Ethnology of Tongareva
The Tongarevan canoe has been supplanted completely by the large sailing boats used in connection with diving for pearl shell and the small outrigger canoes made of sawed planks after the modern Manihiki design. Search failed to locate an old canoe stated to be on Te Puka. The wood of some of the old canoes has been used for house piles in Omoka. Those available were examined, and a certain amount of information was derived from them. A few notes on the parts of a canoe were obtained from the old men, but without an actual canoe it has been impossible to determine how the different planks were shaped. Choris (3, p. 33) figured a canoe with five outrigger booms and states that the “piroques” were made of several pieces of wood bound together and having outriggers (fig. 48).
Figure 48. Tongarevan canoe (after Choris): 1, keel (oa); 2, stern (vero); 3, bow, (isu); 4, gunwale (huatanga); 5, gunwale braces (manu); 6, steersman's seat; 7, raised breakwater; 8, lookout's seat; 9, outrigger booms (kiato); 10, outrigger float (ama); 11, float connecting pegs (tutaki); 12, longitudinal spar (torutoru ama) and spears.
Kotzebue (14, pp. 217, 219) counted about 36 canoes which came out to his ship. Each contained 7 to 13 people. He says:
Their boats are made of several pieces of wood well jointed together with cocoabast cords. Both ends are rounded off, above and below the water, furnished with a projecting spar. They have an outrigger on which their arms are secured…. We did not wait for a boat, which approached us under full sail, from a distant island of the group.
Wilkes (31, vol. 4, pp. 277, 279) says that the canoes each contained from 7 to 16 men. They were made of dark-colored wood, with a light out- page 190 rigger, and were without sails. They were ingeniously constructed of pieces sewn together with sennit, but they leaked so badly as to necessitate constant bailing. They were the largest canoes constructed on a low island that the Wilkes expedition saw.
Lamont (15) gives more detail that will be used to eke out the scanty information obtained locally.
The wood used in making the hull and outrigger of the Tongarevan canoe was tou (Cordia subcordata). Suitable trees about 3 feet in diameter were felled with shell adzes or with large coral heads so split that the inner part furnished a handhold, and the outer, rough part with its sharp circular projections served as a blade that, when struck against the fairly soft wood, nibbled out the scarf. The felled trunk was then seasoned. I understood from my informants that the tou was seasoned by burying it in the ground, and that another timber tree called hano (Guettardia speciosa) was seasoned by soaking it in the lagoon. Lamont (15, p. 151), however, states that the tou (which he spells to) was also rolled into a shallow part of the lagoon, where it was subjected to alternate dampness and heat as the tide flowed in and out. After seasoning by this method the log could be split more readily into the planks used in canoe building. As the supply of timber was limited, the hollowing out of a log into a single dug-out was regarded as wasteful, especially as the larger canoes had to be built up because the timber was not big enough to provide canoes of the dug-out pattern. The use of plank canoes on low islands was thus largely influenced by economic environment.
The lashing material was sennit braid (kaha). Adzes of Tridacna shell were used to split and shape the planks. The edges were marked with a mixture of charcoal and water when the planks were fitted together. Shells with the apical whorls running to a sharp point were used for boring holes. A number of these, picked up near old house sites, have been identified by C. Montague Cooke, Jr., as Terebra maculata, Terebra crenulata and Mitra stictica. Lamont (15, p. 151) says that a piece of sharp stone was used in addition to shell, “assisted by a sharp pointed cocoa-nut stick.” Coconut husk was used in calking planks and plugging lashing holes. A vegetable substance (kana) growing in flat, rounded masses on the coral heads in the lagoon was dried and used like sandpaper for smoothing down the outer side of the planks.
The shape of the canoe as drawn by Choris (3, pl. 12) is shown in figure 48. The native names obtained locally are applied to it. It resembles a Pukapuka canoe in Bernice P. Bishop Museum, and the hull is not unlike the Nanumea type of Ellice Islands canoe figured by Kennedy (13, fig. 59). It is characterized by a long upward slope of the keel toward the bow and stern, each of which terminates in a solid short upward projection.
Parts of the Canoe
The keel (oa) of the Tongarevan canoe is a long narrow piece of wood, convex on the under surface and concave on the upper. Lamont (15, p. 151) states that the keel was about a foot wide. A section of a keel used as a house pile is much smaller than this, but it probably came from near one end. (See fig. 49.) In this piece no groove had been cut on the outer surface at the edges for a covering seam batten, though Pa of Omoka stated that it was usual for all seams to be so covered.
