malie.
The rope is made of strands of three-ply sennit braid but each ply contains as many as eight or nine strands of braid. When the rope has been twisted the further treatment in preparing the shark noose is as follows:
The loop is removed from the horizontal stake on which it was commenced and wrapped or seized with a strand of three-ply sennit braid to form a stiff open eye (
ai niu) Commencing on one side close to the rope junction, the braid end is slanted in a direction away from the junction and held down on the strands with the left thumb while the right hand makes a turn round the strands to cross the braid a little distance from the braid end. The succeeding turns are placed as close together as possible and in the seizing,the end of the braid is covered and thus fixed without any knot. To wrap round is ta'aiorsaloibut the seizing of the rope is termedmatimati. Thematimatiseizing is continued right around the eye. When it approaches the rope junction, a few firm figure-eight turns are crossed alternately around each limb until the temporary junction lashing prevents its further progress. The lashing is removed and the seizing with the continuous braid is carried on over the junction and along the rope.Before continuing the seizing over the rope, however, further steps are taken to insure a smooth running noose. The rope is a thick three-ply twist with marked grooves between the plies. The three interply grooves are filled in (
fa'atumu), each with a strand of sennit braid. Their ends are held in position while the seizing proceeds over them and fixes them. They follow the spiral grooves and are readily kept in their position by the left hand as the right makes the close transverse turns of the seizing. The rope that has to run through the eye is thus made perfectly round so as to run easily and at the same time the seizing stiffens it. The seizing is continued for the length required for the noose to accommodate the largest shark likely to be caught and then ended. In adding a fresh piece of sennit braid to continue the seizing, the ends of the old piece and the new piece are crossed in a slanting direction down the rope. The 'first turn of the new piece thus crosses over both ends, and subsequent turns, by covering both ends, fix them without any knot. In finishing off the seizing, three or four loose turns are made over the left thumb stretched along the rope and after removing the thumb, the end of the wrapping braid is turned back under the loose turns. Commencing with the turn next the seizing, each turn is drawn taut and the end of the braid pulled to remove the slack. The braid is then cut off back on the seizing where it emerges from under the first of the loose turns tightened.
Nowadays, shark nooses are usually made of bought rope of three-ply twist and one end is doubled back and spliced to form an eye. In so'o (to join) and the eye is called tali masanga (tali, to receive; masanga, to act together).
A proper shark rope seen at
Shark nets are now rarely if ever used. Noosing, however, still remains the common Samoan method of catching shark. With such a well-established successful method, there was little incentive to use a hook. My Samoan informants denied the use of a hook and it is extremely doubtful if it was ever used. The wooden hook in the masimasi.
As the noosing method (lepamalie) forms such an extraordinary contrast to the method of using large hooks in marginal Polynesia, it is described here in detail. The shark was lured to the side of the canoe by means of rattles and baits.
Shark rattle (tu'i ipu or lutu). The tu'i ipu is formed of the discarded half coconut shells (ipu) that abound near the cooking house from the grating of coconuts. (See figure 254.)
There is no special care taken in making the rattles. Any shells and any suitable wood serve the purpose. They are used to attract attention by lowering the shells well down into the water and working the handle part violently up and down, care being taken to keep the shells submerged. A sound is made not by the shells clicking together but by the commotion of the water caused by the cups being drawn up and down. The commotion in the water, according to the Samoans, conveys the idea to the shark that there is a school of fish about. As it swims in the direction of the sound, another of the shark's senses conies under the influence of the bait lures. When a shark is seen in the vicinity of the bait lures, the rattle is drawn up.
Bait lures. There are three bait lures, the deep, the float, and the near baits.
A strand of sennit braid is run from the front projection along the middle line of the upper surface, around the posterior peg and back to the neck of the projection. It is looped forward and then lashed to the projection at the neck with the end of another long piece of braid. The long end of lashing braid is left on to serve as a bait tie. Round the curved surface of the float four transverse grooves are cut at fairly equal distances apart. Bait is tied to the mesial longitudinal braids opposite the grooves. Opposite the first two grooves at the pointed end dried bonito heads (pa'o'o) are tied, to the third a fish, and to the last a piece of pork. In these days when kegged beef is so much in favor, a piece of old salt beef may replace the pork.
The bonito pa'o'o bait consists of the head of the bonito from which the skull and under parts are removed, leaving the large opercula connected in the middle line above. (See Plate XLI, B.) These heads are dried and kept as they have a strong, powerful odor which attracts the shark.
The bonito bait is turned down and one side pushed through under the mesial braids so that an operculum is on either side of the middle line. The tie line is brought down the middle line, over the middle line, over the bait, turned to the left and passed around under the float to appear on the right side. From the right it passes to the left over the middle line and over its previous course, and after looping around the left operculum of the bait passes back to the middle line to take a turn around the mesial braids. The second bonito bait is tied in the same way, and then the other two.
The uto float is now baited, but another trailing bait (maunu tafea) is lied to a cord a foot long and the end of the cord tied to the float under the end of the mesial braids. The uto float bait is the most important in attracting the fish. A head fisherman at Papa,
The four names recited are names of kinds of shark. The head fisherman at Salailua, uto float was thrown out:
The float is tied to a long rope and thrown out. A man watches it and reports any movement. Sometimes the bait is taken by a shark. When a shark is noticed making for or biting at the float, the float is drawn in slowly to draw the shark after it. When the shark gets close enough to the canoe to come within the sphere of the near bait, the float bait is drawn into the canoe.
Using the noose. The expert with the noose judges the size of the loop to suit the size of the shark that is now plainly visible near the canoe. He opens out the noose to the required size and, holding it at the eye by the right hand, he grasps the rope with the left hand not far from the right. The noose keeps open owing to the stiff nature of the sennit rope and the seizing. A manila rope is no good because it is too soft and the loop sags. The noose is lowered into the water with the hand above the surface and the loop at right angles to the canoe. The assistant manipulates the near bait so as to draw the shark which follows it into the noose. As the shark's head enters the noose, the expert's right hand carries the loop back until it touches the shark's dorsal fin (tulangongo). The dorsal fin is an anatomical landmark. The shark's lower jaw is set well back and the noose must not be closed until it is behind the lower jaw. Immediately the right hand touches the dorsal fin, the expert knows that the noose is behind the lower jaw, so he pulls the rope taut with his left hand while the right holds the eye of the noose firmly against the side of the shark.
An ordinary sized shark is firmly held in this position in spite of its struggles. It cannot possibly bite while so held. The right hand, by means of the noose, turns the shark's head upward and as soon as the nose shows above the surface an assistant strikes it with a club. If the shark is too strong, the fisherman has to let it go. He must choose the moment when the head of the fish is directed downwards. If he lets go when the head is directed upwards, the shark may snap his hand before he can get it clear.
A man of
When the rope; is let go, it must be held near the tied end to prevent the weight coming on the rear boom and thus snapping it. The shark is allowed to tow the canoe and thus exhaust itself. After a time, three or four sharp jerks will bring it to and cause it to float belly upwards. The shark is drawn in to the side of the canoe. The expert gets hold of the rope against the body very quickly and working his hands downwards on either side gets them into the gills. The head is then lifted up out of the water and an assistant strikes it a couple of blows on the nose with the shark club.
Besides carrying the rope noose and baits, each canoe carries a short club and many carry a wooden spear as well, as part of the shark equipment.
The shark club (fa'po) is a short wooden club from 2.5 to 3 feet long and about 2 inches in diameter. They are round but, owing to the free use of steel tools, some of the more modern ones are four-sided. Heavy wood, such as the manulenga or filofiloa is used. The handle is usually shaped. The term fa'apo means to kill without the victim knowing.
The shark spear (taova'a, from tao, spear, and va'a, canoe) is made of some hard heavy wood, such as filofiloa. It is an ordinary strong stake about 6 feet, 6 inches long and sharpened at one end. If a large shark is caught in a noose and the crew becomes afraid, the spear is thrust into its mouth when it opens. The spear then serves as a gag and the shark is unable to bite. The fish being gagged, the crew have more confidence and the shark is played and despatched in the ordinary way. If the shark is too large to tackle, the spear forms a useful weapon with which to turn it away. A much-used shark spear seen at Papa had cuts all around it from the teeth of various sharks that it had gagged.
The use of the bait float in shark fishing has been described. Wooden floats are used with nets and the line of a squid lure. All are termed uto. A green branch tied to an eel line also acts as a float and is termed fa'autouto (to act as an uto). Two special floats are here described:
The implication is that the secret is out. au was the sharp bone used to catch it, which was obtained from others of its species. The saying was applied to anyone who brought trouble upon himself. I quoted the proverb in au, he explained, was the liver and it was the liver that was used as a bait to an ordinary hook and did not signify the bone at all. Inquiries about the kind of hook baited with the au failed to elicit a satisfactory reply. au also accounted for
Torches (lama) are made of the dry coconut leaflets (au lama) left out to dry in the sun. The leaflets are pulled off the midrib and tied together in small bundles. Torching goes on at night inside the reef. A number of the leaflet bundles are carried and as one burns down another is lit. The taume flower sheaths of the coconut palm when dry are also tied into a bundle termed fausa and used as a torch. Dried candlenut kernels, also called lama, are tuingalama. Both women and men use the torch.
Coconut leaves are used as an individual adjunct in the torch, to form the walls of temporary weirs and as sweep nets and dams in community fishing.
According to a saying the torch was used in catching the sipa outside the reef but no details were obtained.
The long leaf sweep (lauloa). The lauloa method of fishing is one shared by the village community and gives much enjoyment during operations. A long sweep is obtained by attaching coconut leaves to strong vines and, by means of it, fish are driven into a set net. A description of a lauloa sweep at Fangamalo,
The matai heads of families meet together in one of the guest houses and over a bowl of kava decide to have a
lauloa(lau, leaf;loa, long) which is the name given to the method of fishing as well as the means. Thetautaior head fisherman is, of course, present and discusses the tide and time. The meeting decides that ten fathoms of leaf from each family will make a sufficiently longlauloa. The news is promulgated and the head of each family sends one of the young men (aumanga) to the woods to get ten fathoms of thefue vaivine. Others collect green, coconut leaves, split them down the midrib and thin the midrib strips down. The young man returns with a coil of vine and throws it down in front of the family dwelling house. The family head ties a knot at one end of the vine and after measuring off ten full-arm spans, ties another knot. The vine is stretched waist high between two trees and the part between the two knots filled in.The end of the first midrib strip is passed through the left knot and doubled back. The midrib strip is then wound spirally round the vine and two or three close turns taken over the doubled-back end so as to bury and fix it. Every here and there, a leaflet is wound round and round both the midrib and the vine, and the end passed between them and pulled down. The subsequent windings of the midrib keeps the leaflet end against the vine and fixes it. The leaflet windings prevent the midrib strip from slipping. Just before the midrib strip ends, the vine is split (now with a knife), the end of the new strip is passed through the slit and doubled forward. The end of the shortening strip takes some spiral turns round the doubled forward new end and then the new strip continues its spiral turns over both of them and fixes them. The addition of fresh midrib strips is continued up to the second knot, when the end is fixed by passing it through the knot and tying it in an overhand knot. In Plate XLI,
G, the leaflets are shown projecting out from the vine in all directions. The ten fathoms are then coiled and left for the assembling on the morrow. Each family has its ten fathoms in waiting.
The lauloa is worked in conjunction with a meshed net of cord set in the channel. The head fisherman and a few assistants set the net which has two wings forming a "V" with a bag or purse at the junction of the arms. The net has a top rope with wooden floats while the bottom rope is weighted down on the bottom with large stones picked up in the vicinity. The two wings of the net are further lengthened on either side by banana leaves tied to a vine. The vine is weighted down while the banana leaves float vertically, a.)
While the set net is being attended to, all the leaf sections of ten fathoms are assembled on the bench, each section being called a fanganiu. Instead of being loaded onto various small canoes, the presence of a large whale boat enabled all the section to be loaded in one vessl. The boat with its freight of leaves and men is pulled across to the reef at 4 p.m. at some considerable distance from the set net. The first section is dropped overboard after tying an end to another section. Some men seize the other end and carry it out on the reef to the highest dry part, just within the breaking of the sea rollers. This is to prevent any fish escaping between the end and the reef. The boat then pulls slowly back in a curve paying out the leaf, which has the sections joined together in the boat as fast as they are being paid out. (There is always a clear piece of vine at each end.) The ends are drawn together to join up the leafy parts and the vine is tied in a reef knot. The second in command superintends the paying out and giving directions. He is assisted vocally by everyone present. The men and women of the village appear as by magic all along the proposed course of the lauloa, which they judge by the movements of the boat. Men jump off the boat to take up positions behind the lauloa as it is paid out. The leaf is paid out in a long curve and the shore end is carried forward to the level of the reef end. The two ends are dragged forward towards the wings of the set net and the people evenly spaced along the intermediate part push their part forward in the general advance. Most people carry a pole with which they beat the water or rest on when in parts of the lagoon that are deeper than usual. The lauloa is drawn on a falling tide setting towards the open wings of the set net. Thus, by shaking leaves, and shouts, and laughter, the fish are driven forward towards the net. When at the right distance, the two ends of the leaf chain are converged towards the ends of the banana leaf wings. As the line of lauloa contracts inwards the slack is doubled along the margin to reinforce the line in the event any fish are frightened back. The contracting movement is continued until the leaf is closed right up against the wings of the net. By this time all taking part have crowded together, anxious to see the catch. All the fish having been driven into the net purse, the opening of the purse is constricted with the hands. The purse with its load of fish is lifted bodily up into the head fisherman's canoe, which is anchored alongside. The end of the purse is unlaced and the catch emptied out. There is time to draw the lauloa only once on the tide. The fish are taken ashore and subsequently divided amongst the families who had contributed their sections to the lauloa.
Short leaf sweep with mat cone (tu'i.) The tu'i variation of the lauloa was described to me at Salailua, fala floor mats contributed by the village. The mats are stitched together to form a cone (tu'i) and used in conjunction with two sections of coconut leaf sweeps. (See figure 255, b, c.)
The outer ends of the sections of coconut sweep are termed taiao or taiulu. The section which reaches to the beach is termed taiao fonua. As before, the chief fisherman is in command at the receiving (tu'i), while the second in command superintends the sweeping in of the ends. The two sections of leaves are dragged until the outer ends meet. The command, "Soso'o taiao'" (Join the ends), is given and the outer ends of the two sections are brought together and tied. This tie is never unfastened. The people call, "Ua soso'o" (They are joined). Then comes the command, "Talai lau, fai tua fa" (Unknot the leaf, overlap to four thicknesses). The small individual sections are unknotted as the sweep narrows in and sections are overlapped to four thicknesses. As the area further diminishes the slack is taken in on subsequent tu'i as well as the narrowing lauloa. Women armed with scoop nets may catch any fish as the leaf sweeps close in. When closed right in, the unroofed front part of the tu'i is raised as a flap to close the entrance. When fish are plentiful, the chief fisherman may order the scoop nets to be used in scooping fish up into the canoes. If the tu'i can hold all the fish, poles are passed under it and the whole mass lifted up and carried ashore.
Short leaf sweep with scoop net. At Salelolonga, lau) are used to drive shoals of i'a sina into a scoop net ('enu) managed by an expert. The short sweeps from their particular use receive the name of lau i'a sina. They are a modification of the long lauloa sweep, adapted for a particular shoal fish and employing fewer people. In the same district the atule is caught similarly. (See figure 255, d.)
Leaf weirs (tupa), v-shaped, with walls formed of coconut and banana leaves, are used in Salelolonga, tupa and though they belong to the weir (pa) class, they are now described from their being a stationary form of the short lau methods used in the same district. Two weirs, facing in opposite directions, are set close together for the reason that the i'a sina fish, for which they are made, move in shoals towards the east in the morning and towards the west in the evening. Of two such weirs seen, one was made of coconut leaves and the other of banana leaves. (See figure 256.)
The 'enu scoop net with a rigid frame is a little over 2.5 feet wide and fits against the stones on either side of the narrow muli openings of the weir. The stones give definite landmarks for the net to rest against, and also form a more natural opening for the fish to turn into the net. Though the leaves and leaflets of the walls afford ample space to pass through, the fish keep on so long as there is a clear opening in front. The weir bottom consisted of a hard, sandy surface and the i'a sina swim close to it. Hence, though the tops of the leaf walls may not reach the surface, the efficiency of the weir is not affected.
With the 'enu net in position, the fisherman stands beside it on the outside of the weir, holding a cord attached to the lower crossbar of the net opening. When the fish that have entered the weir are forced into the net by the converging walls, the fisherman lifts the opening of the 'enu with the cord. The fish are quickly removed into an ola fish basket tied around the waist, and the net is reset. The process continues until the fish cease to run.
For the morning run from the west, the fisherman must be at the weir before daybreak, as by 5:30 a.m. the fish cease to run in any number. For the run from the east, the right time is at sunset. There is no significance
Dams (puni) are made across tidal inlets or estuaries of streams where the water is not too deep at high tide. The materials used are laupola sheets made from coconut half leaves in the same manner as the laupola thatch sheets. Stakes are driven into the muddy bottom three feet apart and in line, at low tide. The method is a community one. The leaf sheets are plaited and collected but the dam is not made until high tide to allow the fish to come up in search of food on the incoming tide. The dam is then made to intercept them on their return on the falling tide. The leaf fence is formed by tying the sheets to the stakes, there being perhaps three tiers of sheets to get a sufficient height.
A trap also termed 'enu is used in connection with the dam but it differs from the 'enu in being made of 'ie'ie (Freycinetia vine). It has a trap entrance, which prevents the fish from getting out and can thus be left in position without personal supervision.
Openings called fa'atoto'a are left in the dam corresponding to the width of the trap. As many openings are made as there are traps available. If there is a good run of fish the traps may be visited by attendants who empty and reset them. Large traps made of tuafanga vine, may be left in position and the catch gathered after the tide has fallen. The dams are of two forms. Both types are used at the villages near Mulinuu in
A saying is used in connection with the two dams, which indicates united action: "Puni loa, puni mata tongo." (The long dam and the short dam.) It really means that the wide spaces and the narrow spaces must be thoroughly dealt with.
Man lures fish to their destruction by deceiving them with actual food or some imitation of food. Food, when it encloses a gorge or hook or is itself enclosed in a trap, is a true bait for it forms a connected part of the apparatus which secures the fish. Baits used with the sea eel snare and the shark noose are true lures for the fish are drawn into the nooses by the baits which are moved and are not attached to the noose part of the apparatus. The wooden float and the pa'o'o dried heads are used only as lures. The coconut rattle imitates the sound of food in the form of splashing fish and thus lures the shark to its vicinity.
The shell hooks for bonito and other fish which are trolled are really lures for they are made to imitate small fish and thus lead to larger fish swallowing them. As, however, they have definite points connected with them, they come directly under the heading of hooks. In a certain form of a trap, a black stone is used to lure the fish into the trap. The above lures are all used (except the shell hook) in connection with something else. The appliance, however, used in catching squid has neither hook nor bait and constitutes a true lure.
Squid lure. The squid lure (pule ta'i fe'e) is made of a dark basaltic stone shaped like a spinning top (ma'a ta'i fe'e) with two plates of a marine Cypraea tigris) fitted to one side and a long strip of coconut root bearing pieces of coconut leaflet attached to the other. The lure is dangled and jerked about in the water from the end of a line and attracts the squid. The name of the shell is pule, the squid, fe'e, and to attract is ta'i. The compound name of the lure, pule ta'i fe'e, thus means the shell which attracts the squid. For the technique see figure 257.
A six-foot length of twisted cord or five-ply sennit braid is attached to the lure by passing one end under the wrapped median braid on the upper surface of the first pule fao shell by means of the tina wedge and then tying an overhand knot at its end. This is the orthodox fixation and no matter how the lure is jerked, the knot will not slip out from under the median braid.
A float (uto) made of light wood, shaped like a playing top, is drilled longitudinally from the center of the base to the point. The other end of the cord or braid is passed through from the point end and tied with an B.)
Lure fishing is used only by men, just as the stick method is used only by women. The women's method takes place at low tide and lure fishing inside the reef when the tide is in. The lure is used from the small paopao dugout canoes. The fisherman paddles backwards and forwards in the likely parts of the lagoon. Whilst fishing he keeps the canoe moving by paddling with the left hand while the right manages the line of the lure. The lure is lowered to just above the bottom which is clearly seen and kept in motion by constant jerking, which is the movement that first attracts the attention of the squid. When it is jerked violently about, the lure looks ridiculously like a rat in the convulsions of drowning. The sharp jerks also cause the loose second shell to click against the stone and is supposed to represent the squeaks of the rat. That the squeak is not really necessary is shown by the Hawiian lure which has no similar mechanism. The length of line is altered as the water shoals or deepens. Hence, the wooden float is never used as a hand grip.
When an octopus sees the moving lure, it reaches out one tentacle and rests it on the lure. The fisherman who watches his lure in the clear water, draws its steadily upwards. As it nears the surface, the octopus which follows it up still with only one tentacle on it, probably realizing from the increasing light, or lesser weight of the water that it stands a chance of losing whatever the lure represents to it, suddenly pounces on the lure, rests its body on it, and clasps its tentacles around it. This is the psychological moment when the fisherman draws it quickly out of the water and into the canoe. The octopus is seized by the body and bitten between the eyes, to kill it.
Small squids are called ano. A large octopus can best be managed from a canoe in deep water where it cannot rest some of its tentacles on the bottom. In shallow water they are dangerous as they can draw a canoe under. A skilled fisherman if he gets a large octopus on the lure will try and draw it out into deeper water and then pull it into the canoe when it is as susceptible to a bite between the eyes as the smaller ones. If he cannot get deep enough water, he drops his line. When the lure becomes motionless, the octopus leaves it. The fisherman recovers his line by means of the wooden float and thus saves his lure.
la'ei as a squid lure made of ti leaf. The lure is the same as that described, but ti leaf strips are used instead of coconut leaflets to decorate the under part with legs and tail. According to the Safune people (si'usi'u) which attract the squid.
Legend of the Fe'e and the RatThe Tutuilan tale states that the
unga(hermit crab), theve'a(rail) and theisumu(rat) planned a visit to see the red earth cliffs, Lenga-a-Taema, near the western end of the island. The three friends lived between Aoloau and Fangamalo. Owing to the hermit crab being a poor pedestrian, they decided to journey by sea. The rat climbed a coconut tree, gnawed through the stem of a nut and dropped it to the ground. The crab husked it with his claws. Difficulty occurred with regard to splitting the nut open. The crab selected a tree leaning out over some rocks and climbing up with the nut dropped it on the rocks below. The nut split into two halves and the rat and the crab cleaned one of them out to provide the means of ocean transport. The craft was launched and the three comrades embarked. The rail spread out its wings as a sail whilst the crab and the rat kept the masts fixed by holding the rail's legs. With a fair breeze, they sailed merrily along towardsPoloa , but between Fanga-lii andPoloa , they were struck by a hurricane (afa). As the craft foundered, the rail flew away to the land, the crab sunk to the bottom where it was quite at home, but the unfortunate rat was left swimming for its life. As it struggled with the waves, the rat lifted up its voice in lamentation as follows:
|
Ua lele le ve'a, e fai ona apa'au | The rail has flown because he has wings, | |
Ae ngoto le unga i le a'au, a'o a'u nei ua 'au'au. | The crab has sunk to the reef, but I, alas! have to swim. | The octopus (
fe'e) hearing the rat's wailing, took pity and invited the rat to sit on his head whilst he conveyed him safely ashore. On the journey, the rat defaecated on the head of thefe'ewithout the latter knowing it. When safely ashore, the ungrateful rat taunted the octopus by drawing his attention to the insult. The enragedfe'ecould not pursue the rat on land but vowed if ever he caught the rat in the water again, he would avenge the insult.The hate of the
fe'ehas become hereditary and man his utilized it to his own advantage by shaping and decorating a lure to represent the rat. In the story fromSavaii thetuli(plover) takes the place of the rail.
Fish spears were made of a hard wood, such as toa, pangi, asi, and 'o'a. The points were shaped and hardened in the fire. Three types were described but there may be others. The general name for spears is tao.
The one-point spear (tao mata tasi) was made in one piece. The handle was short so that the spear could be easily manipulated under water when used by men diving around the rocks and ma'a piles. It is now replaced by the short spear with an iron point.
The three-pointed spear (tao mata tolu) was a composite one, the points being made of any one of the hard woods mentioned above while the handle or shaft was of any light wood. The points were arranged with a spread and lashed to the shaft with sennit or twisted cord. This spear is better for thrusting or throwing at fish.
The many-pointed spear (tao fuifui) gets its name from the cluster (fuifui) of points lashed to a handle (fuata). The points are ala'a hard wood but olosina may be used. (See Plate XLI, P.) The handle is of the soft fu'afu'a wood, but milo is preferable. The thick end of the handle is tapered clown and four pieces placed round the end. About five turns of sennit braid are taken round them and the other points gradually added to the handle in spiral fashion, the turns of the braid being continued round them. When 33 points have been added, the turns are continued down the handle for a short way and then fixed by passing the end back under three or four loose turns, as in the hafting of adzes.
The spear is used for throwing into the midst of a shoal, such as the young mullet (aua, but 'anae when they are full size).
The bow and arrow method is used for shooting at fish which swim near the surface. The arrows have two or three points of hard wood which are lashed to a shaft of cane (u). The bow is not used merely as a pastime but as a serious method by adult men. In Fitiuta, a bow and some arrows were seen hanging up in a house, so the sport still lingers in this remote village. The bow and arrows are similar to those used in shooting at pigeons and will be described on page 530. Arrows pointed with umbrella ribs brought to a sharp point are used by boys who shoot down on the fish from cliffs or a canoe.
Various forms of scoop nets are referred to on page 476. There are, however, two forms of marine food which swarm in shoals at particular seasons and which are gathered in scoops not made of ordinary netting. The two foods are the palolo and the ingana, a small fry which corresponds to the whitebait of other countries.
Palolo scoop. The palolo swarms to the surface in myriads and all that is required is some form of scoop to dip them up out of the water. Children scoop them up with their cupped hands and empty them into a receptacle made of a sheet of the lau'a'a coconut fabric from the base of the leaves. The sheet is doubled, folded at the ends, and tied with a strip of bark. The little basket (Pl. XLI, D) is simply termed 'a'a from the material.
The palolo though long are very thin. The problem of making a fine enough mesh in netting was evaded by the Samoan who sought a substitute in other material. Two types of scoop were made and both are called 'enu. enu is thus rather general as it applies to some forms of net scoops and also to certain forms of traps.
Both forms of scoop are no longer in use as thin gauze or scrim can now be easily secured from the traders and made into scoop nets which are better or just as effective. At
The palolo consists of the reproduction segments of a sea worm (Nereis) which are freed by the adult whose head end remains in its habitat at the bottom of the reef. The palolo comes wriggling up to the surface in immense quantities on certain parts of the coasts throughout the group, on the last night of the second quarter of the moon, or the following night in the months of October and November. If they appear in quantity in October there may be little in November, and if the supply fails in October, the Samoans look forward with confidence to November. In 1927 they appeared at
On the "palolo nights," they appear earliest in the east at the
A night or so before the palolo appeared an old man stated he could pua palolo, a peculiar indefinite reef smell that is supposed to get very strong immediately preceding the appearance.
The palolo seen were in two colors; green and brown. Cooked, they form a greenish unappetizing mass which, however, has a characteristic salty taste which is quite palatable. Palolo is greatly esteemed by the Samoans. Chief's palolo is cooked with coconut cream. Some is laid aside and recooked again and again with more coconut cream so that it is not only kept preserved but increases in size. With this treatment it is kept till the next season when it is eaten and fresh chief's palolo made.
The palolo season is of importance in the Samoan calendar and some of the months are referred to as before and after "palolo."
The whitebait scoop (fonoti) is made of sennit three-ply braid with the same technique as the sennit baskets already mentioned. (See Plate XLII, A.) Though the sennit basket technique is lost through disuse the sennit scoops are still made at
The technique demonstrated by an old chief of
A piece of sennit braid was tied by one end round a carrying pole and stretched taut round an end of the pole. On the stretched line so provided loops were set up with a knot resembling the netting knot. (See figure 258.)
There are three ways in which the scoop is used:
The season commences in August and may reach to December. The two streams frequented are
The small fish ascend the streams after coming in from the sea. The Samoans follow them upstream and the inland villages get their share. They are said to become a little smaller just after coming in from the sea but to grow larger as they ascend. The larger more mature form is called 'anamangi. The larger fish are not sought after for themselves but are often caught whilst fishing for fresh-water crayfish. My informants had never noticed the 'anamangi in roe or moving downstream at any particular time. Large fish follow the ingana in, and during the season are caught close to the shore. A small sea eel (mango) is caught in the sand below the fall at Puleia. If the sand is scooped up and thrown ashore, several of them are thrown out in it. They bury themselves in the sand tail first. The tail is round and pointed. These small animals also prey on the ingana. When they come in, the mango erect themselves on their tails to catch the ingana. The Samoans call the mango, the 'ata of the ingana, 'ata evidently meaning the natural enemy, as the shark is of the atule.
Legenedary OriginThe
inganawas brought fromFiji bySina with its'ata, an eel like fish called themango, and its guardian, a very large fish calledpa'i tele.Sina came to visit her mother,Le Afine -vave, who lived at Afoasau between Sili and Vaiala. She was accompanied by Ili and Tangoai, both men. They caught a shark on the voyage and landed at Sapapalii inSavaii .Sina sent the men overland with the shark to her mother while she travelled along the shore with her fish.On the journey, the two men ate the liver of the shark. On their arrival at Afoasau, they presented the shark to
Sina 's mother but she, seeing that the liver had been removed, reprimanded them for eating it (fa'asua i le ate). From this incident she named her son Fa'asua-i-au, which is the origin of a title in Afoasau.
Sina left her fish, theingana, for her brother Faasua-i-au and appointed Ili and Tangoai as guardians of the fish for Faasua-i-au.
The kernel of the fruit of the futu (Barringtonia sp.) is grated on lapa coral (Fungia) and used for poisoning pools. It is usually mixed up with wet sand to form balls called maunu (bait). Perforated tin has taken the place of coral as a grater.
The 'avasa (Tephrosia piscatoria) is also said to have been used. Stem, roots, and leaves were all pounded together between stones and made into balls like the futu.
The poisoning of smaller pools is termed oloolo. The method was used in conjunction with nets which were drawn round the rocks or across channels which led away from the pools. Men dived down and placed the poison below the rocks. As it permeated the water, the fish were driven out of the inaccessible crevices and in seeking to escape they were enmeshed in the nets or speared. The poison was used to drive them out rather than kill them. If too strong the fish died in the crevices and many were lost.
Poisoning on a larger scale sometimes took place with the lauloa, where again the object of the poisoning was to drive the fish out of their refuges. In this form the families contributed their share of grated poison.
A man was seen using poison to obtain bait for the modern hook. He spread a lavalava over a small pool and then pushed the poison in under the cloth. The small fish soon began wriggling out of the pool and as the man saw a fish wriggling under the cloth on the margin of the pool he seized it through the cloth.
The principle of the weir is seen in the v-shaped lines of coconut and banana leaves, and the winged nets with a purse in the middle. Walled weirs of stone were known throughout the group but confined to practically one village in each of the three large islands and the
The walls, made of loosely built coral stone, were termed pa and the fish weir, pa i'a. With the exception of scanty notes from
In a) was drawn for me by Sua of Iva. The walls were renovated each year before the season. They are not used now and have fallen down.
Regarding
Walled fish traps are also unknown in
German Samoa (Upolu andSavaii ), except the village ofFalelatai inSouth Aana (Upolu ) where the lagoon is shallow enough to allow the building of such walls. TheseFalelatai walls are also called"pa."TheseFalelatai pasare built only temporarily and after use are pulled down again in order not to stop the traffic of boats in the lagoon. Size varying. Height about 5 feet. Each family or eachfuaiala(division of the village) have their customary place where they build theirpa. Some find it more convenient to make thepaoflaufala(pandanus leaves) as the piling up of stones and their pulling down again means work. The following kinds of fish are caught in theFalelatai pas.
The fish were caught by means of a hand net by the men who are waiting at the entrance of the
pawhen the tide is going out.There are as far as I could ascertain no traditions in
Falelatai concerning these traps (my authority being atulafaleorator of the very best reputation in such things) and, as stated above, in no other village inGerman Samoa are such traps known.
b. He evidently had not heard of the
In
Figure 259, c, from
The traps were visited by
In d, was made by lauloa was drawn across the opening. It was then swung round in whichever direction the shoals went until the lauloa extended from the end of the particular long arm of the weir to the shore. The lauloa was then swept along towards the closed point of the "V" and the fish were secured by hand nets and spearing. In 1920, the pointed ends were in good preservation but I saw nothing of them seven years later. atule.
Both the lauloa.
In e. The long arms were set so that they stood obliquely across the line of the falling tide. There were two widths of wire netting at the deeper part. A circle of wire netting was made round the apex and the two walls prolonged into it. This prevented the fish from getting out as they worked round the netting and could not find the opening. The form may be old but the method of execution is modern. At Fangamalo, large numbers of atule were caught. The advantage of a wire netting trap is that it is permanently set and does not need watching.
From information obtained in both groups, samples of all the known types of Samoan fish traps were secured. Though the manufacture of certain
The material for the lobster pot type, which is the most widely distributed, is the vines or aerial roots of the 'ie'ie (Freycinetia). The roots are collected in lengths, the older roots are discarded as they are brittle and liable to break when bent. The vines are bound in a coil about the size of a motor tire for carrying home. The coil is soaked in sea water and then beaten against the rocks on the shore to denude the vines of the outer bark (pa'u). When cleaned the material is termed sala in 'ie'ie which in sala. It was some time before I found out that they were both Freycinetia. On pointing this out, the Savaiians maintained that the traps were not made of 'ie'ie but of sala. "Don't the 'ie'ie?" they asked.
"Certainly," I replied.
"Then," they remarked triumphantly, "They make them of sala and not of 'ie'ie."
If the traps are not made soon after the sala is prepared, the material is left out at night exposed to the dew to soften it. Three types of trap are made with this material.
A fairly thick, creeping vine (tuafanga) is used to make the large double entrance traps (fangauli). If not used immediately after the vine is brought in, it is kept soaked in water to keep it from drying. In tuafanga was applied to the aerial roots of the 'ie'ie.
Bamboo is used to make fanga'ofe. The lafo creeping plant with long thin stems, and also fine wooden rods are used to make the sea eel trap.
This trap with the longest name (fanga fa'atau tu'u'u) is also the smallest. The type specimen in B) is made of single warps of dressed 'ie'ie (sala) with a single pair twined weft of the same material. The trap is commenced at the bottom. (See figure 260.)
The trap is used by women, in day fishing amongst the coral in the shallow parts of the lagoon, to catch the dark fish (tu'u'u). The woman with an ola malu basket tied around her waist wades out to where branching coral ('amu) is plentiful. To commence with, she places a dark stone about the size of the fish in the trap and lays it on its side near the spot where she sees the fish. The tu'u'u is very aggressive and can be seen darting about ready to fight anything of its size that offers. Leaving the trap on the bottom, the woman remains within reach, with her head submerged, watching the trap. The tu'u'u seeing the black stone which acts as a lure, without hesitation enters the trap to offer fight. The woman immediately places her hand over the entrance and lifts the trap. The first fish caught replaces the stone as a decoy. A short piece of coconut leaflet midrib with a piece of coconut husk fibre tied to its middle is pushed through the lower lip from below and the midrib crossed inside the mouth. The fibre is then tied to the inner end of the trap with the live decoy inside. The trap is again set, closed with the hand as the fish enters and the catch lifted. The tu'u'u seem to have no hesitation in entering, taking no notice of the proximity of the fisherwomen. The trap with the decoy causes the fish to want to fight (fa'atau); hence, the name of the trap in full is fanga fa'atau tu'u'u (the trap which causes the tu'u'u to fight). The women move about among the groves of coral seeking the frequented spots and they catch fairly large numbers. The introduction of water goggles has assisted the method very much for any movement of the fish can be clearly seen. It is almost ridiculous the quickness with which the fish enter the trap and the equal celerity with which the female hand descends over the opening. No such traps were seen in use in
The tu'u'u is one of the fish that is eaten raw. A saying connected with this is as follows.
E uliuli fua le tu'u'u ae otangia (The tu'u'u is black but it is eaten raw).
The significance is that blackness is associated with dirt and low status; the fact that a fish is eaten raw shows that it has an edible status above many other fish that are not so eaten. The saying is meant to drive home the fact that appearances are deceptive.
The self acting traps are built on the principle of the lobster pot. They have a more or less funnel-shaped entrance with the inner, smaller end suspended in space. Being away from the walls, the fish have very little chance of finding it as a means of exit. Some traps are baited and others unbaited. Some are rounded exactly like the lobster pot and have the entrance on top.
The lobster pot type consists of rounded traps made of 'ie'ie aerial roots by single pair twining. The traps are set with the entrance on top. There are three variations of the trap; the fish pot, the crab pot, and the crayfish pot.
The fish pot (fanga i'a or fanga puapua'i). In C) used to catch vertebrate fish is termed fanga i'a (i'a, fish). In fanga puapua'i. The term puapua'i means the smell of the newly disturbed coral sand and rocks such as is produced when the sea bottom is scooped out or cleared of small rocks to form a bed for the trap. The smell of the puapua'i attracts fish to the trap and causes them to enter in search of the food they expect to find in disturbed ground. The puapua'i itself is the bait and no material bait is used.
The prepared 'ie'ie (sala) is used. The straight warp elements (fa'atu) are single while the twining weft pair (ta'ainga) contains two single elements. As the entrance projects into the trap it has to be made first and the bottom last, which is the opposite to the tu'u'u trap described in figure 261.
In the fish pot figured (Pl. XLII,
C), the funnel started off with 18 warps and is 4 inches across the inner opening and 4.5 inches deep. Fresh warps were added till they reached the maximum number of 47 with a maximum trap diameter of 18 inches. The warps were decreased by bringing two together until the trap had reached a total depth of 11 inches and the last turned row left an opening 5 inches by 4 inches in cross diameter. The opening is left patent to form themuliof the trap. The edge of themuliopening is treated in two ways:a, the warps are cut off 3 inches from the last twined row which was closed in against the preceding round of twining. The ends of the weft elements are twisted round to the outside and stayed against the nearest warps, which prevent them from springing back. Each warp element is bent down at right angles, twisted to the right to the inner side of the warp immediately on its right and to the outer side of the next. This is done successively in the same way as the finish of the opening of thetu'u'utrap before it is bound. The first holds securely without further assistance.b, A more elaborate finish is to leave one or both of the weft elements long. After treating the warp ends by methoda, some strips of vine are run around the circumference to cover the warp ends. The warp ends are then lashed with spaced turns around the thickened rim hold everything in position. (See Plate XLII,E.)
The trap is set in likely looking pools or passages from the reef. The spot selected is cleared of stones until the coral sand is reached. The open bottom of the trap is fitted against the sand. Stones are packed and fitted
Some pieces of branching coral are built up over the outer entrance by interlocking their branches and at the same time leaving plenty of room for fish to get through to the opening. The coral house (fale'amu) keeps the big fish away from the trap entrance, which they might break in pursuit of smaller fish. Fish enter the trap to take refuge, from larger fish as well as in search of food. The setting of the trap takes place under water and the fisherman has to dive about collecting and piling the necessary stones.
The trap is visited at low water. The fisherman dives down to see if there are any fish. If so, he clears away the coral house and the stones immediately around the trap which kept it down. The hand is inserted through the funnel opening and grasps the rim of the muli opening, the wrist being bent down to close it. The trap is lifted and the contents emptied into a basket through the bottom opening.
Before the trap is reset it is washed to get rid of any slime or seaweed that may adhere to it. The hand is then waved backwards and forwards over the sand at the bottom to stir it up. This also brings out the puapua'i smell. The trap is then put back and the stones and coral house replaced.
