Distribution

The greatest activities in stonework were directed toward the construction of the stone inclosures which served both religious and social purposes. These structures are termed maraes (Lamont's spelling [15], “mara,” is incorrect.)

The maraes are distributed over the islands. (See fig. 2.) Twenty-four were enumerated, including the abnormal Papa-o-Sokoau marae. A list of the sites, names, and condition of structures, arranged in order from Motuunga east, south, and west back to Omoka, is given in Table 9. The districts in which the maraes of the larger islands are located are given in parentheses.

Two of the 24 maraes listed were not seen, but it is said that the one on Niu-te-kainga is in good condition. Of the four completely destroyed, those at Omoka and Tautau are near the present villages of those names. The sites are known, but all the material has been used in the churches and graves; even the small karaea slabs have been used to mark the edges of the streets which run through the villages. The other two destroyed maraes on Motu-unga and Te Puka are near the two villages established by the missionaries but since abandoned owing to the depletion of the population by the Peruvian slavers. Here again the stone pillars of the maraes have been used for churches and graves. The marae at Kavea was partly destroyed when many of the pillar-stones were removed nearer the sea to form a stone pulpit from which one of the early native missionaries preached open air services. The stone pulpit has the form of a trilithon, with one pillar slab supported on the top of two others sunk in the ground. This postmissionary structure is pointed out lest it form a trap for subsequent investigators. Two other maraes have been used as general cemeteries, and certain other places used as cemeteries may originally have been maraes. In spite of the removal of pillars, the boundaries of 17 maraes were accurately mapped and measured with a chain tape.

The islands are almost flat. On their seaward sides the waves have piled up broken coral and coral boulders into raised banks. Of the 22 sites page 150 seen, 10 are located on the sea side of the islands, most of them on flat ground, as close as possible to the raised banks. These are sites which assure privacy. The marae of Hoenga-waka is on a small, semi-detached islet on the sea side of Tepetepe. The Arahura marae is on the very small island of Vaselu; it served as a second marae to the larger island division of Tautua. The large marae of Hangarei abuts upon the raised bank. On the other hand, 8 maraes are on the lagoon side, including the maraes of Tokerau and Motu-unga. The remaining 4 occupy a midway position between sea and lagoon. The levelness of the ground obviated any attempts at excavating or building up to obtain an even floor, except at Hangarei, where a slight slope led to the raising of one corner 12 inches above ground level.

Ground Plan

With the exceptions of the abnormal Papa-o-Sokoau and the Tongariro maraes, all the maraes have four almost straight sides meeting at angles approaching right angles. Four of the maraes have raised platforms at one end, where there are also higher upright pillars. In the maraes without platforms the end with higher uprights can always be distinguished.

In describing maraes the end with the highest uprights will be referred to as the back and the opposite end as the front; looking toward the back from the front, the sides are left and right; the distance between the two ends is the length, and that between the sides, the width.

Although it may be said that the mares were built on a rectangular plan, the opposite ends of all structures were not designed to be equal. Of 15 four-sided maraes that could be measured accurately, 7 are practically rectangular. In only 2 of them does the difference between opposite ends and sides reach 4 feet in a length of 80 feet. In some the difference amounts to inches only. In the remaining maraes the difference between opposite ends or sides is as great as 12 feet.

Most of the maraes are longer than they are wide. In only 4 out of 15 was the width greater than the length, 1 of the 4 being the large marae of Hangarei. The Arahura (Vaselu) and Te Tohi (Ruahara) maraes are almost square. In some others the difference between length and width is marked.

There is a fairly constant difference in width between the back and front. Leaving out 3 with equal ends, of the remaining 13 maraes 4 are wider at the front and 9 are wider at the back. In the 4 maraes with wider fronts the differences are 1, 2, 4, and 6 feet. The slight differences in the first three may be due to the inaccuracy attending sight planning without a measuring cord and not to deliberate intention. The difference of 6 feet between the front and back of the Hakataungari marae of Hakasusa, now used as a cemetery, may mean that the original boundaries have been altered. page 151 Of the 9 maraes with wider backs the well preserved maraes of Te Reinga, Hangarei, Mahue (Naue), and Rauhara (Naue) have differences of 9, 10, 10, and 21 feet respectively, which indicates that in some maraes the back was made wider deliberately.

It may therefore be stated that the ground plan of most Tongarevan maraes is roughly rectangular with the long axis extending between the front and back, thus conforming, though on a large scale, to the ground plan of the Tongarevan dwelling house. Associated with this is a widening of the back in relation to the front, which reaches its most marked departure from normal in the total increase of width over length.

Lamont (15, p. 111) has conveyed an erroneous impression of size in his description of the first marae that he saw.

Pushing on through the jungle of tall weeds, I suddenly came on an open space of some hundred yards square. It was encircled by tall, flat stones, some six feet in height, though generally much lower, but not more than a few inches in thickness; a sort of “Stonehenge” in a small way.

With no opportunity of verification, Smith (23, p. 91) naturally accepted Lamont's statement when writing on Tongareva, for he remarks: “The maraes, or sacred enclosures, some of which were as much as a hundred yards square…”

From the well known site of the wreck of the Chatham and Lamont's description of his movements it is absolutely certain that the marae that he described is the one named Rakahanga on Mangarongaro. K. P. Emory, who saw the marae in 1926, stated in conversation that he had come to the same conclusion. He gave the dimensions by pacing as 87 feet by 100 feet. My measurements (Table 10) with a chain tape are close to these figures.

The largest marae is Hangarei; the smallest is the well-made and well-preserved marae of Te Reinga. The narrow Hakataungari marae has been converted into a cemetery, and its boundaries may have been altered. The greatest length is the left side of the Rauhara marae, 112 feet. None of the maraes of Tongareva are as much as 100 yards square in area. Lamont's description of “some hundred yards square” must therefore be regarded as referring to the open space in which stood the marae and the tombs that he mentions.

Figure 20. Diagram of model Tongarevan marae: 1, left side; 2, right side; 3, back; 4, front; 5, pillars; 6, curb; 7, raised platform; 8, pavement; 9, house site; 10, small entrance inclosure; 11, small side inclosure.

A rough diagram of the Tongarevan type of marae is shown in figure 20. The two sides are approximately equal in length; the back is wider than the front; the boundaries are defined by limestone uprights; the space between the uprights is filled with a single row of flat coral pieces set on edge. At the back of the inclosure is a rectangular raised platform which stands against the back boundary but does not extend as far as the sides. A few feet in front of the platform there are traces of a disconnected stone pavement. Toward the front of some maraes are the boundary stones of small house sites. At the outer side of the middle of the front boundary there may be some low stones (karaea) set on edge as if to indicate an approach into the marae. (See pls. 5–7.)

Pillars

The upright pillars are cut from limestone strata, which have a natural horizontal cleavage. The strata average about 6 inches thick, but may be as thin as 4 inches and as thick as 10 inches. Some slabs detached by wave action were seen, but most of the marae uprights must have been quarried. Near Titikaveka in Rarotonga rectangular depressions in limestone mark the spot from which slabs were removed with metal tools for building a church. From a broken, free edge of the stratum grooves were cut down through the layer of limestone on either side and at the end, thus detaching a slab which was then levered up. By some such method, but with more primitive tools, the Tongarevans must have quarried the slabs for their marae uprights. The slabs were shaped by trimming the side and the top edges. The lower end which was embedded in the ground and the two flat surfaces of the uprights required no working. The trimmed edges are square and straight, though they must have been cut with shell adzes. Tupou Isaia said that pieces of coral heads such as were used in felling trees were used in trimming the edges of the uprights.