Figure 49. Canoe keel, Omoka: a, section above ground, 17 inches long, about 2 inches thick; b, cross section of narrow end, upper concavity 1.5 inches deep; c, cross section of wide end, concavity 2 inches deep. 1, narrow end of upper concave surface, 5.5 inches wide; 2, wide end, 6.5 inches wide; 3, holes about 0.5 inches in diameter bored through keel about 0.5 inches from edge on both sides. Spacing between holes increases from 2 inches at narrow end to 4.5 inches at wide end.
It was not ascertained exactly how far the keel extended. Lamont (15, p. 151) states: “The keel slopes gradually up at either end till it rises above water-mark, terminating in a solid point called the isu, or nose.” The “isu” is the bow, which Pa maintains included the upward projection and was formed of two pieces. Judging from the Pukapuka and Ellice Island canoes, Pa's “two pieces” probably meant that the keel with the upward projection formed one part of the bow and a bow cover formed the other. Lamont's observation would thus be correct. The term “oa,” as applied to the keel by Pa, is a misplacement in dialect, for throughout Polynesia oa is applied to the raised gunwale.
Pa maintained that there were three tiers of hull pieces above the keel, the order from below being the kape, rau-toru and huatanga (gunwale). Sections of the hull pieces now in use as house piles have the outer surface at both side edges cut down to form batten grooves. (See fig. 50.)
In another hull plank the batten groove was 1.5 inches wide. The hull plank shown in figure 50 belongs to the second or third tier of the hull and must be either a kape or rautoru. Pa could not distinguish them as both these tiers have batten grooves on the upper and lower edges.
The hull pieces were fitted to the keel edges, which had been smeared with a thick solution of charcoal and water. The hull piece was removed, and the irregular parts that bore the impression of color were trimmed with the adz. page 192 The fittings were repeated until the impression was uniform. The hull pieces were fitted to each other at the sides and ends in the same manner. The batten grooves were then cut on the contiguous horizontal edges, and paired holes were made through the planks, clear of the grooves. Before lashing, coconut husk, soaked in water and beaten, was spread over the upper edge of the lower piece to serve as calking.
Figure 50. Hull plank in Omoka house pile. a, end section; b, cross section: 1, plank, 13.5 inches wide and 2 inches thick; 2, 3, side edges, grooved on outer surface for the covering battens; 4, lashing holes cut through plank about 0.5 inches to inner side of batten groove and spaced 3 inches apart—holes on right edge evenly made with cross diameters of 0.5 inches, holes on left much larger; 5, hole on left, cross diameters, 0.75 and 1 inches, perhaps cut with some form of chisel.
Figure 51. Hull lashings of plank canoes. a, Tongareva, outside view; b, Tongareva, inside view; c, Samoa, outside view; d, Samoa, inside view: 1, lower hull piece, or keel; 2, upper piece; 3, 4 (a), wide groove formed by upper and lower pieces, and (c, d), flanges of upper and lower pieces on inner side of end section; 5, 5, paired holes; 6, batten placed in groove; 7, lashing; 8, braid carried obliquely from lower hole to upper hole of next pair to repeat lashing of four turns on outside; 9, free braid going up to next pair of holes. Samoan lashings do not show on outside, and lashing of each pair of holes is fixed and braid cut so that they show as interrupted lashings.
The join is termed hono, and the lashing holes receive the general name of rua (hole). The lashing is termed hau, and the phrase for lashing with sennit is “ka hau te kaha.” When the lashing was complete, the holes were plugged with beaten husk.
Figure 52. Section of join between two planks with seam batten in position, a, b, Tongareva; c, d, Samoa: 1, lower piece; 2, upper piece; 3, 4, Tongarevan batten grooves, each 1.5 inches wide and 0.5 inches deep; 5, holes for lashing; 6, Tongarevan seam batten, 3 inches wide and an inch or more in thickness; 7, lashing; 8, Samoan flange. Samoan lashing does not show outside.
The gunwale (huatanga) forming the top sides is characterized by an inward projection from the top edge forming a broad ledge. This served as a seat for the paddlers, but there are special seats fore and aft (fig. 48). A section of a gunwale, also used as a house pile, is shown in figure 53.
