As throughout Polynesia, bark cloth was the material with which the people clothed themselves. Other material was used on occasion for kilts, but these alternative garments were used at festivities and in dancing.

Bark Cloth.

In the Cook Group, Aitutaki was noted for its bark cloth. Though little may be made now, there are scores of women who are acquainted with the details of manufacture. The wooden anvils and ironwood beaters are quite common amongst the households. Bark cloth is still used as a bed spread after childbirth. The preliminary stages of technique were demonstrated to the author with actual material.

Plants Used. The plant most commonly used is the paper mulberry, Broussonetia papyrifera. There are said to be three varieties, the purautea, anga and aute. Anga is the word most frequently used but anga and aute seem to be interchangeable. Cultivated plants of the anga and purautea were pointed out, and attention drawn to the larger leaf of the latter.

In the genealogy of plants, the ancestry of Purautea was given as follows:—

Atea (m)—Papa (f).
Te Hatu (m)—Anganui (f).
Te Tara  Te Maori  Purautea  Papako.

About the two eldest children there is no existing record beyond their descent. Te Tara and Te Maori, as plants, are unknown to the present generation. Papako, the youngest of the family, yields his bark for the making of cords and ropes and long nets with a big mesh, for catching turtle, shark, stingray and other big fish. It also is used to make a cloth used as a waist cloth, and thus named maro papako.

In addition to the paper mulberry, the bark of the breadfruit, kuru, and the banyan, aoa, are used.

Wooden Anvil, tutunga. The act of beating out the bark is termed tutu and the wooden anvil upon which the page 77beating is done is called the tutunga. Though the industry has ceased to be very active, most of the families retain their tutunga, and sentiment is still strong enough to prevent them being cut up merely as timber for other purposes. The wood used is tamanu or miro. A typical one measured at Vaipae was 12 feet 10 inches in length. It was perfectly rectangular except for a groove on its under surface, as shown in Fig. 70.

Figure 70.
Cross section of wooden anvil, tutunga.

Beaters, ike. The beaters for beating out the bark on the tutunga are made of ironwood and have four grooved surfaces. Sometimes miro wood is used. The handle is shaped, whereas in Mangaia, the grooved surfaces run from end to end without any cutting away for a handle. The grooves run longitudinally and were cut with a shark's tooth fixed to a handle. The beater is called an ike and the

Figure 71.
Bark cloth beater, ike.

page 78grooves, ugao. The grooves may be narrow and closely set, ngao rikiriki, or they may be deep and wide, ngao rarahi.

Each beater may have sets of differently-spaced grooves on each of its four surfaces. The different sets have different names, and each kind is used for a particular stage in beating the bark. There are four sets named and the differences are given in the number of ridges per centimetre, as they were easier to measure.

1. Ngao hoahoa	1 ridge to 1 cm.
2. Ngao papa	3 ridge to 1 cm.
3. Ngao iheihe	4 to 5 ridge to 1 cm.
4. Ngao tahakaieie	6½ ridge to 1 cm.

The typical ike in Fig. 71 is 16¾ inches long. The grooved surfaces for beating are 10½ inches long, leaving the handle part 6¼ inches. The width of the grooved surfaces are 2.1 inches and 2.2 inches. The narrowest part of the handle is 1.7 inches in diameter. The outer end of the handle is flared and cut off at a slant. The four grooved surfaces are ngao hoahoa, ngao papa and two ngao iheihe.

The coarser surfaces of the ike were for the preliminary stages of beating out the bark and the finer surfaces for finishing off. Thus though some ike had the four kinds of grooving, the womenfolk generally had a set of beaters of different kinds.

Dyeing Frames, rakau takiri pahoa. Patterns were printed on the prepared cloth in some cases by means of frames. None of these were seen. A square or rectangular frame was made of strips of the aerial roots of the pandanus or of other wood. To this, the dried mid-ribs of cocoanut leaflets were tied neatly with fine sinnet. They were arranged in two sets of parallel lines, which crossed vertically and horizontally, or obliquely. The frame was called a rakau takiri pahoa. The dye was brushed or dabbed on to the midrib lines and the frame then pressed down

Figure 72.
Patterns on bark cloth.

A—Ngaito. B—Okaoka.

page 79on the cloth. Though the women knew a good deal about cloth-beating, the finer points about dye patterns had evidently not been passed on to the present generation of old women.