Figure 21. Simple marae pillars with clean-cut upper corners: a, pillar from Rakahanga marae; b, pillar from Kavea marae; c, d, pillars from Mahue marae. Position on marae not known for b; others located on back line.

The pillars vary considerably in size. The largest slabs were placed on the back line and the smallest along the front. Also, the maraes vary in the size of the slabs used. The largest pillars seen are on the dismantled marae at Kavea, and the smallest slabs at Hangarei.

The general shape of the pillars is rectangular but of the pillars accurately measured the widths at the top and bottom of only 9 are exactly the same. In 61 the width is greater at the top, and in 20 it is greater at the bottom. (See fig. 21.)

The thickness of the pillars is fairly constant on each marae, the material on each marae being obtained from the same stratum of limestone. In most maraes the pillars are 6.5 inches thick, but the range for the whole series is between 4 and 8 inches, with rare pillars 10 inches thick.

The width at the ground level is more than 50 per cent of the slabs measured averages between 2 feet and 2 feet 11 inches, with a higher percentage below the average than above. Of 155 slabs measured, three are less than 1 foot and only one is 4 feet wide at the ground level. The width at the top is also between 2 feet and 2 feet 11 inches in more than 50 per cent of the slabs measured for this feature. There is a higher percentage, however, above this average than below it, showing a marked tendency to make the slabs wider at the top. Of the series, two slabs are less than 1 foot wide, and one, which is 3 feet 9 inches wide at the base, is 4 feet 3 inches wide at the top. Top and base widths of more than 100 pillars, most of which are large, are compared as follows:

Of the bottom widths 1.9 per cent and of the top widths 2 per cent are less than 12 inches; 29 per cent of the bottom widths and 21 per cent of the top widths are from 12 to 23 inches; 55.5 per cent of the bottom widths and 52 per cent of the top widths are from 24 to 35 inches; 13 per cent of the bottom widths and 24 per cent of the top widths are from 36 to 47 inches; and 0.6 per cent of the bottom widths and 1.0 per cent of the top widths are more than 48 inches.

The tallest slabs are located along the back line, but tall slabs have also been erected on the side lines on the back half of the maraes. The shortest slabs are along the front line. The range in height on the four boundary lines is as follows:

The commonest height on the back line is between 4 and 5 feet, whereas on the other three lines it is between 2 and 3 feet. Of the 67 back uprights observed on 12 different maraes only 6 reached a height of 6 feet or more. On the Kavea marae, however, of 10 uprights which are still standing, two are 6 feet in height and no less than three are 7 feet. These are the tallest pillars seen on any marae but, owing to the dismantled state of Kavea marae, it was not possible to determine to which boundary lines they belonged. It is possible also that some of the largest and tallest pillars from the back lines of many of the maraes have been removed and used for graves. It is improbable, however, that any of the missing uprights were taller than those of Kavea. In the well-preserved Rakahanga marae the tallest back pillar is 5 feet 10 inches; on the perfect Te Reinga marae, 4 feet 7 inches; and on the almost perfect back line of Atutahi, 4 feet 6 inches.

In a number of pillars horizontal flanges were formed at the upper corners by cutting away the sides (fig. 22, a). The upper border is perfectly straight and continous with the flange projection which may be unilateral, but which on most pillars is bilateral. The outward projections range from 1 to 2 inches, though some reach 3 inches. Most of the projections are

Figure 22. Marae pillars. a, pillars with horizontal flanges: 1, 2, unilateral flanges; 3–6, bilateral flanges. 1, Rauhara pillar from back line; 2, Te Reinga, right line; 3, Rakahanga, front line; 4, Rakahanga, left line; 5, Rauhara, right line; 6, Te Tohi, back line. b, pillars with corner notches: 1–4, back line pillars; 5, pillar of left line. 1, Te Tohi pillar with vertical flange; 2, Te Reinga, even unilateral notch; 3, Te Reinga, deep unilateral notch; 4, Te Tohi, bilateral notches; 5, Rakahanga, shallow notch and flange. c, worked pillars: 1, Mahue pillar from back line, bilateral projections; 2, Rakahanga, left, two sloping shoulders; 3, Rakahanga, left, bilateral squared shoulders and bilateral horizontal flanges at shoulders and top; 4, Te Reinga, left, mesial curved notch; 5, Te Reinga, right, perforation evidently a natural flaw and not worked; 6, Te Hara, back, bilateral projections or deep flanges.

page 156 squared from the front, with a perpendicular measurement of from 1 to 3 inches, but in one pillar (fig. 22, a, 5) the perpendicular measurement is 6 inches. There are flanges on both short and tall pillars and on any boundary. One vertical flange was observed (fig. 22, b, 1).

The rectangular notch that cuts off an upper corner is another form of ornamentation (fig. 22, b). Most notches are wider than they are deep. Other methods of treatment are shown in figure 22, c.

The spacing of the pillars is irregular, and apparently no attempt was made to place them at equal distances apart or even on the same line. The pillars of most maraes are closer together on the back line than elsewhere, but in the Rakahanga marae they are slightly closer together on the sides and front. On the back line the average space between pillars is from 8 to 9 feet, but on Te Reinga marae it is 5.7 feet, due to the inclusion of a normal number of 9 pillars in the narrower width. On the Ruahara marae the pillars which remain in position are from 4 feet 7 inches to 5 feet 3 inches apart, and, as the marae is wide, there must have been originally 11 or 12 pillars. Of 5 maraes in which all the back pillars could be accounted for, 4 have 9 and one has 8 pillars. On the sides and front the spacing is still more irregular and ranges from 8 feet 8 inches to 19 feet. In the curved back of the Tongariro marae the spacing ranges from 13 feet 6 inches to 18 feet 4 inches. The spacing on the front line is generally wider than it is on the sides. Not only is there no symmetry in the arrangement of pillars on the significant back line, but even on maraes where there are 9 pillars the middle one is not exactly in the middle of the line. The presence of 8 pillars, so that there could be no middle pillar, as in the Rakahanga marae, also supports the postulation that the marae builders attached no significance to the middle position.

Curb Stones

The front and side boundaries of the marae inclosure marked by three lines of pillars are further defined by small oblong slabs of flat coral (karaea) which are imbedded in the ground in a single continuous row with their long axes horizontal. They fill the spaces between the pillars. The slabs average 6 inches in height, but they may be as high as 10 inches. Besides defining the boundaries, they act as curb stones to prevent the outward dispersal of the coral gravel with which the inclosure is covered, and thus correspond to the stones which define house walls.

For convenience, the lower line of stones will be referred to as the curb, and the coral slabs of karaea as curb stones, to distinguish them from the pillars.