The crab pot (fanga pa'a). Crab pots seen in use at sala treated as one element by both being enclosed in each half turn of the weft twine. The weft consisted of four strips, a pair acting as a single element in the twine. Ordinary fish pots may be used for taking crabs but in places where crabs are abundant, they are made stronger. The opening at the bottom was also present. Even the lobster pot shaped 'enu is sometimes used. In the large shallow lagoon with a soft muddy bottom at
Crab pots differ from the fish pots in being baited. The best bait are the tupa crabs with large red claws which are dug up out of their holes in the beach. These are broken up into suitable pieces between stones. A family were seen baiting a trap by three stages. One woman was tying slip knots on strips of fau bast, with an overhand knot around the standing part. A boy put pieces of crab through the loop and pulled it taut. Another woman attached the bait to the trap by lowering a piece down into the trap on one fau together in a reef knot over the twining in such a way that the bait hung clear of the bottom. The bait was hung from all parts of the roof of the trap.
Three traps duly baited were taken out on a paopao canoe to a still arm of the lagoon at high tide. We went out for about a hundred yards to where the water was waist deep. The fisherman glanced ashore and picked up a landmark. He made a depression in the soft mud with his feet and placed the bottom of the trap in the hollow. He felt about with his feet and if unsuitable increased the size of the hollow. Though the trap was open structurally at the bottom, this fitted into the depression. In addition, a flat stone inside the pot was placed over the bottom opening to serve both as cover and as anchor. Hence, the crab pot was anchored from the inside and not by heaped stones on the outside as were the fish pots.
Each trap had a float consisting of a section of dry coconut husk (pulu) tied to the top of the trap with a strip of fau bark 4 or 5 feet long. The husk was a quarter section of the whole husk and when split in this way to make floats (uto) is termed fa'autouto.
The three traps were set about 20 yards apart and in the same straight line. After setting the last trap, another shore observation for landmarks was made. The line of the traps is important not only for the purpose of picking them up readily, but also to settle any argument that may subsequently arise with another man who may set traps close at hand.
The pots were set in the evening and picked up early next morning. The traps are taken home with their contents as they have to be rebaited. At home, the stone covering the bottom opening is pushed to one side and the crab shaken through the bottom opening. It is astonishing how a large crab will slip sideways through the comparatively small hole when it feels the opening beneath it.
A trap broken in places was seen mended with wide strips of pandanus and banana leaf threaded through the neighboring parts on either side of the hole. The repairs looked flimsy but the pot caught crabs.
Crayfish pot (fanga ula). The crayfish is termed ula and the sea crayfish, ulatai, to distinguish it from the small fresh-water form (ulavai). The trap, used only for the sea crayfish, takes the name of fanga ula. Though vai means water, in which both species live, vai as an adjective always refers to fresh water.
The crayfish pot is of the same type and technique as the fish pot. It is, however, stronger than the crab pots seen at fanga ula shown in Plate XLIII, A, the individual warps vary in the number of strips used, the range being from two to six. The weft pair starts off with two of each weft
In the type pot, the funnel has very little slope, the inner opening having about the same dimensions as the outer. From the outer funnel opening, the bottom is very flat, there being little of the gradual slope seen in the fish pots. This has been purposely done, by doubling a loop forming two fresh warps around every alternate original warp and bending back every original warp abruptly at an angle instead of a curve. Thus, the warps are doubled in number at the outer opening and not gradually as in the fish pots. The weft twine ranges from 1 to 1.5 inches apart but the warps are more widely spaced ranging from 2 to 3 inches apart. The last weft round forms an ellipse, 4 inches by 3 inches in cross diameter. The warp ends cross each other over the opening and a few longer elements are doubled around them and caught under elements to keep them together. The upper surface has the curve flattened more by narrowing the warp spaces and bringing two elements together sooner than in the fish pots.
The crayfish pot is baited with alili (Turbo) which has a hole broken on one side to expose the fish. It is set in the same way as the fish pots by heaping stones around it, but the coral house (fale'amu) is not made over the opening. A young crayfish is sometimes placed in the trap as a decoy for it is said to make a noise which attracts the adults.
The separate funnel traps ('enu) resemble the preceding lobster pot type in the principle of the funnel entrance and being made of 'ie'ie aerial root material, but differ widely in being made in two pieces. A wrapped twine stroke is used instead of single pair twining. In principle, it forms a link between the manipulated trap and the self-acting traps. The bottom and body of the trap are made and finished off at a rim opening as in the manipulated tu'u'u trap. The self-acting principle embodied in the funnel-shaped opening projecting inwards is then made as a separate piece. The funnel piece is fitted to the rim opening of the body and the two lashed together. The body of the trap is thus commenced at the bottom as owing to the funnel entrance being made separately, no complication will occur with it. In the lobster pot trap which combines both funnel and body in one piece, the trap commences with the funnel and ends at the bottom. Two types of trap made with the same technique, but of different shape, are both termed 'enu, which name has also been applied to certain manipulated nets on frames and to the palolo scoop. The tendency to interchange the n and ng sounds is particularly noticeable in the modern use of the term 'engu for these traps. The word 'engu does not occur in
The warps consist of single elements of 'ie'ie root arranged in an outer and an inner series in relation to the passive element of the weft. The weft is compound, consisting of two or three strips of 'ie'ie root and a single length of three-ply sennit braid. The root elements are kept close together and, except for passing spirally in spaced rounds between the two sets of warps, are entirely passive. The braid is the active element which, by passing obliquely over the crossings of the warp with the passive weft elements, binds them firmly together with a wrapped twine.
The method of commencing at the bottom with the outer and inner series of warps, the stroke technique, the body, rim finish, and the separate construction of the funnel are shared by both forms of trap. (See figure 262.)
The lobster pot type of 'enu (Plate XLIII, D) commenced with two sets of five warps in each series. Immediately after reaching its maximum diameter it was narrowed down gradually to its rim diameter of 20.5 inches. The narrowing was brought about by gradually bringing the warps closer together. This technique influenced the shape and gave it the typical lobster pot appearance. The funnel technique and fitting are as described in figure 262 except that it is a little smaller.
The lobster pot 'enu is still made in i'a sina, a fairly small fish. The technique of the wrapped twine was an alternative to the frame net 'enu used in
The domed cylinder type of 'enu (Plate XLIII, E) commenced with two sets of five and four warps in each series. The fresh warps were added quickly, 12 fresh pairs being looped over the first round. By the time the eighth ground had been reached, 72 additional pairs in all had been added. The trap was then 6 inches deep with a diameter of 17 inches. After forming the dome end, very few fresh warps were added, which resulted in the trap continuing in a cylindrical form. Between the 8th and the 27th (last) rounds only four new pairs of warps were added, which increased the diameter from 17 inches to 19 inches at the rim. Of the total length of 28 inches, 6 inches were occupied by the dome-shaped part and 22 inches by the cylindrical part. The funnel and attachment were described in figure 262.
The domed cylinder trap is used at Vaimoso and Vaiusu in puni method of fishing described on page 434. The traps are set opposite the openings made for them. They are laid on the sides and attached to stakes. Fish returning to sea on the falling tide, pass through the openings into the trap. The method resmbles somewhat the use of the net 'enu with leaf weirs in enu of 'ie'ie roots is self-acting.
The fangauli type of double entrance trap (Pl. XLIV) is a large barrel-shaped trap made of thick tuafanga vine. The front end (mata, face) and the rear end (muli, back) have each a funnel whose inner ends are prolonged to form an inner tube passage (tapua) which is blocked in the middle by a partition. On the floor of the tube on either side of the partition is an opening (ala i'a, fishes' path). Fish entering the funnels, pass along the passage and on being blocked by the partition, pass down through the opening into the trap proper.
The technique consists of a single-pair twine around longitudinal warps. Both warps and wefts consist of the same material; the warps are selected from the thicker pieces about 0.4 or 0.5 inches in diameter, while the weft elements are slightly thinner. The warps ('aso) consist of single elements and the weft (filosi) of two. Construction commences with the tapua tube. (See figure 263.)
The funnel (sifa) of the mata end has been formed from 12 original tube warps ('aso tapua) to which three new pairs ('aso fa'angaulua) were added. All these warps project beyond the last weft round of the funnel and enter later into the finish of the trap.
The partition of the tube is formed of three short pieces of vine spaced horizontally across the tube. A strip of vine is doubled around one of the tube warps and two weft lines are made across the partition rods. A round is then worked around all the tube warps on the muli side of the partition. The opening in the floor of the tube on the muli side is defined and the tube and funnel of the muli side made in exactly the same way as on the mata side (figure 263).
The last twined round of the muli funnel forms the technical bottom (muli) of the trap by including the bent back warps which form the afe turn, and the last round of the funnel is also the first round of the body of the trap. (See figure 264.)
The bottom having been turned, the spiral rounds are continued around the 36 warps that now from their direction form body warps. The interweft space remains constant at about an inch but the warps are gradually diverged from an inch apart near the turn to 2 inches at the middle of the trap. From the middle the warps are gradually converged until the diameter approaches that of the mata funnel at the other end. Some of the warps converge to form a pair enclosed by the same half turn of the twined weft.
The warps that run short are lengthened by overlapping a fresh piece for a few inches and including the two in the same half turn of one or two rounds of the weft. The simple join (so'o o le 'aso) is similar to those of the previous traps. The weft join (so'o o le filosi); due no doubt to the greater strain in a heavy trap, is more complicated than those previously used. (See figure 265, a, b.) While the body is being made, an opening for the removal of fish is left on the upper surface about 18 inches from the middle. (See figure 265, c.)
The finish. The decreasing diameter of the twined rounds of the body are so arranged that when they reach the outer end of the mata funnel, the ao) by a series of close twined rows as in figure 266.
Transverse lashings (fa'amau) with pieces of vine around the last set of four or five weft rounds are made by passing the turns around the rim above and a twined row below. They are tied with the ordinary half hitch, or two half hitches, and the end tucked in under the turns. The lashings are spaced to about seven in all and some are continued on from the preceding one by carrying the vine across. The lashings prevent the last twined rows from working over the ends of the warps which are cut off about 2 inches from the last twined row. (See Plate XLIV, A.)
The cover for the exit hole is made of two pieces of vine about 28 inches long which are crossed. Another piece of vine is doubled around one of them and worked in flattened circles with a single pair twine with increasing rounds to form an elliptical-shaped cover. As the twine reaches the cross pieces, it encircles them with a half turn. When the cover reaches the dimensions of 11 by 9 inches, the weft ends are tucked in under some part of the twine. The cover thus has four pieces of vine sticking out. The ends are sharpened. When the cover is applied over the exit hole, the vine ends are pushed down under the twining of the trap. Two stout pieces of vine about 20 inches long tao 'api) are crossed over the ends of the cover and their ends stuck under the warps on either side. This pegs down the door securely.
A loop handle (salatau) is made by doubling two pieces of vine around the mata rim, twisting them together, to form a loop, taking more turns with the individual vines around the rim, and knotting them with half hitches.
The trap obtained was in use at Sapapalii in ma'a tu) with the exit opening above. After propping it in position with rocks and coral, amu (branching coral) is put over the top to disguise the trap. The wet trap is very heavy and if there is a good catch, it is hauled up into a canoe by a rope tied to the handle. The fish caught are funga, ponge, malau, male'i mutalau, lo, and ngatala.
A large trap (fanga tapi) of tuafanga vine is made in fanga i'a).
The bamboo trap (fanga'ofe) resembles the fangauli trap in shape and principle but the stroke technique is an improvement on the wrapped twine used in 'enu traps. The inner tube has a middle partition with an opening on either side of it and the funnel-shaped entrances connect the tube with the ends of the body.
The material used throughout is bamboo. The warps are formed of split bamboo ranging from 0.3 to 0.5 inches in width. The weft passive element, also of split bamboo, ranges in width from 0.5 inches at the ends of the trap to 0.8 inches in the middle. The active wrapping element of the weft consists of the outer skin of the bamboo prepared in thin strips about 0.2 inches wide.
The trap commences with the partition and inner tube. (See figure 267.)
The further technique of the funnels, body, and exit opening with its cover are described under figure 268.
A fanga'ofe trap (Plate XLIII, B) was obtained at fangauli but are not greatly in favor as they rot easily. They are much more easily made than the fangauli and their use is probably fangauli trap but in stroke technique their affinity is with the 'enu traps.
The full turn made with the weft around the inner warp distinguishes the wrapping technique from the more simple wrapped twine of the 'enu traps, in which the half turn is made over the inner warp as well as the outer. The 'enu twine may be termed a "simple wrapped twine" and the fanga-ofe twine as a full-wrapped twine." The full-wrapped twine is used in the anga double entrance traps of
The sea eel trap (fanga pusi) consists of a shallow rectangular box made of vertical and longitudinal sticks lashed together with sennit braid and has a self-acting entrance at one end. (See Plate, XLIII, C.) The making of the box and the lashing technique is described under figure 269.
The self-acting mechanism of the trap has to be placed in position before the top of the box is lashed on. It consists of two parts; a wooden entrance tube and an inner part formed of coconut fabric. The wooden tube is dealt with in figure 270.
The inner part of the self-acting mechanism and the closing in of the upper part of the box with provision for an exit opening is described under figure 271.
The sides are formed of 8 longitudinal rods, 22 inches long and covering a width of about 4 inches. Four crossbars are used, the two end ones being 5.5 inches long and the two intermediates, 4.5 inches long. The end bars are placed so that when the sides are fitted to the bottom, they will be just external to the end crossbars of the bottom. This is rendered possible by the side rods being an inch longer at each end than the rods of the bottom. The bars are further arranged that the intermediates are flush with the top edge so that they project half an inch beyond the lower edge or what is to be the lower edge. Their situation is further arranged that the projecting lower ends will rest against the sides of the projecting intermediates of the bottom. The end bars also are arranged to have a half inch projection above and 1 inch projection below. The lashings are made with two diagonal half hitches. The two ends are slightly different as the front end has the entrance hole. The back end consists of 26 short vertical rods nearly 6 inches long lashed to a fairly thick crossbar (0.6 inch diameter) at the bottom, a thinner crossbar at the top and one intermediate about an inch below the top bar. The two lower cross bars are 12.5 inches long and the upper top bar, 13 inches. As the 26 vertical rods cover a width of slightly over 10 inches, the crossbars project well out on either side to afford supports for lashing the sides.
The front end is slightly wider, the crossbars being 14 inches long. It is also slightly deeper, the vertical rods being 6.5 inches long. Besides the top and bottom crossbars, there are two intermediates, the upper being less than 0.5 inch from the bottom bar in
the middle and less at the ends. The upper intermediate ends flush with the sides of the vertical rods but the other three project as at the back. The entrance hole is formed in the middle between the two intermediate crossbars by cutting the vertical rods as shown in figure 270
b. The hole supports the outer end of a tube composed of wooden laths. The hole narrowed by the tube end is 3 inches in diameter transversely and a little over 2 inches vertically. The box part of the trap is assembled by turning the bottom with the crossbars underneath and then fitting the sides and ends to it with their crossbars to the outside. The projecting ends of the crossbars are lashed together where they fit against each other both at the bottom and the sides. Three thick longitudinal rods are lashed on the outside to the crossbars of the bottom, one at either side and one in the middle. These project for 7 inches beyond the front end of the trap. They afford extra support for lashing at the sides and form a rest to protect the close lashings of the bottom rods from wear against the rocky bottom on which the trap is set. To the long ends of the 3 undermost longitudinal rods a crossbar is attached as a handle. (See Plate XLIII,C.)
The top of the box is now made with two longitudinal crossbars and transverse rods. The two bars are fitted so that they will rest on the inner side of the side walls. The lashings are made leaving a gap in the middle of four inches by not adding any transverse rods in that part. The top is then turned over to place the marginal longitudinal bars on the under side. A mesial longitudinal bar is then lashed to the rods in the true upper surface. Two side longitudinal bars are also added to the upper surface and merely lashed here and there. The top is fitted in position and the projecting ends of the longitudinal bars lashed to adjacent projecting ends. The trap is now complete except for a door to close the upper opening.
The upper opening is the exit opening for extracting the fish. Some wide slats of wood are made and thrust in under the three upper longitudinal bars. They rest on the sides like the other rods forming the top and the longitudinal rods keep them down. As many are made as will close the gap. The trap is baited and the door slats are tied to the middle longitudinal bar.
The trap is set in the lagoon for sea eels. Attracted by the bait the eel finds the trap and enters the tube. When it reaches the fabric part it easily pushes its way through the slit opening, which readily widens in a vertical direction. Once the eel is in, however, it cannot find, let alone pass through, the slit from the inner edge side. The slit is in space and the fish keep going around the sides. The trap mechanism thus acts like a valve which can be passed through from one side by pressure but falls back after the pressure is removed. It is simple and ingenuous and marks a departure from the rigid inner openings of the other self-acting traps.
The material used in the trap described is wooden sticks. The lafo creeping plant with a cane-like stem, used in sugar-cane thatch sheets, is also used for sea eel traps. The old style of trap is rapidly disappearing as ordinary wooden boxes of sawn timber, nailed together, are much easier to secure and prepare. Numbers of such traps were in use in 'enu, the bamboo double entrance trap, and the sea eel trap were obtained in
Nets have the general name of 'upenga. The craft of netting is old, as denoted by the sayings associated with the ancestor Pili: O le 'upenga o Pili a tautau ae fangota (The net of Pili hangs up but it catches fish).
Pili did his netting at night and thus, though people only saw his net hanging up in the daytime, it nevertheless caught fish. Another saying conveys the same idea: O le 'upenga o Pili e fili i le po ae tala i le ao (The net of Pili is bundled up at night but spread out during the day).
The net was spread out during the day to dry but at night it was bundled up for transport down to the sea. Thus Pili did things quietly and unobserved without seeking public assistance.
The net is also mentioned in the early days of Tangaloa-ui for it was with a net spread over the mouth of a cave in
The method of preparing netting cord has been described. The best material is fan songa. Turner (41, p. 167) states that nets were mostly made in the inland villages. This is due to the fau songa growing in greater quantity inland. On occasion the fau tu was used. For special nets, breadfruit bast, paper mulberry bast, and sennit braid were used. Thicker sennit ropes in three or five-ply braid were used for the upper and lower ropes of long nets.
A needle (si'a) and mesh gage (afa) were used. The netting needle, according to kioe, an instrument used by the Hawaiians for mending, but in which there is no split point. The needles now in use (Pl. XLV, A, 1) are similar to those in use in other parts of Polynesia. They consist of a long flat piece of wood expanded at either end to allow for mesial longitudinal slots with narrow slits through the blunt pointed ends. The cord is wound longitudinally between the two slots. The needles are made in various sizes to suit different sized meshes, as the loaded needle must pass through the completed mesh.
The afa gage is a flat rectangular piece of wood scraped smooth, with rounded edges, and of different sizes. (See Plate XLV, A, 2.) Bamboo is a favorite wood as it is easy to shape. afa as meaning united in action in the phrase 'ua afa fa'atasi (from the same mesh stick). Hence, also, the saying, "Ua afa fa'atasi taofi 'uma, 'ua lelei lea mata'upu." (When all the opinions are from the same mesh stick, that matter is settled.)
To commence a net is alu and the mesh is mata'upenga, or mata for short. To lafo is also to make a net. The netting needle is loaded and the required
Good netting depends on the evenness of the meshes. After the needle is passed through the mesh above, the cord is pulled or slackened until the crossing is in the right place on the gage, when it is fixed by the left thumb while the netting knot is made.
The depth obtained forms one end of the net and is referred to as its lautele. The lautele of various nets is distinguished by the number of meshes. The length or 'umi is obtained by adding successive rows of the same number of meshes to the end which has been set up. The method of forming the commencement strip of netting has resulted in two rows of meshes. Meshes are really quadrilateral figures with equal sides. Owing to the nature of the material, if they are pulled in one direction, they lengthen in that diameter and narrow in the other. In making the net, the pull is at right angles to the gage and the long diameter of the mesh is naturally in the same direction. To lengthen the net, the commencement strip instead of being longitudinal to, the netter is placed transversely before him. The pull now comes on the other diameter but, the meshes having equal sides, it makes no difference to the technique and the placing of the gage. (See figure 273.)
In bag nets, any narrowing may be obtained by dropping meshes in the various rows or by using smaller gages. Extra meshes, however, were seen in a hand scoop net at c, d.
The mata 'upenga a Sasaumani. The Tutuilan
Cords and ropes. Most small nets have a cord running through the marginal meshes by which the net may be attached to wooden frames or handles. The cord may be twisted cord or sennit braid. In long nets of the seine type, there is an upper and lower rope usually run through the marginal meshes. The upper rope, because it supports the wooden floats (uto), is termed 'afauto or u'a o le uto. The lower rope is the 'afavae ('afa, rope; vae, foot) or fauvae. To thread in the ropes through the meshes is timata. The lower rope may or may not have sinkers attached.
Floats. Casting nets, short nets set across channels and around rocks, and the long seine nets have floats (uto) attached to the top line. The floats usually consist of tou, fau, or other light wood. The floats are attached directly by the top rope, which is tied around them at intervals as it is threaded through the meshes. Floats are of two kinds; small peg floats and large floats.
Sinkers. Sinkers (maene) are used with the smaller nets, as stated above. In large nets, the weight of material against a rocky, uneven bottom would be a hindrance, let alone the nuisance of transporting the weight of clumsy material. Seine nets are not drawn in the way they are in other parts with a smooth sandy or muddy bottom. The large nets are usually set in position and for this purpose, it is easier to pick large stones from the bottom of the lagoon and place them on the bottom rope or just over it onto the meshes of the net. In a moving net, stone sinkers would catch against rocks and in clefts and seriously retard progress. It is easier in communal fishing for the numerous assistants to push the lower rope along the bottom with their feet in the method known as tolovae. When the rope catches on a rock, the men simply submerge and lift it over the obstruction, at the same time driving fish before them.
The sinkers used with lighter nets consist of two kinds: stones and shells.
Nets range from small hand nets to long seine nets. A very characteristic net is one with a bag or purse in the middle with sides diverged out like the wings of a weir.
Hand nets, shaped to a bag and attached to a wooden frame, with or without special handles, are made in various forms. They are used as dip or scoop nets and as manipulated traps.
The small dip net ('upenga lama or lalama), a small bag net attached to a single pliable rod, bent around through the marginal meshes into the form of a tennis racket with the ends tied together, was used in lama) at night. They were largely used for fishing in the lauloa and stone weirs.
The medium dip net ('upenga sae'e) seen on a'au). (See figure 275.) It had been started from the bottom and increased by additional meshes of the mata Sasaumani type.
In using the net, two men worked in conjunction. The man with the net set it across a likely looking channel while the other drove the fish toward it with a stick. The crossbar acted as a handle ('au). It resembled the hopai (39, p. 290) net frame on a smaller scale.
The long-handled dip net. The principle of the crossbar and the long handle is applied in the sipa but the malolo was taken.) He could not distinguish between the sipa and the malolo flying fish, nor say what form of net was used with the torching. sipa as a small flying fish. Thus, though torching for flying fish was known, it could not have been indulged in to the same extent as in the nearby
The double-handled dip net ('upenga saosao'o) is a net larger than the sae'e net but the distal ends of the frame rods are not tied together. One being used in auauli and talafalu. The net which was attached along the poles for 7 feet was a bag net 5 feet deep and 4 feet wide at the bottom. The net is set across a channel or convenient pool by holding down one pole along the bottom while the other is held up to open the net. The two poles are easily manipulated by one man at the joint. An assistant armed preferably with a pole of light ma'o wood drives the fish into the net. The net possesses the advantage over the sae'e net with a rigid frame in that the poles can be closed together after the fish enter. The driving process is termed so' a.
In 'upenga tu uli. The term uli (to steer) is used in the net name from the fancied steering of the poles of the net.
The scoop nets ('enu). The name 'enu is shared by the midrib scoop used for palolo and the traps described on page 454. A more rigid framework is made by using rods for the bag of the net as well as the rim opening.
An 'enu seen at Aoloau, to which another piece bent into a U was attached. (See figure 276, a and b.) The net was used as a scoop for gathering the fish enclosed by a long net or lauloa.
A similar 'enu at Salailua was made in exactly the same way. The transverse rod of the opening was 40 inches long and always placed at the bottom. The inverted U rod tied to it formed an opening 36 inches high in the middle line. The two crossed curved rods were in this instance tied to the outer side of the opening frame and their crossing at the back called the muli was about 'enu but at the back, another cord was run through the marginal meshes and drawn taut to pucker the net together to close the muli end. A cord was tied to the lower crossbar of the opening. This larger 'enu was used in connection with the tupa leaf weirs described. The opening of the net was placed against the opening at the apex of the weir. It was wide enough to overlap the weir opening and it rested behind the two stones mentioned.
The fisherman stands outside the weir holding the cord attached to the crossbar. When the i'a sina fish have entered the net, the cord is pulled so as to lift the opening of the net above the surface of the water.
Another form of frame was seen in the same district as the above. (See figure 276, c, d.) The net is used purely as a scoop net in connection with long nets or it may be set in a small leaf weir in exactly the same manner as the preceding net. Not only i'a sina are caught with it but also the atule when it moves in shoals.
The arched hand net (se'i) consists of a rectangular piece of netting spread out by four sticks attached to each corner and tied together at the other ends in such a way as to arch them and thus keep the corners of the net taut. (See Plate XLV, B.) A line is run through the marginal meshes of one side and tied to the lower end of two sticks with a clove hitch. It is then run through the marginal meshes of each side in turn and tied also in turn to the lower ends of the other two sticks with clove hitches.
The lashed part of the arched sticks is used as a handle. The net is lowered down into the water and as a fish passes over it, it is drawn up out of the water. The fish caught with it are matu and mumu. It seems to be limited in distribution to the area in tolo matu. The seine net is drawn round to enclose matu fish. Canoes paddle into the enclosed space and the sides of the canoe are tapped with the paddle. The fish swim in to the neighborhood of the canoe on hearing the sound and are then dipped up with the se'i net. The term se'i means to jerk and the name is applied to the net from the quick jerk given to it in drawing up the fish. The se'i is made at the same time as the tolo motu seine net and goes with it as part of the equipment. They are not supposed to be parted. Separating the two is prohibited (sa). The se'i secured was sold by the family whilst the family head was absent in sa restriction was overcome by the vendors themselves.
The arched net with line ('upenga sumu) though not a hand net is dealt with here because it is exactly the same in make as the se'i. The one described in figure 277 was made for me by an old man at
The fish caught was the sumu (genus Balister) which frequents the outer side of the reef. Fishing took place from canoes which came in as close to the reef as possible. The baited net was lowered on a line to near the bottom. When the sumu bites at the bait the jerk is distinctly felt on the line. The net is drawn up quickly, the fish simply resting on the net without any attempt to get away. The net sags as it is drawn up and thus acts like a shallow bag net. A net of this type figured by
The mullet hand net (alangamea) is the most interesting of the hand nets as it marks a novel departure in method. (See Plate XLVI, B.) It is used in catching mullet ('anae) as they jump over a seine net by which they are enclosed. The movement of intercepting anything in the air with a net is known as seu. The netting of pigeons is seu lupe and the alangamea use for matching mullet in the air is seu'anae. The framework is shown in figure 278.
When the net is not being used to intercept fish, the right pole is taken seu 'anae. In action, the net is unwound and the right pole slipped by gradual firm pressure into the fork of the crossbar. The method of use is described later on page 484.
The shrimp net ('u'uti) is much used in the Vaisingango stream near ula vai). (See Plate XLVI, A.) The net has a small mesh (0.4 inches) and consists of a straight piece of net attached to side handles which are straight sticks 0.5 inches in diameter and 44 inches long. The bottom line of the net is tied to the lower ends of the sticks and when stretched apart with the net and sinkers on has a spread of about 5.5 feet. pule (Cypraea mauritiana) about 2 inches wide at the opening. The closed ends of the shells are cracked off so that the long slit opening appears on the other side. (See figure 279.) The bottom line is run up through the marginal meshes at the sides and at the top. The side lines are attached to the handles at intervals with separate pieces of cord. No floats are needed.
The fisherman locates the crayfish and opens out the net holding the top end of the stick handles in each hand. The bottom line is then lowered behind the crayfish and, by sweeping the two sticks round to the front, the crayfish are enclosed. They can then be scooped back into the slack of the net and removed from the water. The inland villages of Mangiangi and Tanung-amanono near the Vaisingango stream are places where the method is still used.
The Samoan casting net ('upenga tili) consists of an ordinary straight net a few fathoms long and about 5 feet deep. (See Plate XLV, D.) It has an upper line with peg floats and a lower line with sinkers. In sangaio) and held in the right hand. As it is too deep to manipulate, it is folded by turning down the upper third. The float line is held in the teeth to stretch the net to its full depth, while the left hand grasps the meshes
The right hand holds the right end of the net temporarily. The float line is seized between the teeth and the weight of the sinkers stretches that part of the net to its full depth. The left hand selects a part of the net a third from the top and transfers the part grasped to the right hand as the teeth let go. The first fold is brought into the right palm between the thumb and forefinger and held by the thumb closing down over it. The float line about 2 feet, 6 inches to 3 feet away to the left is brought in by the left hand and, held by the teeth to stretch the net. The left hand again selects a spot a third down and transfers it to the right hand as the teeth let go. The second fold is passed into the right palm between the fore and middle fingers. The method of seizing fresh folds is continued with regular spacing to the left. Three folds are successively placed between the middle and ring fingers and three between the ring and little fingers. The closure of the fingers is sufficient to keep the folds in position. The last four folds are brought into the right palm on its ulnar side and held by closing the fingers over the palm. The arrangement of folds was 1, 1, 3, 3, and 4, commencing from the thumb side, but if the net is longer, the fisherman alters the number of folds to suit. He gets to know his own net. The number of folds may alter except between the thumb and forefinger which is always a single fold but the order between the fingers must be maintained. To form a circle with the throw, the two ends of the float line must be tied together. If the floats have been dropped backwards, the line must be tied forward or away from the body but if the floats have been dropped forward, the line must be tied at the back or towards the body.
Casting. The fisherman walks along the edge of the lagoon or wades in the shallow water until he sees a shoal of fish. Judging the distance, he swings the folded net backwards and forwards to gather impetus and then, with a curve, to the front of the body and back. From the back swing, he comes forward with the cast. As the hand goes forward, he turns it with the back upwards in pronation as he lets go. The sinker line spreads out in a curve which is restricted to a circle if the ends are tied. The fisherman who demonstrated the method on land, showed his skill by throwing the net over me. The net is thus thrown with a high trajectory into the air and falls fairly vertically round the mark. Subsequent practice showed that the throwing is quite simple if the net has been properly folded.
Some doubt has been expressed as to the 'upenga tili being native Samoan. This is due to confusing the casting net described with the Chinese cast net which is closed at the top, like an inverted bowl. This is a permanent closure whereas the Samoan net is only tied with the float line when being folded. When cast it forms a ring of netting round the fish. The Chinese casting net has become popular in some parts of Samoa but is distinguished by being called 'upenga tili Saina (Saina, China). Sometimes the casting net is thrown without the ends being tied.
The casting net gets the name of tili from the casting process. In Asau, 'upenga mangingi from its being used to catch the mangingi fish which gather in shoals. The net used is about 8 fathoms long and is folded as described above. There are two methods of casting: 1. tili tongi (tongi, to cast) the complete cast with the ends tied to form a ring; 2. fa'asavangatunu, after gathering the folds into the right hand, three are let go as a slack and the end held in the left hand without being tied. When the cast is made, the left end is let go at the last minute so that the net makes a pronounced curve beyond the fish. The fisherman starting forward after his cast drives the fish into the curve of the net. fa'tasavangatunu as to fish with two nets, but in Asau no mention was made of a second net.
Lengths of net with float and sinker lines are used. They vary from short lengths of 8 to 10 fathoms to fairly long nets. The nature of the lagoon bottom over which netting takes place is so irregular from clefts, rocks, and coral outcrop that the usual hauling of a seine net cannot be carried out. Nets are therefore set and not hauled. They form barriers, weirs, or enclosures which prevent the escape of fish. They may be divided into short, winged, and long nets.
Short nets ('upenga fa'alava). Short lengths of 8 to 10 fathoms with pegged float lines and stone sinker lines were used in ordinary family fishing by a small party. At ava and the method of fishing by stretching the net across is tu ava ava. The fisherman dived down to see that the sinker line rested on the bottom and adjusted it into holes, depressions, and around rocks so that no openings were left below the sinker line. The other members of the family, spread out in a curve, worked down towards the net, splashing and beating the water to drive the fish into the net. On the way they subjected rocks to close scrutiny by diving down and feeling or spearing in the crevices. In this manner they caught several fish and others were driven into the net
Short nets were also useful with the artificially made rock heaps. After driving the fish into the heaps, the net was run round it and the sinker line carefully adjusted to the bottom. The stones were then removed by dropping them outside the net line. The fish were speared or caught up in some form of scoop net and the surrounding net prevented their escape, some being caught in the meshes.
The casting net was used for the above purposes quite readily. When opened out across a channel or used round a rock heap, it was an 'upenga fa'alava, but when folded and cast, the same net was an 'upenga tili.
Winged nets. A characteristic net has a purse in the middle called the muli while the two ends are spread out to form a V. The net is set and thus forms a transportable V-shaped weir. It has various names. In 'upenga tali with a bag in the middle and two wings stretching out from it. The name tali (to receive) has an obvious meaning. In Savaii, it is termed 'upenga matalili'i (small meshes) from the fact that the meshes are made small to prevent small fish escaping through the purse. A net examined in Safune, amu branching coral. It was set in an appropriate place and the people worked towards it, breaking the coral with sticks in the tu'itu'i amu process. As the commonest fish in the coral is the tu'u'u, the net was referred to as an 'upenga tu'i tu'u'u, which was merely a convenient way of describing its use.
A much larger net of this type was used at Fangamalo, lauloa method described. The floats were of large branch sections and there were no sinkers attached to the bottom line, large stones from the bottom of the lagoon being rolled over the bottom line when the net was set.
The large net made of breadfruit bast to be described on page 487 was also of the purse type.
Long nets. A long net at uluulu or talaua'au.
In tolomatu) is used in connection with catching mullet. A short one of 26 arm spans was seen hanging out to dry at Iva, tautaulanga. Two pieces were being joined together by running a cord through the marginal meshes of each end. A pointed stick of coconut wood about 3 feet, 6 inches long was used to gather the net on and was called an 'ausi. The pointed end could then be stuck into the roof of the house and thus hang up the net.
At Moataa, 'ausi stakes. Two stakes were used with a man holding each end of each stake. Owing to the large quantity of fish that come into the lagoon, and the length of net required, the fishing operation is a community one shared by the whole village. The net is thus made in sections by the community. At Puapua,
The nets are used to form an enclosure round the fish. The fish are actually caught with the alangamea net described on page 479. After the long net has been set across the direction in which the fish are moving, the fishermen, each armed with an alangamea, take up their positions outside the net and close together. The free end of the alangamea is held tilted slightly upwards towards the set net and the slack of muli bag part is held in the left hand. The fish are driven towards the set net and as they reach it, they jump over it and land in the scoop nets. The scoop is tilted up and the muli slack let go so that the fish slide into it. By bending the heads back, the fishes' necks are broken. They are kept in the slack until 8 or 10 are obtained, when they are emptied into paopao canoes which are close behind the fishermen. Some of the large scoop nets will hold 15 to 20 fish before they need be emptied. The process of catching the mullet in the air with the scoop net is called seu'anae. When the scoop net is opened out and fitted into the fork of the crossbar it is not tied in order that it may be quickly closed when carried to another position.
At Puapua, the mullet fishing takes place between the reef and a ridge of unsubmerged rock running parallel with the reef at a place called Utuutu. A long net is stretched to form a deep U between the rock ridge and the reef with the open part facing to the east. The schools of fish come from that direction and when a sufficient number have entered the open enclosure, tautau 'upenga.
The mullet caught with the tolo matu and alangamea are red-lipped and hence called 'anae ngutu mumu (ngutu, lip; mumu, red). This distinguishes them from the other mullet ('anae Samoa). The red-lipped mullet are stated by tradition to come from alangamea scoop is used for it at Nuuli (
The red-lipped mullet appears first in
In alangamea scoop net is used.
Traditional Origin
Sina and her daughter came fromFiji with a fish calledle i'a a(Sina Sina 's fish). They landed at Sangone inSavaii and travelled overland through the Saleleanga district, the fish following at sea.Sina , who was blind and led by her daughter, travelled through as far as Puapua without anyone offering hospitality on the way. At Puapua, she was entertained byPau andUngalo . In gratitude,Sina gave her fish to the village. To catch the fish, she toldPau andUngalo to order the people to collect mats made of pandanus and coconut leaves. She indicated that the place within the reef, Utuutu, should be enclosed with mats between the ridge of rocks and the reef, leaving an opening towards the east. When the fish entered the enclosure, the open end was to be closed with more mats. The people were to surround the enclosure and catch the fish in other mats as they jumped over. This was carried out successfully.Sina ordered that the catch should be divided amongst the villagers, always after a share for herself had been set aside. The following prohibitions were madesa:No coverings were to be used on the head except lime.
No coverings were to be used on the body except the ti leaf kilt (
titi).In this manner the body was to be exposed to the sun or the rain.
One day
Sina 's daughter came back from the beach saying the fishermen had returned home without leaving out a share for her. Weeping bitterly over this neglect and disrespectSina set out to leave the village, her daughter leading her. The chiefLe Malu , however, met her and finding that she was leaving the village, he begged her to forgive the discourtesy and remain.Sina remaining obdurate,Le Malu abased himself before her by throwing himself face downwards across her path (pa'u fao). He begged her first to rest in his house and then to take his daughter as food for her journey. Such respectful treatment had the effect of inducingSina to stay.
Sina then gave the rule over her fish toLe Malu and delegated toPau andUngalo the task of watching for the coming of the schools of fish in the following words: "'Ia fa'ala ma fa'aua iaPa 'u maUngalo i le va'ava'aina o le i'a pe a sau, 'ae pule 'oe i le i'a" (May the sun scorch and the rain drenchPau andUngalo whilst watching for the fish whether it will come but youLe Malu rule over the fish).Subsequently, Le Malu, in consultation with
Pau andUngalo , decided that Toaloa should bele mata-o-le-i'a(the watcher for the fish).
To this day, a descendant of Toaloa holds the hereditary position of watcher for the fish. He stands to the east of the grounds and signals when the shoals are coming. The story goes on to say that the mats were changed to nets by the order of
Lack of agreement in tradition is seen in the
Two youths of
Tutuila went toManua . On returning on their canoe a man on the coast ofTau called. They took no notice until they found they could make no progress. The canoe would not go forward. They took the man aboard, who was really the god Tangaloa. As they neared the islands ofTutuila ,Upolu , andSavaii , Tangaloa hid the land so that they travelled on toFiji .From this he was called Tangaloa-ufi-nu'u (The concealer of land). At
Fiji , the tale records their success in evading the pointing finger of the King ofFiji , which killed all against whom it was directed. Tangaloa sent the boys back with the'aulosoloso(flower stem of the coconut), the breadfruit (manavenave), and the'anae(mullet). He told them not to bail out their canoe. Opposite Puapua inSavaii , the canoe was heavy with water. They disregarded the command and in bailing out the canoe the mullet was cast out at Puapua. The boys were brothers of Tuiafono.
The use of the alangamea scoop net at va'a) was termed va'a tapa'au. This was covered with coconut leaf mats (tapa'au). A number of them placed outside the tolo matu set net. People then chased the mullet from the inside. When they jumped over the net, they struck the high back and fell down on the bench part where they were secured.
Besides mullet the long net was also used to enclose the matu. It received its name tolo matu from this use. The term tolo is to curve around a net and set it by pushing it forward with the feet as in the use of the lauloa. The net thus encircled the matu and the enclosure gradually decreased in size. The se'i hand net was used for picking the fish up and it formed part of the equipment with the tolo matu net.
Shark nets ('upenga malie.) Shark nets as described in matiata bast. The mesh was large; the D: one with a well marked longitudinal groove on either side was secured by
The bait consisted of two kinds: the maunu seu tied to the top line, and the maunu tau, tied to the meshes at different parts. The bait of fish attracted shark and other large fish which, in trying to secure the bait, got caught by the gills in the meshes.