The only patterns remembered were those of the parallel lines made with the above frames. When the lines ran vertically and horizontally, the pattern was named ngaito, and when obliquely, okaoka. See Fig. 72.

The Beating Process, Tutu.

The bark is removed from the trees or saplings and taken home. Each stage will now be taken under its sub-heading.

Separating the Bast. At home, the inner bark or bast (kiko) is separated from the outer bark (pakiri). This is done by separating them at one of the cut ends with a kahi shell, when the two parts are readily pulled apart.

Soaking in Water. The bast is tied in bundles and soaked in sea water for one day.

Hoahanga Process. The first beating is now done on the wooden anvil. Each individual strip of bast is beaten with the ngao hoahoa face of the beater. This flattens out the strips and brings out the texture of the bark. The process is called hoahanga. The general process of beating bark on an anvil with a beater is tutu, but the particular act of first flattening out the bast is hoa. Therefore this particular part of the beating process is the hoahanga. As coarse heavy ridges are required on the beater to do this preliminary work, that type of grooved surface becomes the ngao hoahoa, the hoa grooves.

Washing. When the individual strips have been beaten enough, they are washed in fresh water to remove the salt from the previous immersion in the sea, and to get rid of sap and green colouring matter, tahe. This cleans and whitens the material.

Draining. The material is left in a rough cocoanut leaf basket, tapora, until next day to allow it to drain. It is then removed and wrapped up in the leaves of the kind of taro called puraka. The puraka has a larger leaf than the ordinary kind but not so large as the wild kape species. It is left for three days before the next stage is commenced.

Alternate Process. Some women do not soak the bast in salt water for a day. They commence the hoahanga process immediately after pealing off the outer skin. The page 80dry strips are beaten, soaked in fresh water for 24 hours and then wrapped in puraka leaves, as in the last process.

Papahanga Process. The material is removed from the puraka leaves, laid on the anvil, and beaten with the medium-grooved surfaces of the beaters. The ngao papa surface is used until the right width, kapu, is attained. The slightly finer ngao iheihe surface is used to further spread out the material. A number of women are engaged along the 12 feet length of the anvil, the edges of each section are overlapped and beaten into a continuous sheet, papa. Hence the process is the papahanga, or sheet process, and the grooved surface of the first beater used is named the ngao papa.

The Finishing Process, hakaotinga. The cloth is finished off, hakaoti, with the finest-grooved surface of the beater, ngao tahakaieie. This renders the cloth manea, or beautiful. Some islands use a perfectly smooth surface to finish off with, but the Aitutaki women did not recognise this procedure.

Drying, ka tauraki ki te ra. The beating technique having been finished, the cloth was spread out in the sun to dry. The edges were kept down with the heavier dark stones known as pohatu Maori. Selected stones were kept for this purpose. The cloth was exposed to the sun for three or four days to render it strong, pakari.

Kinds of Bark Cloth.

Garments of Bark Cloth.

Besides clothing bark cloth was used as bed-covers and as wrappers. It is peculiar that in the island of Mangaia, which is so near to Rarotonga, plaited sleeping-mats were not made, but bark cloth took its place.

As clothing, the garments of bark cloth may be divided into loin cloths, kilts and cloaks.

Loin Cloths, Maro. The term maro is universal throughout Polynesia, and extends to New Zealand and the Chatham Islands. It implies that the material must pass, in some part of its course, between the legs.

In preparing the bast for a maro, only sufficient was beaten to form a narrow width of about two finger spans. The finger span is the largest span between the tips of the outstretched middle finger and the thumb, and was termed an angahono. Two such spans were roughly about 18 inches. The pahoa so prepared was dried and used without further treatment.

It was put on in the usual way. One end was held up against the abdomen. The long end was passed back between the legs, looped round over the right hip, crossed transversely over the first part, passed round the waist above the left hip, and hooked back through the loop made at the back. This straightened up the band passing between the legs. The end passing to the left was tucked in under the horizontal band and so secured the maro. The short end in front was dropped and hung over the waist band with its free end down.

Maro papako. The bast of the papako, roughly made, was used as a maro and worn in the same way.

The maro was worn by men and gave them great freedom of movement in their work.

Kilt, Pareu. A length of cloth was worn round the waist. A short length extending to the knee was worn by the men, and a longer length, extending even to the ankles, was worn by the women. By tucking the outer end in over the waist edge, a belt was dispensed with. Whether used as a kilt or a skirt, the garment was termed a pareu.