Most back lines are destitute of curb stones. In the Rauhara marae (fig. 29) short curbs run in from either side as far as the back of the raised platform. The curb is present in the Arahura marae (fig. 26), but in the Mahue marae (fig. 28) the uninterrupted line of curb stones extends across the back, from 2 to 3 feet behind the back line of pillars. The curb is a constant feature on the front and sides, and where the pillars had been removed the presence of the curbed lines made it possible to ascertain accurately the dimensions of the marae. All the pillars except one have been removed from the Tokerau marae (fig. 23), but the back curbed line remains. The curb stones are continuous, except where short gaps in the line mark the sites from which pillars have been removed.

Platforms

There are raised platforms on four of the maraes, Tokerau (fig. 23), Rauhara (fig. 29), Hangarei (fig. 38), and Te Reinga (fig. 39). The platform is roughly rectangular with the back abutting against the back line of pillars and the two ends extending not as far as the sides of the marae. The platform consists of limestone slabs set on end to form an outer wall to an inclosure, which is filled in with coral boulders. The most perfect platform, at Te Reinga (fig. 39), is paved on its upper surface with small coral slabs laid flat and roughly fitted together. The height of different platforms ranges from 13 to 23 inches except in the Hangarei platform (fig. 38), the front wall of which is 3 feet high. The platforms are from 29 feet 7 inches to 55 feet wide, and from 4 feet 3 inches to 25 feet 9 inches deep.

Lamont (15, p. 122), in describing the ceremony at a marae in Mangarongaro, states: “The whole party then advances to an altar—a heap of rude stones … O Packa … ascended the altar, and, seating himself in front of a large stone …” The description definitely points to the existence of a stone platform, but no trace of such a structure is to be found on Rakahanga, which seems to have been the marae used on that occasion. Neither is there a trace of a platform on the nearby Te Vete marae (fig. 37), which might have been the marae referred to. The material of the stone platform has evidently been removed since Lamont's time, and this may have happened to other maraes.

Pavements and Marae Houses

Short, narrow pavements of small coral slabs laid flat were seen on 5 maraes, Te Tohi (fig. 24), Te Hara (fig. 25), Rauhara (fig. 29), Tongariro (fig. 34), and Te Reinga (fig. 39). Their presence at Rauhara and Te Reinga, where stone platforms exist, shows that they were accessory to the page 158 platform, though it is conceivable that in maraes which had no platform the pavement may have taken their place.

A variation of the stone pavement is the L shaped alignment of coral slabs set on edge in a single row, as seen on Mahue (fig. 28), Hoenga-waka (fig. 30), Te Rupe (fig. 32), and Nukurea (fig. 33) maraes. One limb of alignment is directed in the front to the back axis of the marae, and the other joins the back end, from which it is directed toward the left, except that in Nukurea it is directed toward the right. On the Mahue and Hoengawaka maraes the alignments are in pairs, and on Te Rupe and Nukurea they are single. It is probable that a pair was the usual number, and that the singles point to the removal of material. The careful clearing of all the maraes would probably reveal more of these accessory structures.

Upon the pavements or on the gravel beds defined by the L shaped alignments coconut offerings may have been laid, before the priest ascended the stone platform during the marae ceremony.

The defining curbs of small houses on the maraes were seen at Naue (fig. 28) and Te Vete (fig. 37). Though search was made without success for defining curbs on the other maraes, the overgrowth was so thick in some that they may have been overlooked. In others curbing may not have been used. Lamont (15, p. 162) mentions a small house on the Hangarei marae:

This, I was informed, was the island of “Hangary,” and belonged to our family. A considerable portion of it was occupied by a more extensive “mara” than any I had yet seen, though, from the number of weeds that filled the space and climbed round the huge grey stones, and also the condition of the house in its centre, which was mouldering to decay, it had evidently been long out of use. Anxious to see what the place contained, I was about to enter it, when violent screams of terror uttered by the boys arrested my steps, and I was obliged to proceed with them towards a point whence their cries had been answered.

The person who answered was the priest, Monitu, who, when told by the boys what Lamont had been about to do, at first looked incredulous and then laughed heartily. It is evident that the house was tapu. It is to be regretted that Lamont did not enter to find out what the house contained, but he clears up the matter in his account of a subsequent ceremony on the marae at Motu-unga in which he states (15, p. 180) that an old priest entered the “mara-house” and brought out a long stick with an immense bundle of feathers and other things tied at one end. This was the local god.

It is thus evident that the marae house was a fairly constant feature, that it was used to contain the material representation of the gods, and that it was tapu to those not belonging to the priesthood. It is difficult to see what other sacred objects could have been kept in the marae house, for the Tongarevans had no special religious regalia or such objects as the page 159 temple drums that were stored in the temple houses of Tahiti and Hawaii. Lament made no mention of a house on the Rakahanga marae in his detailed description of the ceremony that took place. The explanation is that on this particular marae the material forms of the gods were made from freshly cut coconut leaves which were afterwards discarded. There was thus no use for a marae house. It may, therefore, be inferred that the groups that had permanent forms for their gods in wood and feathers built houses on their maraes in which to store the tapu representations, whereas the groups that were content with temporary ones did not build marae houses.

In some maraes, for example Hoenga-waka (fig. 30) and Rakahanga (fig. 36), short alignments of coral slabs set on edge were placed outside the middle of the front line, and the space between was filled with clean coral gravel. From their position, they seem to indicate the correct approach into the marae inclosure. They suggest local attempts to embellish the maraes. In a few others which I did not record, these structures are on the outside of the side boundaries. They are spread with clean white coral gravel and might have been used in ceremony. Lamont's account (15, p. 121) of the manner in which coconut husk was placed at certain parts of the marae suggests that these small inclosures were used to receive the coconut husk offerings.

Individual Maraes. (See Table 9.)

1. Kirihuri marae in Motu-unga, on the lagoon side of the island in the vicinity of the ruins of a stone church.

Because of the removal of both pillars and curb stones for a church and graves the original marae lines can not be distinguished. Close to the church is the mound that formed the turtle oven connected with the marae. On top of the mound is a hollow 22 feet in diameter, but the hollow may have been increased by the recent uprooting of a large hala (Pandanus; Tongarevan, hara) tree that had been planted on top of the mound by one of the early missionaries. The presence of the turtle oven indicates that Kirihuri was the marae which Lamont (15, p. 182) mentions in connection with the cooking of a turtle on an elevation of stones. Lamont states that this marae was different from the one at which he had helped to officiate on the previous day, but, though my informants knew of only one marae on Motu-unga, another may have existed. The Kirihuri marae, said to have been established by the ancestor Taruia, is reported to have been large, well kept, and of higher status than the well-known marae on the neighboring island of Tokerau.

2. Tokerau marae in the Rangiriri Division of Tokerau, is on the lagoon side of the island on dry raised ground, free of shrubs and weeds (fig. 23).

All the pillars except one have been removed to form a wharf. The curb stones, however, are in position and plainly indicate the boundaries of the marae. The inclosure is almost rectangular, with but a difference of 2 feet between the narrower back and the wider front. The long axis is between the ends and runs due north and page 160 south, with the back toward the lagoon on the south. The distinguishing feature of the marae is the raised platform, which has four walls of short limestone slabs set upright to form a quadrangular inclosure. The back line is continuous with the back line of the main inclosure and has one upright pillar. The interior is filled with lumps of coral which do not rise to the top of the inclosing walls, some of the stones evidently having been removed. The Tokerau marae is said to have been one of the most important ceremonial maraes of Tongarega.