The net is set outside the reef and at right angles to it. It had to be set in water that was not too deep. The stone sinkers or anchors (taula) had to rest on the ground. Hence the saying, applied to a dreamer who makes impractical suggestions with no sense in them: Lafo le taula i fonua (Drop the anchor where it will reach the ground).
With the first anchor near the reef, and the bottom line resting on the bottom, the floats are dragged under but still serve to keep the net upright. The upper line at each end has a large float, which reaches the surface and indicates the position of the net. The net is set in the afternoon and left until morning as the fish are caught by the gills at night. When the net is set, it is termed fa'atofa le 'upenga (putting the net to sleep for the night). The term is used only with a shark net.
In Asau, 'upenga tanifa) made of breadfruit was used for netting a kind of shark (tanifa) which came into the lagoon in large numbers at a certain season.
Breadfruit bast net. The net ('upenga 'ulu) is made of the bast preferably of the aveloloa breadfruit. The bark from shoots is scraped like the paper mulberry on a board but it is never beaten. When fresh, the bast is easily snapped but when dry and rolled (milo) into twisted cord, it is very strong.
The people at Safune had made a similar net but had given it as a present to the village of muli).
This particular net was said to be made from the 'ulu manu'a variety of breadfruit. It was called an 'upenga 'ulu from the material but the same type was called 'upenga tanifa in Asau from its function. Though the tanifa (a shark with three dorsal fins and a long thin tail like a malauli, caught within the reef in the passage, and close to the village at full tide in season) name is given, all fish that enter are grist to the mill.
Turtle net ('upenga 'afa). The turtle net gets its name of 'afa from the three-ply sennit braid with which it is made. The braid is slightly thicker than the ordinary braid used in lashing houses.
The only net in C. The top line was of three-ply braid of the same thickness as the net material. The floats were the usual round sections of wood, but in tying them with the top line the marginal mesh was caught in with the tie.
The float line was threaded through the marginal meshes before each float was tied on. The bottom line consisted of a well made five-ply sennit braid threaded through the marginal meshes. Stone sinkers were tied on the bottom line by a separate cord.
The net was 24 meshes deep and each mesh was a foot deep. Made in two parts, each about 34 fathoms long, the nets were wound separately on stout poles which served the purpose also of carrying poles hoisted on the shoulders of two men. The village of
The turtle net is used on the rockbound cliff-girt coast west of lauta). The shore ends of the lines then work inwards to join and then advance towards the net driving the turtle into it. The turtle get their heads through the meshes and are caught up in the net. In removing turtle, the front fins are held and the turtle guided in the required direction. In the daytime, the turtle are seen and readily removed. In netting at night, larger turtle are caught. Owing to the darkness, however, the net and turtle are bundled up together and taken ashore.
Methods of obtaining individual fish by the mouth consist of the use of cobwebs, gorges and hooks.
Cobwebs are collected on a forked stick. A length of four to six inches is loosely rolled and tied to a short line on a bamboo rod. The fish caught are a small garfish (ise). The rod is flicked about and the ise in biting at the cobwebs trailing along the surface has its teeth entangled sufficiently to allow of its being lifted out of the water. Sometimes it is difficult to disentangle the fish.
A gorge is a support for a bait, which the fish swallows and is unable to eject. At Tau, it was stated that before metal hooks were used, a piece of wood sharpened at each end and tied at the middle to a line was in vogue. The stick and line were placed in the same straight line and a bait impaled to totally conceal the stick. The bait was swallowed and on pulling the line, the wood got caught cross ways in the gullet of the fish. The only name thought of was matau la'au (wooden hook).
A Savaiian expert said that an appliance was made of two pieces of wood, one long piece and a shorter piece tied loosely to it in such a way that when the line was pulled, the short piece crossed the other. He called it fa'amailei ato. The term ato here carries the meaning of a piece of wood sharpened at both ends while fa'amailei is to form like a trap.
Any doubt of the use of gorges is removed by the existence of actual fau songa. Gorges made of fish bone were also used as shown in the following description by
But the most ingenious method of fishing which I saw at the Samoas was the following: a number of hollow floats, about eight inches in diameter, and of the same height, were attached to a strong cord, at a short distance from each other. To each of these a line was fastened, about ten inches long, at the end of which was a fish bone, made very sharp at both ends, and suspended by the middle; so that when the fish seized the bait, the bone pierced it in contrary directions, and thus secured the prey. The floats answered two purposes; to attract the fish by their whiteness, and to show when it was caught.
The Samoan trolling hooks with shell shanks and turtle shell points mark the highest development in local fish hooks. They leave little to be desired as regards shaping and lashing. Between the gorges and the trolling hooks, however, there seems to have been an experimenting ground from which no successful experiment has spread widely enough to create an established type in common use. A straight point tied at an acute angle to a straight shank has been successfully used but has seemingly remained local. The tilting of fish bone gorges may have led to the idea of tying a fish bone point to a straight wooden shank at an angle. This technique was used in some parts.
Hook with wooden shank and fish bone point. A hook (matau) for catching the mumu fish is described in Samoan by
A fish bone was lashed to a coconut leaflet midrib (
tuaniu). The bait used was theungacrab and thefoleshell fish. The fishing took place from a canoe outside the reef (tua'au) and opposite the channels in it (tuaava). The crab was chewed (mama) and dropped overboard as ground bait. When the fish were attracted they took the hook.
A hook for catching malauli is localized to the village of malauli. The manufacture was the sole right of one family of which Nuu is the present matai or head. Anyone desiring to use such a hook had to give Nuu a present and obtain his permission.
Though there are two definite accounts of this type of hook from different parts, the backward state of hook manufacture is illustrated by such a primitive type being patented and requiring a royalty to permit its use. mumu or malauli type of hook, or to the fish bone gorges described by
Hook, with shank and point of wood. The eel hook (matau tuna) (Plate XLVII, A) is made from two pieces of the hard parts (mata paongo) of the tree fern (olioli) lashed together with fau bast. (See figure 281.)
The bait consists of the large grub (afato) which are about 3 inches long and as thick as the finger. The head part is squeezed to push out the beak which is pinched off. The hook is pushed down into the grub through the opening made and the grub worked up until the hook is entirely covered. The
The fisherman's equipment consisted of a small basket containing a dozen hooks, the grubs tied up in a piece of banana leaf and some coconut husk fibre. When we neared the edge of the lagoon, the hooks were all baited. To the end of the plaited bast carrying the hooks, ordinary strips of bast about 3 feet long had been tied.
The fishing ground was a marshy spot amongst tree roots and dead branches where the stream near Malaeloa, felana or lelepa trees which grew thickly through the lagoon. The small branch with half a dozen leaves was tied to the end of the bast attached to the hook line. The baited hook was thrown into the water with the branch acting as a float (fa'autouto) to help in locating the hook next morning. As it may be towed a little way, it gives to the struggles of the fish, whereas, if tied to a tree the hook might break. The hooks were set in different places, the fisherman making a mental note of his line. As the last hook was set, he marked a tree with his bush knife to indicate the end of his line. The hooks were set in the evening and left overnight. In the early morning we picked up the hooks. Lauofo, the fisherman, could tell by the appearance of the branch floats whether an eel was caught or not. When the leaves were floating on the surface in the same position as when set, he said, "Leai se tuna" (There is no eel). He picked up his hooks as we went along. It was at the end of the line that he called, "
The process of making the hook (matau) is fafau matau. The shank part is the to'o and the short piece with the point, the manga.
On matau tuna is used in the fresh-water lagoon at Malaeloa and in the fresh-water stream at Aoloau.
Dolphin hook. Fishing voyages out into deep water, where not specifically mentioned as alofanga or alafanga in which different unbaited spinners are used, were alluded to as tiunga. In tiunga malie (fishing for shark) the slip noose was used. The expression tiunga masimasi is contantly referred to even now as a form of fishing that was once indulged in. masimasi as a dolphin and Caranx. The Hon. 'iato lima) was used in trolling (toso) a pa hook for masimasi. In Savaii, it was said that the hook was made of two pieces of wood tied together at an angle to form a large hook which was, baited with a whole fish. Exact detail of the hook is lacking but it is evident that a large baited hook probably made of wood was trolled for masimasi.
Other wooden hooks. matau fa'ato'elau. He describes a method of fishing from a canoe drifting before the southeast trade wind (to'elau) with a rod for smaller fish such as the ngatala, as the fa'ato'elau method of fishing. Here to'elau is from the trade wind name. He then goes on to explain the matau fa'ato'elau as referring to a facsimile of the iron barbed hook made of wire, a pin, or also cut wood, where the name means, "according to the example of Ruvettus hooks reproduces to'elau is certainly confusing but the type of hook figured belongs to the Ruvettus hook with unusual features may be correct as the Ruvettus was purposely caught in the Ruvettus according to Ruvettus by accident.
Ruvettus hooks from Nauru but it is more likely to be a masimasi hook, if it is an old time. Samoan hook.
tiunga malie (shark fishing) after dealing with the noose method at length, quotes tangi, and the large wooden hook baited and trolled for masimasi. I would suggest that the Dresden hook is not for shark but for masimasi.
The very lack of hook material in museums is due probably to the poor development of baited hooks and their crude form not appealing to collectors when they were in actual use. The opportunity is now lost as the trade hooks have superceded the baited hooks but have made little difference to the unbaited hooks, which still survive. The active use of the gorge may be taken as an indication that invention in the direction of bait bearing appliances had not reached a stage of satisfactory progress in Samoa.
The hooks dealt with come under the general name of matau, while pa denotes the specialized form which are trolled as "spinners." The pa is a true lure made to represent a small fish. It is drawn through the water without bait. The movement deceives fish into seizing it when they are caught by the point of the hook.
Classes. Samoans divide their pa hooks into four classes; pa tangi, pa'atu, pa ala, and pa seuseu. Each class is distinct in details of manufacture, method of use, and the fish angled for. Field work can supplement the museum study made by
Parts of the hook. Hooks are regionally divided into the shank, bend and point. The point may be provided with a barb. Samoan trolling hooks are composite and consist of two pieces. One piece forms the shank, which is shaped to represent a fish and is termed the pa. Thus, the shank which is the characteristic feature of the hook as against the general matau came to represent the type. The other piece forms both the bend and the point and is called the manga (branch or fork). maga (manga) is the pearl shell shank.
The second piece, though it forms the bend and the base lashed to the shank as well as the point, will hereafter be referred to as the point in preference to barb as used by
The snood is the piece of cord or line attached directly to the hook which together with the lashing forms an essential part of the completed hook.
The hackle consists of something added in the form of fibre, feathers, or hair to represent the tail or fins of a fish and so adds to the efficiency of the lure.
Manufacture of hooks. The making of hooks was expert work and a master fisherman (tautai) was not always a good hook maker. A certain amount of ceremonial is observed in making bonito hooks. The craftsman works indoors seated on a raised pile of mats. When employed by a chief, the chief has to make a special oven of food and send him a basket of cooked food of good quality.
The materials required were shell for the shanks, turtle shell for the points, cord for the snood, fine threads for the lashings and feathers, and strips of fan songa for the hackle.
The tools were the drill (vili), rubbing stone (foanga), and cutting implements. In these days, a saw is used for cutting the shell and a foreign grindstone for rubbing down. In ancient times, stone flakes must have been used for cutting the shell. The Samoan type of drill (fig. 282) is still in common use but the point is steel instead of stone. The disc which acts as a balance is termed tateme, livaliva, or vinavina.
The drill is used by twirling the upright so as to wind up the cords supporting the handle. The winding raises the handle. The point of the drill is placed on the object. When downward pressure is exerted on the handle, the cords in unwinding cause the upright to revolve. Sufficient pressure is used to cause the upright to go on revolving after the cords are fully unwound,
The movement of the drill has been made the subject of many sayings. In them the credit is given to the point. The revolving balance is somewhat unjustly regarded as making a lot of movement or dancing (siva) and doing no work. "Ua tu le matavana i le fmgota 'ae siva le livaliva" (The point formed by the echinus stands up [to work] the balance dances).
temeteme is a variant of tateme.
The tangi is a deep sea fish larger than the bonito. The hook as verbally described was said to be larger than a bonito hook. It was a composite hook made in two pieces and the point was lashed to the shank in exactly the same way as in the bonito hook. No complete pa tangi was seen and it is not figured by B, 11) at pa tangi. The shank, made of lei (whale ivory) is long, with its widest part near the head pointed end, and narrows to the end which bears the point. The front has a flat surface for the attachment of the point, while towards the head, the two rounded sides meet in a median ridge which extends to a point. A transverse hole is bored through under the median ridge in the head part. The back, convex longitudinally and slightly so transversely, is covered with colored pearl shell, neatly fitted and lashed through two paired holes which meet in the body of the shank to allow the lashing to pass through. The paired holes are towards either end of the pearl shell plate. No point was obtained with the shank, but the owner stated that one made of niuvao hard wood had been attached when used by his grandfather.
Tongan hooks with identical shanks and barbed turtle shell points are well established. They are figured by B, 12.
The fact that the type of shank is not elsewhere recorded from Samoa throws serious doubts on the shank being native Samoan. Confirmation is required before it can be accepted as representing the pa tangi of Samoa. On the other hand I have no definite information as to what the Samoan pa tangi was like except that it was a trolling hook of larger size than the bonito hook.
The pa'atu (pa, hook; 'atu, bonito) is a composite two-piece hook with a shell shank and a turtle shell point. (See Plate XLVII, B, 2-5). It is still in
Shell. All shanks are now made of pearl shell, as the shell is readily obtained from other parts and even stocked for sale by traders. Occasionally in a remote village, a person not having pearl shell may fall back on other shell or even wood. As pearl shell does not belong to Samoan waters recourse was had in former times to other material. A tofe bivalve (Perna) and a silicaceous rod-like material (tio) formed in the borings of certain sea worms in coral rock were used and probably other shells as well.
The shell is selected in various shades of color to suit different conditions of water and weather. The pa tio forms a very white hook suitable in cloudy weather. So also is the pa usi. The pa laumilo is yellowish-brown (memea) and the pa ulia has a dark part towards the distal end. The pa lautofe (Perna) is shiny and iridescent. In pa sulu was given to a brownish color (enaena). pa lanulua and pa lupovai as other varieties of shade and material. The craftsmen were expert at producing shades of color by varying the amount of the dark outer surface removed in grinding.
As regards size, the large hooks were named pa no'ono. The smallest size was termed pa maunu as it was said to act as bait (maunu) to attract fish.
Shank and point. The pearl shell was cut across the thick hinge part to include it for the head (ulu) while the thinner inner part of the shell formed the tail (i'u). Head and tail are the terms applied by the Samoans to the proximal and distal ends, and as the shank is made to represent a decoy fish, the terms are useful descriptively. Sizes, vary but a fair range in length is between 65 and 105 mm. (See figure 283.)
The point is made of turtle shell (una laumci) although some are made of pearl shell, generally of the darker part towards the edge of the shell (See Plate XLVII, B, 4.) It is, however, not usual and the points so made have probably been due to lack of turtle shell and not to election. The points tare devoid of barbs and are shaped as in figure 283, c, d, and e.
The hackle is composed of strips of fau songa fibre.
The lashing of the two parts together is termed fausanga. Considerable variation exists in the number of times the binding thread passes through the various holes, the circumferential turns round them, the twists round the cord, and the fixation by half hitches and stop knots. Each combination forms a fausanga which is no haphazard arrangement but one with a definite count. Different experts have their own combinations. Some fausanga which are supposed to be better and luckier than others are kept secret by their exponents. Hooks with the attached snood were sometimes stolen by fishermen in order to study the lucky combination and commit it to memory.
In dealing with the shank, the tail end which carries the point is termed lalo (below) and the head end with the snood is lunga (above.) Lalo and lunga as meaning below and above are also applied to relationship with the snood (ta'a) when the point is in the horizontal position. The fausanga besides including the whole process of lashing also refers particularly to three local lashings that pass through the two holes of the point and the hole through the head of the shank. The first step is to lash the manga point to the distal end of the pa shank and the second to attach the ta'a snood. The lashing to be described was written down by
The point (manga) has been firmly fixed to the shank (pa) by three lashings through the inner hole (fausanga loto) and the lower tail hole (fausanga i'u) while four running loops have been placed in position for the attachment of the hackle (senga). The next stage consists of attaching the snood (ta'a) to the point. (See figure 285.)
The preliminaries of attaching the snood and lashing cord (alaala loloa) to the inner hole of the point having been done, the hook is fixed in an
With the hook stretched taut between his knees, the craftsman has both hands free to complete the lashing of the snood to the head of the shank as in figure 287.
The lashing is quite simple when worked to a formula. The number of turns and knots, and whether the langolango lashings came from below or above, can be varied to form different formulas. The one given was the favorite technique of a master fisherman. Some lashings instead of tying the ends of the threads of the fausanga i'u beneath the loops for the hackle tie them over the end of the hook between the shank and the point.
A bonito hook made entirely of coconut wood was seen by olioli) were used as a substitute.
Rods. The hook was trolled from a length of line tied to a rod, Rods are of two kinds; the long and the short.
The long rod (launiu) of bamboo ('ofe) is about 15 feet 6 inches long, 2 inches in diameter at the thick end and tapering off to 0.75 inches at the other end. The thick end is fitted into a wooden handle. The wooden handle (tu'au) has an upper grooved part for the rod and ends below in a knob
umele). The shape of the handle and the method of lashing the rod is described in figure 288.
A short rod (matila) is made in exactly the same way but with a shorter length of bamboo and a slightly smaller wooden handle. A matila examined had a bamboo rod 8 feet long and the tu'au handle was 39 inches long.
The lashing may differ in having transverse instead of crossing oblique turns. After fitting the rod to the handle, the lashing commenced around the top end of the handle. (See figure 289.)
The hook rest (silinga). About 6.5 inches from the inner end of the groove, a transverse hole is bored through the stock close to the inferior median edge. Through the hole, a twisted cord is drawn to its middle and then the two ends are twisted over each other in the first movement of tying a reef knot to form a series of continuous loops each about an inch long. The loops are continued outward until they reach the commencement of the grooved part of the handle. The loops are really made before the lashing of the handle commences. The ends of the two cords are held against the under surface of the stock, and the transverse lashings round the butt end of the bamboo rod and the handle are also carried over the silinga cords so as to fix them. The hooks are hooked into the loops when the line is not being trolled.
Another type of silinga is formed by stretching one end of the cord taut after it has been passed through the hole. The loops are formed by making open half hitches round the taut cord with the other one. The ends are fixed in the same way.
Attachments. The end of the hook snood is tied to a piece of line which must be longer than the rod. Some were simply tied by placing the two ends together and tying a single overhand knot with the double cord. Another method is to tie an overhand knot around the line with the end of the snood. The end of the line is then tied in an overhand knot around the snood and the two knots drawn together.
The hook is first stuck through a loop in the hook rest and the line drawn taut to the outer end of the rod. At about two inches from the end, the line is tied around the rod with a clove hitch. It is usual to attach two or more hooks, each with its own length of line, to the rod. Another hook is stuck in another loop of the hook rest, the line drawn taut and tied with a clove hitch to the outer side of the first. Two other hooks may be treated similarly, the hooks being hooked into different loops and the taut line tied with a clove hitch just beyond the preceding one. The four slack ends of line are then brought down the rod together and fastened at short intervals with half hitches round the rod. When they come to the end, one which is purposely left longer than the others is tied in a half hitch over all four and fixed to itself with an overhand knot.
The number of hooks attached to a rod may range from one to eight as the fisherman desires. Different kinds of hooks as regards color and shade were attached and all being hooked to the hook rest were out of the way and the lines were all taut along the length of the rod. On the fishing ground, the fisherman unhitched the one he thought suited to the conditions prevailing. If he wished to change, he merely hooked the one that was being trolled into the hook rest and unhooked what he thought more suitable. Thus the "changing of flies" was quick and easy. Also if, while in the midst of a matila rod. More than one was strictly prohibited (sa lava).
There can be little doubt that the long rod with many hooks is a development from the shorter rod with one hook. The name matila (matira) is a widely spread Polynesian word for a fishing rod while launiu (coconut leaf) is a later word used to denote the later development which probably was associated at first with different status in rank. Bonito fishing was a chiefly pastime and the launiu with many hooks was probably restricted to chiefs while others used the shorter matila with one hook. In the Samoan dialect, the older, widespread Polynesian words are the common language and the so called chief's language has coined new words that have a purely local significance. As time went by, the restriction in the use of the launiu disappeared but the restriction of one hook to the matila persisted.
Bonito fishing takes place only from a bonito canoe. The use of the special parts of the canoe can now be followed. When the canoe paddles out the rod with its equipment of lines and hooks, rests horizontally across the booms on the rod rests. When the fishing grounds beyond the reef are reached, the rod is lifted up by the steersman who is also the fisherman. The appropriate hook is unhooked from the hook rest and dropped into the water and the knob projection on the lower end of the wooden handle is inserted into the sennit loop at the back of the steering seat. The rod is then rested on the groove in the rod post (pou 'ofe). In Savaii, we have seen that the rod posts on the stern cover are wider than usual and have two rod grooves the one on the right being at a slightly lower level. In such canoes, there are also two sennit loops attached to the back of the steering seat. The canoe may then carry the two types of rod at the same time, the short matila being on the right and the longer launiu on the left. The lower groove on the rod post supports the short rod at a different angle. It is thus possible for the steersmen to reach round with the right hand and grasp either of the rods. Both rods have to be swung round on the right. The short rod with a shorter line and at a lower elevation can be swung in without interfering with the other. The longer rod at a higher elevation and a longer line can be swung in to the right above the shorter rod without any confusion. In
Two rods were formerly used but they have been abandoned in favor of one. The advantage of two rods is, of course, that there are two hooks out and at different distances from the canoe. The diagonal rod (pu'enga or manu) between the rear boom and the left gunwale of the canoe forms a hold for the left hand as the steersman swings in the rod with the right.
Diversity of opinion exists between eastern and futia. The handle knob is termed umele in muli tuitui in muli tuitui, muli means the end or bottom of the handle and tuitui is from tui, to thread or pass a point into a hole. The term umele is used in futia loop and tied in a loop around the rod post. The umele loop is adjusted so that the futia loop cannot be pushed down by the weight of the rod. Much argument was provoked by asking for the meaning of the well known saying, "Ua o fa'atasi le umele ma le futia" (The umele and the futia have fitted together as one).
The easterners naturally held that it meant that the projection on the lower end of the handle had fitted into the sennit loop and all was ready. The westerners held it to mean that the rod post loop and the rear seat loop had been fitted together and everything was in the right position for the rod. The easterners do not use the rod post loop and therefore the term umele is definitely restricted to the handle knob. The westerners maintain that without the rod post loop or umele the futia loop is pushed down and the rod end jammed under the seat. It is, therefore, less sensitive in transmitting impulses or pulls from the top of the rod to the lower end. The significance of this is that the steersman as he sits on the seat also rests against the lower end of the rod behind him. He does not hold the rod when endeavoring to keep pace with a school of bonito. Both hands are fully occupied in vigorously plying the paddle for he has to steer the canoe as well as paddle. He, therefore, pays no attention to the rod while the hook is merely trailing in the water. When, however, a bonito takes the hook the backward pull comes on the end of the rod.
The rod post acts as a fulcrum and the lower end of the rod is levered forward. The forward thrust is conveyed to the part of the steersman resting against the rod. He immediately drops his paddle in the canoe, reaches around with his right hand, grasps the rod, pulls it towards him and lifts the self-hooked bonito out of the water. With a circular sweep to the right he pivots the rod in the futia loop and swings the fish in from the right into the canoe. To brace himself, he holds the diagonal hand rest with his left hand. The bonito drops off readily as the hook has no barb, or a forward member of the crew removes it. The steersman drops the rod back on the rod post, picks up his paddle and while keeping up the strenuous race with the school, he awaits the next forward impulse of the rod. The rod post loop by keeping the futia loop up makes a more sensitive trigger and is evidently a western improvement. The term umele is more likely to be originally associated with
Methods of fishing. The method of bonito fishing from a canoe with a trolled hook is termed alonga'atu or alofanga. To go out bonito fishing is alo and hence the bonito canoe is va'a alo. In tali tata ridges to protect the bottom lashings from the bailers are just in front of the middle seat. The bonito go in schools (ingafo) in search of food. The lookout watches the sea for signs of the school in the ripple of the surface and heads showing up. Detecting signs he calls back, "Up with the rod." The steersman puts up (langa) the rod and drops (lafo) a hook in the sea. According to
A watch is also kept for flocks of sea birds which pursue schools of small fish to whirl and swoop down on them. The direction of flight is noted and the canoes race to intercept them. If once the canoes can get above the fleeing school of small fish and keep pace with them, the trailing hooks will be amongst the pursuing bonito school. They then take the hook as fast as it is dropped back into the water. It is here that the lightness and speed of the bonito canoes is required. The fishermen make the most of their time while in position for there may be weary miles of paddling before such an opportunity occurs again. Once the school gets past the canoe, the hooks trail in untenanted water.
When the first bonito was landed in a new canoe, the phrase used was, "to'ia le liu" (the hold has been struck). Many sayings are associated with bonito fishing. One is drawn from the fact that bonito chased by a sa'ula
In season, the bonito canoes go out in a fleet during the dark so as to be out on the ground at daybreak. This early departure is termed fa'aao.
On days when the fleet does not set out from villages close to the reef, the bonito canoes are nevertheless kept ready under the trees above high water mark with the rods on the boom rests and the paddles in the canoes. Someone or other is constantly scanning the sea and on the appearance of any flock of sea birds, the cry is raised. The crews dash from their huts, launch their canoes, and are off at full speed to get in the line of the flight.
The work, however, is strenuous and canoes are often out all clay without any luck. They often paddle long distances far from land. It was seeing a fleet of bonito canoes far out at sea that caused
Bonito seasons. The seasons for catching bonito correspond to the breadfruit seasons, of which there are three during the year.
During the month itself, certain days are particular fishing days, and the bonito caught or sought after are named:
The bonito sought at the end of the month when they are scarce are named 'Atu o le sela ma le miti loa. The term sela is to be tired and miti loa, the perspiration which drips down from the nose while the crew keeps on doggedly paddling. Thus, it is the "bonito of weariness and profuse perspiration."
Lucky hooks. The hook that figures prominently in old traditions and and is used figuratively to denote good fortune is the auamanu while that which brings misfortune is the auamala. There is a myth about a rock in a river in fausala).
Mistakes. When a hole is drilled in a hook and is not used in the lashing it is obvious that it has been the result of a mistake or an unsuccessful experiment. A bonito hook in
The hand line trolling hook (pa ala) is smaller than the bonito hook but like it is a Composite two-piece hook with a shell shank and a turtle shell point. It is trolled from a canoe outside the reef with a line but no rod is used. The fishing usually takes place in the early morning and the fishermen have to wake very early in order to be on the fishing grounds when morning breaks. The pa hook is thus named pa ala from ala, to wake from sleep. For a similar reason, the method of fishing is termed alafanga but it must not be confused with alofanga (to fish for bonito).
Shell. The shank is made from various shells. Pearl shell is rarely used. The best types are made from pala'au. This forms the material of the five hooks presented by the fatuaua and foafoa.
Shanks and points. Two types of shank are made: with a rounded head if the kind of shell used is thick enough, or flat throughout if it is not. The B, 6 and 7) is made of pala'au shell cut a rounded head while the ordinary thickness forms the tail. (See figure 290, a-d.) The flat type of shank (Pl. XLVII, B, 8) is made from fatuaua and in such a manner that a thick part of the shell provides sufficient material for other shells which are not thick enough to form a rounded head. (See figure 290, f-i.) The thickness of material influences the direction and number of holes for the lashing of the snood to the head of the shank. In the rounded heads, which may be 10 mm. thick, a single hole from side to side is bored through as in the case of bonito hooks. In the flat heads, which are 5 mm. in thickness at most, a hole has to be bored through from front to back on either side of the middle line to provide entrance and exit holes for the lashing to pass around the snood. In five pa ala in
The points are made of turtle shell 3 mm. thick and of the same shape as the bonito hooks with a long base, bend and sharp points without barbs. pa ala hook with a barb on the inner side of the point but it is not the normal technique and has probably been due to foreign influence. The hooks are still made and it would be an easy matter for a craftsman to copy the metal barb of a trade fish hook. Holes are drilled through the base of the point and the normal number is two, as in the bonito hooks. The point in figure 290 j, has four holes; two for the lashings, a third special hole for the snood, and a fourth without use. The provision of a third hole for the snood is not normal and may be regarded as a later development but whether it took place before native pastors and others introduced innovations from other areas, it is difficult to say. The fourth hole is due to faulty judgment in placing it to the right out of alignment with the two lashing holes which altered the line of pull. On tying the snood to it, the hook proved unsatisfactory and another hole had to be bored in the same line as the two lashing holes. The fourth hole (third in order of boring) thus became useless, serving no purpose except to indicate that mistakes are made. Had the unlashed point found its way into a collection, an ethnologist would be tempted to say that there were three lashing holes and a hole for the snood Departures from the normal that are difficult to explain may thus be due to simple mistakes as well as freaks of genius. (See fig. 290 e, and j.)
The lashing in the usual two-hole point follows that used in the bonito hook but there are usually less turns through the holes and round the shank. The inner lashing (fausanga loto) may not have the transverse langolango turns. The fibre hackle fixed by the outer end lashing though present is smaller and shorter, ranging from 7 to 10 mm. in length. It is not fixed transversely by special sele loops as in the bonito hook. The fibre is laid longitudinally along the back of the shaft and included with the shaft in the" first turn of the end lashing (fausanga i'u). The end towards the head is then doubled back and the subsequent turns of the lashing pass over both limbs of the hackle concealing" the doubled-over part.
The snood is formed by the end of the length of line to be used in trolling and is composed of a three-ply twisted cord of fau songa which is much finer than the snood of a bonito hook. In the two-hole shank it is passed through the inner hole. It is tied with a single knot and the short end is doubled back along the snood to be pushed under a raised ply near the head as in the bonito hook. A long alaala fine thread is tied to the inner hole and wound round the snood for a few turns. A small white feather is laid on each side of the snood with the quills towards the head but not reaching the head hole. The thread is wound spirally round both quills and snood and a couple of half hitches made to firmly fix the feathers. The thread continues the turns round the snood. Above the transverse head hole, the thread passes down
In the flat shank the lashing is similar, as the snood passes between the two head holes. In making the turns over the snood, the thread passes down through one hole and up through the other.
In the pa ala lashing, the thread does not pass directly from the inner hole of the point to the head hole before twisting round the snood as in the bonito hook. Various extra turns to those described are made with the fine thread in some hooks.
In the three-hole point, the inner lashing is made through the innermost of the three holes while the snood is passed through the middle hole. The two limbs of the snood then pass back to be tied in a single knot beyond the head end of the base of the point. In passing back, the snood covers the inner end hole and it may not be noticed at first that there is a third hole.
The line. The fine line which is continuous with the snood is 9 or 10 feet long and is called the matai afo, which really means the head or leader of the fishing line. This is joined to the afo (line) which consists of five-ply sennit braid about 22 feet long. The join is made by tying an overhand knot on one end of the sennit line and then another overhand knot about an inch further back. The end of the fine line is stoppered with an overhand knot and then tied round the sennit line beyond its inner knot with an overhand knot. A couple of half hitches are then made around the sennit line between its two knots and the join is complete. The other end of the sennit line is stoppered with an overhand knot and then tied around its standing part with another overhand knot to form a long loop.
When not in use, the sennit line is wound up in long loops as shown in Plate XLVII, B, 6 and 7, and the fine line wound transversely around one end for a number of turns, tied in a half hitch and finished off in transverse turns around the other end. Another half hitch is made and the hook point stuck under the transverse turns. Numbers of these lines characteristically wound up have found their way to museums and have been confused with bonito lines and hooks. The pa'ala is always attached to a fine fau songa twisted line and the line lengthened by five-ply sennit braid. The bonito hook is never attached to such a line and when not on a rod, is attached to a short snood of much thicker cord.
Feather hackle. More than two feathers may be attached to the snood. The quill ends may reach the head hole and some of the barbs included under the crossed turns over the snood. The tips may cover the point. pa ala hook with the bonito hook. Feathers are not attached to the pa ala and not bonito fishing as
Method of fishing. Fishing' takes place in the early morning, commencing at about 4 A. M., or in the evening from 5:30 P. M. to 7 P. M. The small paopao canoes are used in the trolling which takes place outside the reef. The slip loop on the end of the sennit line is tightened around the big toe to fix the line. Movement but not speed is required as there is no racing school of fish to keep up with. Because it is less strenuous, the method is freely indulged in by chiefs and is regarded as a chiefly pastime.
The line is dropped overboard and the canoe may be paddled with both hands as the line is attached to the big toe and any bite can be felt by that member. The usual way, however, is to pick up the line with the right hand and give it a pull now and again (fa'atata) while the left hand keeps the canoe moving with the paddle. The method was seen in use inside the lagoon at malauli.
Kinds of hooks. A number of names indicating color and material are given to the pa ala as with bonito hooks. pala'au shell are named ulutoto or laveuli, lavelei, and ululalafi. The term ulutoto (ulu, head; toto, blood) evidently refers to the dark markings of the shell at the head. Three flat shanks are named lau, pa ala, sina, and foafoa. The foafoa hook is figured with two pairs of holes through the head. From appearance, the lower pair is misplaced by being too far back from the head point and thus causing side darting when trolled. To correct this, the craftsman bored the second pair very close to the point, unnecessarily close probably owing to the previous mistake.
The pa seuseu (Pl. XI, VII, B, 9, 10), the smallest of the four types, is attached by a line to a hand rod which is repeatedly cast and drawn in as in angling for trout. The motions of casting and drawing in is called seuseu and gives the name of pa seuseu to the hook. Owing to its small size it is also called pa laiti (laiti, small).
Shell. The hook is a two-piece composite hook with a shell shank and a turtle shell point originally. The old time point has now been completely superseded by small trade fish hooks of metal.
The shank owing to its smaller size with no problem of a deeper head is made from any of the shells enumerated with the bonito and pa ala hooks. The edges and left over pieces are also utilized. 'ali'ao (Trochus), alili (Turbo), faisua (Tridacna), fole (Pinna), and tofe (Perna). The fole seems rather thin material but it is evident that in small hooks, any suitable shell was used. Owing to the small amount of material required to form a shank the range of shell material was vastly increased. Different districts utilized the available shell along their coast boundaries. The tupe (operculum) of the alili (Turbo) is also used. The use of the various shells was to get different shades of color.
Shaping. The shell is now shaped with the saw to a long rectangular form and the back rubbed down on a stationary grindstone kept wet with water. Two holes close together and in the same transverse line are bored through at one end with the Samoan drill. The rubbing down of the back passes through three stages: first it is rubbed flat until it approaches the required thickness; then it is rubbed at a slant on either side of the middle line so as to form a median longitudinal edge; lastly the median edge is rounded off. In the last stage, the shell is frequently dipped in water to clean it and held up to the light to see that it has the right shade of color. The different color of the outer surface of the shell is carefully ground off towards the sides in some hooks so as to leave a median streak or a patch near the head which is looked at from the front when it shows up through the inner clear surface of the shell. The required shade being obtained, the sides are shaped to a point at the head end and a narrowing slope towards the tail. The wider head end is for the holes. The thickness of the shank thus varies according to the shade of color required. The shape of the shank varies considerably as shown in figure 291.
Lashing. The original points of turtle shell are rarely seen now and never used. e.)
Though f-h.). A metal hook of appropriate size to fit the shell shank was selected. The end of the three-ply fau songa line was used as a direct snood with pa ala hooks.
The true Samoan point was probably fixed to the shank in the same manner as with the hand line trolling hook with a flat shank and two holes. (See figure 290, f-j.)
Use. The hook with an appropriate length of line is tied to a light bamboo or other wooden rod. The fishing takes place usually between the reef and the shore and this form of fishing is termed aloalo as well as seuseu which gives the hook the alternate name of pa aloalo. Generally fishing take paopao canoe. In paopao canoes were fitted with rod rests on the float and boom to carry the rod when not in use. As stated, the fisherman makes casts with his rod, draws the hook through the water towards the canoe or parallel with the side and casts again. Where the water is not too deep, the fisherman wades about with his rod and puts his catch into a basket tied around the waist. On calm days, he may stand on the outer edge of the reef and fish in the many clefts and small channels with which the reef is seamed.
A large range of the smaller fish were caught such as the ngatala, matamu, malai, matalau, 'ata'ata, umiumia, sungalupe, and patangaloa.
The sexes. The sphere of the two sexes in fishing was clearly defined. Women spent quite a lot of time in the lagoon but the methods open to them were restricted. A woman's ordinary field kit consisted of an ola basket slung over her back or around her waist, a strong mele'i pointed stick for prising up stones and shell fish and a slender rod (la'au sao) for driving squids out of their holes. With this equipment she searched the shallow parts of the lagoon and the dry reef at low tide. The tu'u'u trap and groping under rocks formed other methods. In company with others, the ola tu plaited basket was used as a manipulated trap around the stone heaps and in connection with battering the branching coral amu. She also assisted on occasion in driving with the long leaf lauloa and in these days she may be seen assisting to drive into the shorter nets placed across channels.
To men all methods beyond the strictly female methods were open. In individual fishing he pleased himself but in community methods there had to be a leader who by experience and recognized authority could decide what methods were to be used, when and where, and take command over those engaged. This need created the position of the head fisherman (tautai).
The head fisherman. It does not appear that the Samoans delegated fishing to one particular class who did nothing else. The whole male community indulged in fishing including high chiefs if they so desired. Fishing was a sport as well as a food procuring activity. Some of the community methods were occasions of fun and excitement and corresponded to a combined picnic and sports gathering that took place in that most important Polynesian playground, the lagoon. All experienced fishermen were classed as tautai but amongst them one was elected by common consent to the position of head tautai of the village. In a large village with distinct divisions or groupings, each division might have its own tautai. The heads of families met in a guest house, and over a bowl of kava decided according to the season what form of community fishing should take place, as quantity or shoals of particular fish could only be secured adequately by the cooperation of numbers. tautai naturally assumed command. In the community methods in the lagoon or outside the reef, the head tautai was in absolute command. He was the chief and men of higher rank and authority on shore were inferior to him on the realm of the sea. A clear cut distinction existed between the two spheres of influence. The authority of the tautai did not extend to the land except, of course, in the preliminary arrangements connected with fishing.
The distinction was conveyed in a saying which amounted to a law.
The authority of the tautai is best exemplified by the bonito fleet.
The bonito fleet. The bonito canoes usually go out in a fleet over which the tautai assumes command by right of his position. He selects the grounds to be visted and decides on the movements at sea. If a high chief accompanies the fleet, he does so as a private individual. His social position as a chief, however, is recognized by giving him the first bonito caught on the first morning that he accompanied the fleet. If he is of very exalted position, he gets the first fish on the second morning. Full respect having been paid to his shore rank, he lapses to the position of ordinary fisherman with the others. His authority remains on the land and right from the beginning he obeys the commands of the tautai. The first fish given to a high chief as a purely ceremonial form of respect is termed ngalongia.
When the tautai decides that it is time for the fleet to return to the shore, he raises his paddle as a signal and all obey. Before reaching shore, the tautai further exercises his authority by making a levy (aleanga) on the fleet. He takes from the successful canoes, one or more bonito according to their catch. The fisherman with moderate success goes free if his catch is small as compared with that of others. The aleanga is not for the tautai but to form the material for a community feast for the fishermen in which the unsuccessful member shares equally with his more fortunate fellows. The tautai in making his levy calls to each canoe, "How many?" On the reply he demands the appropriate number which are thrown into his canoe. He cannot always see what is in each canoe. If, however, it subsequently transpires that a fisherman has avoided the aleanga by giving a wrong account, he suffers the penalty of departing from custom in thus not recognizing the authority of the tautai. His bonito canoe is broken up and his fishing gear confiscated to the tautai.
There are two forms of mobilization of the bonito fleet over which the tautai rules. One is the fleet that goes out in the early morning to get on the alei or va'a alei. Watch is kept for the shoals (ingafo) of bonito which follow shoals of small fish and whose presence is shown by the flocks of sea birds which follow. When the flock of sea birds is seen, the signal is given, the canoes are quickly launched, and the fleet is away under the command of the tautai.