Body Garments. It is probable that a strip of cloth was on occasion wrapped round the shoulders, as a shawl or cloak.

Tope. These were described by an old man, who had worn them in his youth, as two strips of cloth over an arm's span, māro, in length. The mid-part of one rested on the back of the neck. The two ends were brought forward over the shoulder crossed over the breast and passed back under the arms and held at the back. The other strip was doubled round the front of the neck, crossed at the back and brought forward under the arms. The four ends were then fixed by a wide band of cloth wound horizontally round the abdomen. The first two strips were termed tope and the page 85third, the tatua or belt. Thus the upper body was completely covered by a garment consisting of three pieces.

Tiputa. The tiputa consisted of a long strip of cloth the width of the shoulders, with a hole in the centre. The head was inserted through the hole and the ends hung down over the front and back of the body. It was thus worn like a poncho. The edges were often cut with serrations, or in such a manner that long narrow strips formed a fringe.

Full Dress. On ceremonial occasions full dress consisted of the pareu kilt, the two tope with the tatua band and over all, the tiputa.

Clothing not Made of Bark Cloth.

Material other than bark cloth did not enter seriously into the clothing of Aitutaki. Bark cloth furnished the every-day clothing, but for rough work or for dances, kilts were made from the leaves of the Dracaena terminalis (ti) and the cocoanut, and from the inner bark of the hau. They all come under the name of pareu.

Ti Leaf Kilt, pareu rau ti. Rau ti are the leaves of the Dracaena terminalis. The kilt is made by plaiting the stalk

Figure 73.
Kilt of ti leaves, pareu rau ti.

page 86ends of the whole leaves into a three-ply braid waistband. Figs. 73 and 74.

Figure 74.
Three-ply braid of ti leaf kilt.

The leaves are about 18 inches long and 3 inches wide. The thick fleshy midribs are cut through near the base of the leaf in such a manner with the thumb nail that when they are stripped off, thin strips of the stalk are left attached to the leaf. Fig. 75a and 75b. The stalk ends of the leaves are then plaited into a three-ply braid. Each leaf is added from below on the under side of the braid. Fig. 75c. On completing the required width of the kilt, the ends are plaited on for a short distance, divided into two lots, and tied with a reef knot.

Figure 75.
Technique of ti leaf kilt.

The three-ply braid not only serves to attach the leaves together but also to attach the kilt round the waist. The actual appearance is shown in Fig. 74. These kilts are quickly made and cast aside after their temporary use is ended. The golden colour of the partly dried leaves gives a pleasing effect in the dances in which they are mostly used. The old name for this kill is titi. The same name is used in Samoa, where Turner¹ says it was the only page 87garment that either sex thought necessary during the day. Stokes² described it in Rapa. It probably has as wide a distribution as the plant.

Cocoanut Leaf Kilt, pareu nikau. A similar kilt was made from the leaflets of the cocoanut. My informants summed up the technique in the words, "Ka toetoe te rau, ka hiri." The leaflets were split with the thumb nail on either side of the midribs, torn off and the butt ends plaited with a three-ply braid. Its use was also temporary.

Hibiscus Bast Kilt, pareu kiri hau. Here we have another use for the bast of the Hibiscus tiliaceus, hau. Long thin strips of the inner bark, which splits readily after soaking in sea water, are attached to special cords with the effect shown in the two examples in Fig. 76.

Figure 76.
Hibiscus bast kilts, pareu kiri hau.

The three-ply braid plait at the waist is here abandoned. The strips of material are attached to two two-ply twisted cords of oronga or kiri hau. The two cords, somewhat longer than the circumference of the waist, are sketched between two sticks, posts or trees. Any uprights are used and there were no special sticks made or reserved for the purpose.

Figure 77.
Single pair twine in above kilts.

The method of attaching the strips has already been described by the author³ in a kilt from Rarotonga with the same technique as in Aitutaki. The middle of the strip is placed below the cords. The two ends are brought up on either side and passed down between the cords as in Fig. 78a.

Figure 78.
Attachment of bast strips to two cords.

In some kilts a more secure attachment is made by passing the ends back through the loop made by the middle part of the strip below the cords. Fig. 78b. The strips are looped closely along the cords by one of the above methods until the required width of the kilt is attained.

In a kilt obtained In Rarotonga, the second of the above methods was used with only one suspensory cord. Fig. 79. From the figure, it is seen that the knot made was a lark's head. The kilt is shown in Fig. 81.