Figure 23. Tokerau marae: a, general plan; b, raised platform. 1, curbed back line continuous with back wall of raised platform; 2, standing pillar in back wall of platform; 3, left curbed line showing gaps, pillars removed; 4, right curbed line with gaps; 5, front curbed line with gaps; 6, raised platform; 7, back wall of raised platform; 8, limestone slab in front wall, 4 feet 3 inches wide and 1 foot 7 inches high; 9, limestone slab, 3 feet 7 inches wide, 1 foot 1 inch high; 10, limestone slab, 3 feet 8 inches wide, 1 foot 2 inches high; other wall slabs narrower. Platform does not occupy exact middle position on back line.

3. Nui-te-kainga marae on the island of Nui-te-kainga on the north side of the lagoon was not seen. It is reported to be well preserved and much like the others.

4. Te Tohi marae in the Punua division of Ruahara, on the sea side of the island close to the raised bank, with the back toward the sea (fig. 24).

The long axis of the marae is between the ends. The sides are approximately equal in length, but the back is 4 feet 3 inches wider than the front. The back line has 6 pillars standing, 5 of which are flanged or notched. Another flanged pillar is broken. Standing pillars are fairly close together. Long intervals indicate that at least 4 have been removed, making about 11 pillars on the back line. The two end pillars abut against the side lines, and there are no curb stones between the pillars. The front line is marked by curb stones, and wide, short pillars are standing. The right side is curbed and has three pillars standing, two of which have unilateral horizontal flanges. One pillar is broken off and two others are missing, making a total of six pillars for the right side. The left side line is obscured by a thick growth of ngoso shrubs. The pillars have been removed, but the line is curbed. Toward the back there is a short, narrow pavement of flat karaea slabs, and near the front on the left are three coral slabs set on edge to form an open rectangle. The inclosed space is page 161 covered with white coral gravel. It looks like a fireplace, but there is no evidence of charcoal or burned coral in this or similar inclosures. Offerings in connection with the religious ritual were probably laid upon them.

5. Sivalau marae on Ruahara. The second marae in the Sivalau division on Ruahara was not examined.

Figure 24. Te Tohi marae: 1–7, pillars on back line; 8–10, pillars on right side line, average space between pillars on right when all standing, about 10 feet; 11–13, low pillars on front line; 14, short pavement; 15, open rectangle of two side slabs of coral 10 inches long and end piece 12 inches long set on edge and projecting upward about 10 inches. Pillars 1–11 range in height from 2 to 6 feet, in bottom width from 1 foot 8 inches to 2 feet 10 inches, in top width from 1 foot 8 inches to 3 feet 5 inches. Average space between pillars 2–5, 5 feet 1 inch. Pillar 2 shown in figure 22, a, 6; pillar 7, figure 22, b, 4.

6. Te Hara-taurekareka marae on Rukutia, on the sea side of the island with the back to the sea, close to the raised shore, where rocks have been heaped up close to the back but distinct from the boundary of the marae (fig. 25).

The full name of the marae is Te Hara-taurekareka-te-sau-a-tonga, which means “The beautiful hala growing in the south.” The marae is said to have been built by page 162 the ancestor Turua. The sides of the marae are approximately equal, with but 1 foot different in length. The long axis is transverse, with the back nearly 5 feet narrower than the front. The pillars have been taken away from the front and sides to form a wharf, but two pillars remain standing on the back line and two others have been broken. Two fallen pillars are near the right side. A small pavement is located about 9 feet from the back line and to the left of the middle line. On the right of the pavement stands a short pillar. The sides and front are curbed, but no curb stones have been laid between the pillars on the back line.

Figure 25. Te Hara-taurekareka marae: 1–5, pillars on back line; 6, pavement; 7, short pillar on right of pavement; 8, piled up rocks clear of back line. 1–3, broken pillars; 4, pillar shown in figure 22, c, 6; 5, pillar with horizontal flange on one side. Spacing between pillars 1–4 ranges from 8 feet 6 inches to 10 feet 10 inches. Three pillars missing from back line.

7. Arahura marae, on the sea side of Vaselu, a small island planted with coconuts and lying north of Tautua, from which it is separated by a shallow channel (fig. 26).

The marae is almost square, with a slightly longer transverse axis. It is unique in having a small quadrangular inclosure projecting back mesially from the main back boundary, which is toward the sea. One of the old men of Tautua termed the small inclosure the raukava and stated that it was the part of the marae occupied by the au mana, or people with authority, namely, the priests (tohunga). Pillars had been spaced along the boundaries, but some had been removed for graves, which were located within the marae. The remaining pillars are low. The boundaries are curbed.

8. Papaki-reia marae, Tautua, is now occupied by the village church and cemetery of Tautua. Not even a curb stone of the original structure remains in position. The marae followed the fate of those of Omoka, Motu- page 163 unga, and Te Puka, where close proximity to a Christianized settlement led to the use of the limestone and coral material for churches, graves, and the curbing of village roads.

Figure 26. Arahura marae: 1, back quadrangle; 2, pillars on each side of quadrangle at back line junction; 3–7, pillars on left side; 8–10, pillars on right side: 11–13, pillars on front line; 14, paved path; 15, graves.

9. Mahora-kura marae on Kavea, a short distance inland from the lagoon side of the island, the waterfront of which is built up with stone to form a wharf (fig. 27).

Figure 27. Mahora-kura marae: 1, pillar set obliquely on left; 2–5, pillars evidently forming one boundary line; 6, 7, pillars forming another line at right angles to previous one; 8, 9, two pillars forming another line; 10, pillar lying on ground; 11, house foundations of flat coral slabs set on edge and 6 inches above ground (left wall of two slabs each 2.5 feet long, gap of 1.5 feet; right wall of two slabs, 3.5 and 2 feet in length, 1 foot gap; back wall of two slabs, 2.5 and 3 feet in length, 2 foot gap on right); 12, pillar 2 feet wide by 4 feet high on open side of house foundations. Pillars 1, 2, 3, 6, 7, 9, erect; the others down.

Mahora-kura must have been an imposing structure before it was dismantled. From this marae pillars were taken to construct a stone pulpit on the sea side of the island, where open air services were held after the advent of Christianity. The remaining pillars are large and massive and are of greater average size than those on any of the other maraes. Part of one boundary line can be distinguished, but the four erect pillars seem to be out of alignment. The foundations of a house that must have been within the marae precincts are present. There are no curb stones.

10. Mahue marae, Haitaki division of Naue, on the sea side of the island with its back toward the sea (fig. 28).

The boundary lines of Mahue marae are unequal, the longest being the right side, and the shortest, the front. The back is longer than the left side and 10 feet 4 inches wider than the front. The pillars are in position, 7 on the left, 6 on the right, 5 on the front, and 9 on the back. Curb stones mark the places between the pillars on the sides and the front. A back line of curb stones is separated by a distinct space from the alignment of the back pillars. Two L shaped arrangements of stones are on either side of the middle line toward the back of the inclosure. Two of the back pillars have bilateral horizontal flanges, and one has an upper corner notch on the right.