Limitation of power. The tautai rules over a community acting together. His authority does not apply to the individual acting away from the community. This is illustrated by the individual canoe termed tulialo or va'a tulialo. While the fleet is out, a chief, seeing a passing flock of birds, may send a bonito boat out to try its luck. Should the tulialo canoe come up with the fleet it does not come under the authority of the tautai as it did not enroll, so to speak, with the fleet by mobilizing with it. This is recognized by the tautai for he cannot demand the aleanga levy from it. The individual canoe is responsible with its catch to the chief who sent it. In this way, in spite of the law limiting the sphere of influence of the land authority, the chief still exercises some authority but it is very limited and does not clash, in the Polynesian mind, with the sea authority exercised by the tautai. The tulialo canoe does not exist as far as the tautai is concerned and it is only accident that brought it in the vicinity of his fleet.
Observances. In addition to main principles, there are a number of observances that must be carried out and so become established as custom. Such are those based on hospitality, but enforced by custom for those who are not innately hospitable.
Fishermen, on coming in, must give a fish or a portion of fish to anyone they meet in the water of the lagoon or on the shore. These people, of course, have not been fishing, and are termed tui atua. The share given to them is termed tufaanga sa tui atua; sa means the member of the tui atua, and is not prohibited or sacred. By this observance, the people who could not go out are assured of a share. As the fleet comes in they go down to meet it and obtain the tufaanga sa tui atua as their right. Custom saves them from the opprobrium of being regarded as mendicants.
The custom applies equally to men of rank such as chiefs and talking chiefs. They have only to meet the fisherman anywhere on his journey to his house and the fishermen have to recognize their superior position by giving them of the best.
Prohibitions. There are always prohibitions in fishing as in other activities to avoid bad luck and form an excuse for ill success though the latter reason is not verbally expressed.
In bonito fishing, as in netting mullet, the fishermen wore nothing but ti leaf kilt in olden days, and now wear nothing but a cloth kilt. The upper body must be bare. Nothing must be worn on the head except lime which is used as a protection from the sun. The taumata eye shade is allowed. Other prohibitions exist such as not spinning the paddle in the air, leaning back in the seat or stretching the legs over the topsides. The last is termed sapoliu. There are also pre-fishing prohibitions: the craftsman making a bonito hook must be seated on a pile of mats and not on the floor and no noise must be made in the vicinity while he is making or lashing. A means of magnifying the importance of the hook maker and insuring it by the threat of non-success to the hook if the respect is not duly paid can be seen in this last prohibition. Many other prohibitions have probably been in force, but have not survived the clash of cultures.
As regards food prohibitions during fishing, the Samoans did not have any as evidenced by the fact that the cooked food fu'efu'e or lafoa'i was taken out on bonito fishing expeditions.
Status of bonito. That the bonito had some status amongst fish is indicated by the use of special words such as were created around the rank of chieftainship. The common name of bonito is 'atu, but it also has the poetic and honorific names of pau and pa 'umasumu. Large as applied to a bonito is not tele but sumalie. In counting them, they were grouped in tens expressed by prefixing tino to a unit as tinolua (twenty bonito).
Hook obtaining custom. The custom of obtaining ready-made material from some one else applied to bonito hooks. A master fisherman could call on another with the view of obtaining hooks from him. Such a visit was malanga fanga. From such visits, the hooks in active use that were tied to the rod and stuck in the silinga were exempt. The total number of hooks sa'ana on the rod may have been an inducement to a fisherman to have extra lines on his rod, not only for variety as already described, but to save them from a visiting fisherman on a malanga fanga. Should the visitor announce his wish to see his compatriots' hooks, the tautau basket containing the spare hooks has to be taken down and emptied out before the visitor. To thus cause a fisherman to turn out his spare hooks is termed fa'ausu. The visitor handles and examines the hooks and either directly or indirectly expresses a wish for one or two. The owner makes the best of a bad job and gives them to his visitor. As Tufele of talanga.
Shark fishing. There is a large species of shark called naiufi that is regarded by fishermen as the king of sharks and treated with ceremonial respect even while planning its capture. If not prepared to noose it on first meeting it, the fisherman makes a speech addressing it as a chief of the highest rank in the terms, "Afio mai lau afionga." He apologizes to it for not being able tautai gets another bigger canoe perhaps and a better crew if the first one is frightened. As these sharks are said to remain in one area for some time, he returns the next night. The baits are put out, the shark attracted to the canoe, and after much ceremonial speech on the part of the tautai the naiufi is noosed with the shark rope. The shark is a vigorous fighter and the canoe may be towed miles first in one direction and then in another. Always, however, according to the tautai it returns to die at the spot where it was noosed. Hence the saying, "'O le i'a e ngase lava i le mea no lavea ai" (The fish will assuredly die by the thing that caught it).
The thing, besides meaning the distinctive object, also carries in the saying the idea of the place where it was noosed. Hence the experienced fisherman will never cut loose no matter how far the fish tows him away, for he is confident that it will tow him back again.
It is a deservedly great honor to kill a naiufi. As the canoe comes in, the shell trumpet is sounded from it and the canoe parades backward and forward before the village. The owner meets his canoe at the landing with a fine mat and touches the head of the fish with it. The mat goes to the tautai and the fish to the canoe owner and the village chiefs, amongst whom it is ceremonially divided. The tautai who has noosed a naiufi is forever established in authority. The event may even have a bearing on the appointment of his son to succeed him after he has retired. In an argument between two aspirants, the decision in favor of one is clinched if it can be said, "His father caught a naiufi."
After snaring two of the ordinary species of shark, the tautai allows his crew to use the noose.
Distribution of fish. The individual fisherman has the right to his own catch, subject, of course, to the inroads that may be made by obeying the laws of hospitality. In bonito fishing with the fleet, the catch is subject to a levy for the communal feast of the fishermen. In community fishing with narcotics, leaf sweeps, and nets, the head fisherman gets first pick and the catch is divided up into heaps corresponding to the number of families engaged. Heaps are usually put out for privileged people, such as high chiefs, visitors, and in these days, the village pastor. Shark and turtle are divided up with special parts assigned to those with hereditary titles.
Hereditary rights and titles. The position of head fisherman while not strictly hereditary was often transmitted from father to son or a close member of the family on account of their better opportunity of acquiring expert knowledge. Information of practical value as to the habits of fish, favorable or unfavorable weather signs and suitable nights, months, and seasons were acquired by long experience and transmitted orally to succeeding generations of the same family. The acquisition of such knowledge gave a member of the tautai family the extra qualification that fitted him for succeeding to the safa'ausu. The conferring of a special title shows the honor with which the position of chief fisher was regarded, but whether the title was inherited or not by a particular family is not clear. The pride taken by a family in holding such positions would, however, lead to reciprocal efforts between father and son to retain it in the family.
Certain privileges in connection with fishing were, however, held by some families. In the tu'i method of fishing at Salailua, tu'i were restricted (sa le i'a) to a certain family. A cord was tied around the tu'i towards the tail end. Of the fish on the entrance side of the cord, the head fisherman got the first pick and the rest were divided up among the villagers.
Some of the privileges, like many of the chiefly titles, are associated with a traditional origin of a mythical nature. Such is the position of le mata-o-le-i'a (the watcher of the fish) held by the family of Toaloa in Puapua, pule) over the fishing arrangements in connection with her mullet and the right has been exercised by the family ever since.
ingana fish to pule) over the ingana is still recognized. If a man wishes to angle for the big fish that have followed the ingana shoals in close to shore, he must send a fine mat to Faasua-i-au, who in turn gives it to Hi and Tangoai, the hereditary guardians of the ingana.
Another example of an inherited right is furnished by the family of Nuu in malauli hook made from a fish bone tied at an angle to a piece of wood. Anyone wishing to fish for malauli with such a hook made his request with an accompanying present to the head of the Nuu family.
The special monopoly exercised by the high chiefs of
Snakes are found in ma'a) in the clearings and dealing with them as in the stone heap method of fishing. The deep ola tu basket used in fishing was also used for snakes but owing to its function, it was called ola fai ngata (snake catching basket). The open basket was held against one side of the stone heap and the stones gradually cleared away from the other thus driving the snakes into the basket.
They were cooked in leaf wrappings in the earth oven and tasted like eels or sea eels according to my informants. They became very fat and the greasy leaf packages after cooking looked as if they contained fat pork. The fat (nga'o) of the snake receives the special name of pipi and the fat condition is termed tai taeao.
The snakes according to the Asau people are non-poisonous as many have been bitten by them without any poisonous results.
Hunting animals is restricted to rats, wild pigs, and the flying fox. In fowling, there is a larger variety of bird life which thus calls forth a greater range of methods and equipment. It is more convenient to deal with the animals first and then with fowling methods.
Both the rat and the wild pig are caught by baited spring traps. In trying to secure the bait, a trigger is moved which releases a running noose by which the animal is secured.
Rat trap (va'a 'iole). The rat is named 'isumu, 'imoa, and 'iole. The trap itself is va'a, and each of the above rat names may be added to va'a as a qualifying word. The name in common use is va'a 'isumu, but in va'a 'iole which is held to be the older term. (See figure 292.)
The baited trap with the spring set is placed in the runways made by rats. If in a good runway, the rat will enter the tube even without the attraction of a bait. On entering the bamboo tube, the rat finds the bait or stick obstructing its way. In getting past or nibbling at the bait, the bait stick is moved to one side. The movement of the stick upsets the balance at its upper end and the horizontal tie stick is released. This allows the spring to fly up. The loop is pulled up from the floor and the rat is caught against the roof of the tube by the tightened loop. For a more elaborately lashed rat trap see Plate XLVIII, A.
Pig trap (mallei pua' a) is still occasionally used in
Flying fox net. In pe'a) was caught in a fowling net with a long handle. The bats lived in a tree down the side of a cliff. In the evening when they went out to feed, they flew up over the cliff and passed through a narrow defile. In this defile, the hunter watched near the cliff. As the flying foxes came up over the edge they were fairly close to the ground. The hunter then intercepted them with the net which he swung (seu) in the manner of netting pigeons (seu lupe). The method was, therefore, termed seu pe'a and is purely local.
A curious seu method prevails in catching flying foxes with a vegetable hook appliance tied to a long handle. The angaoso is a creeping plant provided with curved thorns forming natural hooks which curve towards the root end. A number of small branches are cut off in about 30-inch lengths. About 16 are tied to one end of a long handle with sennit braid. (See Plate XLVIII, B.) The object is to hook the wings of the animal and by tearing them bring the animal to the ground. They are caught thus in three ways:
They are very fond of the flower below the banana bunches of the fa'i pata kind. The hunter conceals himself near likely trees and allows the bat to alight on the fruit before he hooks it.
The hunter takes up his position between trees laden with fruit and intercepts the flying fox as it comes to feed in the evening.
The seu a lunga method' of netting pigeons from a platform in a tree is also observed with the flying fox in
The birds caught were the iao, parrakeet (senga), rail (ve'a), manuo, tern (ngongo), dove (manutangi), and pigeon (lupe). The iao and parrakeet are caught with snares by children for fun and the latter are kepts as pets. The rail was shot at with bow and arrow for sport. The manuo and tern are both eaten. The manuo is caught with a noose and the tern by a net. The trap was used for the dove and both bow and arrow and net for the pigeon. Both moa vao) are also trapped.
Set snare. The usual slip knot snares with single thick threads of coconut fibre are set on the flowers (funga) of the coconut, usually by children. The iao is caught merely for fun and also the Samoan parrakeet. The parrakeet becomes very fat in season and is eaten. Its feathers are of no use for mats, being green with only the slightest touch of red on the head. It is, however, kept as a pet and tamed. A perch is made for the pet and the flesh of young coconuts (niu aleale) is the proper diet. Coconut fibres knotted together to form a long line were tied to the leg and the birds flown (fa'alele senga). The sport was not indulged in particularly by chiefs, but by boys and common people.
A parrakeet is often caught accidentally by the sticky fluid oozing from pu'a berries. The feathers become so sticky at times that the bird cannot fly.
Manipulated snares (sele). A slip knot snare made of strips of alava from the coconut leaf butt and tied to a long bamboo pole was used to catch the young of a seabird (fuao). The fuao breeds amongst the steep cliffs near Sanata in alava strips insure the noose being kept open until it is placed over the heads of the young birds. The cliffs were alluded to as being five coconut trees high. The birds were eaten and the feathers used for pa ala hook hackles.
Traps consist of one self-acting spring trap and two manipulated traps. The spring trap is used for wild fowls. One of the manipulated traps is used also for wild fowls and the other for the dove.
Self-acting fowl trap (mailei moa.) Domesticated fowl (moa) which have gone wild are caught by a mailei trap which is intermediate in form between the rat and pig traps. The material required is a springy sapling, two stakes about 14 inches long with short pieces of branch left at one end to form a hook, a rod as thick as the finger and 18 inches long, three short thinner rods and sennit braid about 6 feet long. (See figure 294.)
An enclosure is made around the trap with an opening on one side of the noose and a bait on the other closed side. The fowl on entering has to cross the trigger to reach the bait. In doing so, it touches the trigger stick which on moving releases the tie stick and the spring flies up, catching the fowl in the noose.
Manipulated fowl trap (fale moa pa'u). A frame built of wood and likened to a house (fale) is propped up on one side by a stick; bait is placed under the house, and a line tied to the prop, the end of which is held by a concealed observer. A decoy rooster may be used also. When the fowl walks under the house to get the bait, the string is pulled and the house falls (pa'u). The trap is also termed fale moa tali, tali meaning to await.
Much superstition exists in
Decoy dove trap. The Samoan dove (manutangi) was caught in a trap like the tu'u'u fish trap on a larger scale and manipulated on the same principle. The trap was made of 'ie'ie roots duly denuded of bark. The single pair twined technique and the general shape was similar to the fish trap. It had to be large enough, however, to hold a decoy bird and the wild bird which was attracted into the trap. The opening was formed as in the fish trap.
The decoy dove (manu fonua) was preferably caught young on the nest and reared as a pet. Adult birds injured in fighting have been caught and tamed (fa'alata). They were fed on cooked talo or breadfruit mashed and squeezed into pellets the size of small cherries. A cage of the type mentioned above was made for them and hung with the opening upwards. A perch (tulanga) of cane (u) was placed across the inside of the cage. A string of two-ply fau bast termed a lauvae was bifurcated at one end and a part tied to each leg while the other end tethered the bird which is thus enabled to fly about and return to the cage. The bifurcated part of the cord was termed manga lua and the single, long part, autasi.
The fowler selected a part of the forest where wild doves were abundant. The cage which also formed the trap was hung up on the branch of a tree within reach of the ground or it was hung to a prepared horizontal perch set on forked uprights. The fowler concealed himself close at hand. The decoy by its calling attracted the wild bird (manu vao) which lighted near the cage. A good decoy attracted the wild bird to it and like the tu'u'u fish it entered the trap to fight the decoy. A well trained decoy would get above the wild bird in the cage and spread out its wings to close the opening by which it entered. The fowler immediately the wild bird was well within the trap called out in a sharp voice, "'Ae, 'ae." The wild bird startled by the voice crouched down and in a flash the fowler reached the trap and placed his hand over the opening. From this method originated the saying: "'Ae 'ae lea manu ua ulu" (Shout at the bird which has entered). It is no use shouting at what is not in the trap. The application is, therefore, be satisfied with what you have got.
A bad decoy sometimes uttered a warning note which kept the wild bird away. This warning note (tangi to'ia) is applied to orators who estrange their hearers. Hence the phrase, "Ua fa'atangi to'ia le launga" (The speech has taken an inimical tone). Some decoys brushed the wild bird away and prevented it from entering the cage. Such a bird is termed manu tafi manu and the term is applied to a chief who estranges people. In another saying, "'O le a ngase manu vao 'ae ola manu fonua" (The wild bird will perish but the tame bird will live on), the wild bird represents evil minded people from other villages and the tame bird, the people of the home village.
Among chiefs the trapping of doves became a favorite sport and competitions termed fa'atau manu tangi were held for the greatest number caught.
The use of the bow and arrow has been mentioned for shooting fish but a use that was more important in the estimation of the people was the shooting of pigeons. This higher appreciation of pigeon shooting found expression in sayings and references in speeches, whereas fish shooting is not drawn upon.
The bow (aufana). Most Samoan bows were made of fisoa wood and the bowstring of five-ply sennit braid. For ornamentation a piece of the braid was sometimes wound spirally around the bow. The bow string is called fu'a which also means the strand of a rope. A groove was cut transversely around part of the ends to give a grip to the bowstring. (See Plate XLVIII, C, 1.)
The arrows. The arrows are termed u or u fanafana from the name of the u cane of which the shafts are made. The shafts seem long for the bow, being 4 feet long in the general type of arrow and 3 feet in a short type made for a specific purpose. The cane is thin, being 0.3 inches in diameter at the tail end and increasing slightly towards the head end. The end that fits against the bowstring is cut off level at a node and is not feathered.
The arrow points are made of a hard wood, such as pau or olioli and are attached singly or in twos or threes. The wood is rounded to about 0.2 inches diameter but from 6 inches to the point it is left square in section. This is tapered off to a four-sided point and small nicks are cut on the four edges. In the single pointed arrow (Pl. XLVIII, C, 2) a long strip is removed from the shaft back to a node and the near end of the point let into' the natural hollow of the cane. Sennit braid is used for the lashing in close turns with the end turned back under three or four loose turns which are then drawn taut in turn and the end pulled to remove the slack.
In the two-pointed arrow (Pl. XLVIII, C, 3) no longitudinal strip is removed from the shaft but a point is laid on either side and then lashed. The three-pointed arrow (Pl. XLVIII, 4) is dealt with similarly by spacing them evenly around the shaft end and lashing. The three points are shorter than the others. The ends diverge so that the actual points are separated from each other. The long arrows receive names from the number of points as, u mata tasi (1), u mata lua (2), and u mata tolu (3).
A short arrow used for shooting birds near a house is fitted with one long point which has a wider, somewhat flattened, serrated part 1.5 inches long at the actual point. (See Plate XLVIII, C, 5.) It was lashed with crossing spirals and receives the special name of u ta'afale. The sport relies on the knowledge that pigeons frequent the pools left in the beds of rocky streams as the streams dry up. Such a frequented pool is chosen and a convenient perch placed across it close to the water. If there are other small pools at; hand, they are covered over so as to force the pigeons to the one hole. A
The fowler's house (fangai) is just big enough for the archer with sufficient room to draw the bowstring back. The smallness of the archer's house is applied as a figure of speech to sauali'i is a title and fangai represents the little island of
Method. A decoy pigeon may be used to attract wild pigeons to the spot but the sport is more often carried on without them.
The long arrows which would be awkward to manipulate inside the house, are left with their points outside and the shaft end just inside the openings of the front wall.
When, therefore, a pigeon lights on the perch, the archer lifts his bow, lays the end of the arrow on the string and draws his bow. The arrow thus comes to the bow without causing the trouble it would in a narrow house if it were inside.
For a clear shot at a bird on the perch, the archer selects the three pointed arrow (u matatolu) or the double-pointed one. Should, however, a bird alight in the branches of a neighboring tree and sit there undecided as to whether it will alight on the perch or not, the archer is tempted to shoot at it without waiting. As it sits in the tree branches, both the double and three-pointed arrows are not desirable as they are more likely to catch on a branch. He, therefore, selects the long arrow with a single point (u mata tasi). Hence the saying,
The moral is to get busy before the other person has had time to make up his mind. To use the long, single-pointed arrow, the bird must be in a favorable position for the use of the bow. Sometimes, however, a bird perches somewhere near at hand where the long arrow cannot be used owing to the cramped position of the archer. It may be above him or to the side. It is then that the short arrow (u ta'afale) is used. It can be used from within the house should the bird be in an unfavorable position. The short arrow is thus likened to a person who can deal with difficulties: "Ua se u ta'afale" (Like the short arrow).
If luck were against the archer, he remained in the fangai all day without getting a shot. The head of the three-pointed arrow remained outside exposed to the sun. In the evening he returned home leaving the arrows and bow which were wet by the dew during the night. It he were again unsuccessful
In other words, with no opportunity of using the bow, he therefore had nothing to offer.
Remarks. The bow and arrow were mainly used for the pigeon (lupe) but, on occasion, for wild fowls, the rail, and the flying fox. The rail (ve'a) has been mentioned in a proverb quoted by
ve'a was not sought after for food as were the pigeons, wild fowls, and flying foxes.
Enough has been said to show that the bow and arrow had their most important use in fowling. Shooting fish was less important. No information was obtained as to whether shooting as a mark or for distance entered into any game. The lack of a saying indicating this would intimate that it was not. Its use as a weapon of war was emphatically denied. Roogevein (27), the Dutch explorer who was the first to see the island of tuinga headdress worn by a chief's taupou were sufficient to create straw hats and a white woman. The bows may therefore, be interpreted into orator's and chief staffs or even spears, while the arrows were imagined.
A net attached to a handle and termed 'upenga seu was used to catch the pigeon (lupe). Decoys were used to draw the wild birds within reach of the concealed fowler who by a dexterous sweep (seu lupe) of the handle caught the bird in the folds of a bag net attached by a frame to the handle.
The following information was mostly obtained from the chief Etau of Aapo, an inland village of seu net but they constantly use the language of fowling in a metaphorical sense with little actual knowledge of the fowling method from which it sprung.
The pigeon of Samoa is still to be heard constantly in the forest as its natural habitat remains undisturbed by the woodfeller clearing a farm in the wilderness. A large assortment of trees provide berries such as the manaui, talie, 'asi, tavai, 'an'auli, anume, puapua, taputo'i, pipi, aoa, aumanongi, manu-lenga, ma'ali, malili, musoi, mati, and mameloa. Where there were plenty of berries, there were plenty of birds. Each crop brought in its own brood as the Samoans say, "'O le fuanga ma le foanga" (The crop and its brood).
Decoy pigeons. Young birds taken from a nest were more easily taught as decoys, but sometimes older birds caught with the net were kept for training. Captured birds were scratched (fafai) over the eyes with the claws of another pigeon. The birds were kept in a cage and also allowed out with a two branched cord attached to the legs (tauvae). A perch termed a tula consisting of an upright with a bend or a branch was provided for them. They were taught to sit on the branch and coo to attract other birds. They were also taught to respond to the pulls of the cord on the legs, to fly straight up or to either side in response to a movement of a stick on which they sat and come down again on the perch in response to a pull on the string. They were pets as well as decoys and chiefs tended them with affection. Their diet was carefully attended to, cooked talo or breadfruit chewed (mama) and rolled into pellets (mama lupe) the size of the usual berries.
A Savaiian story of the origin of the pula'au kind of talo brings out the importance attached to feeding the decoys.
The chief Pulu-seu of Faasaleleanga on going out netting in the morning told his wife
Sina -vai-o-le-malama to chew somemama lupefood for his decoys against his return.Sina had only one cookedtaloin the house and while eating it forgot to leave some pellets for the birds. The omission was serious enough to causeSina to run away from home rather than face the consequences. At Nofoa, she was assailed by labor pains and lay down under a tree by the wayside. Overcome by pain she called out to the chief Mani-laulau, who was passing, for assistance. He duly tended her and she was successfully delivered of—amama lupepellet of chewedtalo. The pellet was buried and grew up into thepula-auvariety oftalo.
The myth mechanism of punishing a sin in a form that directs attention to the particular error is a natural process of thought in Polynesia but my informants failed to tell me what kind of talo it was that led to
The morning feed was given at about 8 A. M. and the time was called fanga i lupe le la. The phrase has something to do with the sun waiting for the pigeons. Children who sleep in in the morning are aroused with the cry, "'Ua fanga i lupe le la."
A chief preferred to have more than one decoy. They were called as in the case of decoy doves, manu fonua in distinction to manu vao, the wild bush pigeon.
Netting platforms (tia seu lupe). A good place was selected in the forest usually on the ridge of a spur where the flown decoys could be seen. A space was then cleared and levelled to form the platform (tia) on which the fowling houses to conceal the fowlers could be erected. Ridges that had an upward slope had to be cut down at one end and the spoil used to build up the other end. Unworked stone was used to build up the sides of the earthwork. One was seen near tia were built in localities much frequented by birds. They received proper names and some become famous in local annals.
Fowling houses (fale seu). The tia remained for all time but the fowling houses were freshly built each season. The houses were merely a shelter of green vines to conceal the fowlers but they received names according to position on the platform. A fully equipped ground tia had four houses set as follows: towards the descending slope end was the fale mua (first house) also termed the fale va-ai (lookout house). At the uphill end was the fale matua (principal house), or simply matua. To the left side looking downhill was the falelele (flying house) and on the right the palalau. Between these houses was the central clear space that allowed a sweep of the net from any of the houses near which the pigeon flew.
The houses in Aopo district were made of the laua vine which is very leafy and thick enough to provide wooden framework and leafy cover in one.
A length of vine was curved in an arch longitudinally towards the center of the platform, with the ends stuck about 4 feet apart and the top of the arch higher than the head of the fowler sitting on a low seat. Another arch was similarly made about 2 fee from the other. A third transverse arch was laid transversely over the middle of the other two and the ends brought vertically down to the ground. One or two more turns might be taken to add to the main framework but the half in front of the middle transverse arch was not covered above. This was to allow the fowler to rise quickly to his feet to make his sweep with the net. The sides and back of the house were also covered with the leaves of a large bush fern called
aulauta. The sticks for building thefale seuwere termedaufale. The house described was 4 feet by 2 feet ground plan. A round house was formed by placing the middle transverse arch under the two longitudinal ones and spreading out its ends before insertion into the ground. The main principle of the fowling house was to give as much concealment as possible from the sides and the back but to have the front part of the roof open so that the fowler couldstand up instantly. Across the front a low crossbar was tied to the lower ends of the arches.
The netting seat (nofoanga). The house is incomplete without a seat made of a solid section of a tree and about one foot high. The Samoan in ordinary life sits on the ground cross-legged and finds no necessity for raised seats. In fowling, however, a seat was used to enable the fowler to assume the erect position as quickly as possible. The seat was used for quick rising and not for the ordinary purpose of rest. Well-made seats of dubbed-out timber, fitted with legs were also made and indicate the importance with which the sport was regarded. A nofoanga, in the
The fowling net ('upenga sen). A. This is one of Etau's nets left him by his father and it has been in actual use. The handle was also old but had been trimmed up and restained with charcoal and water.
The net ('upenga) is long and narrow. The frame ('a'au) consists of two rods of tough pliant asi wood sharpened at the lower ends for insertion into the handle sockets. At the proximal end, the meshes are collected into three groups and the cords carried downwards for lashing to the handle.
The handle (na'a) made of fau is only 5 feet, 2.5 inches long with a diameter of 2 inches at the proximal end. It expands to a width of 4.5 inches at the upper end where a slot 2 inches deep is cut in to form a fork.
The lower end of the net is drawn down and the free cords at this end tied around the cross lashing between the prongs of the handle.
The free ends of the cords at the upper ends of the net are tied to the ends of the rods on either side and one cord is brought across to tie to the other rod end so that the ends are kept at 6 inches apart.
The net as thus set up on the frame is 9 feet 5 inches long, 6 inches wide at the top, 27 inches wide at the widest part and narrows down to 7 inches at the handle. The net has a bag of 30 inches at the upper end, but the middle and lower parts are quite shallow.
The side lines of the net are gathered up with one hand at regular intervals of three meshes. The thin top ends of the rods of the frame are pushed through the gathered third meshes. The other side is similarly treated and the net sides spread down along the rods. The characteristics of the net are the shortness of the handle and the length of the net. The total length of the net and the handle is 15 feet. The method of lashing the pigeon net to the handle is shown in figure 296.
The eye shade (taumata). Before the fowler can take his seat in the fowling house he has to screen his head which is uncovered by the absence of the roof in the front part. An ordinary eye shade of coconut leaflets (taumata) is made. (See figure 124.) Some green leaves of the aulata fern are laid longitudinally over the head and then tied on with a circumferential strip of fau bast. This conceals the hair. The eye shade is then placed over the head and conceals the eyes. Some aulata may be run through the shade as well.
Two forms of netting are used; one from a platform on the ground as described (seu lalo), and the other from a platform made in a tree.
Ground netting (seu lalo) was best in the early morning. The birds of that period were called to'anga o le taeao. Fowlers arriving late on the tia missed the early birds. Hence the saying applied to anyone who is late for meals or the ceremonial speeches: "'Ua a sau 'ua te'a le to'anga o le taeao" (You have arrived when the early morning flock has gone).
The fowler took his decoys with him. The tula bent stick perch was stuck in the ground to one side of the house entrance. A stationary decoy was placed on the perch to attract wild pigeons towards it that had been attracted within sight. A flying decoy with a long cord attached to its legs was taken inside the house. It obeyed signals made by pulling the cord, or moving the perch. The signalling by movement was called tafili and hence the particular decoy was named manu tafili. A special perch was provided for it consisting of a rod about 3 feet long with a hole bored in each end. The near end to the fowler was tied with a short length of sennit braid to an upright of the house. To the far hole was tied the end of the long tauvae string attached to the bird. The perch rested on the ground and the front crossbar of the house.
Having arranged the two decoys, the net was laid on the ground outside with the end towards the center and the handle resting on the crossbar of the house. The fowler duly comouflaged with the eye shade and fern leaves, sat down on the seat ready to pick up the net handle and spring to his feet. This stage is referred to in the saying, Fale seu o lo'o ainga," or "
The fowler then flew his decoy. He did so by holding the near end of the movable perch and by moving it to the right, left or upwards, he indicated by the actual movement itself the direction he wished the decoy to fly. The pigeon flew in the direction indicated by the movement of the perch. A well-trained pigeon would obey the slightest move conveyed in this manner. Wild pigeons which alight out of reach of the net must be attracted by the fluttering decoy, but care was necessary lest any obvious movement of the perch startle it away. Hence, when a speaker wishes to show that some concealed motive of a previous speaker is known to him, he uses the saying, "'Ua 'atangia tanga tafili" (The movement of the decoy is seen).
Sometimes the long cord (fau) became tangled as the decoy flew about. A saying is used to denote an assembly who are not working together but are confused in their plans: "'Ua numi le fau" (The line is tangled).
By flying about on the end of the string, the decoy attracted wild pigeons to the platform. As they flew over the open space they also saw the stationary decoy and came down into its vicinity. The fowler who could judge direction of flight, as a pigeon approached within range, picked up the net handle and rising to his feet, swept the net through the air (seu) to intercept it.
When a pigeon alighted out of distance, the fowler studied it. He noted the direction in which it was facing, for in starting its flight it must go in that direction. Also, a pigeon in starting its flight always swoops slightly downwards before it rises. The fowler studied just whereabouts the line of its first flight would be. He then startled it with a hissing sound and as the pigeon langatila, with the net held straight up like a mast, fa'aifo, a down sweep, and fa'aifo i tualima, a back-handed sweep.
The captured birds were not necessarily killed at once. They were put in an ola basket or in a small covered stone enclosure (fale lupe) near at hand. The birds which were to be eaten had the long feathers of the wings and tail plucked (futi opa) while the ones to be kept for training as decoys were not plucked.
When two were netting in the ordinary way and both had been successful, it was the correct thing for the second fowler to say, "Fa'afetai, mau lupe oe, mau iupe a'u" (Thanks be! You have got a pigeon, I have got a pigeon).
The fowlers rested at a camping place a little away from the platform where they had their food. The place was called a malolonga (resting place). A little distance away from the tia seen on the ridge near malolonga marked by the oven site and cooking stones.
Tree netting (seu-a-lunga) was an individual pastime as only a small platform could be built in the tree from which but only one person had sufficient space to sweep a net.
A good-sized tree, usually tavai or manaui, was selected and a platform of branches also called tia was made in it. The tree in which his father had had his tia was shown to me by Etau. He in his turn would have maintained the platform in the same tree but for the advent of guns.
Ladders (ala 'i le tia). The trees used had the lowest branches high above the ground so a ladder or way (ala) had to be built "to the platform" ('i le tia).
Two poles a little apart were tied in an upright position against the tree trunk. One pole was tied, the rope carried across to the other pole and tied, and then encircled around the tree back to the first pole to which it was tied again. The fowler then worked up the poles, attaching wooden rungs between the poles and every here and there carrying a turn around the tree trunk. Fresh poles were joined on and rungs added until the branches were reached.
The tree platform with the house was called tia seu a lunga or fongatia. A suitable place was selected where natural branches if possible formed cross beams.
Timber was hoisted up with ropes and cross pieces tied in position to form the platform or
pae. On this one fowling house about 2 feet wide was made in the same way as on the ground except that the ends of the arches were lashed to the cross beams instead of being stuck in the ground. This was covered with green leaves. When dry the thatch was said to beua afu le lau fale, and the thatch had to be renewed to match its. Some of the branches were cleared to get a sweep with the net and a particular branch in a good position was selected against which to lean the net. The net was leant in an upward slanting position. The selected branch was called surround- ings surroundings aupale.
Platforms were often in neighboring trees. Though the tree tops were close together, the way between was long as one had to descend the ala ladder, walk across the ground to the other tree and then ascend the ala. Things which appear close are often far apart in reality and the tree platforms were thus used in a saying applying to such conditions:
Method. One flying decoy was used as in ground netting. Two stationary decoys might be used. They were tied to the branches above on the aupale. The flying decoy was sent upwards and flew high above the tree top. It attracted birds towards the tree and as their wings were heard, the decoy was drawn down. As the wild birds flew above the tree, they saw the stationary birds sitting on the aupale branch. To fly towards them, they had to cross the clear space and here they were met by the sweep of the net which came forward from the aupale branch to meet them.
When wild birds were plentiful they kept coming above the tia and the expression applies to abundance of anything: "'Ua malu maunu le fongatia" or, "'Ua lavalava le fongatia" (The tree platform is full of birds).
On the other hand when no birds were about, no sound of wild wings broke the silence. The condition is applied to times of dearth: "'Ua lilingo le fongatia" (The tree platform is still).
Other birds such as fuia alighting on the tree near the platform, were regarded as common (vale) or undesirable and no notice is taken of them. A speaker whose rank or status may not be quite up to the standard of those privileged to speak before an assembly, may excuse himself with the saying: "'O le a sosopo le manu vale i le fongatia" (A common bird is about to alight on the tree platform).
Phases of the moon.
It will be noted that similar terms were used in connection with bonito. The experts of Aopo,
Remarks. The tree method of netting pigeons had the advantage of being conducted at the natural level at which pigeons move about. It was easier to attract them into the tree top than onto the ground. On the other hand, there was less room for sweeping the net and the space admitting of but one fowler, the competition and greater fun of the ground method was absent.
Birds from different forests will come together to a common forest where particular berries are especially plentiful. Their unity is short lived for as the cause which united them is disposed of, they return to their different habitats. The Samoan philosophizing on the transient nature of the unity which has brought people of different districts together compares it to the berry seeking pigeons.
The tern (ngongo) was caught in a net similar to the pigeon net on the edges of the high cliffs along the south coast of seu ngongo.
Decoys. Trained decoy terns were used. The Samata people held that only white ones were used. funga fish in the battle of the birds with the fish. The decoys were given the fluid from drinking coconuts but if there was a shortage of nuts, the white bird was given coconut and the black bird ordinary water.
Method. The fowler sat naked without an eyeshade in a clear space near the cliff edge with his net on the ground before him. The decoy (maunu) had a string about two fathoms long tied to its leg and it flew about to attract the tern. As the tern came up over the cliff edge, the fowler caught it with his net.
When the tern is caught, it cries out "A." In the discussions of men, talk must be indulged in first before a decision is arrived at. Metaphorically, when a decision has been reached, the ngongo is caught and cries out "A." Hence the saying, "'Ua fafunga le A pei o le faiva o seu ngongo" (The A has been arrived at as in the sport of netting tern.)
Food. The Samoan in his quest for flesh food as an ina'i complement to the carbohydrate basis of his diet obtained little assistance from hunting in a country devoid of native wild animals. Rats, which had an appreciable value in the diet of tu'u'u trap with a live decoy might be employed for catching the wild dove. Both the tu'u'u fish and the dove are aggressive towards a live decoy of their own species. The fowler made a larger trap of the same material and by the same technique. He placed a decoy within and hiding close at hand, he placed his hand over the open mouth of the trap after the wild dove had entered. He thus followed in every detail the methods observed within the lagoon. The use of the fisherman's hand net to intercept the leaping mullet in the air was also adapted as a method of catching pigeons and both are included under the term seu.
Sport. Though fowling was primarily carried out through economic motives, certain psychological factors came into play. The skilled fowler was not only an economic asset to his family but by virtue of his success, he was a social asset as well. His skill created a reputation and led to emulation on the part of others. Thus competition took place to secure the reputation of being the best fowler in the village or district. The villages became interested in the success of their particular champions and pigeon netting progressed beyond a purely economic process to become a sport in which the glory of victory over others became the main incentive. The chiefs stepped in and pigeon netting became monopolized by those of high rank. Chiefs could command the organizing of labor to build the earth platforms faced with stone. They had the leisure and time to carefully train decoy birds. They could pay skilled craftsmen to make the best nets. By increasing the size of ground platform, as many as four men could compete together. They could thus compete for the one bird which they all saw approaching. This element
Competitions. Competitions were held between two chiefs or two pairs. The two end fowling houses were occupied in a single competition, and in a foursome the lesser skilled partner occupied the side house on the right of his principal. The competition was for the first bird caught or a number agreed on, usually two. In a competition for the first bird, a competitor might throw (velo) his net at a bird out of reach in the hope of securing it before an opponent got a more favorable sweep at it.
Chiefs often travelled to another island to meet an opponent worthy of his net. Some competitions have been made the subject of song such as the historic meeting between Fao of
Ulumu speaks:
Fao, assisted by a well-trained decoy, won the contest.
Fao speaks:
Ulumu, true to the best traditions of sport, replies:
Social influence. The elevation of netting to a sport monopolized by chiefs is illustrated by two oft quoted sayings attributed to Laauli, a half-brother of lafulafu)"? Laauli replied, "Lafulafu a tama seu ngongo" (The untidiness of a man who has been netting tern). The girls remarked that his hair was sparse and thin (valavala). Laauli again replied, "Valavala a tu manu" (Sparse and thin like the birds). As the sport of netting could be indulged in by high chiefs alone, Laauli indicated by his replies that his dirty and untidy appearance was due to his exercising a privilege pertaining to his exalted rank. The sayings are used to indicate that people should not be judged by outward appearance.
The association of high social status with pigeon netting has reacted on technique in providing the long neat net used and the lashed seat used in the fowling house. The netting seat appears to be the only made seat used by the Samoans and while a section of tree trunk was quite sufficient for ordinary purposes, the neat seat with legs lashed to projecting lugs with a decorative design probably required the incentive of high social status to call it into being.
The competitions in which high chiefs were the participators, led to the full resources of village social organization being called upon. Visitors of distinction had to be suitably entertained by feasting and dancing. malolonga camp. The chiefs carried on the sport while their people prepared the entertainment for them for the intervals between netting. Polynesians are never happier than when they are feasting and dancing. Pigeon netting by providing a cause for such gatherings must therefore rank high as a social institution and its purely economic status occupies a secondary place.
The social influence of the sport is reflected by the large number of sayings derived from it. Both sayings and phrases used in connection with pigeon netting and archery became incorporated in the classical language used by orators. No other phase of Samoan life has contributed so much to enriching the language used by scholars. It is interesting that fowling should have had a similar effect on two languages widely separated for what falconry was to the language of the Court in England in
General features. Samoan horticulture is not very intensive. The people grew enough to supply their own needs with something extra to comply with the levy so often made for the entertainment of travellers and guests and special festivals. The tuberous food plants cultivated were the yam, talo, sweet potato, and arrowroot. The sugar cane, kava, and ti were also cultivated; sugar cane principally for its leaves to furnish the lau thatch for houses, the ti to furnish everyday clothing, and kava not only for personal use as a beverage but to supply the tungase presentations to visiting chiefs. The paper mulberry for clothing and the lau'ie, laufala, and laupaongo (kinds of pandanus) for the various mats were also planted. The banana, breadfruit, and coconut also received some attention. The cultivations were made in clearings in the forest inland of the villages and usually on the uplands above sea level. In
The bush clearings are now readily formed by using the heavy introduced bush knife which performs all sorts of functions from whittling small wedges and cutting up pigs to felling fairly large saplings. It is now the inseparable companion of the man who walks into the forest. Formerly, the bush clearings had to be made with implements of stone. The small scrub and light saplings were ringbarked. The whole was then fired and it mattered not that a large tree remained standing here and there in the clearing.