Figure 79.
Attachment of bast strips to one cord. a. Anterior. b. Posterior.

The waist band is further defined by running a twined row across the strips an inch or so below the waist cords. Two or three of the doubled vertical strips are treated as a single warp. A long cord is doubled round the left marginal warp, which is enclosed in a half-turn of the two parts of the cord. The anterior cord, as it turns to the back of the next warp, passes above the posterior cord as that in turn comes forward to pass in front of the next warp. They again change position on the third warp, and so right across the width of the kilt. Thus each vertical warp of two or three double strips is enclosed by a half-turn of a single-pair twine. Fig. 80.

Figure 80.—Single-pair twine technique.

There is usually one line of single-pair twining as in the two kilts in Fig. 76, but in the Rarotongan kilt in Fig. 81, there are two rows. As the single-pair twine is useful for comparative purposes, it is shown up more clearly in Fig. 77.

Figure 81.—Hibiscus kilt ornamented with coloured seeds.

Further ornamentation in dance kilts was obtained by attaching a plaited band of pandanus strips to the waist border above and to a twined row below. The plaited band was attached by piercing the material with an au needle of ironwood and then pushing a continuous cord through the holes thus made. The lower twined row kept the lower edge of the plaited band taut and prevented wrinkling. The plaited band was either plain or had a design worked in colour after the manner of the sleeping mat borders.

Strings of coloured seeds or shells were also attached to some kilts as additional ornamentation, rakei. Though the kilt illustrating this is from Rarotonga, the Aitutaki method was exactly similar.

Armlets and Leg Bands.

Armlets and leg bands of ti leaves were used at dances and festivals. The ends of the leaves were plaited together with a three-ply braid as in the case of the ti leaf kilts. They were then tied round the wrists, upper arms ankles and below the knees.

In the drama played to depict the coming of the first discoverer of Aitutaki, the principal character, Ru, was completely dressed in Dracaena leaves. Besides kilt, armlets and leg bands, he wore another kilt as a cape and had a band of large leaves bound round the head with the tip ends projecting upwards. It gave a picturesque appearance, but the man himself was hardly discernible for vegetation.

Head-Dresses, Pare.

Anything worn on the head came under the generic term of pare. They varied from a garland of leaves to the elaborate head-dresses of the ariki class, which were built up on an actual skull cap made of sinnet.

Temporary pare made of leaves were used as sun screens when the people were working on their cultivations. Whatever suitable plant grew near was quickly plaited or twisted into a band or wreath that gave the necessary protection to the head. The people were very fond of wearing flowers and greenery and even when sun protection was not actually needed, a pare was worn for decoration. At dances, feasts and festivals, pare of this type were appropriate. Pare rauti, braided with the leaves of the Dracaena, were worn at dances. Pare of the leaves of the page 91ngatoro plant were also used as a sun screen. Leaves of the maire fern were so used.

Pare tainoka were made of the tainoka creeper that grows on, the beach. This plant is common on the islands in the lagoon. The long creeping stems with the attached leaves are simply twisted round to form a wreath. As the stems are of a yellowish colour, the pare is ornamental as well as useful.

Wreaths made of flowers, the drupes of the pandanus and the red berries of the porohiti interspersed with scented herbs were also worn at feasts, umu kai. These are purely decorative.

For the pare of a more permanent nature, the basis seems to have been made of sinnet.

Pare kaha. Kaha is a cord of sinnet fibre. High conical caps, somewhat resembling the Egyptian tarboush, were formerly made of kaha and thus received the name of pare kaha. They are not now made on Aitutaki and a specimen could not be obtained. They were said, however, to be of exactly the same technique as those of Atiu, which are still made. The pare kaha seen from Atiu was made by coiled weaving but the details of technique belong to the material culture of that island.

Pare huka rau. This was said to be a pare ariki, the head-dress of a high chief. None were to be seen. In Rarotonga, however, the pare ariki of the Makea family was formed of a very old pare kaha to the outside of which a wooden framework carrying feathers was attached. In the historical story of the famous Marouna, some pare of this kind must have figured. In order that he might speedily go from Rarotongs to Aitutaki to the aid of his grandfather, Maeva-i-te-rangi, he is said to have bought the canoe "Te mata-koviriviri" from Angainui with a part of his head-dress, tetahi manga i te pare. It seems likely that he gave some part of the wooden framework carrying the feathers to indicate the transmission of some of his power or authority over land, or other material benefit.