Figure 28. Mahue marae: 1–9, pillars of back line; 10–16, pillars on left side; 17–22, pillars on right side; 23–27, pillars on front line; 28, two L shaped alignments with transverse limbs of two coral slabs and longitudinal limbs of two slabs on the left and three on the right, set on edge. Pillars 3 and 5 bilaterally flanged; pillar 8 notched. Height of pillars 1–9 ranges from 3 feet to 5 feet 2 inches; bottom width, 1 foot 10 inches to 3 feet 3 inches; top width, 2 feet 2 inches to 3 feet 10 inches. Space between pillars 1–9 ranges from 6 feet 9 inches to 9 feet 4 inches. Back curbed line is 2 feet 2 inches behind pillar line on the left and 3 feet on the right. Pillar 6 shown in figure 21, c; pillar 7, figure 21, d.

Figure 29. Rauhara marae. a, general plan: 1–7, pillars on back line (1, tallest of series, 6 feet 7 inches high, flange on right; 2, pillar with flange on left, 6 feet 4 inches high, set at back of platform (12); 3, 4, broken pillars at back of platform; 5, broken pillar included in back wall of platform; 8, double flanged pillar on right side; 9, pillar included in right wall of platform; 10, back line of curbstones extending from left corner, passing 2 feet behind corner pillar (1) and 2 feet 4 inches behind platform (12), and turning in at right angles on right to meet back wall of platform; 11, back line of curbstones commencing 1 foot 3 inches behind right back corner of platform and continuing to back of left edge of right corner pillar; 12, platform formed of four walls of limestone slabs set vertically (height of walls, 1 foot 7 inches, on back, 1 foot 6 inches on front), inclosure filled in with coral boulders but not quite to top, giving impression as in Tokerau marae that some of material has been removed; 13, short curb line in front of pillar completing an inclosure floored with white coral gravel, thus raising it above surface of marae: 14, small inclosure; 15, paved open inclosure defined on right by curb extending from back curb line to left back corner of platform and covered with white coral gravel; 16, narrow transverse pavement with left end in line with left end of platform; 17, rows of two curbstones set on edge; 18, angle on left boundary; line runs to right of angle for 2 feet 7 inches and turns at obtuse angle to run forward to meet front line at left corner; front boundary curbed, with no spaces for pillars. b, inclosure (a, 15), enlarged scale, formed by back curb line (10) and front and side lines of curbstones set on edge: 2, left side, 3 feet 4 inches long; 3, right side, 4 feet long; 4, two large curbstones set parallel with left side; 5, two large curbstones set parallel with right side; 6, two curbstones, closing front of inner inclosure, open at back; 7, skull and bones; 8, skull and thigh bones.

11. Rauhara marae, Ivirau district of Naue, on the sea side of the island with its back to the sea (fig. 29).

The long axis lies between front and back. The sides are uneven in length, and the back is 21 feet 4 inches wider than the front. In the uneven length of all four boundaries Rauhara resembles the neighboring Mahue marae. It is like the Tokerau page 166 marae in that there is a raised platform at the back. The platform inclosure is not quite filled to the top with coral stones, giving the impression as in the Tokerau marae that some of the material has been removed. On the left of the platform is a small inclosure containing human bones and skulls and covered with white coral gravel. The pillars at the back are in line with the back of the platform. Two pillars have single horizontal flanges, 2 have the ordinary squared ends, and 3 are broken. The uneven spacing leads to the conclusion that 1 pillar is missing, which makes a total of 8 pillars at the back. A back line of curb stones lies outside the pillar line at both ends, but evidently the intermediate part at the back of the platform has never been filled in. All pillars on the left side and all but a double flanged pillar on the right have been removed. The front line is formed of higher curb slabs which form a continuous line throughout without any spaces for pillars, which were apparently not used on this line. A narrow transverse pavement is situated in front of the platform on its left.

12. Hoenga-waka marae, Tepetepe, district of Hoenga-waka, on the sea side of Tepetepe, separated from it by a depression which may be covered at high tide (fig. 30).

Figure 30. Hoenga-waka marae. 1–9, pillars of back line: 3, pillar with double flange; 6, tallest pillar; 7, space occupied by missing pillar; average height of pillars, 5 feet 2 inches; pillar 2 is 9 inches thick, and others, 6 inches. 10–13, left side pillars: 10, 11, down; 12, 13, standing. 14–19, right side pillars: 15, down; others standing. 20–26, front line low pillars: 21, down. 27, L shaped curbed lines with a short limb 2 feet long extending to left from longitudinal limb. 28, small inclosure, open on left, 2 feet 5 inches wide and 2 feet deep. 29, right L shaped lines with short transverse limb 2 feet long. 30, three short lines of curbstones on outer side of front line: middle line shorter; fine gravel spread between lines.

The back of the marae is to the sea. The long axis is from front to back and the back is wider than the front. The sides are also uneven with the right side longer than the left. The back line is composed of 9 pillars, 8 still standing, and 1 missing. No curb stones are present on the back line. The spaces between the pillars range from 7 feet to 9 feet 7 inches and average 8 feet. The right side has 5 pillars standing and 1 down, and a space of 36 feet is without pillars. Judging from the average spacing, probably 3 pillars have been removed, which would make a total of 9 pillars for the right side. The left side has 2 pillars standing and 2 down; the others have been removed. The front line has 6 pillars standing and 1 down, making a total of 7. No raised platform is present, but there are two L shaped lines near the back, similar to those in the Mahue marae. Curb stones extend along the sides and front, and three short lines of curb stones form an approach to the front line.

13. Te Puka-nui marae, Te Puka, on the sea side of the island with its back toward the sea (fig. 31).

This marae, though its name signifies “big Te Puka,” is a small structure. Its long axis is between front and back. The sides are equal, but the back is wider than the front. All the pillars of the back, front, and sides have been removed to form graves which are located both within and outside the inclosure. The curb stones forming the front and side lines are in position. A few thick limestone slabs about 10 inches high are left on the back line.

Figure 31. Te Puka-nui marae: 1, back line; 2, left side; 3, right side; 4, front line; 5, 6, graves within the marae; 7, 8, graves outside marae.

14. Punaruku marae, Te Puka. The site is marked by the ruins of the church and the cemetery close to the lagoon edge of the island near two small fish ponds.

Tradition stated that Punaruku marae was built by the ancestor, Mahuta, who came from Tahiti in his voyaging canoe, Waimea, and settled down at Te Puka. A village established by the missionaries on Te Puka included the ancient marae within its boundaries. As a result, the marae was completely demolished, as were the maraes of the villages at Omoka, Motu-unga, and Tautua. The pillars of the marae were page 168 used for graves and were also incorporated in the walls of the Christian church. Even the curb stones of the marae were removed and used to define the edges of the village road. Hence, no original stone remains in position to mark the boundaries. My informants stated that the marae had its long axis running east and west and that the highest stones marking the back line were on the lagoon side, as they are in the Tokerau marae.

The Punaruku marae was described by Lamont (15, p. 235) as follows:

The following day …. I was shown the mara, celebrated throughout the group for its extent, the size of the stones, and for some peculiar religious qualities. In the centre were several tombs of great Tepuka chiefs, long since called to their fathers. The large stones forming these structures would not have made contemptible monuments for some of our own illustrious dead. One in particular was pointed out, supposed, as well as I could understand, to be that of the founder of their race, the original Mahuta, who came here with his wife Ocura, bringing in his great canoe cocoa-nuts and other plants for the earth, fish for the sea, and birds for the air. As he is universally admitted to have landed in Tepuka, we may infer that he came from the southern islands and not from the land beyond the sky, as the ancient faith of the islanders would lead us to believe.