The implements used in preparing the ground were the digging and the planting sticks.
The digging stick (oso) was of hard heavy wood, between 5 and 6 feet long and about 2 inches in diameter at the thick end which was sharpened. The bark was peeled off and it was devoid of any foot step or ornamentation. It was jabbed into the ground with both hands and the soil loosened by levering up the point. With such a crude implement, only the actual parts where the seed was to be planted were dealt with.
The planting stick termed oso to (to, to plant) was thicker, with a blunt, rounded point. This was thrust down into the loosened ground and levered from side to side to enlarge the hole. Lack of information as to weeding implements implies that special implements had not been devised.
The cultivation was visited from time to time and the spreading creepers or young growth cleared away. Under normal conditions, the people divided their attention between tending their cultivations and fishing in the sea and
The yam (ufi, Dioscorea sp.) formed the first crop in newly prepared cultivations. The plants were grown from seed tubers which were planted close to the trunks of the trees left standing or besides large branches to give the vines support as they spread. The yams are easily damaged by bruising in transport or planting and do not recover from rough handling as the talo does. The yam crop was thus uncertain and often failed. Owing to this greater uncertainty and the greater care needed, it is not now grown so extensively as in former times when people were more careful in utilising all their available material. The easily procurable flour and bread has provided a substitute that has led to less activity in certain directions. The spreading vines of the yam are termed tolo.
One of the numerous
Sina married the king ofFiji . Her brother Pili went to visit her but hid in the bush being evidently doubtful of his brother-in-law. He asked the birds and trees how he could get a meeting with his sister. Thepalaiyam offered his services and sent one of histolovines toSina 's door. AsSina came out, her foot caught in the vine. At that time there was a shortage of food inFiji soSina followed along the vine to find the tuber. There she found her brother. Thepalaiyam is very long.Sina broke off a piece and returned home. Every day she returned to meet her brother and broke off another piece of yam. (A yam broken off is termedmatanau.) As she had followed up (tuli) the broken-off tuber, the tuber was calledtuli matanauHence repeated attempts at meetings or consultations to effect unity is referred to as, "Tuli matanau, le ufi oSina " (Tuli matanau, the yam ofSina ).
Yams are to be seen growing wild in old cultivations. In times of scardly, they are sought after. The season for planting is June, July, and August.
The talo (Colocasia antiquorum) was the staple crop of Samoa. Various kinds are distinguished by names. The following five were enumerated at Tanga,
The semi-wild talo, talo pula'au was given a mythical origin from the chewed pigeon food (mama lupe) described under pigeon netting. It is edible but not fit to place before guests. Hence has arisen an apologetic expression for the poorness of the food placed before visiting chiefs:
When dug up, the tops of the talo were cut off with the stalks in one piece. The outer wider leaves were removed. This formed the seed (tiapula). Following the custom of collecting mats and other articles, people short of planting material visited other villages ceremonially and were supplied.
The tiapula were planted in holes (lua'i) made with the planting stick which was then called oso to tiapula. The hole was always made much larger than the plant. tiapula are planted singly against the side of the hole which is kept patent. The talo cultivations are along the banks of a stream which often overflows and fills the holes with silt. Children clear the holes by scooping out the silt with their hands. The idea of the large hole is to allow the tuber to expand laterally. If the holes are not cleaned out, the talo grows long and thin.
There are two forms of planting; the dry planting in the bush cultivations, and wet planting (loiloi) in swampy lands. The swampy lands are limited in distribution and it may be said that dry planting is the more common form. Fairly extensive wet cultivations were seen at talo swamp was formed between the hills and the raised beach. The water accumulation was natural and not due to artificial irrigation. The area was divided off into plots (fuinu'u) by cut drains and pathways. The drains (alavai) contained stagnant water and had been cut to lower the water surface and not for irrigation which was impossible owing to the one level of the cultivation. The paths had large stones placed in the damper parts as footways.
Mulching was used by not only spreading the cut weeds round the growing talo, but by using coconut leaves and even old mats spread over the ground between the talo. This, by keeping the sun off the ground, not only kept the ground from getting dry but also restricted the growth of weeds.
Light fences of sticks were run along the sides of the drain and helped to keep the coconut leaves at the edges from slipping down into the drain. Scarecrows were also used.
In some rocky parts, the holes had to be made in between the stones and were necessarily shallow. This condition exists at Asau,
The planting season for the talo extends the whole year round. It is usual, when a new cultivation is not contemplated, to cut off the tops as the talo are dug up and plant them in the same place.
Wet planting is confined to natural swamps and seepages and the alavai are merely drains. The advanced system of irrigating terraces by leading the water down through irrigation ditches from the stream at a higher level is not part of the Samoan agricultural system. No built-up terraces were seen and it was stated that they were not known.
In the new cultivations, the talo followed the first or second crop of yams.
The ta'amu species of talo is the large form with most of the bulbous part growing above ground, known as kape or 'ape in other parts of Polynesia. It is grown and used as a food which is not restricted to times of drought. The following names were collected: laufala, lauo'o, usolenga, famai, funga lata, fanga. The last two are not eaten. Two kinds (tonga and niukini) were introduced from
The sweet potato ('umala, Ipomoea batatas) was seen growing at
Arrowroot (masoa, Tacca pinnatifida) is cultivated from the stalks of the plant cut off in lengths after the mature root is dug up for use. In Samoa, according to pia was abandoned because it had another meaning which he characterizes as obscene. The term masoa was substituted to satisfy the sentiments of prudery newly acquired from foreign teachers.
Sugar cane (tolo) was planted principally to provide leaves for thatch but is also grown for eating purposes near the houses in heaped up mounds called tapu'e.
The kava ('ava, Piper methysticum) was planted by preference in rocky places around the backs of houses or back in the cultivations. It was planted from the branches ('ata) and when the roots were dug up, a branch was always planted to provide for the future. Hence the saying, recommending wise provision at the time of attending to present needs:
E sua le 'ava 'ae to le 'ata (Dig up the kava root but plant the branch).
The ti (Cordyline terminalis) was easily grown from small side shoots. They were planted in the neighborhood of the houses and near at hand as the leaves were in constant demand for clothing. It may be owing to its extensive use in furnishing the titi kilts that it was not used so much to wrap
The paper mulberry (u'a, Broussonetia papyrifera) was extensively cultivated from cuttings by the women folk. On more than one occasion I saw a woman with a bundle of short cuttings, which she was going to plant in the bush cultivation on the following morning. Fresh crops were planted from time to time to keep up the supply. Some may be seen growing near the houses but the large stock is back in the bush usually dotted here and there amongst the other plants of the garden.
Species of pandanus (laufala, laupaongo, lau'ie) were planted for mat material, though the wild growing pandanus may be used on occasion. The undergrowth is cut. The old leaves are usually plucked off, and the growing heads may be bound round the base with a strip of leaf.
The banana (fa'i) is planted as food not only for the ripe fruit but also for cooking. Not much care is taken in selecting good plants or taking care of the cultivation. Holes are made and the plant stuck in to take its chance.
The Samoans divide bananas into the soa'a (plantains) and fa'i. There are three kinds of soa'a, distinguished by the fruit: soa'a, smallest and longest; sulasula, intermediate in size; fa'i puta, large and short.
Of the other bananas, a large number of varieties are recognized as the following list from
The breadfruit ('ulu, Artocarpus incisa) grows in the villages and affords shade as well as fruit. It is also grown back in the cultivations. A large number of varieties are named as in the following list by
The maopo has no divisions in the leaves, which are used as plate for foods such as fa'ausi and tafolo.
There are four seasons (fuata). The fuata starts with the flowering of the fruit and it is two months before the fruit is properly mature for eating. The approximate months of each season are: 1. Fuata mafu i ato seu (about March); 2. Fuata ta'oto (May). After the westerly winds, the trees which have been blown down, still bear in the prone position as long as a root connection with the ground is maintained. Hence the name of ta'oto (lying down) is used. 3. Fuata tu fa'amanu (August). Again after the seven winds, the leaves are blown off the trees and the fruit shows up in the stalks like birds (tu fa'amanu). 4. Fuata a le tau (October).
The coconut (niu, Cocos nucifera) seems to be planted anywhere regard less of order, distances apart, or elevation from the sea. Many seem to have been taken up on the hillsides to show that someone has had a cultivation in a seemingly inaccessible spot. Many trees have grown up close together from the fallen fruit being allowed to take care of themselves. Carelessness as to progressive planting exists and the people take up the attitude in many instances that the coconuts will take care of themselves. An increasing reluctance to use the climbing bandage has resulted in steps being cut in the trunks to afford toe holds. The trees deteriorate and when a storm lakes place, the trees snap off at one of the cut steps. Various names are applied to different kinds of coconuts but a list was not collected.
The trees are all privately owned, but no objection is raised to travellers taking drinking nuts so long as they do not abuse the privilege. Special trees set apart for copra or other purposes are marked with a sign termed a tapui to indicate that they are prohibited. Such prohibitions are also made on the trees devoted to assisting the income of the church. When the mature nuts are collected for commercial copra, they are stacked up around stakes to which they are fastened. (See Plate L, B.)
Other plants, such as the maile lau li'i for wreaths and the au'u'u for fish poison, may be planted for use.
Scarecrows (aveau) were seen in the talo cultivation at maile (rough food platters) of coconut leaves placed back to back and inserted vertically over the end of sticks stuck upright in various parts. As the wind blew they rotated a little, giving sufficient movement to startle the manuali'i (Porphyrio samoensis) which frequent the talo patches and destroy newly planted tiapula.
An owner may protect his coconut and breadfruit trees by putting up notices on the trunks. A notice consists of a piece of coconut leaf placed longitudinally against the trunk with the butt upwards and three or four leaflets from either side brought around the tree and tied in a knot. Sometimes the notices are more elaborate and are made with a stake beside the path as a warning to travellers. A mature coconut may be suspended to the trunk of a coconut tree by a strip of the husk torn down from one end. In olden days, some of the tapui were rendered virulent by some ritual which caused sickness or death to those interfering with them.
The origin of the coconut tapui is attributed to Nafanua who delivered
Samoan tradition abounds with myths concerning the origin of various plants. The talo, coconut, and kava according to one, were obtained by Losi from the Tangaloans in the heavens after a struggle in which they were given to him to rid heaven of an unwelcome visitor. The paper mulberry and the pandanus were brought to Samoa by Fulualela, a Fijian chief, as part of his daughter's dowry. Various myths occur in different parts, all bearing witness to the spread of useful plants by human agency.
The comparatively large area of cultivable land in proportion to the population may have been responsible for the absence of intensive cultivation that is evidenced by the irrigation methods of producing talo which exist in some parts of Polynesia. If the ancestors of the Samoans were acquainted with irrigation terraces, some practical reason such as lack of necessity must have led to their being abandoned.
Though the horticultural implements were confined to two forms of pointed sticks, the presence of the planting stick (oso to) is all important and marks the gap that exists between food gathering and food producing peoples. Its importance was fully appreciated in ancient times for the early ancestor Pili, in dividing land and spheres of influence among his sons, gave the western part of
Toys made of coconut leaflets were made in the same way as at pe'ape'a), spinner (moamoa), and bull roarer (Iango mumu). The bull roarer from its humming noise takes the name of the buzzing fly. A primitive Jew's harp of a piece of coconut leaflet midrib held between the teeth and one end played by the finger is seen amongst children. Toy bracelets and leaf canoes are made to play with. Coconut shoes (vae ipu) are made by running a strip of bark through the eye of a half coconut shell or a hole knocked through the shell. A ball somewhat square is also made of coconut leaflet with a covering check plait.
Swings (taupenga) are used, and a skipping rope (tafue).
Children were seen using stilts in
Sliding (fa'ase'enga) without apparatus is indulged in.
Tobogganing. A course was sometimes made on a down slope with pieces of lapalapa or coconut leaf midrib laid across the course and set into the ground. Another piece of midrib from the butt end was used as a toboggan. Sitting astride of the curved part, one end was held in the hands land the primitive toboggan coursed down hill over the transverse pieces on the course. Nothing in the way of dubbing a toboggan out of wood was known. No name in particular was given to the sport beyond fa' ase' enga.
Surf riding. Surf boards made from the sides of old paopao or bonito canoes were used. The length was measured from the finger tips to the breast and cut off. They were used in this position. No elaborate long ones were used, and standing on the board was never attempted as its small size was against such attempts. Surf riding was also known as fa' ase' enga.
Tops (moa) shaped like European whip tops were made. A small coconut was used with the muli end sharpened to a point. The top was spun by winding a strip of fau bast round it and then jerking it sideways. The fau strip was called 'afa in spite of its not being coconut fibre. The side jerk is termed se'i. Sometimes the top is jerked downwards (tongi) to strike another. The spinning of the top is vilivili and the following refrain is chanted as it is spun.
Stone was never used for making playing tops though stones shaped exactly like tops were used to form squid lures. Tops were only spun with the bark strip. No whip was used to keep them going.
Teetotums. A game similar to teetotum spinning is described by Turner (41, p. 128). A coconut was spun in the center of a seated circle of players and whoever the three eyes at one end of the coconut pointed to had to pay a forfeit. The forfeit usually consisted of performing some manual task. It was also used as the method of casting lots designating one to accomplish a particular duty, the group was unwilling to undertake voluntarily. It was also used to point out a thief when one of a party was charged with a crime.
Jackstones (sapo) was played by children, especially girls, with five stones about the size of ordinary marbles. The stones were termed 'ai which is a general name applied to small articles used in games. The principle of the game was to toss a stone and while it was in the air to execute various movements before catching it in its descent. A large number of figures were gone
The game tapalenga (from tapale, to strike) is played in the water between two sides. A short piece of light wood (uto) and two sticks (au tapale) form the equipment.
The water must be shallow enough for the players to stand on the bottom and use the sticks. Two imaginary goal lines (tingi) are agreed on. The players hold a stick in each hand while the uto has to be struck over the goal line. With the left hand stick, the player strikes the water near the uto so as to throw it up in the air on the splash. While the uto is in the air, he strikes it with the right hand, stick in the direction of the goal. The game is simple, but has two oft quoted sayings connected with it: "Le uto 'ua tingi" (The float has reached the goal). On reaching the goal it is returned, hence: "Taliu le uto" (The float comes back).
When visitors pass through a village on their way to another, the saying is used to invite them to call in on their return.
Kites seem to have been unknown to the Samoans. No names, proverbs, or sayings were known to the talking chiefs consulted. When asked questions about kites they said the children flew candlenut leaves to imitate pigeons while chiefs flew tame piegons on a long string.
Flying leaves. The leaf used is the candlenut leaf (lau lama) and the attempt to make it fly like a piegon is fa'alupe. The game is thus fa'alupe lau lama.
The sides of the leaf are pinched off to make them parallel with the leaf midrib, and the stalk is removed. When a wind is blowing, the leaf is held in the palm of the right hand with the upper surface away from the palm and the stalk end towards the little finger. The thumb and forefinger press in slightly to hold the leaf. The arm is held downwards with the palm upwards. The cast is made by bringing the arm sharply upwards against the wind and letting go the leaf as the back of the hand comes uppermost. The leaf starts off with the stalk end inclined upwards and forwards so that the wind catches it like a kite. The leaf may fly fairly straight, but it usually undulates in waves and darts off to either side as it volplanes along.
Owing to its erratic course, the leaf is often likened to a chief of uncertain mind. His talking chief in remonstrating with him will adjure him not to fa'alupe lau lama, or, in other words, not to allow his mind to dart from" side to side like a candlenut leaf flown as a pigeon.
Angry women, giving vocal expression to their wrath, often spread out their arms and turn from side to side to add force to their remarks. This action is known as fa'alupe lau lama. Spectators beyond hearing, seeing the." action, know that the person is so angry that she is "flying a candlenut leaf."
Tame pigeons used as decoys were also flown (fa'alelenga lupe) on a very long cord for recreation. The cord was three-ply sennit braid, but made very thin, so as not to be too heavy for the bird. The coconut husk was not beaten on an anvil but the fibres were separated in layers with a piece of flat pointed wood as in the making of strainers from laufao Beating on an anvil bruises the fibres and renders them weaker. For a thin cord, extra strength was needed. The cord was wound in figure of eight turns on the palm or back of the left hand round the thumb and little finger. The coil was removed, doubled at the crossing, and half hitches made round the doubled end as in the i'o fanga method. The inner end of the cord was pulled out and the cord tied round the pigeon's leg.
A piece of the thin bark cloth known as lauu'a was rolled round the right forefinger. The coil of cord was held in the left hand and as the pigeon flew, the cord ran out of the coil, and over the right forefinger, which was protected by the wrapping from the friction of the running cord.
The cord was made as long as possible. As it neared the end, the cord was given a couple of tugs as a signal for the trained pigeon to return. The return signal was called pao. As the pigeon flew back, the right hand brought back the slack to the left hand in coils.
The old village of
Hornell (16, pp. 71-76) collected six string figures in Samoa during his brief stay at
If an excuse for imparting unverified information is to be made, it is in the fact that the Samoans have not retained many figures, though as Hornell found, the few are widely known. Hornell's natural surmise after his success in vaepato (duck's foot) after the most widely known figure.
Terminology. The terms used follow
Position 1 is with the string looped over the thumb and little finger of each hand with the string passing over the palm between them.
Opening
A, the orthodox Polynesian opening, is made by inserting the index fingers proximally under the palmar strings of the opposite hands and withdrawing them with index loops.The term
navahoing, introduced byHaddon and followed by Hornell,Handy , and others, consists of slipping the proximal of two loops over the tips of the fingers holding them.
Duck's feet (fa'avae pato). This figure was obtained in three ways, one being like Hornell's (16, p. 72) and the other two slightly different to each other.
The figure was obtained at
The last two movements are the same as in the previous except for releasing the thumb.
This results in three loops from either hand crossed in the middle.
An assistant passes the right forefinger above the ulnar crossing, below the lowest crossing, and above the radial crossing.
Continuations from ipu figure: a, laulau—before assistant lets go, she holds the two middle strings running antero-posteriorly which results in a figure called laulau; b, fa'amavaenga—the assistant lets go and the figures run or separate (fa'amavaenga).
By two players.
The girls demonstrating the figure sometimes made mistakes by getting the wrong order. It was interesting to note that they really thought the wrapping and rubbing of one string round the other was necessary to success. When the figure was purposely done without wrapping and rubbing, their surprise was unbounded.
In addition to Hornell's method of the vaepato (16, pp. 71, 72, 75) the fiti figure and the moenga were obtained.
An indoor game (fiti) is played by flicking light rods off a roll of matting with the finger and thumb. The rods are about 18 inches long and range about the thickness of a lead pencil. The wood is mosooi, fu'afu'a, or the discarded aumafuti sticks after peeling off the bark of slender paper mulberry sticks. The number is large, but not fixed, varying with the number of people playing.
An ordinary pandanus leaf floor mat is rolled transversely and tied with sennit braid. The roll is set on end and a rod stuck upright in the folds to form a boundary at either side to prevent the bundle of jackstraws, which are laid on top of the mat end, from falling off. (See Plate XLIX, B.)
The game is played between two sides. All the players of one side follow in succession. Each man flicks off as many sticks as he can. If he misses, does not remove a stick off the end of the mat with a single flick, or removes more than one stick with one flick, he is out and the turn passes to the next. Each stick flicked off counts a point. When all the sticks are flicked off before a side has finished, the whole bundle is put back again on the mat. When all players of one side are out, their total score forms the number that the other side has to surpass.
The game sounds simple but the sticks get crossed on the mat and it becomes extremely difficult to flick one off without removing another.
Another indoor game played with discs of coconut shell pitched towards the end of a long narrow strip of matting is termed lafonga or lafonga tupe. The game is played by adults and particularly chiefs, though the young men of the village usually have a community set of discs with which they play amongst themselves.
The discs are called tupe but in the game individual pieces are referred to as 'ai. The total number called au lafo, consists of two sets of five. Each set ranges in size from 2.5 to 3.5 inches in diameter, in the smaller four to a large one from 4.75 to 5 inches in diameter. (See Plate XLIX, C.) Thick coconut shells (niu mafiafia) are selected for making the discs, the thickness being 0.2 inches or slightly more. In the four smaller discs, the edges are rounded off, but in the large ones, they are cut straight like a kava bowl. The discs are circular and the concavity naturally varies with the size. The large discs are formed from the side of the shell and not the ends. The outer convex surfaces are studded with small holes arranged in patterns and lime filled into them to make the patterns stand out. The large disc invariably becomes cracked across and is joined together by opposite pairs of holes with lashings of fine cord. The joins add to the value of the disc as it indicates age. Craftsmen making sets for sale, deliberately break the large discs and join them with lashings to increase their sale value.
The mat used to provide the marks is about 6.5 inches wide, plaited in check with wide pandanus leaf wefts, and several feet long.
The game is played by opposing pairs who take up their positions at the ends of the spread out mat. One pair pitch alternately up to the other end the mark being the far edge of the mat. The two sets of discs are differently colored, one being light and the other dark. The large disc is named to'e 'ai (remaining piece) because it is reserved for the last to displace the opponent's disc that holds a scoring position. The discs must remain on the mat to score. After the opposing pair at one end have thrown their to'e 'ai, the score is counted and the other pair compete from their end. The score to win was ten according to one informant, but others said it was one hundred.
Much excitement prevails over the game and the loser was always supposed to hide his feelings and force himself to laugh. This results in the saying: "'O le fa'a'ata'ata o lafonga" (The forced cheerfulness of lafonga).
Narratives occur in which stakes as high as life itself were played for a, p. 112) records the solo or recitative song of Liufau in which is described how Liufau and Lua-le-manga had to play for the lives, of their men and themselves against the king
Disc container. The discs were kept in a larger coconut shell cup (fafao) When the game was over the discs were put away in the fafao which ended things. Hence, the saying to denote finality: "'Ua atoa tupe i le fafao" (The discs are all in the container).
A disc (te'a) was thrown for distance in the game of te' anga. The word te'a in the various forms of teka (te'a (ke'a ('ti'a. The disc, however, is thrown in the
The disc was made of coral stone or green breadfruit. The breadfruit, preferably the ma'afala kind, was cut to a transverse slice about 1.5 inches thick and about 4 inches in natural diameter. The disc was trimmed round the periphery to make it a quarter of an inch thinner. The stone disc was made of punga coral rock. Generally a natural waterworn piece was used, but some were trimmed. The disc seen was a waterworn piece, but had been used as a te'a. It is very crude as compared with the grained and polished ulu maika discs of moli u'u) were used instead of discs. The oranges were jerked behind the back without a bark strip.
The throw. The disc was thrown with a strip of fau bark wound closely round the periphery and the end twisted round the right forefinger. The disc was held between the thumb and middle finger while the forefinger embraced the adjacent part of the periphery. The player measured off by eye an appropriate running distance from the mark. Standing at a right incline, he made preparatory swings across the body with both arms, the right with the disc crossing above the left. Then, with a series of steps, he ran sideways down to the mark. He reached it with his left' side towards it. The right arm swung round the body from front to back and delivered the disc with a sharp jerk from behind the back in the direction of the course. As the disc left the hand, an upward lift was given to the bark strip which gave the disc a top spin forward. The disc bounded forward at first in a series of jumps and then ran evenly on its rim. The throwing strip of bark is called tafau. Occasionally the disc was thrown without the strip of bark.
The ground is the usual village road. The part from which the disc was thrown was termed panga and the actual mark from which the disc was delivered was the ulupanga.
The game was played between two sides. All of one side threw before the other commenced. All throws made by one side, which surpassed the longest throw of the other, counted. The calls and methods used were similar to those used in dart throwing. Various songs were sung while waiting and after a good throw. Such a one used as a chorus in reply to the leader's shout of "Mua o" was given me by the aged expert, but without translation.
When a team scored ten points, it had scored an ulu. Arrangements are made beforehand to compete for so many ulu. If a game is for five ulu, the team which first gets three ulu wins the game as the remaining two ulu cannot affect the result.
te'a muli.
The wide-spread Polynesian game of throwing darts for distance is found in Samoa. Here, however, the dart is called ti'a as against teka in te'a (Samoan form of the ward teka) is applied to the throwing disc. The act of casting the dart is tati'a and the game, tanga ti'a. Dart throwing must not be confused with spear throwing (tolonga).
The dart. The ti'a dart was originally of cane (u). The talo patch.
The wood commonly used is fu'afu'a. The dart, after it is peeled, is about as thick as the little finger. It was rubbed smooth between two short pieces of wood termed suni. Each player carries these pieces and rubs his dart between casts. The dart has a specific name according to its length:
The ulu toa was longer than the velo and as its name implies, it had a head (ulu) of toa wood attached to the shaft, thus resembling in principle, the darts of
The ground was an ordinary clear space in the village on the road. The part at either end from which the darts glanced off the ground was termed panga ti'a. A slight rise was preferred and the loose sand scraped away to
The sides of the course were known as tapula'a or afenga. Hence the saying, "'Ua tulia afenga" (The sides are occupied by standing people), which denotes that everything is ready.
The game is played between two sides, the number depending on mutual arrangement. In the game, one side is called auti'a. Played between two villages, the game is a tavasanga. All the players of one side threw a dart each before the other side commences. The side waiting beguiles the time with song such as one commencing with "Aue lulu, aue moa." The team leader throws first and follows up his dart so as to exhort and encourage his team from the other end. He is usually a mediocre player, but a humorist gifted with eloquence and an adept at posturing and grimacing. A judge and markers are appointed.
The throw. The dart is held at the thinner end between the thumb and middle finger with the forefinger over the end. The throw (ta) is made by taking a short run, turning sideways and throwing the head of the dart in a slanting direction to glance off the surface of the ground. An upward slant in the ground thus assists the dart in rising and hence the sloping of the panga. The dart is thrown at the ground in two ways: a, veto, with an overhand sweep; b, tasali, with an underhand motion.
Success depends on the skill with which the striking angle of the head with the ground is made, so as to cause it to ricochet off the surface. The correct action is called fa'amasau, but if the head misses and only the tail touches, it is pa i'u. If the dart misses the ground altogether, it is termed fa'alele and the throw disqualified. If the head strikes the ground at too abrupt an angle (sulu panga) it fails to rise. Where it strikes the ground twice, it is panga lua. When it does not rise, but runs along the ground like a lizard (mo'o), it is termed sulumo'o.
The flight. The perfectly thrown dart strikes the panga a glancing blow, rises gradually with a trajectory like the flight of a golf ball and flies straight and true down the course accompanied by yells of approval. Such a dart is called ti'a ulu tonu as against the ti'a ulu afe, the dart which turns to the side and thus loses distance. Hence the saying: "'O le ti'a ulu tonu lou finangalo" (As the straight flying dart are your thoughts).
A similar thought is expressed by using the term seu from pigeon netting, where the high flying piegeon cannot be caught: "O le ti'a e le seua lou finangalo" (As the dart which cannot be intercepted are your thoughts).
The word ulu (head) used in some of the expressions applies to the head of the dart. The dart that flies too high is ti'a ulu manu and the one that dips too quickly is ti'a ulu tofu. The object is to pass the other throws which is expressed by the verb 'ausia while to pass beyond the furthest dart is 'ausiti'a.
The contest. The longest throws are marked by official markers. All long throws count until they are surpassed. When the second side throws, all the darts not reached by their furthest throw count to the other side. On the other hand, all throws that pass the furthest count to the throwing side.
The leaders exhort their men with shouts of "Maui! Maui!" In a close finish, the leader of the side which got the judge's decision shouted out a saying from the game of tapalenga; "Le uto 'ua tingi" (The float has reached the goal). His side replied with the chorus of "Tingi o." After the leaders of either side throw, they go forward to the other ends where they watch results with the judge and markers. The teams, however, remain at the casting end until all have thrown when they go to the other end to pick up their darts and then cast the next round from that end. At the end of the round the judge gives the number of points scored by the particular side. The leader of that side immediately transmits the results to his team, who were crossing over, by wild capers and the yell of "Mauia!" (meaning a point) His team replied with "Maui," and the leader repeated "Maui" for each point scored in the round, his team replying after each call. In casting, when a dart went ahead of the best recorded, the leader yelled, "Mua" (in front), and his team shouted back, "Mua o." Before casting, a player anxious to excel used the following incantation, whilst he twirled the head of his dart on the ground:
It was a coincidence apparently, that the thrower who recited the above to me, immediately after made the winning throw of the round. Later, how ever, it did not prove effective.
On changing ends, the winning thrower of the last round has first throw. The game goes on until one side reaches the number decided on. No special form of counting such as in
A winning cast about which there is no doubt has given rise to the following saying: "'O le mua e le fuatia" (The winning cast is not measured).
The saying is used to accompany presents in return for hospitality when the receiver depreciates what he has done. The reply infers that his kindness cannot be measured, it is so outstanding.
Protests. When two throws of opposing sides are so close as to render a decision difficult, the judge calls it a draw and asks the teams, not the two
puketa as an exclamation of triumph used in the rame of tanga ti'a, which is derived from puke, an interjection meaning catching you. He states it is the only word with k until the recent corruption of t into k.
Dart thrown with a cord. The dart (ti'a) termed ti'a tafau (ta, to throw; fau, hibiscus bark) was thrown as the name implies with a strip of fau bast. The bast strip was tied in an overhand knot at one end and two methods of applying pressure over the knot to hold the cord on the dart are shown in figure 305.
The use of the throwing strip gave extra length to the throwing movement and thus acts on the same principle as the throwing sticks of the Australian aboriginals. The forward pull against the knot kept the bast turn round the dart in position, but when the dart passed forward, the pressure was removed and the throwing strip automatically released itself without impeding the flight of the dart. The dart so thrown was glanced off the ground in the same way as the darts without the throwing strip. Any of the darts could be thrown with the knotted strip but the short tapu'u dart had to have a coconut leaflet midrib stuck in the end to furnish a tail to make it fly truer.
In another method the throwing strip tied to a handle was wound spirally found the dart after fixing the knot. The dart was laid on the ground and jerked forward with the handle. Both methods were used in
Children's dart game. A forfeit game (tapongo) is played with ti'a darts by boys. The poorest thrower in each round is termed mona. He is struck by each of the other players with their darts. The mona has the privilege of first thrower in the next round. The mona of each round pays the forfeit of being struck. The game continues for eleven rounds. A refrain is sung as the penalty is being enforced, a number accompanying each stroke.
Other numbers follow with their particular accompaniments.
A military exercise or game (tolonga; tolo, to throw a spear) was generally indulged in as practice when war was impending. It was regarded by others as an ominous sign. If news arrived that a particular village was playing tolonga, other villages that had, reason to suspect such activities were at once on the alert in taking defensive measures to provide against attack. The game was played by opposing sides casting spears at a target.
The target was formed from the stem of a young coconut tree, dug up with the rounded part of the root. The root end with the bulbous lower part of the stem was trimmed up as the actual mark and the other end buried in the ground to raise the target about 6 feet from the ground. A transverse ring (fusina) was cut below the bulb. The sides above the ring or boundary was called the fa and the cut off circle of the root which faced upwards was the niu sina or bull's-eye. A spear landing in the bull's-eye was the highest possible achievement in the game and was metaphorically applied to civil life. This originated the saying: "'Ua tau i niu sina 'ae le tail i fusina" (It has landed in the bull's-eye and not on the boundary). If a man marries the daughter of a high ranking chief, he has struck the niusina, not the fusina.
The spears used are of two kinds; au velo fua, and 'ape-too.
The au velo fua are the ordinary spears used which range from 10 to 14 feet in length as suits individual players. They are usually made of olasina wood. One end is thicker and is sharpened.
The 'ape tao ('ape, to pluck out, and tao, spear) is so named from its being made larger and heavier than the others and used to dislodge opponents' spears that are stuck in the bull's-eye. In such cases it is thrown first by the second team and at a slant or crosswise to knock the scoring spears out of the target. When not so needed it may be thrown as an ordinary scoring spear.
The game. The number of players in each team is left to mutual arrangement. The starting mark is about 15 yards from the target. All the players of one side throw before the other team commences. As the bull's-eye faces upwards, the object is to throw the spear with a high trajectory so that it may fall fairly vertically on the mark.
To score, the spear must remain stuck in the mark after both teams have thrown. The spear in the most central position scores as do all those on the niu sina that are nearer the center than any of the opponent's spears. When a member of the first team scores in good position, the following players of that side attempt to protect their scoring spear by sticking spears in front of it and on the side of the vertical portion (fa) facing the throwers. The command given by the leader has also become a saying applied to any preventive measures: "'Ia puni lava le fa" (Fill up the front of the target).
If the front of the target is filled with spears the ends of which stand obliquely upwards, it is extremely difficult for the following team to knock out the scoring spear with the heavy 'ape tao.
The target chip. A curious score is obtained when a spear pierces the edge of the fa and carries away a piece of the target with it. So long as the chip which is termed laufa remains imbedded on the spear point, it scores a point. Hence: "
The chief is the spear, and his status and title, the target chip which clings to him.
On the other hand, when a chip is knocked off without sticking to the spear, it is of no account and is likened to a chief without home or supporters: "Laufa le ainga" (A target chip without relations).
Laufa is a talking chief's title in fono) of thirteen talking chiefs under the
Other sayings. taosala as a spear placed high up in the hilt of the target and tusala as to stand in the wrong place. An old man in a village whose contemporaries have all passed away is referred to as, "'O le tao 'ua tu'ua i le fa" (The spear left in the target).
The side boundaries of the throwing ground are called 'aulape. Before the game starts the 'aulape are defined by the people of the village not taking part in the contest. This preliminary gathering is used metaphorically to denote that all is ready for any project that is being planned: "'Ua ta'oto le 'aulape o le nu'u (The side boundaries of the village have been formed).
The above is a variation of the saying concerning the afenga boundary in the dart casting game.
Club, matches (aingofie) were much in vogue at some of the important gatherings, but have now been entirely abandoned. They consisted of hand to hand fighting and formed an excellent opportunity for the warriors to, display their skill and prowess.
The club used was made of the large lapalapa midrib of the coconut leaf. The butt end was used for striking and the thin, curled side edges were trimmed off. As the butt was liable to split, it was bound transversely with sennit braid. The thinner part formed a natural handle from which the leaflets were removed thus making a club with an expanded head. The club was about 4 feet long and was used with both hands.
The game. nafa). The challenge was delivered by the warrior brandishing (oli) his club with the head upwards or causing it to quiver (vilivili) as he struck light downward blows on his left hand and then drew the club over the palm. The challenge was similarly accepted.
The contestants then paraded between the two opposing forces. The seconds took their position to the left and slightly to the rear of their principal's opponent. fefulitua) facing their own parties who saluted them by raising their clubs. After receiving the salute, they turned and leaped back into position.
Owing to the heavy nature of the clubs, there was none of the light footwork and quickly changing positions that marked
The stance is of the utmost importance, and the object of each fighter as soon as he comes within reach of his opponent is to dig footholds, the left foot forward, and much of the chance of success in the combat rests in the fortune of being able to establish oneself firmly. The skilled fencer will not leave his foothold, if satisfactory at the beginning, until his enemy is disabled or gives ground.
From the above data, there seems to have been little variety in the scope of attack and defence.
The fight went on until one was disabled or lowered his hands (taulalo) in token of defeat. The victor tossed his club in the air and raised the ailao yell whilst his party shouted "tue" in triumph.
Sham fights. a, pp. 55, 56) gives the following data on sham fights at religious festivals:
In the Aana district of
Upolu , they took place at the centralmalaeofLe Ulumoenga at the dedication of Aana to the god Le Fee. InAtua ,Upolu , they took place during the festival of the "Carrying ofAtua toTupua -le-ngase," first at Moamoa in Falefa and then at Falepapa in Lufilufi. A picked company of Atuan men called Tulanga-a-Sasavea were champions of the district and challenged any visitors to single combat at the Moamoa gathering. On the following day at Falepapa, another picked company called Tulanga-a-Sasavai challenged the visitors. The warriors who had contended with each other the day before exchanged their ti leaf kilts in token of good will.
The stand-up fighting from stationary footholds, with heavy clubs capable of a limited number of strokes, is interesting as indicating an early form of military exercise in which endurance and strength were the main factors to success.
Material objects. From the point of view of material culture, Samoan craftsmanship produced little in the objects connected with amusements and games. The people had recourse to the natural material supplied by their environment and expended but little effort in effecting improvements on nature. Children obtained their toys from the leaves of coconuts, their swings and skipping ropes from vines, and small pebbles from the house flooring furnished them with jackstones. The stilt and the surf board which would have required some technique are both doubtful. The coconut was used for tops and teetotums while the wooden top used did not progress far in the way of evolving types of any value. Adolescents and adults did little better. The candlenut leaf furnished the flying pigeon, the discarded aumafuti sticks of the paper mulberry formed jackstraws and the butt end of the coconut leaf midrib provided toboggans and the dummy clubs used in sham fights. Even the important games of dart and disc throwing had recourse to simple sticks and slices of breadfruit or naturally rounded pieces of coral. The pitching discs were the only objects upon which craftsmanship was expended to any extent and they were, as a result, regarded as of some value and carefully preserved. The spears also took some trouble in making. In spite of the simple nature of the material used, they fulfilled the object of giving pleasure and enjoyment to the community.
The social value of the community games was important. The smaller competitions in the same village brought the young people and adults together
Under the heading of musical instruments will be grouped together all instruments or objects from which sound is produced for the purpose of giving time to dances, for giving notices and warnings, and for making sounds for the pleasure conveyed by the sounds themselves.
Ceremonial speech, food, and presents marked festive and important occasions. The natural accompaniment was singing and dancing. Dancing in the form of the siva and the poula contained a number of figures composed of different movements and postures. These were performed by groups representing various villages, or in the local gatherings, various divisions, or family groups in the village itself. Usually after combined figures, individual dancing was indulged in. In all group dancing, the movements were made in unison and faultless time was the criterion of excellence. A method of beating time was a natural accompaniment of dancing. This was done by singing, clapping hands, and beating time on some object. Hence, every dancing party had a small orchestra, to provide the time, not only for the dancing, but incidentally the clapping of hands and the singing. The orchestra remained seated cross-legged on the ground behind the dancers. When individual dancing took place, the rest of the dancers remained seated and added their voices and hand clappings to that of the orchestra. The time beating instruments are very simple.
Rolled floor mats (fala). The ordinary coarse pandanus floor mats are simply rolled to form a hollow cylinder and tied with a strip of bark or braid to prevent it coming unwound. This simple type of drum is then beaten with two light sticks and the rhythm of the sharp taps gives excellent time. To beat the mats is tafua and the phrase for beating the mats to commence the tafua le fala. The leader of the dance will call "Tafua le fala," and the dance commences. The time is perfect and various flourishes are introduced. One or more rolls may be used and the companions of those beating time also keep time with hands and voices.
Rolled mats with bamboo (tui' tu'i). A number of pieces of bamboo varying in length from 4 feet downwards, with one end open and the other closed by a node are wrapped up in a floor mat with the open ends enclosed in the mat. This improved instrument is beaten with two sticks in the same way as the simple mat roll, but the hollow bamboo rods give a different sound. Different sounds may also be produced by beating on different parts of the instrument. Turner (41, p. 125) states the open ends of the bamboo were enclosed in a mat bag but he probably meant the wrapped mat. A mat bag requires a special plaiting technique and no such article was mentioned during my inquiries about plaiting. The Samoan does not usually make a special object if something in common use will serve his purpose. Both the mat roll and the bamboo roll were expedients used for the occasion and neither was a permanent arrangement reserved for beating time only. When the dances were over, the cords were unfastened and the floor mats resumed their ordinary functions.
Bamboo lengths ('ofe). Sometimes the preceding instruments were further assisted by a few of the orchestra using various lengths of bamboo, with one end open and the other closed. The closed end was thumped on the ground in time with the beating on the mat and the various lengths naturally emitted different sounds. The example of the short length in Plate L, A, 3 has the upper open internode split in a number of places to add a rattling sound to the usual booming hollow sound of the whole column.