Pare kura. The pare kura was the head-dress of a chief. It consisted of a pare kaha to which feathers such as those obtained from the small island of Manuae, were attached. Though kura means red, the feathers on a pare kura were not necessarily red. They were probably red in the period and place when and where the name was first applied. When migrations took place to regions where red page 92feathers became unprocurable or difficult to obtain, other feathers were used of necessity, but the head-dress retained its name of pare kura from its continued association with chieftainship. The feathers were sometimes tied to the midribs of cocoanut leaflets and attached to the sinnet skull cap.

There was another pare concerning which little information could be obtained beyond its name. The hare karioi already mentioned as having stood at Vaitupa, was built by a tahunga expert named Rahui. Rahui wore a head-dress called a pare kauhatu and hence was known as Rahui-pare-kauhatu. What form of head-dress it was my informants were unable to suggest.

Rarotongan Head-dress. Though the head-dresses of the chiefs of Aitutaki have entirely disappeared, the author had the privilege of examining the pare ariki of Makea Ariki of Rarotonga. It is the only surviving chief's headdress on the island. The sinnet cone-shaped cap that forms its basis is shown in Fig. 82. This was covered with feathers tied to light rods and gave the imposing appearance evident in Fig. 83.

Figure 82.
Sinnet foundation of Head-dress of Makea Ariki

Figure 83.
Head-dress of Makea Ariki. with feathers attached.

The feathers are of cock's plumage and the long red tail feathers of the tropic-bird, Phaeton rubricauda. The feathers were changed with each new ariki but the sinnet cap had been in the Makea family for several generations. As the feathers shown in Fig. 83 had been recently put on for the raising of Makea Tinirau to the ariki position, we raised no objection when, with his innate courtesy, Makea removed them to enable a closer study of the really old sinnet cap to be made. The rods carrying the feathers had been sewn on with strong cotton and there was no detail of old technique that could be destroyed.

The sinnet cap had been made in Atiu. In Fig. 82, it will be noted that the lower part consists of horizontal rows that had been done by coiled work. This is the present technique of the Atiu sinnet caps already referred to as pare kaha. The upper part, however, consists of vertical rows of alternate black and natural brown sinnet. This work is no longer known. The technique appeared to resemble the two-pair interlocking weft used in the better class of New Zealand garments. The view of the cap is from the side. The rows from the front and back can thus be seen meeting in a line that extends upwards from the middle of the sides to meet at the apex.

From the construction of the Rarotongan pare ariki with a sinnet cap and feathers tied to sticks, it would seem to correspond to the Aitutaki pare kura.

Footwear.
Sandals, Tamaka.

Sandals did not form part of the wearing outfit of everyday use. The soles of the feet were hard enough for ordinary progression, but even they had to be protected against the sharp points on the coral reef, where so much time was spent in procuring fish and shell-fish.

Sandals, tamaka, are made from the bast, of the hau, which is separated from the outer bark after immersion in sea water, and then dried.

The bast is twisted into long single-ply cords by means of a short stick. One end of a strip is held with the left hand and the end of the stick twisted a couple of times round the single bast strip. Fig. 84. By rotating the stick with the right hand, the bast is twisted on itself into a rounded cord. As the strip twists, the stick keeps travelling back along the untwisted part. A fresh piece is added by page 94overlapping the ends for a few inches and continuing the twist over them.

Figure 84.
Twisting single-ply bast cord.

When a sufficient length to form one sandal has been twisted, the bast strip is split into three equal parts and plaited into a three-ply braid, fresh pieces being added as required. The other end, which has been left long before the twist commencement, is also split and plaited into a three-ply braid.

The twisted part of the long cord goes to form the longitudinal elements of the sole of the sandal, and the braided ends go to form the transverse elements and the curds for tying on the sandal.

Figure 85.

A. Now sandals, upper surface. B. Sticks used in manufacture.

C. Old sandals, under surface.

Implements. Two sticks not quite as thick as the little finger and 14 and 16 inches long respectively, are required. Both are pointed at one end and the shorter one, usually of iron wood, has the blunt end split for a couple of inches. See Fig. 85B.

Technique. The longer stick is held between the big and second toes of either foot. The feet, which are kept apart a little wider than the proposed width of the sandal, rest on the outer sides with the soles turned inwards.