Lamont's inference about “the ancient faith of the islands” shows lack of understanding. The human history of Mahuta is quite well known to his descendants, and “the land beyond the sky” is the land beyond the horizon.

Figure 32. Te Rupe-tangi-rekareka marae. 1–9, pillars of back line: 2, 3, pillars with double flanges; 3, tallest pillar, 4 feet 6 inches high and 3 feet 2 inches wide at bottom; 8, space of missing pillar; average space between back pillars, 7 feet 9 inches. 10–12, three remaining pillars on left side. 13–17, four standing pillars on right side: 17, pillar broken; spaces between pillars range from 10 feet 9 inches to 19 feet. 18–22, four standing pillars on front line: 20, pillar broken; spaces between pillars range from 10 feet 10 inches to 16 feet 3 inches. 23, L shaped figure of single rows of coral slabs with short limb 2 feet long.

15. Te Rupe-tangi-rekareka marae, Atutahi, on the seaward side of the island, with the back line close in under the raised sea bank (fig. 32).

The meaning of the marae name is “the pigeon that cooes sweetly.” The long axis is from side to side, with almost equal sides and the back slightly wider. No curb stones were found between the pillars, and as no coral slabs were seen, the marae boundaries seem not to have been defined by the usual lower lines of curb stones. The back line had, originally, 9 pillars. The 2 end pillars abut against the sidelines. Each of the other three sides has 5 pillars. The row of pillars forming the front line is curved. Near the back line is the L shaped formation of low coral slabs set on edge. The back line and the right side are overgrown with ngoso.

16. Nukurea marae, Vaiari, on the lagoon side of the small island with its back toward the west and not toward the main lagoon (fig. 33).

The long axis is from front to back. The marae boundaries are much overgrown, and the positions of the pillars were only approximately ascertained. Near the back some lines of low coral slabs have been set on edge.

Figure 33. Nukurea marae. 1–4, remaining pillars of back line: pillars between 1 and 2 removed, leaving space of 31 feet 10 inches; 2, tallest pillar, 5 feet 2 inches high and 3 feet 3 inches wide at the bottom, with a notch 10 inches wide and 7 inches deep on top left corner; flanged pillars 3 and 4, down. 5–9, standing pillars on left side with one missing between 5 and 6. 10–13, standing pillars on right side: 11, pillar with double flange. 14–18, standing pillars on front line: 14, pillar with double flange; 17, pillar with single flange. 19, single line of curbstones. 20, longitudinal line of curbstones with two short limbs extending to right at right angles.

17. Tongariro marae, Hakasusa, in the Hakasusa division of the largest island, on the lagoon side, with its back to the lagoon (fig. 34).

Tongariro marae shows a marked departure in ground plan, the back and front boundaries being curved. The back is much narrower than the front. The pillars on the back boundary are widely spaced and, with the exception of three, are not high. page 170 None of the pillars is flanged or notched. On the sides the pillars are low and irregularly spaced. Some have probably been removed. The poor supply of limestone in the immediate neighborhood is reflected in the use of low pillars and of coral slabs on the front boundary. The marae inclosure is in a hollow with the back sloping upward. Near the back boundary is a short line of coral slabs set on edge with a limestone pillar set upright on its right. Across the middle longitudinal line and about 24 feet from the back line is a narrow pavement. Two low pillars are also set upright within the inclosure near the right side and toward the front. Four Triton shells were found at the base of one of the front pillars. The marae is said to have been built by one of the brothers, Oriaitu or Umutoro, who lived eight generations back from 1900. A song used in connection with this marae is given on page 216. An informant stated that there was a marae named Tangaroa in the Takurua district of Hakasusa. There may have been a confusion in names between Tongariro and Tangaroa.

Figure 34. Tongariro marae. 1–7, standing pillars on curved back line: excluding pillar 2, the six pillars average 4 feet 2 inches in height; 7, tallest pillar, 5 feet high. 8–11, standing pillars on left side. 12–14, short pillars on right side. 15–18, short pillars on curved front line, Triton shells lying at the base of pillar 16. 19, line of low coral slabs. 20, pillar on right of coral slabs. 21, transverse pavement of double row of coral slabs laid flat. 22, low pillars set upright within marae.

18. Hakataungari marae, Hohonu district of Hakasusa, on the lagoon side of the island, with its widest end toward the lagoon (fig. 35).

The marae has been converted into a cemetery and the pillars have been removed, leaving only the curb stones to delineate the boundaries. It is said to have been constructed by Turua, who flourished seven generations back from 1900. Another version of the legend states that it was made by Pokaipuni, a contemporary of Turua. Within the marae are stones marking graves. One grave has an erect, double-flanged pillar.

Figure 35. Hakataungari marae, all pillars removed: 1, wider, lagoon end, probably the back; 2, grave with upright pillar 4 feet high, 3 feet 3 inches wide at bottom, 3 feet 7 inches wide near top, horizontal flange at each upper corner; 3, grave, without pillar.

19. Rakahanga marae, Mangarongaro, in about the middle of the Mangarongaro district near Te Toto and not far from the reef where the Chatham was wrecked, but nearer the lagoon side of the island (fig. 36).

The Rakahanga marae is without doubt the first one that Lamont (12, p. 111) saw and described as a sort of “Stonehenge.” Some doubt was expressed by my informants as to whether the marae was named Rakahanga or Awanui. Its long axis is between front and back, at right angles to the lagoon shore. The back, which lies toward the sea, is slightly wider than the front, and the sides are almost equal. The marae was mapped out by K. P. Emory during his visit in 1924. The back, which corresponds to Emory's “western side,” contains, as he observed, the largest pillars, which number 8. The total number of pillars on the left is 9; on the right, 9; and on the front, 7. The right side pillars are of simple rectangular shape, but four left side pillars are flanged and notched, and one front pillar has a double flange. Curbing between the pillars is present on the front and sides but absent on the back line. There are no indications of a pavement, platform, or house site.

20. Te Vete marae, Mangarongaro, near the lagoon, with its long axis at right angles to the lagoon shore and its back toward the sea (fig. 37).

The pillars of the back line have been removed for the innumerable graves that are dotted about. The marae has equal ends and equal sides. The broken lines of curb stones indicate the boundaries clearly. There are foundations of a house within the inclosure, and as these are nearer the lagoon end, it is presumed that the back of the marae was toward the sea.

Figure 36. Rakahanga marae: 1–8, pillars of back line, four broken and four upright; 9–17, pillars of left line, four flanged and notched and 3 fallen; 18–24, pillars of right line, seven recctangular and two missing; 25–31, low pillars of front line, one double-flanged; 32, four coral slabs set on edge. Pillars 9 (fig. 22, c, 2), 10 (fig. 22, a, 4), 11 (fig. 22, c, 3), and 13 (fig. 22, b, 5) flanged and notched.

Figure 37. Te Vete marae, equal ends and equal sides: 1, back line with pillars removed; 2–4, pillars standing on left side; 5, 6, remaining pillars left on right side, 5 feet 4 inches and 5 feet 3 inches high respectively; 7, front line, no pillars; 8, house site defined by curbstones and floored with coral gravel.