The smaller wooden gongs were also beaten to give time to the dances. Nowadays discarded biscuit or benzine tins are preferred. It rather detracts, from appearances to see the men and women arrayed in their best, garlanded with wreaths, and glistening with coconut oil going through the various, postures and movements in perfect unison to the time produced from an empty kerosene tin at the back of the serried lines. However, the tin requires no physical effort to hollow out, it costs nothing and gives a better sound than anything evolved by native culture. Incongruity gives way before utility.
The instruments made from the section of a branch or trunk of a tree, hollowed out through a fairly narrow longitudinal opening, which does not quite reach the ends, are wooden gongs. These hollowed dugouts with ends cut off at right angles to the long axis are beaten with one or two sticks and emit a louder sound consonant to the hollowing out. Three well-marked classes are distinguished by size; the pate, the lali, and the longo. All these nafa) is held to be true Samoan.
The small hand gong. The pate is the smallest of the gongs. Two varieties are shown in Plate L A, 1 and 2; one with the ends cut off square, and the other with one end produced to form a handle.
The inside hollowing follows the elliptical section of the wood, the narrow longitudinal opening being along one end of the ellipse. The hollowing stops a little way from the ends, which are also slightly hollowed from the outside. Different notes which the musician utilizes are thus produced by beating over the middle and over the ends.
pate was introduced from pate was brought with it. The pate is now extensively used throughout the islands to give notice as to school hours at both the Government and Missionary schools. In pate resting on the fold of the left arm whilst the right hand beats on it with a single stick or 'auta. The pate has thus become associated with school notices in distinction from the sounds made with other instruments for different purposes. The sound of the small instrument conveys notice to the small people.
The pate is sometimes used to call pigs (vala'au 'ai pud'a) when an owner takes food to the pig enclosure. The enclosures are fairly large and covered with natural growth, but the sound of the pate soon brings the pigs at full speed to the food that it announces.
The pate may be beaten with two sticks at dances. In the true home of the pate, it is essentially the instrument for marking time in the dances.
The medium-sized gong (lali). The lali are made in the same way as the pate with both ends squared. They are made out of larger sections of tree trunks and are thus stationary instruments on land or in canoes. The true lali are used in pairs which have a slightly different note purposely tested during the hollowing out.
The pair figured in Plate L, C, belong to a Savaiian village. Each lali is beaten with two sticks ('auta). They are placed under a shed in a central place in the village and are beaten by two men who blend the different notes of their instruments as they play. They are used to call the village people togther for some meeting connected with village affairs.
The Samoans state that the lali was introduced from lali is fafangu. The beating of the lali to call meetings together is a fa'aali'i.
The large gongs. The longo is a great hypertrophied gong made from the trunk of a large tree. A very large one in longo was made from a talie tree that grew in the district of Vaatia. Thirty men tried in vain to drag it to the sea, but the party, reinforced to 70 men, were successful with great difficulty. Ten trunks of mosooi, each as large as the supporting pillar of a round house, were made into a raft and the talie log placed on it. The raft sank. Two fautasi boats with 12 and 14 thwarts respectively, were sent to float the raft. Divers attached a strong rope to the raft. A piece of stout timber was stretched over the two boats and the rope hauled over this cross piece. After the raft was hauled about half way up to the surface, it was towed, thus submerged, to Fangasa, an operation which took two days. It was left outside the reef, buoyed up. On the third day, it was dragged on to the reef. After landing the log, the longo took 14 days to make.
The longo is beaten with a heavy beater ('auta), which is thrust against the inside of one edge of the opening. The original beater was of olasina wood; the present one is of ala'a.
This huge and unique longo was obtained by negotiation through longo would soon decay through exposure, wished to have its life prolonged in an institution where it could be taken care of. The longo is named 'O le sui fofonga o le Atua (The Voice of God), so called from its being used to summon the people to church.
The Fangasan people maintained that the longo was termed lali originally and after the course of time the name became longo. Some maintain that the idea of the very large wooden gong comes from longo is a development from the smaller lali. The very large form became associated with churches to serve the function of church bells and the name of the type became longo. Whilst very large lali may have been made in the past, it is certain that in Samoa the manufacture was increased by the adherents of the new faith, each village desiring one for their church. The accomplishment of their desire was rendered possible by the use of steel tools. Throughout Samoa, the longo is "the Voice of God" that summons the people to worship. It is an old time voice, strengthened by modern methods, whose sound is used in the interests of a new faith.
The Samoan gong. As to whether there is a modification in structure between the Samoan gong nafa and the lali, it is difficult to say as no examples of the nafa were seen. An old man at Taputimu, nafa, was commissioned to make one, but owing to his ill nafa did not materialize. It was a dugout of wood belonging to the wooden gong class, but was played differently. Two sticks were used and various tunes and rhythms were produced by the expert who showed off his skill by beating the sticks together and tossing them in the air in time to the tune he was playing. Evidently a greater range of play was associated with the nafa than with the lali.
Many of the Tutuilan people claimed the instrument as a true Samoan one in distinction to the introduced lali. They spoke of it as tangafa, in which ta is a prefix meaning to beat and ngafa, an illustration of the modern tendency to mix up the ng and n sounds. nafa as a native drum but ngafa bears no similar meaning.
nafa as a Samoan drum made of a hollowed log and now copied by the longo which he states was derived from fa'a-alii as another name for the nafa, but the word should be fa'aali'i, which simply means to honor as a chief and thus designated the purpose of the instrument and not the instrument itself. In this usage, it bore a similar likeness to the Tongan lali which apparently it much resembled in form except that it was shorter.
From lali and longo and origins from lali and longo is of modern date, but that before the development of the large church longo, they were probably synonymous. nafa is the true Samoan gong and the others were introduced.
The use of the word drum, so commonly applied by
The true drum is a marked feature of marginal Polynesia with the exception of itulasi with shark skin stretched over it was formerly used, at the same time it was said to have been introduced from elsewhere. The lack of definite information concerning it bears out its foreign origin and the fact that it made no headway.
Trumpets (pu) may be made of shell or of wood. In Samoa two types of shell trumpets were used and a doubtful type of wooden instrument. The shell trumpets have the name of their particular shells, but when a hole is bored into them, they become pu from the Samoan word pu, meaning a hole pu, while satisfactory in the Samoan dialect, may not be acceptable to the many other Polynesian dialects in which pu is used to denote trumpets.
Triton shell trumpet (pu faofao). The faofao is the widely used Cynatium tritonis with a hole a little over 0.5 inches in diameter chipped through the third whorl from the end. (See Plate L, A, 4.) No mouthpieces of wood were used, the trumpeter applying his mouth directly to the shell. The sound carries a considerable distance. The trumpets were used on the canoes returning from deep sea fishing to announce not so much their return, as the fact that they had made a good catch. They were also used by travelling parties voyaging by canoe to warn the villages of their coming, and to make a display. War parties also used the shell trumpet.
Cassis shell trumpet (pu foafoa). The foafoa is Cassis comuta and to make the trumpet the apical whorls are cut off. (See Plate L, A, 5.) The foafoa is found more commonly in Samoa than the faofao. According to a head fisherman at Papa, foafoa were fished for on the sandy sea bottom outside the reef. A bait consisting of the cooked underground stem of the ti was weighted with stones and let down to the bottom. This attracted the shell fish which came to feed on the cooked ti. The fisherman returned and if the shell fish had been attracted they could be seen from the surface. He then dived for them. The faofao was also caught in this manner but not so often.
The pu foafoa is used for the same purposes as the Triton shell trumpet. In modern times, the shell trumpet forms the official announcing instrument of the village magistrate (pulenu'u). Any regulation or by-law is promulgated amongst a meeting of chiefs who are called together by a crier sounding the trumpet as he passes through the village and calling the place and time of the meeting. It has become a habit on hearing the sound of the trumpet to listen for the announcement which follows. One morning a week the trumpet may be heard followed by an exhortation to the various families to go forth and bring in their quota of rhinoceros beetle to the village magistrate, which the law demands as a measure for suppressing the pest. Thus the shell trumpet has merged with the elements of the new culture, and seems assured of a prolonged period of activity.
Wooden trumpet (fa'aili niuvao). fa'a-ili-niu-vao as a pipe producing louder sounds than the various smaller pipes or whistles. "It was formerly much used by parties of warriors on their march, or at their general musterings and reviews—aungaau." The niu vao from which it is made is one of the Samoan wild palms.
Instruments which are used for other purposes beside the actual production of the sound itself have been described, but leave a number of instruments in which the sound itself gives pleasure or interest to the producer. Anything that makes a noise gives pleasure to children and the use of a primitive form of Jew's harp and whistles satisfied their needs. Adults attempted to play tunes by means of flutes, pipes, and a sounding board.
Jew's harp (utete au lama). This primitive instrument consisted of the midrib from a dry coconut leaflet (aulama) which, broken off into a convenient length, was held by the left hand against the teeth and vibrated by the right hand so as to make a chattering sound against them.
Whistles (fa'aili) or small trumpets were made by children out of various leaves by winding narrow strips in a spiral form. The leaves of the banana, ti, and pandanus were used, which gave the plant name to the whistle as fa'aili laufa'i, fa'aili lauti, and fa'aili laupaongo.
Bamboo flute (fa'a'i). A flute was made of a piece of bamboo into which four to six holes were bored: In fa'a'i and it was played a good deal by girls. Actual tunes were attempted and as it was evidently used to play love songs between younger people of both sexes, they got more out of the instrument probably than a skilled musician of another culture. fangufangu:
The flute,
o le fangufangu, made of bamboo, was a favourite instrument with the young, and from it they produced a variety of plaintive notes.
Pan's Pipes (fa'aili 'ofe). Pan's pipes were made of five pieces of thin bamboo of varying lengths. These were bound together with a lashing termed fausanga selu (comb lashing), probably the wrapped twine used with combs. It is mentioned by
Sounding board (pulotu). pulotu or fa'a-alii-la-iti "a small instrument used to accompany a solo, was formed by fitting loosely a thin slip of board into a bed of close-grained wood. It was beaten with two small sticks, and although the sounds produced could not have been very pleasing, it was used exclusively by the higher chiefs, some of whom were considered to excel both in this instrument and in that of the Nafa."
Some forms of kilts were specially made for the siva dances, and other activities that took place during festivals and food presentations. At these dances, model clubs were usually carried by the village maid and manaia chief of the parties. Hence the opportunity for display in the dances gave a certain amount of stimulus to the kilt making and model club industries. At the night dances of the poula type, the stiff ceremonial relaxed somewhat and extra clothing was not so much in evidence. Different forms of dancing were indulged in, and in many the various methods of making a noise, thus adding to the efforts of the orchestra, were in vogue.
A piece of coconut leaf midrib was trimmed and the ends split and termed a sasa lapalapa. These were used by the dancers and struck against the thigh and against the sasa of a neighbor in the various evolutions of the dance.
In some dances, each performer carried a stick in either hand with which time was beaten in company with the orchestra. The sake is the dance in which the sticks are used and is said to have been introduced from
Two half coconut shells (ipu) are used as cymbals, really for marking time, in a form of dance termed fiti and perhaps introduced.
ofe) of bamboo which they blew while dancing the siva-a-'ofe.
Samoan methods of producing sound were materialized in instruments which display no great effort at craftsmanship yet they met the needs of the various age groups within the community. Children derived pleasure from the sounds produced by the jew's harp, whistles, and toy bull roarers, all of which were easily made from the leaves of plants. Adolescents expressed the yearnings of their age in the love songs played on simple bamboo flutes, while adults obtained a range of notes from the more carefully made Pan's pipes. Chiefs, as was their wont, maintained the distinction of class by monopolizing the sounding board.
Community needs in the dance were met by the mat bundle and bamboo rods, with which rhythmic sound produced by beating time was the essential factor. The material instruments had no special status and are thus being readily replaced by the empty benzine tin. Special craftsmanship expended its skill on the nafa gong. The gong stimulated emulation in the way of skill and introduced extra movements for the sake of display. The gong attracted the attention of the chiefs and was evidently on the way to join the sounding board as a class monopoly.
Before the advent of the pate, lali, and longo types of gongs, their functions of calling the community together must have been exercised by the nafa. It was a stationary instrument for the carrying of gongs in canoes is attributed tautai head fisherman coming in from the sea with a successful catch of shark, the travelling party grouped around a chief of rank, and the army of attack were all heralded by blasts of the trumpet. The trumpet called into activity the full resources of social organization and the village was galvanized into action that resulted in laughter or in tears.
The absence of the nose flute and the skin drum mark a gap in Samoan culture. Though the bamboo flute and Pan's pipes have now disappeared with their simple scale of Samoan instrumental music, it is interesting to note the survival of various sounding instruments which have been assigned new functions in the changing culture. The pate gong calls the children to school, the pu trumpet announces the edicts of the Government, and the loud boom of the longo calls the faithful to prayer. The instruments of a neolithic age are being exercised on behalf of the education, law, and religion of the new culture and the appeal is expressed in the forms of sound associated with the past which the Samoan people are called upon to forget.
The weapons (au'upengfa) of Samoa include clubs, spears, and slings.
The clubs are all wooden, cut in one piece out of the solid. The wood preferred was pau, but ironwood (too) was also used, and in saitamu was utilized. The clubs were originally cut out with stone adzes, but the early introduction of steel implements has probably resulted in some elaborate forms created for trade and ceremonial.
After being shaped, the clubs were rubbed smooth (olo) with 'ana, a species of nullipore used as a pumice stone. A further polish was imparted by rubbing with a smooth stone or a shell.
The clubs consist of a number of types. The main characteristics and origins of the types may be best understood by following them through the initial stage of construction. After selecting and cutting his timber, the club maker adzes out the timber into the rough shape of the particular type. This must be of the maximum thickness and width of the particular club as edges, spikes, teeth, and all the individual characteristics have to be cut out of the solid for nothing is added or joined on. The first rough stage may be termed the structural foundation of the club.
The structural pattern. For certain types of club, the structural pattern already existed in the growing plant. A suitable sapling formed the foundation of the billet clubs and batons. The sapling dug up by the roots formed the foundation of the rootstock clubs, and particularly thick, expanded roots may have influenced an early form of mace and throwing club. After cutting off suitable lengths, the craftsman went on with the second stage of elaboration in whittling down the grip and dealing with the two ends.
For the clubs with a bigger head or a wider blade than a natural sapling could provide, the maximum sized timber had to be cut and adzed into the rough structural pattern before the second stage of elaboration could be proceeded with. The rough shape had to fulfill particular requirements and follow a certain plan for the various types. Each type of club thus had a particular structural pattern which, while providing scope for the elaboration of the type, also entailed certain limitations as to shape. The craftsman cut out his structural pattern according to plan, and the rough shape was associated in his mind with some natural object. He adzed it to that rough shape and in doing so expressed the process by using the word fa'a (made like or to the shape of). These structural patterns were made to the shape of the pandanus fruit (aufala), the coconut leaf stalk or midrib (lapalapa) the banana leaf (laufa'i), the paddle (foe), and the lobe of the ear (lautalinga). Whatever the idea that gave birth to a type, the first stage of construction followed one or other of the above structural shapes. Both the coconut stalk and fa'aaufala (mace), fa'alaufa'i (bilateral toothed club), and fa'alautalinga (ear shaped club).
The head, or blade. The head of the mace, or of the throwing club is merely a larger section of the material and further treatment deals with cutting out spikes or rounding off the head. In the other clubs which are adzed down or split to provide the width to form a laterally expanded blade, one form of treatment in the rough shaping stage persists through all types of clubs. The coconut stalk, banana leaf, paddle, and ear lobes are bilateral with an equal spread outwards from the middle line. The blade is thickest in the median longitudinal line. With the timber or plank flat on the ground, the craftsman adzes the blade outward from the middle line with a slightly downward slope towards the lateral edges. On this side, there is, therefore, a distinct median longitudinal edge formed by the junction of two surfaces which are on different inclined planes. The plank is turned over, and the process repeated so that the blade is made much thinner at the lateral edges than in the middle line. This is ordinary craftsmanship to make the lateral edges thinner or even sharp so as to inflict an incised wound, and to strengthen the blade by keeping the middle part thick. The important point is that in all clubs the median longitudinal edge is the orthodox technique, and is not afterwards rounded off. If the occasional thick lateral edges which may form a narrow surface are omitted, the blade has four surfaces and is lozenge-shaped in cross section, the short diameter extending between the median longitudinal edges on either side, and the long diameter between the lateral edges. The term median longitudinal edge is used in preference to ridge, as ridge may convey the idea of a specially raised part which rises above the adjoining surfaces. The edge is distinct and characteristic, but it is merely the meeting of two surfaces. It may be exaggerated by making the surfaces slightly concave on either side, but is distinct from a carinated or raised ridge, which occurs in a few clubs as a specialization of the median edge.
From the occurrence of the median longitudinal edges on opposite sides of the clubs as orthodox technique it is seen that the craftsman followed the coconut stalk and paddle forms in general shape alone. The coconut stalk at the butt is transversely concave on its upper surface and convex on its lower. The Samoan paddle is transversely flat or slightly concave on its back and the median longitudinal edge of the front extends only a short distance down from the handle. Both coconut stalk and paddle clubs have a median longitudinal edge on either side extending to the far end of the blade.
Distal end of club. The treatment of the far end of the club is thus readily understood when the structural pattern of the club is known. Clubs derived from the coconut stalk pattern cannot have a functional point, whereas,
The shaft is merely the continuation of the general shape in round billets, but in other clubs it narrows off the lateral expansions and median longitudinal edges to the round section of the grip.
The grip in all clubs is rounded to suit the grip of the hands. It is usually 1.3 inches in diameter, but in some, it may be slightly elliptical. The hypertrophic toothed and hook clubs are also thick in the grip. The Samoan clubs may have some narrow bands of carving round the grip, or a few narrow bands of sennit braid, each band consisting of not more than three turns of braid knotted to form rings. Most grips are perfectly plain. The close braid wrapping characteristic of Fijian clubs is not present. (See figure 306.)
The proximal end of the handle is usually flared. The hook and tooth clubs and occasional odd clubs are of the same thickness at the end as at the grip. Fijian clubs are characterized by absence of flaring, the end being of the same diameter as the grip, but terminated by a distinct flange which forms what
Flanges from both
The absence of the flange in true Samoan clubs is an important diagnostic point.
The lug. After cutting out the characteristics of the club type, the Samoan made a lug at the end of the handle to carry a cord for hanging the weapon up. The lug is formed by cutting down the proximal end square with the long axis, but leaving a mesial piece ranging from 0.2 to 0.4 inches in thickness and projecting up to 0.8 inches. The projection was trimmed in a number of shapes. (See figure 307.)
Even where the grip is perfectly round, the flared end is usually of two different diameters, the greater diameter generally following the transverse spread of the blade or head. The lug usually follows the greater diameter of the flared end. In some clubs owing to a prominent mesial longitudinal edge on the blade, the greater diameter of the flare follows it and becames vertical.
The hole is usually bored straight through the middle of the lug at its base.
The cross rib. In shaping the structural pattern of certain clubs, the craftsman made the material thicker in certain parts. These were afterwards cut down into cross-ribs extending transversely, or obliquely across the blade from the middle line.
Very little detailed information regarding the names and uses of clubs was obtained in the field. Most of the old authentic clubs have disappeared from Samoa into private and museum collections. The study of weapons is thus a museum study which needs compiling and checking from the museums throughout the world. Museum collections again are much confused by containg clubs made for sale and clubs derived from private collections in which the true locality of the clubs is extremely doubtful. Clubs from
The povai is a billet of heavy wood which resembles an American baseball bat in appearance. There are two varieties, the round and the four-sided in cross section.
Round billet. In round billet clubs the distal ends are domed, some coming to a blunt point, and the greatest diameter is at the part where the head turns off to form the domed end (Pl. LI, 1 and 2). These clubs average in diameter about 2.7 inches at their head, diminishing to about 1.6 inches at the grip. The, proximal ends are flared and furnished with lugs. Bands of carving encircle the clubs and are inlaid with coral lime. The club shown in Plate LI, 1 was used by King
While the heads of most billet clubs are domed,
The four-sided billet has two sides narrower than the other two. (See Plate LI, 3). The widest part is distal where the curved end commences. The four surfaces of the head gradually narrow in a convex curve to form a blunt point. The club surfaces narrow gradually towards the shaft which is rounded and runs into the narrower grip. The proximal end is flared and provided with a lug. Bands of carving inlaid with lime form ornamentation.
The rootstock club, so named by
The mace clubs (Pl. LII, A, 10, 11) are related in form to the rootstock clubs but in making it the craftsman carved the solid wood to resemble the pandanus fruit with the spikes or knobs resembling the pandanus keys. Hence, structurally and in appearance, it became fa'aaufala (fa'a, made like; aufala, pandanus fruit).
A, 10, 11 are in
The mace clubs include long two-handed and short one-handed varieties. Few of them exceed 30 inches in length, but
The
The
The coconut-stalk name used by
The structural pattern was adzed out as already described with an expanded bilateral blade diminishing in width towards the proximal end where it ran into a rounded shaft. The technique of adzing from the middle line towards the side edges left median longitudinal edges, yet even here the influence of the coconut midrib pattern is apparent. The coconut midrib in the part used as a weapon has a marked convexity on one side while the other is concave merging into a flat surface. In cross section, the well made club in Plate LI, 7 does not form an even lozenge, but one surface is much flatter than the other.
The distal end is cut across to comply with the structural pattern. In most clubs, the end follows the pattern literally by being concave from side to b.) The concave or blunt end of the coconut clubs is a distinct feature. The slight projection that sometimes occurs is never carried out to form a curved point such as occurs in some Tongan clubs with a tendency to a coconut stalk blade.
Ribs. In thick clubs, the lateral edges are correspondingly blunter, and in some, no attempt has been made to form a sharp edge. Such clubs are not so liable to nick at the edges. In clubs with thinner edges, the blade is strengthened by having raised ribs running across it at intervals. The ribs form wide bands raised above the general surface of the club and projecting at the side edges. A rib at the distal end is constant and widest, while the others diminish in width towards the shaft. The ribs may be left as plain raised bands (fig. 308, c) but they are more usually serrated by cutting parallel V-shaped grooves over their surfaces in line with the long axis of the ribs. The ribs are thus converted into a series of sharp-edged ridges crossing the blade and projecting beyond the lateral edges as sharp points corresponding to the number of ridges. The ribs not only strengthen the blade, but the projecting points improve the efficiency of the club as a weapon. The direction of the ribs conforms to the type of the distal end, which naturally influences the distal rib which bounds it. The other ribs take their direction from the distal one. With a straight distal end, the ribs are straight from side edge to side edge. With a concave or angled distal end encroaching inwards on the middle line, the distal rib has to follow suit and the grooving results in a series of chevrons with the apices directed proximally towards the handle. This is the commonest form. With the distal end projecting outwards in the middle line, the ribs also follow suit and chevrons are formed with their apices distal.
Subtypes. The treatment of the blade with, or without ribs, and the type of ribs results in four subtypes of coconut leaf club: a, without ribs; b, with straight ribs; c, with oblique ribs with the chevron apices proximal; d, with oblique ribs with the chevron apices distal. The types of distal ends and ribs are shown in figure 308.
In all subtypes, the flared handle and the perforated lug are constant features. When the lug is broken, an oblique hole may be bored through under the flared edge of the proximal end.
Variants.
The shaping of raised ribs by cutting away the wood between them is technically termed tongi, or tongitongi. lapalapa which is the name of the coconut leaf midrib forming the structural pattern. In uatongi from the technique of carving the ribs. The term lapalapa applies to the plain subtype representing the original coconut stalk and uatongi to the later developed subtypes with ribs.
The eight-spiked club (taiavalu), with four spikes on either side, must be clearly distinguished from the bilateral-toothed type with more than eight teeth on one side. No examples of the taiavalu occur in b), and f). The club is thus established in literature, but has been confused with the much more common bilateral-toothed type to which the name of talavalu has been wrongly applied.
The distal end of the eight-spiked club is wide and concave and has been aptly termed crutch-shaped by
The spikes. Four spikes, separated by distinct spaces, project from the blade on either side making eight in all, and thus giving the club the name of talavalu (eight spikes). In c) the spikes range from 2.5 to 4 inches in width and the spaces between them range from 0.5 to 1 inch. The spikes were bevelled from their middle longitudinal line to sharpen their projecting side edges. The median longitudinal edge thus formed with each spike is carried on to meet the median longitudinal edges of the blade at right angles. The bevelling commenced on the spikes is carried inwards in decreasing depth until it disappears at the median longitudinal edge of the blade.
In the clubs figured by talavalu club was made from the coconut stalk structural pattern. c) to the coconut stalk pattern in the absence of a pointed head and proximal shoulder. The presence of the "crutch finish," he recognized as "characteristic of the coconut stalk clubs and in that connection is explicable structurally; the objection to such an ascription rests on the absence of serrate edges from that type; yet it might prove possible to discover intermediates which could connect the sparse teeth of this piece with an overdevelopment of the bands in such a club as 3 J."
The club 3 J referred to is a coconut stalk club with ribs. My Samoan informants in describing the ribbed coconut club (uatongi) stressed the fact that the ribs were serrated and projected beyond the edges of the blade. The natural sequence would appear to be that the plain ribs used to strengthen coconut stalk clubs became grooved and thus serrated on their lateral projections. The use of such serrations was carried a stage further by making tala spikes on a coconut stalk pattern with the spikes spaced in a manner similar
Name. talavalu as eight-spiked (tala, spike; valu, eight). He was undoubtedly referring to the true eight-spiked club described. talavalu as applying to the many-toothed club with a pointed end, questioned valu as meaning "to scrape" or "to rasp" instead of '"eight." He considered that talavalu to explain an aberrant eight-spiked form, whereas it was
The bilateral-toothed club (fa'alaufa'i) has been confused in name with the talavalu. The distinction, both as regards name and characteristics, are marked. (See Plate LI, 4 and 9.) On making inquiries in talavalu club, based on talavalu but not a talavalu, called a fa'alaufa'i." On drawing a bilateral many-toothed club, it was pronounced to be a fa'alaufa'i and not a talavalu. A clear distinction exists as to three types of lateral projecting points; the uatongi, tala, and nifo. (See figure 310.)
The teeth. From the Samoan distinctions given, the close points of the fa'alaufa'i will be referred to as teeth (nifo) and the spaced single points tala) confined to the talavalu. The characteristics of the bilateral-toothed club are a pointed distal end and a proximal shoulder with the intervening part of the blade on either side occupied by teeth set close together. Owing to the length of the blade being fully occupied by teeth, they number more than eight on either side. The fact that the close setting of the points constitutes them nifo (teeth) and not tala (spikes), together with their number being more than valu (eight), renders the club name of tala valu inapplicable to this type of club.
Structural pattern. The new features of a distal point and a proximal shoulder, so distinct from coconut stalk and eight-spiked clubs, is due to another structural pattern having been used in the manufacture of the type. The club name of fa'alaufa'i (fa'a, like; lau, leaf; fa'i, banana) indicates clearly that the structural pattern was likened to the banana leaf and figure 311 shows plainly how such a pattern accounts naturally for the special features of distal point and proximal shoulder.
Owing to the median longitudinal edges on either side of the blade, the point is pyramid-shaped, or forms what
The drawing by talavalu is retained. (See figure 311, d.)
Structural pattern. Theoretically speaking, Samoan paddle clubs should resemble Samoan canoe paddles. It has been seen, however, that the structural pattern of clubs, while following the general outline shape of the pattern motive, does not conform to it in cross section for technical reasons. Thus Samoan paddle clubs have a median longitudinal edge on either side that extends for the entire length of the blade. It has also been shown that the craftsman purposely leaves thick parts on the structural pattern to convert into cross-ribs if it suits him. The commonest form of paddle clubs are thus; characterized by cross ribs which have arisen out of structural technique independent of the pattern motive.
In the coconut stalk and banana leaf structural patterns, the general outline conformed fairly closely to that of the original pattern motive. In the paddle structural pattern a departure takes place from the outline shape of the canoe paddle motive. In the canoe paddle, the greatest width of the blade is nearer the shaft than the point and the canoe paddle shape is ovate.
In the commonest types of Samoan paddle club, the greatest width of the actual blade is nearer the distal than the shaft end. The club blade is, therefore not ovate like its original motive, but is obovate. Where the blade is too narrow to be termed obovate, the blade is oblanceolate instead of being
The paddle motive. The question is, did the Samoan craftsman recognize the paddle as a motive for a structural pattern, or are we merely attributing to him what we think? At amuamu) made of ifilele wood was in great favor at Fangaloa in
The significance of the chant is in the third line.
The chant was that of the amuamu club, but the objects that the chanter was rubbing down (olo) were in the structural pattern form of a paddle (foe). Unfortunately the subject could not be followed up at Fangaloa itself, but the chant shows that club makers recognized the paddle as a structural pattern for clubs.
Types. There being no full sized clubs available, a selection of various forms attributed to Samoa are reproduced (fig. 312) from
carinated and serrated clubs as two types distinct from paddle clubs. Until more study material is available, it is proposed to group all clubs derived from the paddle structural pattern under the heading of paddle clubs.
Such clubs may be divided into four types: 1, without cross ribs; 2. with short cross ribs; 3. with medium length cross ribs; 4. with wide cross ribs.
Paddle clubs without cross ribs. This is a theoretical group as regards Samoa based on the idea that the structural pattern might be retained without the addition of a crossbar.
Paddle clubs with short cross ribs. The cross rib is situated at the junction of the blade and the shaft. The rib merely defines the commencement of the blade where it diverges outwards from the shaft. The rib is so short that it merely form a rim (fig. 312, c) and there is thus no necessity for a proximal shoulder. In the type (fig. 312, b) which is doubtful, the rib is widened longitudinally. Again, while a, b, c, and d) clubs which may be regarded as belonging to it, but two (a and b) are probably from
Paddle clubs with medium cross ribs. The effect of lengthening the cross bar immediately affects the shape of the blade. While in the structural pattern, the thickening for the rib must be left across some part of the blade, the effect of the rib in the subsequent shaping is to limit the weapon blade to the part beyond it whilst the part on the proximal side becomes a shoulder connecting the cross rib with the shaft of the club. The width of the cross rib influences the width of the blade and its consequent shape into two main varieties.
Narrower leaf-shaped blade (Plate LII, A, 1). The medium width across rib is shown in figure 312 e and f. From the cross rib which defines the blade, the blade slopes inward and then gradually outward to form the "leaf shaped" blade so characteristic of the ancient Greek swords. The shape is oblanceolate, but in some, the greatest width is near the middle of the blade and gives it a somewhat elliptical form. figure 312 f is carinated.
With wider straight edged blade (Plate LII, A, 2). The wider cross rib and blade is shown in figure 312 g after h) after e) is carinated, and the straight edges spring at once from the
In some clubs, the surface on either side of the middle line is cut down before being slanted outwards. This converts the median longitudinal edge into a carinated ridge. The carinated ridge is confined to the blade and is not continued on the proximal side of the cross rib. f and h) are merely an accentuation of one feature and are best regarded as a variety of type 3, as "carinated" cannot well be applied to the clubs of the same type, but with no carinated ridge.
a) which shows a leaf form blade springing from a wide shoulder a little internal to the outer margins. He (10, vol. 1, p. 70, No. 2) also shows another club b) in which the point has evidently been broken off and the club trimmed with two points.
Paddle clubs with very wide cross ribs. This division includes the type termed serrated clubs by g) figures one in the b) he relegates to
The range of measurements of
The extraordinary lengthening of the cross rib is seen. The widest cross rib in Type 3 was 6.5 inches and the shortest in Type 4 is 25 inches, practically four times as long. The club in the e. In the absence of authentic examples from Samoa, the very appearance of the club shows it to be outside the range of thought of the Samoan club maker. Two of the Fijian clubs are shown in figure 313 c, and d.
In summing up, it may be said that the Samoans specialized on the paddle club of Type 3. It evidently became a favorite club type with the warrior and hence with the club maker. Owing to their specialization, Type 1, if ever used as a weapon, disappeared, Type 2 was being displaced, and Type 4 was never evolved locally or adopted from
Structural pattern. The clubs shown in Plate LII, A, 4-6 are termed fa'alautalinga (fa'a, shaped like; lautalinga, a toadstool, or the lobe of the ear). lautalinga and hence termed the type mushroom clubs. Looking at the matter from the point of view of the craftsman's structural pattern and the object to which he compared it, the toadstool is much less likely to form the name motive than the lobe of the ear. The toadstool I never saw or heard of during my sojourn in Samoa, but the ear is always present and frequently referred to. The craftsman in adzing out his structural pattern shaped it like two ears placed together, as the club was bilateral, and with the lobes forming the curve to the external angles.
Three clubs, selected from the eight weapons in A, 4-6) show the varieties. (See fig. 314.) The length of the head is from the level of the external angles or lobes to the distal end. The proximal thickness is also taken from this point.
From the figure, the shorter length is apparent for the clubs are full sized weapons. The head increases in thickness to the distal end, which in one club is 2.4 inches. The surfaces of the club are sloped outwards from the middle line so as to form a marked longitudinal edge which becomes more a, Aberrant form of ear-shaped club with typical angles (1) which slope in to an expanded blade (2) ending in shoulder (4); the proximal end is flared and has a suspensory lug (3) length, 46 inches. The club is in the A, 4). A, 5). A, 6.) The clubs (Table of Measurements (in Inches)bcd
a). This aberrant form shows an ear-shaped head with the lateral curves prolonged proximally to form a blade somewhat like a coconut stalk club. The ear-shaped club is a specialized short club which is more likely to have sprung from a long type of club than to have commenced short. The origin of the type from a long club seems further supported by the fact that it is made from a split plank and not from a round sapling or bough like the knobbed clubs and the maces.
All the clubs are carved on the head and the lateral expansions from the shaft.
The nifo'oti club is marked by a row of teeth on one side which are of the bevelled type seen in the bilateral-toothed clubs. The close set bevelled teeth of the nifo'oti club give it its name (nifo, teeth; 'oti, to cut). The toothed side of the nifo'oti club has little to distinguish it from the side of a fa'alaufa'i club, except that the teeth are longer and, owing to the nifo'oti club being much thicker in the median longitudinal line, there is a concave slope from the middle line to the points of the teeth. In the type figured (Pl. LI, 8) there is a slight space between the individual teeth. The bevel, however, of the sides of the teeth, and from the middle longitudinal edge of the blade are the same as described. Owing to the greater thickness and length of the teeth, they are more spiked in appearance, whereas the fa'alaufa'i teeth are flatter and more saw-like. As in the fa'alaufa'i, the toothed side has a proximal shoulder of the same nature. Beyond the distal tooth, the blade, after sharing in the tooth bevelling on the proximal side, is brought back square to the middle line. It is simply a tooth sharpened on the proximal side only.
The other side of the club is laterally expanded as if to provide another set of teeth which, however, are never made in the orthodox club. The distal end is further expanded as if forming one side of the head of an ear-shaped
The 5 nifo'oti clubs in
The nifo'oti is thus a hybrid compounded from other clubs. The toothed side may be regarded as a fa'alaufa'i club with the distal point cut off, or a coconut stalk club with the side cut into teeth. The untoothed side is a coconut stalk club with the distal end adopted from the ear-shaped clubs and the external angle exaggerated into a hook. Three well-shaped hybrid clubs; one figured by a); and one in nifo'oti has been derived is more probably the coconut stalk than the banana leaf, and that the proximal shoulders have been due to the wider expansion of the blade to provide bevelled teeth. (See figure 315.)
A more modern development is the steel bladed weapon with a marked hook. (See figure 315 d.) This weapon formed the favorite club in the modern Samoan wars and also received the name of nifo'oti.
In discussing the significance of the term nifo'oti, nifo-oti' (nifo, teeth; oti, dead), but closer examination resolved the name into nifo'oti (nifo, horn; 'oti, goat), from the resemblance of the hook to a goat's horn. nifo-oti being translated as "tooth of the dead," for, from a linguistic standpoint it would mean dead tooth. The "goat's horn" meaning, he also rejected as being too modern and implying that the weapon had no name before the introduction of goats. He takes 'oti as meaning "to cut," and nifo'oti as being applied to the teeth (nifo) with which the head was cut off ('oti). The function of the hook as given by
The hybrid nature of the club indicates that the nifo'oti club is a later development than the three established clubs from which it derived elements, of form. The hypertrophied form without distinct proximal flaring and suspensory lugs together with its clumsy nature and lack of balance, all point to the ceremonial usage which accompanies the display of peace rather than the utility of war. The ear-shaped clubs and the composite ear-shaped and coconut stalk clubs (fig. 315, a and b) could both be used in warfare with good results. The exaggerated wooden nifo'oti club would be a misfortune to any warrior who had to use it in actual combat. Until well-balanced wooden nifo'oti with distinct hooks that were used in war before the advent of the blubber knife, can be described, the club must be regarded as a fairly modern development made for show purposes. The natural sequence would appear to be that hybrid clubs of the type depicted in figure 315, a and b, were toothed on one side and were termed nifo'oti from the fact that the teeth could inflict a cut. The cutting off of heads was beyond the scope of such a club and there was no incentive to produce the angle of the ear-shaped end into a hook. The advent of the blubber knife, however, provided not only an edge to sever the head, but a ready-made hook to carry it away. The hook carrying idea having been supplied, the angle of the ear-shaped end of the nifo'oti was shaped to form a hook. Through the hook motive, the wooden club then reacted on the steel weapon and provided it with the name of nifo'oti, although it had no cutting teeth (nifo'oti). The blubber knife became the functional nifo'oti used in war and the wooden nifo'oti underwent the hypertrophy associated with the ceremonial observances of peace.
The club with the long lozenge-shaped blade (Pl. LI, 5) and five bilateral pairs of spikes at the junction of the blade with the shaft was collected by myself in Samoa. Others were seen and are used in dances. The name of anave was given but the type seems new and as if introduced.
A club with a longer narrower blade, without spikes, but with two narrow transverse raised bands with serrated edges is catalogued as Samoan in
Maces. The short mace (Pl. LII, 11) has already been described on page 589.
Batons or short billets were used as single-handed weapons. The one in Plate LII, A, 7 is old and well polished. The specimen in Plate LII, A, 8 is thicker and has had sennit braid wrapped in loose spirals round the grip. It has a piece projecting from the proximal end that was too narrow to pierce with the ordinary sized hole. Both batons have a large hole bored transversely through the handle for the supporting cord.
Knobbed throwing club ('olo). The knobbed throwing club was described by q) figures one from Samoa and the club in A, 9) is also from Samoa. The Samoans call the club 'olo and it was evidently well known. The club was thrown with the handle in the hand but as it went forward, the ends reversed and the object was struck by the end of the handle. The club figured has four bands of carving round the handle and shaft. The handle increases in diameter towards the proximal end without marked flaring. The knobbed head is smooth and the distal end projects as if the handle had been inserted through the knob and projected beyond it for 0.2 inches. The proximal end of the handle is domed.
Spears (tao) were made of pau wood and also of mature coconut wood. They were scraped, polished, and oiled. The long spears were usually from 9 to 10 feet long. Shorter spears of 6 to 7 feet were also used as staves which could be converted into weapons of offense or defense. The general name for spear is tao. A chief's spear or staff was called uai. A short spear used as a staff could be referred to as to'oto'o.
The spears were made in one piece including barbs. Barbs are termed tala. One end only was pointed, the other end being blunt and without any differentiation into grip or counterpoise.
Barbless spears. Some spears were, merely scraped to a point and hence termed tao valu fua (valu, scraped; fua, alone). Here fua means without barbs. ato as a stick sharpened at both ends and used in ancient times as a spear. This is the exception to the usual one pointed spears.
Barbed spears. The barbs noted may be divided into simple and compound barbs.
The simple barbs are made by cutting nicks in the head of the spear as in figure 316, a. In this form, the head is shaped to form a projecting longitudinal edge in which nicks are cut with a forward obliquity. The type of barb is present in the spear in Pl. LII, B, 3, which conies from Malaeloa, fa'atala lau paongo (made like the spikes of paongo pandanus leaf). The longitudinal sharp edge on which the barbs are cut is prolonged forward towards the point. The spear was used as an orator's staff and hence called to 'oto'o fa'atala lau paongo. The barbs are on one side only of the head.