One end of the twisted bast, near the braid junction, is passed over the forward stick from the inner side of the right big toe, brought round the stick between it and the second toe, up over the dorsum of the big toe and round to the right between the big toe and the stick, as in Fig. 86. In the figure the reader is looking down at the sticks as if he were carrying out the technique. The twist round the big toe fixes one end of the twisted cord. The main part of the cord is now to the left. It is neatly coiled for ease in passing backwards and forwards.

The second stick is held parallel to the first and at the distance required by the length of the sandal. In Aitutaki the second stick was not sharpened and was held by the pressure of the feet near the two heels. In Rarotonga, a Mangaian, who demonstrated the technique, held the stick with his left hand.

From the right big toe the coil is taken back under the posterior stick. It passes up over it and goes forward to pass under the anterior stick, and then back under the posterior stick. Thus the coil passes round and round the two sticks, but always passes under and returns over each stick. The coil passes round the posterior stick fourteen times. From the method of passing the coil round the

Figure 86.
Sandal, setting up the longitudinal warps.

Figure 87.
Sandal, heel loop and first weft row.

page 96sticks, it will be seen that the lengths of the cord passing forward cross those coming back. After the 14 turns, the posterior stick lies above 14 alternate lengths of the cord and below 14 alternate lengths. We have thus 28 longitudinal lengths of the cord set up between two sticks in two sets of alternates, just as warps are set up for check weaving.

The next step is to provide the crossing elements which will correspond to the weft in weaving. The braid end of the cord that is twisted round the right big toe, is unwound and brought back under the front stick towards the posterior stick. On the left side the braid end of the cord forms the continuation of the last longitudinal length that passed back over the front stick. The posterior stick separates the two alternate sets of longitudinal lengths, which we shall now term warps. The posterior stick provides a shed through which the braided ends may cross from either side as wefts. The turns round the posterior stick define the back margin of the sandal. Before the regular crossings of the two braids take place, a heel loop has to be provided. This is provided by the right braid which, in Fig. 87, is denoted as A. The right braid, A, is passed across from right to left along the shed provided by the posterior stick. After emerging from the left side, it forms the loop, H, and re-entering on the right, again passes through to the left. The left braid, B, now passes along the same shed to emerge on the right.

The posterior stick is now removed. The forefingers from each side are run along the shed made by the anterior stick, and pulled back to push the crossings of the two sets of alternate warps hard up against the transverse braids, which form the first weft line which has taken the place

Figure 88.
Sandal, second weft row.

page 97formerly occupied by the posterior stick. The braid B from the right side, and A from the left, are now passed across through the new shed provided by the fingers in the last movement, Fig. 88a.

The fingers of both hands now work in from each side, the thumbs pressing down the warps that are above the crossed braids, A and B, and the fingers picking up the warps that are below them. This crosses the sets of alternate warps and provides a new shed for the wefts A and B. These are promptly passed through from either side as in Fig. 88b.

Again the upper series of warps is pushed down with the thumbs and the lower set raised on the forefingers to provide a fresh shed for the wefts A and B, which are passed through from either side and pulled taut. This process is continued as long as it is possible for the fingers to do it.

When the wefts A and B pass from the fifth to the sixth crossing, the short posterior stick is passed through the loops made by the weft braids at the side margins. It is left in this position to prevent the loops being drawn taut against the margins, Fig. 89. These two side loops are to be subsequently used in lacing the sandals.

Figure 89.
Sandal, side loops for lacing.

As the uncrossed parts of the longitudinal warps get shorter towards the toes, it becomes impossible for the fingers to separate the alternate sets. This is now done with the shorter stick, which is slipped out of the side loops. The point of the stick is pushed down between the first two warps on the right and passes under the first lower warp. This is levered up and the point passes over the next upper warp to pass down to lever up the next lower warp. In this manner by working across to the left, all the lower page 98warps are levered up and the others pushed down. The stick now occupies the correct shed for the next weft row.

The thin end of the braid on the right is now inserted into the cleft on the blunt end of the stick, which is also on the right, Fig. 90a. The stick is now pulled through to the

Figure 90.
Sandal, weft technique at toe end.

left and carries the braid with it. The braid A is detached from the stick and the slack hauled through to the left.

The left-hand braid weft B has to pass across the same course as A. This is very neatly done. The braid weft A, close to where it emerges on the left margin, has one of its three plies loosened and pulled up to leave an opening between it and the other two plies. Through this opening the fine end of the braid B is passed, as in Fig. 90b. The loop of the braid A on the right margin is then pulled and the part carrying the end of B is drawn through. When B emerges on the right margin it is detached from A and its slack pulled taut. A, which, owing to its length, still has its end on the left, is now pulled taut from the left and the weft line is completed.