21. Hangarei marae, Hangarei (fig. 38), on the seaward side of the island, with its back resting against the raised bank close to the seashore.

The long axis is transverse, and the back wider than the front. The back boundary of the inclosure proper is irregular, formed of three interrupted sections which come forward toward the right, making the right side shorter than the left. A raised stone platform extends back from the middle part of the back line, and its height above the marae floor rises with the ground at the back. Inclosures defined by coral stones set in the ground are situated on either side of the stone platform, but they do not extend outward as far as the side boundaries. The middle and right of the back boundary are formed by upright blocks of limestone which set off the lower level of the marae floor from the raised ground at the back. Two tall pillars are located on the short left section of the back boundary, but the rest of the line has no pillars. The left side is poorly defined toward the back, but toward the front half it is well curbed and has four pillars. The right side is well marked throughout by curb stones and carries 11 short pillars, 3 of which are broken. The front line is well curbed with higher stones at each end. Five short pillars remain in position, two of which are broken. A depression in the ground on the front right corner has been filled in with stones to raise the level to 12 inches above the outside surface. Coral gravel is laid over the stones. About 13 feet outside of the front line a single row of stones is set in the ground parallel with the front line, and from its left end another row extends away from the marae at right angles. They form the boundaries of a path leading to a turtle oven which is 39 feet from the left half of the marae front line. The Hangarei marae is said to have been built by Amosia, 13 generations back from 1900.

22. Te Reinga marae, Motukohiti, at Ara-a-hupu, not far from the boundary between Motukohiti and Omoka near the lagoon side of the island, with its long axis parallel with the lagoon shore and its back to the north (fig. 39).

Te Reinga has a perfectly preserved raised platform, and though small, it is the best preserved of the Tongarevan maraes. The longer sides are nearly equal and the back is wider than the front. The back line contains 9 pillars, 2 of which are notched. The left side has 8 pillars, 2 being broken. There is a curved notch in the middle of the upper edge of the pillar nearest the back corner. The right side has 7 pillars, 1 a single horizontal flange, 1 perforated, 1 broken, and 1 missing. The raised platform is roughly rectangular. The walls are formed of large, irregularly shaped slabs of coral. The inclosure so formed is filled with pieces of coral to the top of the walls to form a level surface which is paved with flat pieces of coral fitted together. In front of the platform are two narrow transverse pavements bounded on the right by low coral slabs set on edge. The front line of pillars contains four low pillars widely spaced, and to the outer side of the middle four short lines of coral slabs are set on edge with coral gravel spread between them. Te Reinga is said to have been built by Ponakino, a son of Matakunui, who flourished 14 generations ago.

23. Marae of refuge: Te Papa-o-Sokoau marae, Motukohiti, close to the Omoka boundary and about equidistant from sea and lagoon (fig. 40).

The marae consists of a raised stone platform ranging in height from 3 feet 9 inches to 4 feet 6 inches and forms an imposing structure. The ground plan is irregular and roughly resembles a headless human body. The walls defining the boundaries of the figure are formed of large coral blocks set on end and firmly imbedded in the ground. The interior is filled in with coral boulders to make an even surface which page 174

Figure 38. Hangarei marae. 1–3 sections of interrupted back line: 1, left section; 2, middle section defined by limestone slabs which in middle part of section form front wall of raised platform (6); 3, right section defined by low limestone slabs. 4, pillar 6 feet 5 inches high, 2 feet 3 inches wide at bottom, and 2 feet 10 inches wide at top. 5, pillar broken, 6 feet 9 inches high and 10 inches thick. 6, raised stone platform, differing from platforms in Tokerau, Rauhara, and Te Reinga in lying outside of back line (front boundary formed of coral boulders built up above the back line (2) of limestone slabs, 3 feet above marae floor; coral boulders forming platform somewhat scattered, due to operations of seekers after land crabs that seek refuge under stones). 7, left back inclosure above level of marae floor defined on left by single line of low stones, with similar line at back and no traces of house foundations. 8, right inclosure above level of marae floor, no traces of house foundations. 9, gap. 10, left side with few small pillars standing. 11, broken pillar 3 feet 4 inches wide and 2 feet high. 12, right side, number of small pillars standing, height ranging from 18 to 26 inches except broken pillar (13). 13, pillar 3 feet 6 inches high. 14, front line, three standing pillars and two broken. 15, left corner with higher coral slabs. 16, right corner raised and filled in to height of 12 inches to level marae floor. 17, line of stones. 18, line of stones. 19, turtle oven on mound of accumulated discarded coral used in cooking, 130 feet in circumference and about 6 feet high with hollow at top containing charcoal and ashes from cooking fires.

page 175

Figure 39. Te Reinga marae: 1–9, pillars on back line, average space between pillars 5 feet 6 inches; 10–17, pillars on left side, average space between pillars 6 feet 6 inches; 18–24, pillars on right, pillar 19 missing, pillar 22 broken; 25–28, pillars of front line, very low, one being 11 inches high; 29, four rows of coral slabs set on edge at right angles to curb; 30, platform (height ranges from 22 inches on left to 23 inches opposite pillar 6 and 20 inches on right wall); 31, left pavement; 32, right pavement. Pillar 7 is notched on right (fig. 22, b, 3); pillar 9, notched on left (fig. 22, b, 2); pillar 10, curved median notch (fig. 22, c, 4); pillar 21, perforated (fig. 22, c, 5); pillar 24, simple flange.

is covered with a layer of coral gravel. Long recesses are formed between what correspond to the arms and the body, and another shorter recess lies between the short legs. There are two other recesses on the body, that on the right being connected externally with the recess between the right arm and body. Upon the surface of the platform seven rectangular inclosures are defined by coral slabs set on edge. The seven rectangular spaces, which resemble grave inclosures, contain white coral gravel in shallow layers that rest on the coral boulder filling of the marae. The gravel was turned over and searched, but no bones or other objects were present in any of the inclosures. The larger inclosures may have been used for the exposure of dead bodies which were afterward removed.

This structure shows such a marked departure in ground plan and construction that it would not have been classed as a marae were it not for the assurance of Pa, who is not only the oldest man on the atoll, but owns the land surrounding the page 176 structure. Pa stated that the marae was not used for the same purposes as other maraes but corresponded to a city of refuge. Any person who committed a crime and was in danger of losing his life was immune from attack if he gained the marae and remained there. The marae was named after Sokoau, a daughter of Tangaroa. Sokoau was killed by her husband, Tonu, for infidelity. The husband cut up her body and distributed the pieces among his people, who consumed them. To avenge her death, Sokoau's two brothers killed Tonu. According to Pa, this is the only record of cannabilism in Tongarevan tradition. The horror with which it was regarded led to the application of the name, Sokoau, to the refuge marae to remind avengers of her story and of the disasters which followed the gratification of the blood lust for vengeance. The term papa refers to the platform of stone, but Papa-o-Sokoau may be freely translated, the “Refuge of Sokoau.”