Another type of simple barb is made projecting out from the head and usually with a curve backward. This type restricted to one side of the head is shown in figure 316 b.
The simple barbs may be arranged in transverse rows around the head. The spear in Plate LII, B, 1 has long backward curved barbs arranged in three longitudinal rows and in transverse rows of threes. To get this formation the head is cut out of the solid in triangular form in section. The points of the longitudinal rows thus coincide with the longitudinal edges and the intervening material is cut away. Towards the point of the spear in Plate LII, B, 2 the barbs are arranged in four longitudinal rows and in transverse rows of four. Here the spear head is cut out in quadrangular shape and the angles correspond with the points of the barbs. Where the diameter decreases towards the point of the spear, the barbs are short with the usual backward obliquity. If made further back on the head, the barbs become Monger and curved. The point beyond the last barbs usually maintains the triangular or quadrangular character of the head until it rounds off into the actual terminal point.
While, the barbs are sloped backward to render penetration easier and extraction more difficult some of the furthest back barbs, as in the proximal end of the head in Pl. LII, B, 2 may be directed forward.
The compound barbs in Samoan spears consist of three points with their bases together and all cut out of the solid. The middle element of the three tala-o-le-lo (spine of the lo fish) are generally arranged in three or four longitudinal rows and also in transverse rows. Again the spear head is cut triangular or quadrangular and the points are formed on the angles by cutting out the intervening material. Sometimes the simple and compound barbs are intermixed. Even the spears in which the compound barb is the main theme are furnished with simpler barbs towards the point as the thickness of the head diminishes. The compound barbed point gives its name of tala-o-le-lo to the spear itself and such spears are the most prized. They were made out of both pau and coconut wood. (See Plate LII, B, 2.)
The elements of the compound barb are triangular in shape with the bases attached to the head end and the apices forming the free points. (See figure 316, c.) The fore and aft elements have median longitudinal edges which made the barbs lozenge-shaped in cross section.
Slings (ma'ata) were extensively used but no information was obtained as to technique except that they were made of the bast of fau or fau songa. The sling stones were not worked but were selected from waterworn pebbles (ala) of appropriate size. Some stones (ala papa) were said to burst on striking. At Fitiuta,
Forts and refuges ('olo). Many villages were protected by stone walls termed 'olo. Some places of refuge also termed 'olo were situated back in the hills but owed their protective qualities to natural inaccessibility to attack. No information was obtained concerning special defences erected by the refugees. In 'olo. The people took refuge there after defeat or to defend themselves against an overpowering force. Stones were piled up at the top of the paths and dislodged upon a pursuing enemy. In the back country of
Pointed stakes (su'i). Some villages, such as su'i pierced the bare feet of infantry warriors. They retarded the landing of hostile forces and impeded their mobilization for attack against defensive walls.
Armour. No field information was obtained concerning armour. If used at all, it must have been a foreign diffusion that did not become established in Samoan culture.
Practically all clubs, except perhaps the bilateral-toothed clubs, the maces, and batons were carved in bands across the blade, shaft, and grip. Where more extensive, panels were formed on the blade and, as in the ear-shaped clubs, the head may be completely covered on either side. The area covered, however, is not so extensive as on Tongan clubs. Most of the authentic old Tongan clubs of the coconut stalk and narrow bladed paddle types were completely covered with carving from the distal to the proximal ends. The paddle club in Plate LII, A, 2 is attributed to Samoa but the extensive nature of the carving which includes even the proximal round end of the handle, makes it more likely that the club is from e, f, and g) leaves little doubt as to its foreign origin.
To carve by cutting away the surrounding surface to leave ribs or ridges in raised relief is termed tongi or tongitongi which corresponds to bas-relief or cameo carving. An experienced Samoan carpenter termed the method of carving clubs, vane, because the motifs were cut below the general surface. Thus vane is the opposite of tongi and corresponds to intaglio of which Samoan club carving consists. The chief carving motif is the small triangle termed fa'amuli'ali'ao (fa'a, like; 'ali'ao, the Trochus niloticus; muli, the point formed by the apical whorls). The same term is applied to the triangle motif in tattooing which, however, is much larger than the triangle in carving. The chevron motif if deliberately cut out is termed fa'avae'ali (like the legs of a bamboo pillow) and a wavy line is fa'anga'ai (like the intestines). Small squares are termed fa'amu (like a draughts board) but as the game of draughts was introduced, the name and motif alike are modern. No other motifs could be given by my carpenter informant regarding the carving of clubs. The instrument used in carving was a shark's tooth. Short parallel lines arranged in panels to represent skeuomorphs of plaiting were not present in the Samoan material available but together with cross hatching were commonly used on both Tongan and Fijian clubs.
The range of Samoan club carving is thus very restricted and is of diagnostic importance in recognizing clubs that have been wrongly labelled as Samoan. All weapons which depart from the use of the simple triangle motif have to be checked very carefully if attributed to Samoa. Such simple departures as the serrated crescent and circle (fig. 317, e-g) immediately cast doubt on the Samoan origin while the simple motive (fig. 317, h) which h) and from the extensive carving combined with the "unusual diamond section of the stem and haft" (5, p. 58), little doubt exists that the club is Tongan. The other club from which the rubbings were made may also from the extent and type of the carving be more correctly attributed to
Foreign influence. Careful analysis of club forms would show that Fijian technique influenced Samoa very little. The Samoan club makers went l) as a paddle club without a crossbar but on examining a very old Tongan club of this type belonging to Professor Wood-
Samoan characteristics. Samoan clubs are characterized by their short length. The two-handed clubs, except the shorter ear-shaped clubs, range between 3 and 4 feet in length and rarely exceed 53 inches. The majority of them are heavy and were used for crushing blows. The method of fighting from footholds used in club fighting with coconut leaf midribs, probably indicates a very early form of fighting in which heavy weapons were considered the most effective weapon.
The lighter paddle clubs in which a thrusting point is added to bilateral cutting edges marks a distinct advance in club technique. The lighter the weapon, the quicker the strokes and the more skill required in parrying and in footwork. The transference of the point to the lower end of the club was never adopted in Samoa. The flared proximal end with the suspensory lug, distinguishes Samoan clubs from those of parts of Polynesia in which a stabbing point was formed below the hand grip at the proximal end. The double functioning ends form the highest development in wooden clubs. Such clubs increased the variety of strokes and guards and quick footwork was necessary to success. The light double-ended club reached its highest development in
The definite use of the bow and arrow in fowling and fishing emphasizes its not being used as a projectile weapon in war. The sling was the definite projectile weapon in Samoa as in other parts of Polynesia. It evidently gave satisfaction as proved by its continued use and consequently there was no place for the bow in warfare.
The acceptance of the teachings of Christianity has been so marked, and extended over such a long period, that information in the field regarding ancient religious regalia and sacred objects is difficult to obtain and lacking in detail. The Samoans have been influenced by teachers of their own race to become ashamed of ancient religious practices and to purposely deny even a theoretical knowledge.
Stone structures. Various natural stone formations that have religious or superstitious significance have been mentioned. Anything in the nature of the stone marae of
Wooden temples. The use of wooden houses with a thatched roof as a religious structure is discussed on page 70.
Sacred objects. Besides a material incarnation in some animal, bird, or fish, some of the Samoan gods seem to have had a material representative of some kind that was kept in the temple, treated with care, and consulted.
Turner (41, pp. 23-35) gives a number of these as follows:
A number of other gods had similar representations. The material objects were kept in the village temple.
A number of other stones representing gods were situated near particular villages and offerings placed on them at particular times, or when passing. Owing to some historical incident associated with ancestors who became gods, certain objects, such as plaited coconut leaves, coconut leaf baskets, and white bark cloth, were worn or treated with respect as being the symbol of the gods.
Idols. The objects representing the gods mentioned above were not carved in any way to represent a human figure or god. Any object considered the representation of the god obtained sanctity and power by association of ideas.
A carved wooden idol is figured by
Head coverings are not worn on ordinary occasions, but wreaths, banana leaves, bark cloth, and a special headdress of human hair, are worn as circumstances demand.
Wreaths (pale). These are made of flowers, leaves, and creepers on festive occasions. Even on ordinary days, anyone who has anything special to do such as making kava or poi for visitors, carrying stones for the church floor, or similar duties, often plucks the neighboring greenery to make a wreath for the head.
Banana leaf hat (pulou lau fa'i). Though going without protection so much, the Samoan does not care to have his head wet when working in the bush in the rain. A section of banana leaf with the stem behind is placed over the head and tied on with a strip of bark passing round the head circumference. The pointed tip end is then doubled back over the head and the tip tucked down into the bark tie behind. Plucking the banana leaf, or any leaf, for that matter, is 'oto. Hence a saying applied to action which has become unnecessary: "A toe 'oto lava, 'ua la" (Why pluck again when the sun is shining?)
Turbans of bark cloth (faufau tu). A strip of bark cloth siapo was twisted round the head sometimes by fishermen but it could not be used in bonito fishing or mullet netting owing to special prohibitions. On occasions of war, the warriors often wore head coverings of cloth as a distinguishing mark against themselves.
Human hair headdress (tuinga lauulu). The tuinga headdress is a development associated with rank and status. Only certain families are allowed the privilege of wearing tuinga, and the honor became hereditary in the title with which it is associated. In important festivals, the tuinga is worn by the chief's official taupou and could also be worn by his son or manaia. The tuinga shown in Plate LIII, B was worn by Fa'apu'a, the taupou of Tufele at
The headdress consists of bleached human hair supported on a foundation of bark cloth and embellished with a lave upright framework decorated with feathers and a forehead band of shells. Sometimes an additional cover of feathers is used.
The type of hair preferred is frizzy (mingimingi) and thick (pi'ipi'i). In olden times, women cut their hair short and men wore it long. A girl with hair of the above type was allowed to let her hair grow long for making a tuinga for those who had the right to use them. A chief desiring hair (lau ulu) approached the parents of the girl and made arrangements for the supply.
The hair is tied together in tufts (Pl. LIII, A, 2). The process of tying is called fa'atavaitui, and the tufts so tied are fua or fuatifuati lauulu. The extra thick fibres called tuatua are selected from the hanks of dressed coconut fibre and used to bind one end of the tufts into a closed loop or eye. (See figure 319.) The tufts are strung on a cord passed through the eyes of the loop. The hair is bleached (fa'aenaena) by two methods: a, lime—the tufts are rubbed well with coral lime or soaked in a wooden bowl in a thick mixture of lime. They are then hung up on the cord exposed to the sun and rain in the open air. Some hair is sufficiently bleached in three months, but dark hair requires relimeing at the end of three months. b, Salt water—the long cord carrying the tufts is tied to a pole at both ends so as to stretch out the tufts. The pole is dipped in sea water and one end stuck in the ground to allow the tufts to bleach in the sun. The dipping in sea water and exposure in the sun is repeated daily until the hair is thoroughly bleached.
When sufficiently bleached, the hair is cleaned with laumea leaves (u'u laumea) and fresh water in which the leaves form a lather. This removes the lime or salt. The tufts are set up on a number of three-ply twisted cords of fau songa. The tuinga examined in
The lave is an upright frame, now made of five slender wooden rods kept together with three crossbars. (See fig. 320.) The rods which originally consisted of coconut leaflet midrib wrapped round with lauu'a bark cloth, are now usually wrapped round with some gaudy red foreign textile.
Side cords are attached to the lower end of the frame to be used for tying the lave in position.
It seems likely that the wider framework diverging outwards and upwards is a modern development to provide a framework for the introduced glass mirror. The more elaborate form of midrib comb is known as selu tuinga, and was used with the tuinga headdress as its name indicates. Such combs or midrib framework decorated with feathers including the long feathers of the tava'e'ula, a species of tropic bird (Phaeton rubricauda), were the precursors of the more elaborate form of modern lave.
The forehead band (pale fuiono) was formed of a length of three-ply fibre braid to which was attached a width of bark cloth about an inch wide to form the foundation for the shell ornamentation of fuiono which Pratt
Nautilus pompilius. The pale fuiono in
The outer end of the nuclear piece is cut off square and directed downwards on the band. The outer whorl piece is cut off square at the end, directed upwards and rounded off below. The nuclear piece is drilled through from the side at its upper part making two holes and drilled with one hole in the middle line near the lower squared edge. The pieces are sewn on to the band with one continuous thread through the upper pair of holes whilst separate knotted loops are passed through the single holes below. The lower pieces have a hole bored through close to the edge on, either side at about the middle. They are attached by a continuous thread run along the back of the band with long stitches and with short stitches passing forward through the material and through the hole and back again before it makes another short stitch through the hole in the side of a neighboring shell. The lower pieces have the natural convexity facing outwards so that the concavity of the straight cut upper end fits over the lower part of the nuclear piece. The nuclear pieces thus have the appearance of large beads. (See Plate LIII, A, 1.) The
The cords carrying the hair tufts, the lave, and the forehead band are the essential decorative parts of the headdress, but each is independent of the other and are stored away as separate pieces. The headdress is only put together on the head of the wearer and taken to pieces again immediately after use. The assembling requires assistance.
The foundation (pou) is formed of a sheet of the thin lau u'a form of bark cloth. This is laid over the head and tied circumferentially round the head by a cord passing over it round the back of the head, across the temples and round the forehead. The cord comes low down above the outer end of the eyebrows in order to keep the sheet on firmly and prevent it being pulled upwards. When tied, the outer margins of the sheet are folded upwards and twisted to form an upward projecting knob on the top of the head. This knob is termed the pou (post) and gives support to the headdress. It must be large enough and high enough to allow the cords bearing the hair tufts to be tied to it.
The cords carrying the hair are wound round the knob and tied to it (noanoa le lau ulu) in such a way that the tufts are to the back and sides.
The lave is then placed in position in front in the middle line and the side cords passed round the knob below the hair and tied behind.
The forehead band now crosses over the lower end of the lave and the end cords are also taken round the knob beneath the hair and tied at the back. Two bands of pale fuiono may be used. They cover both the lower end of the lave and the edge of the bark cloth covering so that they conceal as well as embellish. The part of the band covered with shells is 10.5 inches long which is quite long enough as the ends are covered by the hair at the sides. All the ties of the cords of the various parts are completely concealed by the hair tufts which hang down at the back and sides over the cords. See Plate LIII, B.)
The above completes the true tuinga but a cover (ufi) may be used, consisting of the long feathers of the tropic bird or shorter feathers tied to coconut leaflet midribs (tuaniu). These may be stuck into the hair near the knob or tied to the lave. A smaller set of the red parrakeet feathers arranged in strings as in the titi'ula kilt may also be added over the top of the hair. Often the titi'ula is wound round the knob instead of hair.
The tufts of hair are lastly combed out with coconut leaflet midrib so as to fluff the hair out at the back and sides.
The drawback of the tuinga as a headdress is that it is not permanently put together. It is an addition of various elements that have not developed into one structural combination. The attachment of the bark cloth foundation by a circumferential cord leads to constriction of the blood supply causing headache, pain, and even fainting. Immediately the ceremonial is over, the wearers of tuinga disappear, glad to have them removed instead of flaunting them for as long as possible, which would occur if the headdress were a better organized structure.
The toilet of the body, and especially the hair, receives much attention. After the day's work in forest or sea, the fresh water pools or streams, where available, are in much demand. The Samoans say they cannot sleep at night if the sea salt is not removed by fresh water bathing, as it causes itching of their skin. In places like
Most people have a little toilet basket ('ato ta'ele; ta'ele, to bathe) in which is carried the soap or its Samoan equivalent and the husk of a particular kind of coconut with a soft fibre. (17, p. 28). The soap material consists of the toi (Alphitonia) and the fisoa (Columbrina asiatica) which form a lather. The body and head are lathered and rubbed with the coconut fibre which is termed pulu ta'ele. The leaves are also rubbed against stones (ma'a tamea) to produce the lather. In addition to leaves, oranges and limes are now much in favor for washing the hair.
A clay (u'u), found in the beds of streams, is also used with fresh water as a soap and often left on the head for some time.
Chewed preparations (tuitui) of candlenut kernels, with orange leaves was rubbed into the hair to scent it. The term tuitui is now applied to any chewed preparation used for a similar purpose, such as ifiifi seeds (Parinarium) mixed with coconut oil and langaali. The bulbs of the mumuta are also chewed to form a tuitui.
Coconut oil (lolo), usually scented with sweet smelling leaves or flowers, was much used for rubbing into the hair and over the body after the bath, when the person had returned to the house to don a change of garment. In all dances of the siva posturing type, coconut oil was rubbed over the body so that the dancer entered the arena with the body gleaming with the wet oil.
The oil was prepared from the meat of the mature nut, grated and exposed to the sun. Leaves and flowers could be bruised and broken up with it. After sufficient exposure to make the oil run, the nut material was placed in a to wringer similar to that used in expressing the 'o'a dye from the scraped bark of the 'o'a tree. The wringer which, from its specific use in this instance, is termed to tau lolo was twisted with the usual pole and the oil ran down into a wooden bowl placed below. The oil, if already scented, was stored in coconut shells and corked. If not scented, the scenting material could be added to the prepared oil.
Lime (namu) from coral blocks burnt over a fire in a pit has been largely in use for mission houses and similar types of introduced buildings. The use of lime, however, as a mixture for plastering over the hair is ancient. It was the correct head covering in mullet netting and bonito fishing. On land, protection from the fierce sun in the hot season was obtained from the shelter of the houses and the shade of trees. The Samoan works usually in the early mornings and has a strong objection to exposing his body to the heat of the noonday sun. He is up and away attending to his duties while foreigners are still in bed. When the latter rouse to their activities, the Samoan returns from his plantation and sleeps during the period of heat. The higher culture is thus apt to wrongly accuse the Samoans of laziness as tourists and journalists see them during their resting period and not during the hours that centuries of tropical experience have laid down as the sensible period for physical
Dressing the hair. The women wore their hair short; the men kept their's long. After cutting, the women kept their hair short by singeing the ends with lighted bark. Besides using the pomades described and lime, women also stiffened their hair with breadfruit gum (pulu). After the use of pomades and lime, washing with limes helped to change the hair color to a brown which was greatly desired. Turner (41, p. 122) says the women wore a small twisted lock in a curl with the end lying on the left temple. tutangi ta method, restricted to virgins, consisted of shaving the middle of the head from the front backwards and allowing the hair on the sides to grow long and hang loosely over the shoulders.
For men, the hair was cut or trimmed by regular attendants (songa) who acted as chiefs' valet and barber, besides cup bearer, trumpeter, and messenger. The songa used hinged cockle shells as forceps for pulling out the hairs of the beard. The thick fibres of coconut husk are also used by making an open overhand knot and closing it over a single hair by pulling the ends. The hair is then jerked out. Both processes were slow and painful. The beard was also shaved off with sharks' teeth. The hairs growing from the inside of the nostrils were also plucked. The long hair was tied in a knot called fonga and worn usually a little to the right side of the crown. It could, however, be worn to the front, back, or sides, as fashion directed. fonga. Young men occasionally cut their hair
The body, besides being rubbed with coconut oil was, on occasions, rubbed with turmeric. In death ceremonials, when the turmeric kept for a father by his children (lenga o le matua) was handed out to the mourners, they proceeded to rub themselves with it.
Flowers worn through the hair or over the ear are termed sei.
Ribands. The white, silky looking material obtained from the outer surface of the young inner leaves of the coconut in long riband-like strips is tied in bunches to a piece of coconut leaflet midrib to form the ornament sei milamila which is stuck in the hair.
Pieces of coconut leaflet midrib were used by the Samoans to comb out the hair in a similar manner to combing out the lauulu hair tufts of the tuinga headdress but they did not seem to have been combined into permanent articles for combing. Permanent articles termed selu were made for decoration. The teeth of the selu combs formed the lower ends of a decorative framework for sticking in the hair. Combs were made of both green and dry coconut leaflet midribs and of pau wood.
Green coconut leaflet combs (selu tuaniu) were temporary ornaments made from the green leaflet midrib (tuaniu) and were worn by the rank and file during any special activities, such as carrying stones for a house platform. Such combs were made in two forms; a horizontal row of triangles, or a long vertical form.
The triangular form (Plate LV, A, 7) consists of green midribs partly split to embrace a cross piece, with the split portions worked into triangles above the cross piece and the lower unsplit ends forming teeth for insertion into the hair. A number of triangles so formed were pushed together on the cross piece with their bases touching. (See figure 321.)
The long vertical form (Pl. LV, A, 1) consists of a number of long pieces of green midrib bound together with the long filamentous ends of other green midribs in three bands about 2 inches apart. The lashing of the bands consists of wrapped work technique in the upper band and half of the middle one while the rest is finished off with a wrapped twine. The technique is described in the next type of comb.
Dry coconut leaflet combs (selu tuinga) were made of dry midribs which were scraped and rounded and fastened together with strong single coconut husk fibers in wrapped twines which form good examples of Samoan fine work. Decorative variety was obtained by using hair in the bindings and in post-European times by the addition of glass beads.
The type comb (Pl. LV, A, 5) was commenced at the narrow top end with 9 pieces of midrib each 7 inches long. The first lashing row was a fixation lashing formed of a combination of wrapped work and a wrapped twine. (See figure 322.)
About 1 inch from the first lashing, the primary set of nine midribs was reinforced by another nine midribs placed individually on the right sides of the previous elements. The 18 midribs were lashed together in exactly the same method employed in the first lashing except that the passive element of the weft was in front. The wrapped work is made with the long ends of the midribs towards the worker. (See figure 323.)
Owing to reversing the sides on the completion of each course, the binding turns which are always made towards the right, show in opposite directions on each course and thus give a chevron effect. The hair strands are crossed on each other at the end of each course. When sufficient depth has been obtained, the horizontal hair strands are fixed by tying them together with a reef knot, generally in the space between the marginal and the premarginal midribs to hide the knot. Variety for decorative effect is obtained by change of material. In the comb described, 14 courses were made with hair, then 6 courses with thick coconut fiber, followed in turn by 4 courses of thicker strands of hair (with hair also as the binding material), by 8 courses of hair with coconut fiber binding and ending up with 4 courses of coconut fiber. The two horizontal elements, in this instance, were finished off with a reef
The other varieties of tuinga comb shown in Plate LV are based on the same principles of construction as the one described but are rendered more ornate by the addition of beads. In some of them, not all the upper midrib ends reach the apex.
In the comb (Pl. LV, A, 6) the two middle midribs and each marginal one reach the top but a short midrib extending from the lower end of the decorative lashings is added to the outer side of the marginal midribs. The other midribs forming the comb are cut off obliquely a little above the main cross lashing. The main cross lashing before the commencement of the decorative courses consists of wrapped work over threes for one course which is followed by a wrapped twine. The decorative part is embellished with rows of white and blue beads added to the binding element on every third course after it makes the crossings to either side. The midrib pairs that project upwards to the apex are wrapped with a spiral of coconut fiber on which beads are strung at intervals. After stringing a bead, the fiber passes between the 2 midribs and thus fixes the bead in position.
In the comb (Pl. LV, A, 2) the two middle and two marginal midribs from either, side are carried up to the apex. The other midribs are cut off straight just above the main cross lashing. The lashings consist of wrapped work without the wrapped twine. (See figure 324.)
The comb (Pl. LV, A, 8) is taller and more elaborate in the arrangement of beads.
The tuinga combs are interesting from the use of simple wrapped work and wrapped work combined with a wrapped twine which establishes this form
The selu tuinga as the name implies was associated with the tuinga headdress of human hair and was probably used more before the elaborate lave frame with mirrors was introduced into the decorative scheme. The term selu tuinga is sometimes applied to the long vertical comb made of green midribs (Pl. LV, A, 1) which on account of its length was sometimes used as a framework for feathers with the tuinga headdress. The association of the comb with the headdress has probably led to the wrapped work technique being used with the rather inferior material of green midribs.
Wooden combs (selu pau). Wooden combs cut out of pau wood conform somewhat in shape to that of the dry leaflet combs. Of two combs figured, one (Pl. LV, A, 3) is cut short at the upper end but a middle vertical projection carries out the idea of the triangle. A longer form (Pl. LV, A, 4) projects well upwards and maintains the triangular shape more clearly. Holes, circles, and semicircles, with radiating spaces, have been cut through the wood and the triangle motif is carved in intaglio. The general shape of the comb and the use of intaglio carving come within the sphere of Samoan craftsmanship but the perforated work creates the suspicion that the carved wooden comb is a modern development rendered possible by the acquisition of steel tools.
The ears according to my informants were never pierced and consequently special ear ornaments were unknown. Turner (41, p. 122) states that the
Necklaces of various kinds come under the general term of 'ula. They were made of flowers, leaves, fruit, seeds, shells, and whale ivory.
Flowers. The flowers generally used were those of the mosooi (Canangium odoratum), pua (Plumiera), langa'ali (Aglaia), singano (male Pandanus flower), suni (Phaloria burnettianes), pipi, moa fa'i (petals of banana flower), pualulu, teuila (ginger), nu'anu'a (Nelitris vitiensis), and 'alo'alo (flower of ngatae, Erythrina indica).
The flowers or the petals are threaded on to a strip of fau bast to the end of which a piece of coconut leaflet midrib is attached as a needle. It is usual to space the flowers with scented leaves and much time is spent in blending various flowers and leaves. The girls of the family and the village taupou not only make the 'ula for themselves but also for the guests who happen to be staying with them. Any festivity or ceremonial occasion is marked by the wearing of flower necklaces.
Leaves. The lau maile (Alyxia) is a favorite material. The maile vine is picked and the bark together with the leaves removed from the wood. A fern (langasese) and a strong-smelling shrub (usi, Evodia hortensis) are among those used.
Fruit. Of fruits, fala (pandanus fruit keys) is the most popular. In 'ula fala seemed the favorite decoration of talking chiefs and I was frequently decorated with them after the speeches were concluded as a mark of attention from one talking chief to another. The polo and poloite which, includes both Solanum and Capsicum and matalafi, a smaller Capsicum were much used. The seasea ( Eugenia sp.),
Seeds. Red seeds (lopa) were much in use so that lopa is also used as a name for necklaces made of seeds. Another red seed used is the matamosi-mosi. The dark seed in the fruit of the pu'a (Hernandia peltata), the dark seeds of the Tantania, and the river reed (sangasanga) are also used.
Shells. Land shells (sisi vao), sea shells (sisi tai), and fresh-water shells (sisi vai) of various kinds are used. Of the sea shells, the small cowries (pule) and a spiral shell (pangea) are among those used.
Holes were formerly bored through seeds with a sharp thorn and through shells with the drill. These necklaces are more permanent and are often taken from island to island. Both seeds and shells if obtained from necklaces do not necessarily belong to the island where the necklace was procured. Natives themselves may perpetuate an error by saying the shells are obtainable on their island from a resemblance to a local species.
Whale ivory necklaces. The most valuable of all necklaces ('ula) were those made of whale ivory (lei) and hence called 'ula lei. (See Plate LIV, D, 2.) The material was obtained from whales teeth which were ground down into long curved, pointed pendants. Those in the necklace figured range in length from 3.3 inches at the sides to 5.6 inches in the middle, without allowing for the curve. The greatest width is from 0.4 to 0.5 inches. The thick end is cut off at a slant towards the concave side. All are curved and brought to a long slender point. In section they are fairly round. They are bored through transversely about 0.3 inches from the thick end and in such a way, that when strung on a cord, the concave sides are all on the same surface. The method of attaching to the cord is shown in figure 325.
After the last pendant is fixed, the cords are knotted. The necklace is hung round the neck by the cord. The convex surfaces of the pendants rest against the breast and the points project outwards from the body.
By making the figure-of-eight turns over the back of the pendants, the lashing is not noticeable when the necklace is worn. The necklace is worn by high chiefs and the village maid when in full dress regalia. An 'ula lei necklace is really essential to complete the costume on dress occasions.
The boar's tusk was worn as a breast ornament.
Leaf armlets and anklets (taulima and tauvae) are made of banana leaf or ti leaf when festivities are on. In serving any special foods such as poi and taufolo, the server usually marks the occasion by wearing a wreath necklace, or a piece of the above leaves tied round the arms or ankles. In the siva and poula dances it was usual to wear armlets and anklets of ti leaf or lau maile.
Seed anklets are made of the large seeds of the ifi lele (Afzelia bijuga). The large, flat, elliptical seeds are cut in half and holes drilled through the ends. Pieces of the thin lau u'a cloth are threaded through and knotted at the end which is left in the hollow interior of the seed. The strips are then plaited together in a three-ply braid, the ends of which are tied round the ankle. The seeds rattle against each other as the wearer dances. Single seeds to each cloth strip is a style said to be for males; two seeds to each strip, for girls. Though said to have been in vogue for some time, many maintain that they were introduced from
Boar's tusk armlet. A boar's tusk drilled to carry a cord was worn round the upper arm. (See Plate LIV, D, 1.) White cowrie-shells were also worn on the upper arm in the same position as the boar's tusk.
Garments, such as the fine mat, the various forms of kilts and bark cloth material worn as a skirt and a belt, have all been described. It is again stressed that these articles were not used as ordinary clothing, of which there was little need, but chiefly as personal adornment to distinguish rank during festive and ceremonial occasions.
Human hair belt. The belt of human hair figured in Plate LIV, D, 3 in
Such belts were made in
Both high chiefs and talking chiefs used wooden staffs (to'oto'o) as walking sticks on ordinary occasions, and as a symbol of rank during ceremonial conducted in the open air. The staffs are long rounded sticks, smoothed and polished. One presented to the B, 3) In to'oto'o fai launga (staff with which a speech is made).
Fly whisks (fue) were made of a number of short lengths of sennit braid attached to a wooden handle. The technique was demonstrated by Mikaio of matofi hank and twisted as the braiding proceeded without making the usual fa'ata'a rolls. In this way, thinner braids were formed than the usual sennit braid for lashing purposes. (See figure 326.)
Fly whisks are of two forms; those used by high chiefs, and those by talking chiefs. Formerly the high chief's whisk consisted of a few short lengths of sennit braid lashed to a long handle but it has lately become the custom to use the hair from a white horse's tail and to regard that as the exclusive type for a high chief. The horsehair is braided in three-ply and attached to the handle in the same way as the coconut fiber. The talking chief's whisk is the symbol of his office. In the story of the ancestor Pili, he gave the middle district of B.) On official occasions, a talking chief dresses in bark cloth with a fine mat skirt on the outer side and an appropriate wreath round the neck. He carries his orator's staff in his right hand and his heavy fly whisk in his left.
C in full dress with the orator's fly whisk in position.
Fans (ili) are made of coconut leaf, pandanus, and wood. Coconut leaf fans are divided into two kinds according to the treatment of the leaf which results in different shades of color.
The brown coconut leaf fan (ili aulamalama) is made of ordinary dried coconut leaf (aulama). The tip end of the leaf is used and the leaf midrib forms the handle. The leaflets are first plaited in check over the midrib to cover it and then continue on the opposite side from which they originated, in the check technique. From the midrib the plaiting spreads out to form a widely curved edge. The example shown in Plate LV, B, l is narrower than the average. From the side edges and the far end, the leaflets are doubled back into the body of the fan.
The white coconut leaf fan (ili tea) is made from the young unfolded leaf (moemoe) which on bleaching in the sun assumes a whiter color. The whiter color (tea) gives the fan its name. The leaf midrib is also utilized and the leaflets plaited over it to diverge to the opposite sides. The plaiting is so arranged that the near part of the fan is carried out at right angles to the handle to attain the widest diameter and then curved forward to a mesial point (Pl. LV, B, 3) or to a bifurcated end (Pl. LV, B, 2). In the ilitea, the wefts are much narrower than in the previous type and additional separate wefts have to be added.
In the bifurcated fan (Pl. LV, B, 2), the additional wefts are added in a bundle on either side of the midrib and covered by the wefts springing from the midrib stem. The wefts belonging to the midrib form the lower part of the fan with a check plait. The added wefts then diverge out above them and continue the far end of the fan in a twilled-two plait. Colored material has also been added. The rim is formed by a particular check technique, but no opportunity occurred of seeing the details of plaiting fans. In the pointed fan (Pl. LV, B, 3), the check plait is pursued throughout and the crossing elements are added as individual wefts at the proximal base. Openings are also made as a decorative element. Both fans have wooden handles lashed to the butt of the coconut midrib stalk.
Pandanus leaf fans (ili fala) are made of pandanus wefts with a check plait and of the shape of the example in Plate LV, B, 4. A wide strip of bamboo is used as a handle which is placed in position and covered with the plaiting.
Wooden fans (ili pau) are made of pau wood with a saw after the manner of the wooden combs (Pl. LV, B, 5). They are highly thought of by the Samoans but like the combs, they must also be regarded as a modern development rendered possible by the acquisition of steel tools.
Use. Fans are used not only to cool the user but to keep away flies during meals. Female attendants fan the guest during meals as part of the courtesy devolving upon hospitality and at the same time they keep a wary eye on flies attempting to share in the food. The fanning motions are thus divided between the guest and the food.
The material for decorative adornment was obtained mostly from the native flora of the islands. The Samoans have a deep-rooted love for flowers and leaves not only for their beauty and shape but even more so for their sweet scented odor. Wreaths and necklaces were further supplemented by fruit of appropriate size and color and by seeds. The leaves and bast of plants supplied the kilts and garments which cannot be dissociated from decoration. Coconut leaves supplied the material of most of the combs and fans. Sennit fiber entered into the fly whisk and furnished material for binding. Wood entered into the lave frame, handles, and the combs and fans of doubtful antiquity. Beyond vegetable matter there was little scope in material. Small shells from land, sea, and fresh water furnished necklaces while the lack of pearl shell prevented its use as a material for breast ornaments. Red feathers that were universally prized by all branches of the Polynesians were scarce and practically monopolized for fine mats. Other feathers were utilized for the few kilts made and in association with the headdress but featherwork as a whole was very restricted in both scope and technique. Human hair figured in the important headdress, in tuinga combs, and the doubtful belt. The pig's tusk furnished armlets and breast ornaments while the teeth of whales supplied material for the valued necklace.
The technique in ornamentation did not progress very far. Wrapped work combined with a wrapped twine and figure of eight turns were utilized as lashing techniques. The poor progress in decorative technique is exemplified by the separate elements of the tuinga headdress. The seizing of the hair tufts is a wide-spread technique used by the Marquesans with tufts from old men's beards and by the lave frame, and forehead band into a permanent structure. The high status of the headdress and the physical inconvenience suffered by the wearers should have supplied sufficient incentive, yet Samoan craftsmanship stopped short with the individual elements. It may be that the tuinga
The social value of decoration and its accessories is evident. While flowers and leaves were worn by individuals on any occasion through love of them, the more lavish use by the community marked events of social importance. Even for a solitary guest, the person preparing a bowl of kava will often tie a leaf of ti or even a strip of banana leaf around the head or neck to add importance to the occasion. Flowers, leaves, fruit, seeds, and shells were readily available to all classes of society but the pig's tusk and whale's teeth owing to their scarcity and the necessity for more technique in converting them into ornaments, became associated with those of superior rank. The rarer ornaments were the ornamental adjuncts of the superior types of garments, and became a form of wealth, in themselves indicating higher social position.
The staff and the fly whisk also distinguished social position. Thus, the staff distinguished the head builder among the members of a guild who were erecting a new house. The diplomatic talking chiefs restricted the length of their staffs to the height of their own bodies and the superior status of the high chiefs was distinguished by a longer staff. The talking chiefs soothed their own vanity, however, by creating a more ornate fly whisk which not only distinguished them from the high chiefs as a special class but drew particular attention to their high office as public orators. In making such a departure, they were supported by the traditional dictum of the ancestor Pili that the fly whisk symbolized oratory. The tuinga headdress through being restricted to the holders of certain titles was a further distinction of class within class.
The Samoans are the one branch of the Polynesians who have retained the full tattoo designs of the past as a necessary form of decoration in the present. Tattooing is actively carried on throughout the group with the exception of tatau, the operation ta tatau, and the expert craftsman tufunga ta tatau. Females are tattooed on the legs with sparser lines and on the wrists and hands. Female decoration is not called tatau, but is termed malu.
The implements and material required by the expert are a set of tattooing combs kept in a special receptacle, a tapping mallet, mortar, pestle, candlenut soot, and a palette for the pigment when operating. A small wooden bowl of water in which to rest the combs and an assistant with a supply of damped bark cloth to use for sponging complete the preparations for the operation. (See Plate LVI, B.)
Tattooing combs. The au tattooing comb consists of a serrated bone attached transversely like an adz to a handle. In olden times, Turner (41, p. 89) says a piece of human bone from the pelvis (os ilium) was used to form the comb. They are now made of boar's tusk ground down into plates not exceeding 1 mm. in thickness, ranging from 5.5 mm. upwards in width, and from 17 to 31 mm. in depth. The width of the plates varies with the width of the curved sections of the tusk. The vertical axis of the plate corresponds with the long axis of the tusk. The upper edge is bevelled on the anterior surface.
The bone plate is lashed to a plate of turtle shell which makes the connection with the handle. The turtle shell is ground down to an average thickness of 2 mm., though in a few instances it may be as little as 1.5 mm. or as much as 2.5 mm. The width and depth varies with the variety of comb. (See figure 327.)
In olden days the teeth must have been cut with thin chips of stone. Nowadays a steel table knife with the edge roughened to act like a saw forms an efficacious instrument. In sawing the bone plate with knife or chip, the plate, from the support it receives on the midrib vice, is not liable to break or snap. The plate is divided into from 1 to 1.5 teeth per millimetre. In shaping the teeth, the two end teeth are always made wider than the others to prevent their breaking off so easily, but the points are made as fine as the others.
In these days, the teeth are sharpened with a piece of school slate ground to a thin edge from both surfaces. The instrument is rubbed between the teeth to sharpen the points and clear the inter-teeth spaces. The length of the teeth is about 5 to 6 mm.
There are four types of tattooing instruments of different widths for different functions.
The handle ('au) is made of fau and ranges in length from about 9.5 inches in the smallest instrument to 11.75 inches in the largest. The handle is rounded or elliptical in cross section for most of its course. In the smallest instrument it is 0.3 inches in diameter and 0.4 inches in the largest. The proximal end of the handle is cut away on the under surface for 3 to 3.5 inches for about half its thickness. This gives a flat under surface for the grip and renders the handle less liable to roll in the hand during the operation. The distal end is cut away slightly on the under side to form a slot for the upper end of the turtle shell plate. When the plate is fitted in position, its front surface lies flush with the end surface of the handle. Some handles are made of thin bamboo.
The handle lashing forms a neat piece of work. The thicker tuatua coconut fibres are used and look like fine copper wire. (See figure 328.)
Some of the lashings are very simple but the vertical and lateral turns through the hole are the same. In many of the au tapulu there are two holes for the lashing. In these, one half of the vertical turns and one set of lateral turns pass through each hole but the principle is the same.
Instrument container. The word tunuma which signifies the Samoan container for holding the implements must not be confused with tuluma, a wooden box with a lid cut out of the solid and said to have been introduced from the tunuma is a cylindrical wooden vessel open at both ends as in Plate LV, B, 4. This form is made of a section of pandanus trunk which is easily hollowed. An old tunuma that belonged to the father of Faioso, the young tattooing artist, has been cut out of solid wood and is wider at one end than the other. This is the characteristic shape of the tunuma and not the evenly cylindrical. It was also unique in being covered on its outer side by a close check plaiting of the single thick coconut husk fibres, which overlapped the upper wide margin and descended to within an inch or so of the lower narrower margin.
The instruments are packed away inside the container with the teeth turned inwards to the middle. The handles protrude down through the smaller end and are packed round with thin bark cloth inside the container to prevent the combs turning and becoming injured. A number of instruments may be packed away as the combs are turned in at different levels.
The mallet (sausau) is made of a piece of dry coconut leaf midrib cut to a length of about 23 inches with cross diameters at the thick end of 0.6 and 0.5 inches. The handle part is cut away like the instrument handles to give firm hold to the fingers and prevent it turning in the fingers. (See Plate LVI, B, 5.)
The pigment is prepared from the soot of the candlenut in exactly the same way as for dyeing bark cloth. The material is kept in a coconut shell named pupu lama from its use.