The stick between the toes is removed. The last loops which were kept patent by that stick are each twisted with the fingers and placed over the point of the cord-carrying stick. The two weft braids are crossed in the manner described above, and the actual making of the sandal is thus completed.

All that is now necessary is to tighten up the work. Starting from the heel end, the braids A and B are drawn taut from the loops at the sides. Each row is done in turn, page 99the succeeding row drawing in the slack from the row above. By pulling the wefts simultaneously on each side, the warps are drawn tightly together to form a compact sole. The two side loops are left at a convenient size and the rows below them tightened until the whole sandal is firm and all slack is drawn out into the two long braids A and B, which emerge from either end of the toe margin. See Fig. 85, which shows the heel loops above, the smaller side loops, and the two braids below.

Wearing. Both the heel loop and the two front cords are called kaveivei tamaka and the two side loops, taringa (ears). In putting on the sandal, the two front cords are passed between the first and second and the fourth and fifth toes respectively. They cross the dorsum of the foot to pass through the heel loop, recross on the dorsum to pass through the side loops, and are then tied round the ankles.

The sandal may be reversed to distribute the wear over both surfaces. In fishing the sandals are often reversed to avert bad luck.

Remarks. The sandal technique is the technique of check weaving. The stick held between the toes is surely the most primitive beam over which the warp elements could be arranged. The posterior stick supplied the other part of the loom for stretching the warps taut. The fingers and later the posterior stick, took the place of the heddle in changing the alternating sets of warps to provide the shed for the crossing braids which were true wefts. The fingers again, by pushing up the weft rows to bring them close together, performed the functions of a comb. The work proceeded forwards or away from the body and again supplied one of the characteristics of loom weaving.

Comparisons with New Zealand. ³

Again, the severer climate of New Zealand effected drastic changes in the material culture of the Maori. The clothing that was quite sufficient and suitable in a tropical climate could not give satisfaction in the colder and wetter conditions of their new home. But this the earlier pioneers had to find out as the result of experience. Besides cultivable food plants, the colonising expedition of 1350 brought the paper mulberry plant under the name of aute; but the aute did not do well. Both it and other local bark substitutes were found unsuitable and were abandoned as clothing. For sentimental reasons, however, the aute was page 100still grown and used in making kites and for other minor purposes. The little bark cloth that was made, was evidently beaten in the Polynesian method, for two four-sided crubs with longitudinal grooves were dredged up from under 6 feet of gravel in the Whangarei Harbour. They were not made of iron wood, but one seemed to be of kauri branch and the other of manuka, both local woods.

On abandoning bark, the Maori found a more suitable substitute in the scutched fibre of the Phormium tenax. In some of the most primitive forms of rain cape, the flax was soaked for days in water to soften the inter-fibrous material, which was then easily scraped off with a shell. As the Aitutaki people first soak their bark, whether for cloth or ropes, in water, it seems reasonable that this was also the earliest stage in the Maori preparation of flax material. Later, as he got to understand his material and know where the most fibre was situated in the leaf, he improved his method to cutting across the back of the leaf, and either pulling the fibre away (takiri) or splitting off the back portion which contained little fibre (haro). The use of shells in the scraping and cleaning of fibre is universal throughout Polynesia.

The four-sided iron wood beater (ike) for felting the bast strips together on a wooden anvil (tutunga) gave way to the smoothly rounded stone beater (patu muka) for beating the flax fibre into softness on a large waterworn stone. But whereas the former was the actual manufacturing process, the latter was merely a stage in the preparation of material.

When the Maori was forced to abandon bast and use fibre he had to find a totally different technique for dealing with his new material. His inspiration could not come from the technique of bark cloth but he had to review the processes used in making clothing other than bark cloth. Judging from Aitutaki as a type of the islands whence he came, the only elements of technique that he could have brought with him were the three-ply braid of the Dracaena leaf kilt band, the attachment of hanging strips to one or two cords stretched between supports, and the single-pair twine of one or two rows in the hibiscus bast kilt.

At first, it is probable that for his hanging strips he used rough strips of flax fibre poorly scutched after soaking in water, such as are still used in the para type of rough rain cape. Later on, for improved cloaks, he subjected page 101the well scutched and pounded fibre to the hiro process of twisting into loose two-ply cords on the bare thigh. The name, however, changed from hiro to miro.