Figure 40. Marae of refuge, Te Papa-o-Sokoau. a, ground plan: 1, body 2, right arm; 3, left arm; 4, right leg; 5, left leg; 6, longitudinal distance to point 4, 47 feet 6 inches; 7, 7 points between which transverse dimensions are 40 feet 8 inches; 8, recess; 9, recess, communicates with recess on its outer side; 10, recess closed on all sides; 11–17, small inclosures; 18, highest point of marae, 4 feet 6 inches from ground; 19, point at which height is 3 feet 11 inches; height gradually decreases to 3 feet at lower end. b, outer wall facing front at point a, 18: 1, slab, 21 inches high; 2, slab, 22 inches high; 3, layers of slabs loosely fitted on flat uprights.

24. Omoka marae, Omoka, on the site now occupied by a church. The marae has been dismantled—the fourth marae completely destroyed by the modern villages. The curbstones marking the village road and the limestone slabs in the cemetery were taken from it. Lamont (15, p. 175) was present at a ceremony conducted at this marae.

Another marae named Saeha (rage, jealousy) was said to be situated on the Omoka side of the boundary of Motukohiti, and yet another near the southern point of Motukohiti at Parahatea.

Function Of The Marae

The marae was a social as well as a religious structure. It is traditionally recorded that each of the voyagers, Taruia and Mahuta, built maraes on his arrival in Tongareva. Subsequent mares are attributed to Turua, Umutoru, and other chiefs. When independent communities developed it became necessary for them to have community maraes. The chief of the community initiated the movement, as he had the control of community labor and could direct the preparation and transport of material to the selected site. The builder of a marae not only served a community need, but he added to his individual prestige and left a monument to his achievement. Though the community did the work, the building of the marae was associated in the traditional records with the chief of the community in a phrase such as Na Turua i po te marae (Turua constructed the marae). The word po is the technical term used to denote the digging of holes for the pillars and applies generally to the whole structure.

In the Cook Islands some maraes were reserved for religious ceremonials, others for purely social purposes, and others, again, were used for both. Tupou Isaia stated that Tongarevan maraes were used for the following purposes:

The first two of these functions are religious and the other two social, but the religious and social functions cannot always be separated. In the eating of turtles (p. 91) a ceremony in which the status of the gods was recognized had to be performed before the turtle could be eaten. In the terms of Western culture, there was grace before meat, but the grace had to be recited at the marae.

Similarly, Lamont and his companions were given coconuts to eat after the religious ceremony on the marae had taken place. In the ceremony attended by Lamont only three coconuts were used, and these were given to the three leading persons in the white community, probably in recognition of their status. It is evident that the eating of coconuts that had figured in the marae ceremony was the Tongarevan ceremonial form of recognizing status. From Tupou Isaia I understood that on occasion regular feasts at which large quantities of nuts were consumed were held.

Invoking the gods (pureanga atua) was a necessary part of all marae ceremonials, but occasions probably arose when the religious observance was the sole theme of the ceremony, not influenced by the social factors of eating page 178 turtle and drinking from coconuts. The appointment of priests (hikianga taura) is interesting, for no mention is made of a similar ceremony in regard to the ariki, or high chiefs. In the Cook Islands and other Polynesian areas the raising (hikianga) of the ariki is a dignified ceremony and must if possible be performed on a marae associated with that function in the historical past. The failure of Tupou Isaia to mention such a procedure is in keeping with Lamont's account (p. 49) of the election of a successor to the ariki Opaka, in which there was feasting, but no marae ceremony. Opaka, however, functioned as high priest and it may be that the ariki enjoyed a dual capacity and was only “raised” for the religious duties of his position.

Comparisons

The Tongarevan marae is simple in structural plan; the flatness of the ground created no complications. For defining the rectangular inclosure a curb of coral slabs at the front and sides met all requirements. The exact definition of these boundaries and the erection of pillars on the front, sides, and back are constant features of the maraes.

There is a considerable affinity between the maraes of the northern Tuamotuan archipelago described by Seurat (22, pp. 475–484) and those of Tongareva. Both have raised platforms made of limestone slabs, a filling of coral material, and limestone pillars on the back line. However, the raised platform of the Tuamotuan marae faces an open court that is not clearly defined and not embellished with pillars on the front and sides. In the sketches of Tuamotuan maraes reproduced by Emory (4, p. 110) the pillars at the back are irregular in shape. In general, the Tongarevan marae shows more care in definition and better craftsmanship in the preparation of material.

The maraes of Necker Island described by Emory (4) show some affinity with those of Tongareva in that a raised platform faces a rectangular court and the back line of the raised platform is embellished with a row of uprights spaced at intervals. Because the pillars are made of basaltic blocks they are much shorter. In Emory's type marae (4, fig. 25, a) it is a coincidence that the back row consists of nine uprights, the average number for Tongareva. The importance attached to placing the highest pillar in the middle is, however, not recognized in Tongareva. The platform of Necker Island extends the full width of the marae—perhaps because it is built on the uneven or sloping ground. To construct a raised platform on sloping ground it is easier to cut a terrace the full width of the inclosure than to excavate the ends. In Tongareva, where the ground is flat, extra work and material would have been involved in extending the platform to the sides of the marae. Labor and material must play a considerable part in modifying page 179 the structural pattern. A median upright on the platform with two others below it on the court floor and pillars set on special sites on the floor are features absent from Tongarevan maraes. On the other hand, no uprights are spaced along the front and sides of the Necker Island marae.

Emory (4, p. 709) draws attention to the resemblance between the small Tahitian inland marae and those of Necker Island. The large raised platform of the Taputapu-atea marae in Raiatea shows the same technique of construction as those of Tongareva. Huge limestone slabs rising over 8 feet above the ground have been set on edge to form a rectangular inclosure which has been filled in to a height of 8 feet with coral boulders and rocks. Some of the limestone slabs have fallen down, revealing another row of limestone slabs within, but of less height. It is thus evident that the Taputapu-atean platform was originally smaller and lower. Doubtless, as Taputapu-atea increased in fame the larger limestone slabs were erected on the outer margins of the original structure, and the inclosure thus enlarged and deepened was filled in to the present height. In spite of its size and fame, the structural technique and pattern is that of the simple, low platforms of Tongareva.

It is apparent that an ancient marae structural pattern consisted of an open court with a raised platform at the end, formed of limestone slabs set on end, and filled in with loose material. At the back, tall uprights that may have had some religious significance, or may have been purely ornamental, were set up. On Tongareva progress has proceeded in the direction of defining the boundaries of the court with curbstones and extending the stone uprights to all four boundaries. I consider the Tongarevan pillars purely ornamental, placed there to add dignity and impressiveness to the structure. Even today, when Polynesian communities wish to improve the appearance of their villages they erect stones at intervals along the sides of the village road. Their ancestors did the same with their maraes. The shaping of the pillars was a further advance by the Tongarevans. Emory (4, p. 110) draws attention to a pillar in the Tuamotuan marae of Ramapohia that is cut to represent the human form and is called ofai tiki (representative of the deity). In Tongareva two similar pillars, but much more carefully cut, appear to be wholly ornamental. (See fig. 22, c, 2, 3.) Both are on the left boundary of the Rakahanga marae, not on the back line where the priests officiated when in direct communion with the deity. The inhabitants do not know whether or not they represented gods or ancestors.

Detached and special stones, whether in the court or on the boundaries, were significant in the marae ceremonies which have been dissipated by Christianity.