A mortar (Pl. LVI, B, 1) is prepared from a coconut shell which is cut off above the equator. It is termed the ipu tu'i lama, as the lama soot is ground (tu'i) in it. The required quantity of soot is poured into it from the shell container and ground up fine with a wooden pestle (tu'i) made of fau wood. (See Plate LVI, B, 1.) When the dry soot is rendered sufficiently fine, water is added to form a thick pigment.
The palette (ipu tu'u lama), is also formed out of a half coconut shell (ipu) over the opening of which a talo leaf or a mamala leaf is tied with a strip of fau bast. (See Plate LVI, B, 3.) The pigment is dipped up on the pestle and dropped on the improvised palette. In order that the half shell may rest on the ground, the bottom or muli is chipped off or the shell may be set up between three stones to keep it level. Nowadays the talo leaf is usually tied over the end of an empty beef tin because of its flat bottom.
Vessels. A small wooden bowl containing water is used for the instruments when the operation of tattooing is going on. The handles of the instruments are rested on the ground with the combs leant over the rim of the bowl and their teeth thus submerged in the contained water. As the operator finishes with one instrument, he leans it over the rim of the bowl and picks up the particular instrument he requires. The bowl is not a special part of the artist's kit but is a suitable sized bowl obtained from the household.
Another bowl of water is used for washing and dampening the cloth, serving as sponge or towel.
Towels (solo) are formed of old siapo cloth (ta'afi) that owing to use is frayed and soft. (See Plate LVI, B, 8.) A piece is held by the artist wound round the handle of the instrument in use to keep it from slipping. The bare handle is never held.
Another solo is held by the assistant. It is usually dampened overnight and is spread over the skin in front or below the part being tattooed. The assistant stretches the skin taut with the solo. Every now and again as the artist lifts his instrument, the assistant wipes off the blood and extra pigment with the towel. He then quickly moves it down away from the tattooing edge and turns the cloth to expose a clean part as he sees fit. The assistant is also called solo and his main duty is to sponge the tattooing.
Preliminary procedure. For the children of ordinary chiefs (matai), the father consults the artist he desires and they arrange the date for commencing the operation. On the appointed day, the artist visits the village and commences work without any ceremonial.
If a chief's son, the father visits the expert and offers a fine mat to seal the bargain. The mat, if accepted, is called the fusi ta. The expert, on the day appointed by himself, visits the village and takes a large pig (le momoli o le tufunga) as a present to the family of the patient. The pig is divided up amongst the chiefs and talking chiefs of the village. The act of respect adds greatly to the expert's social status, especially if he is a commoner. A speech (launga) is made to the expert by the village talking chief and appropriate references are made to the craft of tattooing. The ceremonial drinking of kava takes place and the tattooing artist has the honor of the first cup. By virtue of his occupation, he ranks with builders as the companion of kings (angai o tupu). The usual feast follows. If all the ceremonial takes place in the morning, the expert may commence tattooing in the afternoon, but if too late, he announces to the gathering that work will commence on the morrow.
Samoan tattooing is rectilinear. The surface covered is divided into areas with its boundaries and orthodox form of treatment. Each area is done in regular order. A number of thicker lines or bars are spaced over the back and each has its proper name. Between the bars and in other areas various forms of lines and secondary motifs are used which also have their names. The secondary motifs may be introduced into panels set in the wider dark bars or areas. It will help the study to enumerate the various secondary motifs with their names and forms.
Lines. The general name for lines is aso, whether they are used as motifs themselves or to form boundaries. When acting as boundaries of an area they receive a specific name associated with the area or their function. As a motif they are divided as in figure 329.
The term fa'aila is applied to some smaller motifs let into the dark areas formed with the au tapulu instrument in the wide bands on the back or any of the filled in parts of other areas. To define it, a rectangle is usually formed first which is enhanced in various ways and the part outside the rectangle is then filled in solid. A birth mark or a spot on the skin is termed ila. The term fa'aila has come to mean window and it is more in this sense that it is used in tattooing. A rectangle set in a dark band or dark area is compared to a window-like opening set in a dark wall. The use of the term is fairly modern. Therefore, as applied to tattooing it marks the period of greater enhancement and decoration in the craft. Where the older school had plain bands and dark areas, the younger school lightens them up by the introduction of fa'aila. The motif within the fa'aila rectangle may consist of any of the smaller motifs. If a motif is sufficiently large and striking in itself, it may be introduced into a dark area without any surrounding rectangle. The motif
Measuring the back (ano le tua). Apart from the execution of detail, a person's tattooing is judged by the height at which the lines from the top
The tattooing takes place in a house set aside and generally with the wall screens let down on the side facing the public street. The father and local chiefs are present at the initial steps. The patient kneels on both knees with his back towards the artist and the light. He is instructed to keep the back straight and rigid with the head up, elbows bent slightly out from the sides and forward with the hands clenched and raised. The artist has all his instruments arranged beside him with the combs resting over the rim of a bowl of water. The prepared pigment is on the talo leaf palette. The tattooing artist, Faioso, demonstrated the whole process on a young man. To get the first line he used what he termed mafaufau (judgment). He could give no anatomical landmarks or approximate measurements of the parts of the body to get the level. No, it was a question of mafaufau that came through serving an apprenticeship to a master craftsman. It was easier to draw the line than to explain exactly where it should be drawn. The same mafaufau persists throughout the operation.
The six lines completing the measuring (ano) of the back (tua) finished, the patient stands up and shows his back to his father and the other chiefs present for their approval. A discussion takes place as to whether the position of the figures is too high, too low, or correct. If agreed to at once, the actual tattooing commences immediately. If not agreed to, further argument and discussion take place the artist naturally adhering to his opinion as expressed by the marking. A difference of opinion is termed utunga lua. When unanimity prevails, the tattooing proceeds.
A large roll or bundle of bark cloth is laid on the floor and the patient lies chest down with the fold under his abdomen. The fold thus elevates the small of the back and stretches the site of the operation. The artist sits on the head side and the assistant on the opposite side towards the legs. The artist uses the mallet with his working hand so the other hand which holds the tattooing comb must be on the side nearest the patient's body. (See Plate LVI, A.) The artist rolls a piece of bark cloth loosely round the hand, takes up the appropriate instrument for making lines (au songi aso laitiiti), dips the points of the comb in the pigment, lays the points on the right end (from himself) of the base line (fig. 330, a, 1) and gives it a sharp tap with the mallet. He lifts the comb to continue a connected line and with quick taps he soon runs the line across. The hand with the mallet is held a little distance away so that it is the end of the fairly long mallet that strikes the handle of the instrument above the comb. The tapping is by finger and wrist movement. The swing of the mallet end describes quite a long arc and the tap is made
The back motifs are filled in as shown in figure 330b. Before filling it in, however, certain motifs are made as in the pula tele to relieve the wide dark band. The right side of the intermediate lines, the pula tele and the canoe are then made symmetrical with those of the left.
The younger generation of tattooing artists maintain that they do better work than the older school because they introduce more ornamentation lines and break up the monotony of the wider dark bands. This is best shown by taking the simpler older design in figure 330c, copied from
The enhancing of wide bands by the modern school is shown in figure 330d.
Then follow the lines (5 and 6) made parallel with the fourth lines (4). but both starting from the base line (1) on the) left and finishing at points on the right that are equidistant from the spine. The outer ends of line (5) correspond practically with the continuation downwards of the post-axillary line. The figure is known as the
pula tele. b, The base line (1) is completed and the lower lines of thepula tamatriangle (2). The triangle is then filled in (tapulu) with one of the wider combs. Between the triangle and the upper line (a, 6) of thepula tele, the space is filled in as the taste of the artist or the patient directs. In this case, there is first a line of dots (tongitongi) as (3), then twoaso tongitongidotted lines (4 and 5) followed by another row oftongitongidots (6). This covers the space and the boundary lines of thepula tele(7) are made and the part between filled in (tapulu). Before filling in certain figures are formed to relieve the monotony of a plain wide band.In filling in the above, only the base line and the
pula tamatriangle were completed. All below the triangle were made on the left to the middle line. The left half of the canoe (8) is also filled in.c, The canoe (3),pula tama(1) andpula tele(2) are all plain and unrelieved by any introduced motives. Between thepula tamaand thepula tele, there is only one sigle line, whereas inb, there are two lines of dots and two serrated lines that from the point of view of the younger Samoan give greater decorative effect.d, Enhancement of the left side of the canoe (va'a). Here the large serrations (1 and 3) are known asfa'atala laupaongo(spiked like pandanus leaves). The rectangular figure internally enhanced (2) is thefa'ailaalready mentioned. The long, narrow triangles (4) are known asfa'asingano(made like the long petals of the male pandanus flower). Any variety of motives may be introduced in a dark field and the general term for such treatment isfa'aila. The outer lines (5) are thefa'aulutaolines that run from the upper angles of the canoe around to the front of the body.
Lower back motifs. The artist works on until he considers that he has done enough for the day or stops earlier if the patient shows signs of excessive pain and exhaustion.
From the pula tele, the tattooing works down on the left in the middle line. To the outer side, the motifs extend out to about the post-axillary line. When a certain amount has been done on the left side, the right side is done in the same way so as to make the design bilaterally symmetrical. The order of motifs in one of Faioso's designs is shown in figure 331.
The parts between the wider bands are filled in as the artist pleases. There is no set rule as to the number of finer lines between the wider bands. Thus in the simple design in figure 330 c, they are filled in with a varying number of fine parallel lines (aso laitiiti).
Owing to the lower oblique lines of the pulatama triangle (the upper part of the patterns) the lines of tattooing run obliquely outwards and upwards from the middle line of the spine. The second tafani runs out to about the anterior superior spine of the ilium. In some designs, a wide vertical band called the ivitu is run up the middle line to divide the two sides. In any case, the two sides diverge towards the lower end owing to the anatomical separation of the gluteal muscles. As the divergence takes place, the part between is filled in with solid tattooing (tapulu) for as far as is convenient.
Some artists work down as far as the second tafani and then complete the anterior side margins (aso fa'aifo) but others keep on down to the solid thigh lausae as described. When the latter course is pursued, the tattooing
Figure 332 shows a very elaborate back design drawn by Faioso himself in pencil and copied in the figure in every detail.
Front of the body. The lines (aso) terminating (fa'aifo) the design on the anterior part of the body are curved forward from the end of the back base line and descend on the anterior surface of the abdomen to meet the punialo unit of the design. The punialo is a triangular unit in the middle line with the base upwards and the apex resting on the pubis, the point being indeterminate and lost as it were in the pubic hair. It is usually tattooed in youth some years before the age of regular tattooing is reached. In modern times, it is used as a test. Young men desiring to be tattooed, yet fearful of the pain, have the punialo tattooed and if they can stand the pain, they go on with the full tattooing later. Many young men are to be seen with the punialo unit alone. They could not summon up the courage to go on with the major operation. This, of course, exists in modern times when youth can please itself. In olden
fa'aifo and the continuation of the back motifs round the sides to the front are shown in figure 333.
Back of the thigh. With the patient on his face, the upper boundary of the thigh is formed by the talitu and the line of fa'aila tautau already tattooed and corresponding practically with the lower fold of the gluteus muscle. The popliteal space at the back of the knee is a special region (atingivae) bounded by a transverse line above the level of the joint. A little to the inner side of the middle longitudinal line of the thigh, a line termed the aso fa'amuli'ali'ao (aso, line; fa'amuli'ali'ao, long narrow triangle motif) is tattooed from the transverse line above the knee joint to a point below the gluteal fold. This important line divides the back of the thigh above the popliteal space into outer and inner areas. As the name indicates, the line forms a base upon Trochus ('ali'ao) shell triangular motif is introduced. The long narrow triangles are made singly or in pairs (moelua), with their bases formed by the longitudinal line and their apices directed outwards. The spaces between the triangles are filled in solid (tapulu) and are included in the outer area (lausae) which is continued outwards to the front of the thigh. The inner area is filled in with lines parallel with the longitudinal boundary line and like it are termed aso fa'amuli'ali'ao. The inner back area of the thigh receives no special name, the name of the dominant line motifs being evidently considered sufficient. The treatment of the two posterior thigh areas and the popliteal space is shown in figure 334.
Front of the thigh. The back of the thigh having been dealt with and the tattooing complete from the canoe motif on the back to the atingivae behind the knee, the patient is turned to lie on his back. The lower boundary of the completed tattooing (see fig. 333 b) is formed by the anterior continuation of the talitu base which ends internally in the comb (selu) motif. A certain amount of the outer surface of the thigh has been dealt with from the back. The serrated lower border of the dark lausae area is brought around to the front and continued obliquely upwards to the inner side of the lausae. The inner anterior part of the lausae area is termed the auanga and is embellished with long panels (fa'aila o le auanga). The remaining area is filled in with the tapulu instrument. The dark lausae area thus extends over the back, outer side, and front of the thigh, being bounded above by the talitu, on the back by the longitudinal line carrying the Trochus (shell) triangles, and below and anteriorly by the line above the atingi vae which is continued upwards and inwards over the front of the thigh. The large dark area so formed has been referred to as the tapulu by Marquardt (20), and tapulu refers to the dark filling in with the tapulu instrument. All the wider bands, the canoe, pula tama, pula tele, tafani, and saemutu are filled in with the tapulu instrument which, from its size, fills in the space more quickly. While the large thigh area may be termed tapulu from the extensive use of the tapulu instrument lausae and the auanga is a special part of it.
The part between the lausae and the longitudinal lines on the inner side of the back of the thigh is divided into two areas by a line which rises on the outer side of the knee cap ligament and runs obliquely upwards to the inner part of the thigh. The outer of the two areas is the fusi (belt) and the inner is the ulumanu (bird's head). The belt area is continuous with the outer part of the atingi vae or popliteal area. The line of tattooing ends below the knee joint. The manner in which the three areas of lausae, fusi, and ulumanu are dealt with is shown in figure 335.
It is natural to assume that tattooing as serving a decorative purpose, receives more attention in the parts that are likely to meet the eye. Of the thigh areas described, the upper and external parts are the least artistic while the lower front, inner, and back parts receive considerable attention in the use of various motifs.
The Samoans themselves give the reason that the darker upper area is covered by the kilt but that in sitting cross-legged in the kava circle and meetings with the kilts tucked up to free the knees, the lower anterior and inner surfaces of the thighs with the belt and bird's-head areas were well exposed. Good tattooing had thus an excellent chance of being displayed and admired. The Trochus (shell) triangles and the atingi vae were seen from the back and the common habit of wearing the kilt short at the back appears to originate from the natural vanity of displaying the decoration of those areas. The extra attention to detail by the younger school is well shown by Faioso's chart of the thigh areas. (See figure 336.)
The pubic area (punialo). The term punialo under the form of pungialo has been given by ngongo or white tern motif. It has been seen that the aso fa'aifo lines descend to terminate or form the boundaries of the alo or belly. The tattooed triangle base upwards is put in to close (puni) the (alo) space and hence the name of puni alo. From its name, the punialo was probably at one time always filled in last of the important areas but later became used as a primary endurance test though the name is quite applicable when made by itself.
A pattern is shown in figure 333 b, 2. Serrated lines and other motifs may be used and it is for filling in the interior of the triangle that the cross-hatching termed fa'a'upenga (like a net) is used. After finishing it, any side lines that fall short are produced to reach the sides of the triangle.
The tattooing was finished with a mark made over the navel (pute) which showed that the work was done. (See figure 333 b, 16.) To the
tufunga it was the last tap of the tattooing comb but to the patient it was the symbol of endurance and manhood, for no youth gets the navel mark until his full tattooing is completed.
From following the tattooing operation through its various stages, what appears complicated on first glancing at a completed design is really simple when dealt with in areas. Though the expert relies on mafaufau (experienced judgment), he makes use of anatomical landmarks by judging distances between points though he may not be able to express himself in words. A certain area has to be filled in and if the back is extra long an additional saemutu band fills the space. The lines are straight though some may show a curve from following the body contour. The aso fa'aifo curving in from the side to the pubis, the lines of the belt, and the bird's head over the front of the thigh show some curve, but there are no great difficulties involved. It may be said that the boundaries of the sectional areas are the same in all designs. The variation occurs in the different decorative treatment of the thicker bands, the dark areas, and the enhancement between the main lines of the design. There is much difference in appearance created even by using a serrated edge to the inner anterior boundary of the dark lausae as compared with a straight edge. In the wider bands of the back, the addition of fa'aila motifs at the ends of the canoe and the various bands makes considerable difference. Unenhanced bands are simply referred to by name but when specially treated receive the name of the motif as a qualifying word with fa'a preceding the motif. Thus the second tafani band in figure 322, 7 is enhanced with a large caterpillar motif, and the band is termed tafani fa'aanufe. Similarly, the addition of caterpillar, centipede, and other motifs between the main bands depend on the artist. Some are certainly an improvement while others have a tendency to become overdecorative and thus confuse the main lines of the design. These enhancements are secondary additions and should be considered in a full design in their proper proportion. In old time tattooing, the design was probably in the nature of figure 330, c throughout; the wide bands being unrelieved by fa'aila or other motifs, and the lines between the bands simply parallel thin lines placed close together.
The more elaborate the design, however, the longer it takes to complete. Thus, a straight line is quickly made with the wider implements, but if the line has to be serrated, it has to be gone over again with the narrowest comb. Every enhancement within a band, or on the dark spaces, takes extra time and care and the artist has to be paid accordingly; not only has to be paid more but he also eats more food. Hence enhanced designs may only be secured by those who can pay. The skilled tufunga ta tatau were like their confreres of the building guild. They created more elaborate design for those who could
aso talitu of a chief and a talking chief. I received no information, but the statement conforms with the general policy of the Samoan talking chiefs. They were the councillors and advisors. They seem to have deliberately created a number of artificial distinctions between themselves and the high chiefs in order to flatter them. When it came to material advantage in the distribution of food and fine mats, the distinctions still existed but the material advantage was on the side of the talking chiefs. Though some such distinction may have been created as a political move by talking chiefs, as regards other distinctions in special parts of the design being reserved to districts, these are more likely, if they occur, to have developed with a craftsman and a school rather than that districts used special designs as a particular badge. The craftsmen used the designs they had been taught or to which they added extra decoration and went wherever their services were paid for.
A number of names of different parts of the tattooing design have been erroneously put on record owing to the modern Samoan tendency to use k instead of t and ng instead of n in their present spoken speech. Such words as kapulu, pungialo, sai muku and kafangi should be tapulu, punialo, saemutu, and tafani.
The small triangular motif (pula tama) below the canoe on the back is also given as pe'a (flying fox) by umangi as the tattooing on the lower belly, which must be an alternative for punialo. Marquardt (20, Taf. IV, 13) gives fa'a vala as the name of the fa'aila motifs on the inner anterior surface of the thigh, known as the auanga part of the lausae. The wide bands on the back below the tafani were given to me by Faioso as saemutu but both Marquardt and saimutu and saimuku.
The meanings of pula, tafani, and saemutu were not obtained.
The operation usually takes place when the youth is about 16 years of age. The time taken depends on the fortitude of the subject. The artist can finish the whole operation in three or four days. Such periods are endurance tests, of benefit neither to the subject nor to the operator.
A good deal of scabbing takes place after the operation and much of the outer skin is shed by scaling. Coconut oil and turmeric are rubbed over the skin to assist healing. Some of the youths have to lie up as they can only
From the point of view of modern surgery, the preparations and carrying out of the operation seem to invite danger. The cloth used for sponging and applying directly to the bleeding surface is old bark cloth that has been softened by use. It is soaked over night in any available fresh water contained in an unsterilized wooden bowl. The skin over the operation site and the instruments receive no special treatment. There seem to be few septic complications in spite of the lack of asepsis and antisepsis.
The tattooing of women being far less elaborate and ceremonial than that of the men is not dignified by the name of tatau but is termed malu. The malu is a lozenge-shaped motif placed in the middle of the popliteal space and being the most important single motif is also used to include the whole process. There is thus no confusion in terms as the malu motif is confined to women.
As the patterns and full design are quite simple, the tattooing of a girl is often used as an opportunity for a student to try his prentice hand. This is also rendered possible by the fact that there is no fusita (fine mat) passed or any of the ceremonial that marks the tattooing of the male. It is often sufficient reward for the novice to have the opportunity of practice and to be well fed during the period occupied by the operation. When the tattooing is finished he may get a fine mat and a feast with pork. In these days, he may receive a little money. The reward depends upon the wealth and generosity of the girl's father. For the daughter of a high chief, who is to become the village taupou, it can be readily understood that an expert artist would be requisitioned and his reward greater.
The wide bands and series of fine parallel lines (aso laitiiti) are not used. The design consists of the individual motifs that were used with men for enhancing between the bands and filling in the larger areas without bands. These motives are arranged in horizontal, vertical, and oblique rows.
Of the motifs used with men, the centipede, caterpillar, tern, and the bent knee are much in vogue. The full design of a woman's tattooing was obtained for me by
Only two instruments are used; the narrowest au fa'atala, and one of the medium au songi for making lines. The very wide au tapulu is never used.
The first motif tattooed is the lozenge-shaped malu at the back of the knee. The patient lies first on her face to expose the back of the leg and then on her back to expose the external surface of the thigh. (See figure 338.)
The anterior, internal, and external vertical rows having been formed with diagonals between them, the patient is turned over on her face again and the back surfaces of the thigh and knee (alo-i-vae) are dealt with. The fusi upper belt is the last part tattooed.
The artist may vary the general routine order to suit himself.
The hands. Both men and women tattoo the hands. A band of some motif is run round the wrist and forms the fusi (belt). On the dorsum of the hands and the fingers, dots, stars, tern, bent knees, and other motifs may be distributed at the artist's taste.
Scarring. Scars were burnt on the arms or chest to form ornamentation. A lighted dry coconut midrib or a piece of lighted bark cloth were used. The scars were also made as a sign of mourning. Young children often do it out of bravado to show how they can stand pain.
The tattooing artists formed an influential guild in their day. The patron deities of the craft were Taenia and Tilafainga. In the myth recorded by
Even now it is customary for a number of youths to be tattooed at the same time. The parents share the expenses of employing the artist. In olden times, according to tulafale (talking chiefs) to be tattooed at the same time as the chief's son so as to share his pain (tale-i-lona-tingd). The chief's family bore the expense of feeding and paying the operators and in addition made presents to the boys who had shared the pain. It is said that the tattooing of "the sharers of pain" was often carelessly done and was looked down upon as 'O le ta tula-fale (talking chief's tattooing). Be this as it may, we have here further evidence of the astuteness of the talking chiefs. They, as the councillors, were responsible for custom and ceremonial to a large extent. Hence, in the custom of sharing the pain, the talking chief got his son tattooed free of cost and was paid with a fine mat as well. It is probably this less decorative form of tattooing that led to the distinction originally between the tattooing of a high chief and that of a talking chief as mentioned by
When a number were tattooed, a large shed was often built specially in a village open space to form an operating theatre. Sham fights and various, games were played during the gatherings. In present times, young friends gather around with modern musical instruments and may be seen in the operating house endeavoring to cheer up the patient with melodies that have been diffused from foreign music halls.
The artists as belonging to a skilled guild were formerly highly paid in fine mats and bark cloth. The payment took place after the completion of the operations. Following the payment of the artists, the families of the boys who shared the pain received their recompense.
Lulu'unga-o-le-tatau). The night before, the artists and their assistants with lighted torches went through a number of motions until the torches were all put out at a given signal. A coconut water bottle was dashed to pieces in front of the newly tattooed youths, the torches relighted and a search made for the plug. The finding of the plug was important as its loss denoted death to one of the tattooed party.
The next day all the newly tattooed were sprinkled with water from a coconut by one of the operators. The sprinkling (lulu'u) ceremony had to be done over each person tattooed irrespective of rank but the breaking of the coconut water bottle was clone only for a high chief.
In olden days custom, personal status, and the approbation of men and women were the incentives to undergoing the operation. In modern times, approbation is still sought. The fear of pain is overcome by the keen desire to bear the marks of manhood and to be able to hitch the kilts a little higher at the evening dances and so demonstrate superiority over the untattooed. As maturity follows and the young man succeeds to the position of matai (chief), he has the satisfaction of feeling that he is truly one of the elect for he can bare his knees with assurance as he sits cross legged before his wall post in the circle of the titled.
A comparative study of Polynesian material culture, though tempting as each group of islands is investigated, cannot be dealt with comprehensively until the field survey of the whole Polynesian area has been completed. Polynesian culture has been affected by the neighboring cultures of
The fly flap. It is difficult to dispose of culture traits fairly in the brief space allocated to them by parallel columns. The listing of the fly flap (19, p. 454) illustrates this. The fly flap is given as present in
Bowling. It is difficult for an authority on one Polynesian locality to cover the whole area by means of brief comparative tables and avoid error. One difficulty is the question of what elements in the cultures of the various localities are truly comparable. If the question of origins is being studied, it seems an error to compare the plank canoe of one locality with the five-piece canoe of another when the two forms exist in each locality. Another difficulty is the question of exact data. Valuable data may not be available in print even of localities concerning which the information has up to now been considered as sufficient. This is true with Samoa as shown by the listing of bowling in important in not used in the other five. The natural deduction is that bowling a flat stone disc with an underhand throw is a local development in te'a has been shown in this work to be present in Samoa in the form of a slice of green breadfruit or a disc of selected coral rock. It was thrown with or without a strip of hibiscus bark. Though pua in some of the islands of the neighboring ulu maika is applied to the polished stone disc when ulu is the Hawaiian name for breadfruit.
Throwing cord. The distribution of the throwing cord in Polynesia must also be revised in the light of more recent information.
The task of separating what is known of Polynesian material culture into distinct Negroid, Caucasian, and Indonesian divisions, as attempted by
Territorial relationship of culture.
Territorial designations. Samoa has been referred to in various writings as central and in others as nuclear Polynesia. It is more accurate to regard it as forming a locality in
End posts. The rectangular house with the ridgepole supported by end posts is regarded as characteristic of to sunu'i type of Samoan house which shows that the ridgepole supported by end posts was also an old type of Samoan construction.
Rounded ends. The rounded ends or apses, now characteristic of Samoan houses, have been shown to have been obtained in the simple form by parallel principal rafters laid between the end rafters of the middle section and a curved wall plate. In the lave). In the better houses of the lave locking joint are unknown. The Samoan apse is curved longitudinally as well as transversely by arranging the heights of the transverse arches. The
Rounded houses. The round house has been listed for the
Roof framework. The roof of the middle section of Samoan houses is curved by using thin flexible rafters of coconut wood. The horizontal purlins are on the inside of the rafters. As the rafters are hung from the ridge-pole and the purlins braced apart by collar beams, the wall plate plays a minor part and the wall posts are put in last. In the
Furniture. The wooden seats with four legs cut out of the solid which are characteristic of the Society and
The rectangular house with the ridgepole supported by end posts, with straight, rigid principal rafters supported by the ridgepole and a wall plate, with wall posts that have to be put in to support the wall plate before the
In
In Samoa, little use is made of pounding cooked food with the result that no food pounders of a permanent type, either in wood or stone, were made. The adz-shaped breadfruit splitter is characteristic. In the east, each locality is characterized by stone food pounders well made and typically shaped. The breadfruit splitters are wedge-shaped without the Samoan type of handle. While the Samoan tripod tree trunk type of coconut grater frame was seen in
In Samoa, the wall screens and better food platters, both plaited in twill, are characteristic. The wall screens occasionally used in the
The Samoan floor mats are coarse as compared with eastern mats. In the makaloa mats reach the highest development in fine work. In Samoa, fine plaiting was directed towards mats used as clothing on state occasions.
Kilts. The ti leaf kilt with a three-ply braided waist cord was common to both areas as the usual garment for rough wear. Kilts made of strips of bast for use in dances are also common to both areas as are also the methods of attaching the bast strips to a single or a double suspensory cord. In the further development of such garments, however, Samoa confined itself to a plaiting technique in check or twill with an ornamental finish in three-ply braid tails in the 'ie tutu pupu'u textile kilts. There is no trace of the single-pair twine being used in garments though the technique was much used in fish traps. In
Tagged garments. The Samoan shaggy garments made with a check plait and completely covered with tags of fibre are characteristic of Samoan culture. The
Fine mats. The Samoan fine mats used as ceremonial skirts and a unit of value, mark a high development in fine check plaiting. Though no such mats were evidently made in malo girdle in
Featherwork. Samoan featherwork though not extensive was important as the valuable fine mats were not complete without it. The feathers were not woven into the edges of the fine mats as tuinga headdress consisted of strings of feathers knotted in the same way. The technique of knotting feathers to a thread was described by
Bark cloth. The paper mulberry and the manufacture of bark cloth are found throughout Polynesia. Different methods depending on the presence or absence of felting occur between the eastern and western areas. In upeti tablet frame has been described in detail and constitutes another important difference between the methods of east and
Types of beater. From the difference of the two methods of beating the bast, the marked difference in the types of beater from the two areas can be readily understood. The Samoan method did not require the close parallel grooves as a finishing stage in a felting process and consequently they were not cut on the beaters as they were unnecessary. The paper mulberry was carried to
Stone structures. In Samoa, the marked feature in stonework is the absence of stone religious structures corresponding to the marae of the east central area and the heiau of
Stone adzes. The lack of applied effort in working stone is reflected in the technique of the Samoan adzes. The commonest types are the least carefully worked in the Polynesian area. The chipping is coarse and there are no evident signs of bruising or pecking in the finished adzes. Grinding in the common types is reduced to a practical minimum. Though Hawaiian adzes also show the minimum of grinding as compared with other eastern localities, the chipping technique is perfect as shown by the control of the marked longitudinal curve so characteristic of that locality and the fine regular chipping is in marked contrast to the coarse and somewhat irregular work characteristic of the common Samoan types.
In cross section, the commonest forms of Samoan adz are quadrangular and not triangular as A. Of the triangular forms, only one specimen was obtained of the reversed triangular form with the wide surface in front that constitutes the common
Another point of marked difference is afforded by the development of the tang. In eastern culture, both central and marginal as listed by
For comparative purposes, canoes must be dealt with in two distinct divisions; the dugout and the plank canoe. The simple dugout consists of a hollowed-out section of tree trunk. In some islands, where trees of sufficient length of straight trunk were or had become scarce, the dugout hull may be formed of two or more sections joined together transversely. In
Technique. An outstanding difference in technique lies in the method by which the separate gunwales and planks were joined together. The most widely spread and evidently the oldest method consisted of boring holes right through the wood from side to side in opposite pairs near the edges. The lashing thus showed on both the inside and the outside of the canoe. After lashing one pair of holes, it was usual to carry the braid along the inside of the canoe to the next pair of holes and thus form a continuous lashing. The other main method of boring holes through projecting flanges formed on the inner side of the plank edges has been described in full in this work. After lashing each pair of holes, the braid is fixed and cut off. The lashing is thus interrupted and shows on the inner side only of the canoe. An intermediate form of lashing consists of boring oblique holes from the same surfaces of two planks so as to meet on the edge surfaces that are fitted together. In this method the planks must be fairly thick but there are no projecting flanges. The lashings show on one side only. The three forms of lashing may be referred to as the right-through, the flange, and the oblique methods.
Simple dugout. The simple dugout while common throughout Polynesia is characterized in Samoa by a deep vertical or even concave cutwater at the bow with a sharp projection forward at the upper end. The usual dugout in the eastern area is marked by a long gradual slope upwards to a pointed bow. In some islands, however, the sharp cutwater is seen but it is not so deep usually as the Samoan. On the other hand, the gradually sloped bow is present in the smaller left canoe of the double 'alia of Samoa. The transverse join of the dugout hull is not seen in Samoa while it is common in the east.
Five-piece canoe. The five-piece canoe was present in Samoa in the 'iato lima type while the five-piece bonito canoe with a dugout hull is supplanting the plank bonito canoe in some parts of the group. The bow and stern covers are flat or conform to the slight upward curve of the ends. There is no break-water, forward bow or upward stern projections. Ornamentation takes the pule shells attached to knobs on the upper surfaces of the bow and stern covers. The forward and upward projections at bow and stern for ornamentation that were present in the taumualua type of canoe alone were an innovation to that type which was first built by an American in 1849. The gunwales in the bonito canoe are attached by the flange method of lashing and the covers by a combination of the flange and right-through methods. The eastern five-piece canoe is characterized by the right-through method of lashing. As
Plank canoe. The plank canoe in Samoa is characterized by the flange join with interrupted lashings. The gunwales may have an alternating series of oblique lashings with the lashings showing on the outside. The eastern plank canoe, is characterized by right-through lashings of the continuous type though they may be interrupted here and there. This feature was observed in a plank sailing canoe from
Holes were bored into the keel and planks at even distances apart, and the men set to work in the following order: Hatu, the chief of
Hiro 's artisans, worked on the outer side to the right of the canoe, andTau -mariari, his assistant, worked on the inner side;Memeru , the royal artisan of Opoa, worked on the outer side to the left of the canoe, and his assistant, Ma'i-hae, worked on the inner side. Each couple faced each other, fixing the planks in their places and drawing the sennit in and out in lacing the wood together; and the canoe soon began to assume form, the bows facing the sea. To make the work light, they sang.
|
| Te Pehe O Hiro ( | The Song of.)Hiro |
E aha ta'u, e | Tane e,What have I, | O Tane ,|
| Tane , atua no te purotu e? | O Tane , god of beauty?|
E 'aha. | 'Tis sennit. | |
E 'aha o te hui o te ra'i, | 'Tis sennit of the host of heaven, | |
E 'aha na'u e | Tane e!'Tis sennit for thee, | O Tane !|
E tui i roto, e puputa i vaho, | Thread it from the inside, it comes outside, | |
E tui i vaho, e puputa i roto. | Thread it from the outside, it goes inside. | |
| Nati hua, nati mau.Tie it fully, tie it fast. |
Seam battens. The right-through method of lashing in repairing split planks has been described for Samoa. In
Outrigger. Details of the various forms of outrigger construction in the different parts of the eastern area require to be recorded and analyzed before full comparisons can be made between east and west. If
Peg lashings. Another curious development in
The sail. The difference between east and west in the setting of the sail has been remarked by tila attached to their edges. The apex of the sail was fixed forward of the mast. A rope was tied to about the middle of the upper tila and passed through a special support at the masthead. By means of the rope the sail was hauled up into position with its long axis oblique.
Fishing methods employing narcotizing, spearing, sweeps of coconut leaves, walled traps, and nets made with netting needles and mesh gages are similar in principle in both areas. In Samoa, the purse net with two wings set like a V-shaped weir is common. Trolling for bonito with an unbaited hook with a pearl shell shank attached by a line to a bamboo rod is similar in principle in the two areas but there are variations in the shape of the hook point, the hook lashing and the method of setting the rod. Albacore fishing with a double canoe and a crane which is a marked feature of the masimasi and in mahimahi. Apart from the masimasi hook and the poor specimens described, Samoa is devoid of baited hooks. The shark hook was unknown and the method of noosing prevailed in its stead. Deep line fishing was absent and there is no authentic evidence to show that the deep sea Ruvettus was ever caught on a long line with a baited hook. The Samoans were thus surface fishermen. In the east, varieties of baited hooks occur and the shark hook is common. Ruvettus has been shown by Ruvettus fishing. Samoa is rich in varieties of fish traps but detailed technique from the various parts of the eastern area is lacking.
Traps. The principle of the bent sapling spring, the running noose and the trigger seen in the Samoan mailei trap to catch fowls and wild pigs was used in the fowl traps of the
The bow. The knowledge of the bow and arrow as shown by
Net. The pigeon net of such importance in Samoa occupied no similar position in the eastern area. The dove decoy cage also seems confined to the west.
Terraced irrigation. In the eastern area, the terraced irrigation of the taro (Colocasia antiquorum) forms a marked feature. The water is tapped at a higher level of the stream and conducted down by a main irrigation channel to flats formed by bends of the stream or to where the narrow stream defile opens out into a broader valley. The lower margins and sides of succeeding terraces are built up and faced with stone. Some areas are large in extent and must have involved considerable labor and ingenuity. Terraced irrigation naturally requires streams with a fair fall. The method exists not only in the east central area but also in the marginal localities of
Common to east and west. Of seven games tabulated by
Bowling and coasting which lafonga played with coconut shell discs is present in tupe.
Eastern area alone.
Western area alone. Of Samoan games not recorded from the east, I can only think of the water tip-cat game of tapalenga and the fiti game of jackstraws.
Shell trumpets of Triton and Cassis shell and the mouth flute are found in both areas. The true drum made of an upright hollowed log with shark skin stretched over the upper end is characteristic of the east and absent in the west. The eastern drum was important in the ritual connected with the
The wooden gong made of a hollowed log with closed ends and a narrow slit-like opening is characteristic of the western area. The history of the slotted gong in Samoa is peculiar in that the true Samoan nafa form has disappeared whilst the two larger forms present are said to have been introduced from pate form is also stated to have been introduced from pate gong is present in the east central area. A larger well carved form known as kahara in the takurua ceremonies at Atia-te-Varinga, which
As
At the present time, the comparative study of weapons can best be made in Museums whither most of them have gone. One feature however, may be remarked on here. Samoan clubs are marked by a variety of forms but they are all short, rarely exceeding four feet in length. They are mostly made for striking with one end whilst the proximal end is flared and blunt. The characteristic weapons of the Society and
The data from Samoa would indicate that canoes with topsides or planks joined together by the right-through lashing and with the use of the covering
Certain culture traits, such as the marae type of religious stone structure, stone figures, stone food pounders, upright drums covered with skin at the upper end, and the nose flute, passed into or developed in
The striking features distinguishing Samoan material culture from the general culture of
The builders' guild in Samoa seems to have had more set rules of organization than their fellow craftsmen of
While the kava plant reached a, p. 327) believes the ceremony connected with kava drinking to be adapted from the Buddhist ceremonial tea drinking. Ceremonial kava drinking and the guild of builders (Sa Tangaloa) are both associated with Tangaloa. As neither the elaborate kava ceremony nor the trade unionism of the Tangaloa builders reached
Plate XXXVI. Front, back, and right side views of adzes: A, Type I. 1. (B.8937) 6 oz. (See fig. 173.) 2. (C.820) 15 oz. (See fig. 174.) B, Type II. 1. (C.788) 15 oz. (See fig. 179.) 2. (L.2029) 7.5 oz. (See fig. 180.) C, Type III. 1. (C.586) 5.5 oz. (See fig. 185.) 2. (C.796) Broken adz, weight 5 oz. (See fig. 186.) D, Type IV; 1. (L.2186) 5 oz. (See fig. 187.) 2. (C.812) 3.5 oz. (See fig. 188.)
Plate XXXVII. Front, back, and right side views of adzes: A, Type V. 1, (L.1491) 10 oz. (See fig. 192.) 2. (C.807) 11 oz. (See fig. 193.) B, Types VI, VII and VIII. 1, Type VI. (C.825) 5-5 oz. (See fig. 195.) 2. Type VII (B.8940) 15 oz. (See fig. 197.) 3. Type VIII. (C.584) 4 oz. (See fig. 199.) C, Chisels. 1. (L.1566) Length, 94 mm., width, 31 mm. 2. (L.1565) (See fig. 211.) 3. (L.1483) (See fig. 212.) 4. D, Hafted adzes and haft. 1. Haft with projecting heel, single lozenge lashing. 2, Haft without projecting heel, adze butt entirely on toe of haft, chevron lashing. 3, Haft with projecting heel, double lozenge lashing. 4. Haft showing high shoulder and plane level surface below for adz butt, toe angle not so acute as in other three hafts.
Plate XXXIX. A, Section of hold of bonito canoe (Pl. XXXVIII, C) near aft boom showing connecting flanges with sennit lashings on either side of keel and also between lower and second tier; end join between two pieces of lower tier; strengthening ribs in a lower tier section to right of boom; wide gunwales with single lozenge lashing over the aft boom; pu'enga handhold between left gunwale and aft boom. B, Bow section of above canoe; showing bow cover with mesial line of ovulum shells attached to raised stands; fore boom with gunwale lashings and lashings to inner and outer pairs of connecting pegs which are inserted into the float close together; vertical suspensory attachment passing under float; short middle boom; some lashings between gunwale and side piece showing externally; gunwale with wide upper surface and external raised rim.