It seems likely also that he originally used the two-cord commencement for attaching his hanging strips, as the method has survived in a tussock rain cape found in a cave in the marginal area of Otago.

With the fibrous strands suspended, it required no great mental effort to continue with the single-pair twining technique, not for one or two rows as in the hibiscus bast kilt, but for as many rows as the type of garment required. Here was a simple technique for providing a close fabric. The cape is a kilt with more rows of single-pair twining and the cloak is a longer continuation of the cape.

It was found that the first twined row or weft line was quite sufficient for suspending the vertical elements. The two-cord commencement, being unnecessary, passed out of fashion in the more progressive technique of the North Island. The ends of the vertical strands were afterwards made into a waist or neck band, which varied in type with the particular garment. The commonest band in a kilt or rain cape is the three-ply braid which finds its prototype in the Dracaena leaf kilt common throughout Polynesia. Owing, however, to the combination with twined rows, the three-ply braid of the Maori garment is an end process.

The Aitutaki woman, in making the hibiscus bast kilt, tied the supporting cords for her commencement between any two supports whether trees, posts or stakes. The kilts were used only for dances and festivals. They did not take long to make once the material was prepared. The Maori woman though she abandoned the two-cord commencement, still had to stretch the two ends of her first twined row between supports in order that she could continue the other twined rows beneath. To her, the weaving of garments was no spasmodic effort, but a regular occupation that had taken the place of the bark cloth beating of the tropics. Each garment required many sittings and weaving took place within the house. Special supports were therefore needed that could be set up and taken down at will. Thus special pointed weaving sticks, with knobbed or carved upper ends, the turuturu, were made and became a necessary part of every woman's equipment.

The method of getting a black dye with mud is interesting owing to its wide distribution. Both the page 102Aitutakian and the Maori soaked their material in a bark infusion. In Aitutaki, the material was afterwards soaked in the mud of a taro swamp whilst, in New Zealand, the soaking took place in any swamp where the mud was particulary black. In the restricted area of Aitutaki, every wet or swampy part is converted into a taro swamp and assisted by irrigation. There are no swamps that are not taro swamps. In New Zealand the taro is grown in dry soil. There are no taro swamps, but plenty of other swamps. Therefore any swamp does that provides mud with an iron salt to combine with the tannic acid from the bark infusion. Tannate of iron is formed, which is the black dye in both cases.

From the simple beginnings we have indicated, was evolved an increasingly improved technique, that in the ramifications of its details became peculiar to New Zealand. Some other elements such as the method of attaching rain tags, may have come over with the simpler technique, but they are certainly not to be found in the Cook Group. Climate did not demand more than was there provided. Necessity forced the Maori to exploit the possibilities of flax fibre, and once he had started with the rudiments of a useful craft, he went on to improve his simple twining stroke to the two-pair interlocking weft, and he introduced various forms of ornamentation. Where his Island kinsmen expended their decorative efforts in dyeing and staining the available bark cloth with various designs, the Maori, using flax fibre, invented or adapted the wrapped twined stroke to taniko work to produce coloured geometric designs in the fabric and so gave expression to his innate artistic sense in the material with which a useful craft provided him.

Of head-dresses, the Maori seems to have forgotten any technique he may have had ere he left Hawaiki. The famed kura head-dress that was cast into the sea when the colonising canoes sighted the red flowers of the pohutukawa, is remembered in song and story, but no description of technique has been handed down. Wreaths, plaited fillet bands and feathers worn in the hair retained the name of pare. The wreath of green leaves became a sign of mourning and was worn only at funerals. The ornamental whalebone comb and individual feathers of the huia, the kotuku and the albatross became the insignia of chieftainship and the elaborate feather head-dresses of Polynesia page 103disappeared. Feather head-dresses on a woven foundation of flax fibre have in recent times been worn by an odd old man at Maori receptions to overseas visitors. These are really inverted feather baskets and are ridiculous from an ethnological point of view.

Sandals were worn in certain districts only. There is no affinity in technique to those of Aitutaki. The Maori, working so much with flax, naturally plaited them in that material with a check or twilled two stroke. The Cordyline australis was also used. A heel loop and side loops were provided and the method of tying with two long cords from the toe end was exactly similar to the Island method. The sandal was called paraerae and the Island name of tamaka was applied to a round cord of four strands corresponding to the taura puna.