Figure 53. Gunwale (huatanga), vertical part 11 inches deep on outer side, upper surface of gunwale 4 inches wide, both parts 2 inches thick, and horizontal ledge projecting inward 2 inches from side. a, inside view; b, cross section: 1, vertical part; 2, inward projecting ledge; 3, holes 3 inches apart and bored through 1 inch from lower edge.
The stern piece (vero) was, from Pa's description, in one piece with the keel, and probably solid. The upward projection at the aft end was termed the maramara by Pa. The steersman's seat mentioned by Lamont (15, p. 152), shown in figure 48, 2, appears to be a widening out of the stern piece just behind the aft boom.
The bow piece (isu) has the same upward projection as the stern piece. Choris (fig. 48, 7) shows an upward projection of the fore part of the hull where it meets the bow piece. This corresponds to the wave guard of the Pukapuka and Ellice Islands canoes. It is probably this raised part that forms the second element of the bow piece that Pa stated was formed of two pieces, as against the one piece of the stern. The part aft of the wave guard is expanded into a seat for the lookout, whose duty it is to direct the steersman when a coral head or shoal is approached.
Lozenge-shaped projections, also termed maramara by Pa, were made on the upper surface of the bow and stern pieces and are distinct from the markedly raised projections at the fore and aft ends, respectively, of the bow and stern pieces. There are two on the bow, and four arranged in two pairs on the stern.
The outrigger is placed on the port side. The outrigger booms (kiato) are made of tou wood and range in number from 3 to 5, according to the size of the canoe. They are lashed to both gunwales with sennit braid, the braid passing through holes bored through the gunwale below the site of the boom. The booms are straight and project beyond the port side of the canoe (fig. 48, 9).
The float (ama) of tou wood is a long spar which Choris shows to be about the same length as the raised part of the hull. (See fig. 48, 10.) He page 195 shows the aft end with an upward curve, but no mention of this was made by Pa. The float projects well back behind the aft boom and is not cut off close behind it, as it is in Samoan canoes.
The attachment between the straight booms and the float by means of four connecting pegs (tutuki) to each boom is indirect, but how the pegs were lashed was not ascertained. Pa stated that a suspensory cord (ua) was used as an additional connection between the boom and float.
A longitudinal spar (torutoru ama) is lashed over the outer ends of the booms and served as a rest for the paddles. The spar is shown in figure 48, 12, but the extra width is probably due to spears, which were also carried in this way across the booms.
In size the canoes ranged from the ordinary fishing canoes with 3 booms to the large war canoes with 5 booms. The largest canoe seen by the early navigators carried 16 men, but it is possible that none of the large war canoes went out to the ships in the short time they were in the Tongarevan waters. The war canoes were drawn up under shelters and were refitted for launching before warlike expeditions. The news that a particular island was refitting its war canoes was equivalent to a declaration of war and put enemies upon guard. Lamont (15, p. 195) states that a chief, whom he names O Pai Tangata, selected 30 warriors and launched his war canoe. As Lamont accompanied the party and only the one canoe was used, he evidently means that the war canoe carried the 30 men. He speaks (15, p. 346) of three war canoes containing 60 men on another occasion.
Figure 54. Model paddle (hoe), a, surface toward paddler (front); b, side view; 1, handle, round in section, approximately same length as blade, with slight anteroposterior curve and convexity at front (handle may be proportionately too thick in model); 2, blade; 3, straight oblique shoulders of blade; 4, sharp side edges, expanding but slightly from shoulders; 5, blunt end, with no thickening or reinforcing projections.
The paddle (hoe) was made of tou. A small model made for me has a long curved blade. (See fig. 54.) Lamont (15, p. 152) states: “The paddle is long, the blade narrow and curved.” The curved blade marks a difference page 196 from the long, narrow, straight blades of near-by Manihiki. In the shallow water between the inner lagoon reef and the shore the canoes are poled as well as paddled.
The mast is termed tira and the sail, ra. Kotzebue (14, p. 219) states that he saw a canoe approaching from a distant island under full sail, but he did not wait for it. It is curious that the only sail mentioned by Lamont (15, pp. 242–243) was a makeshift one of coconut leaves and used within the lagoon:
I now for the first time saw a Penrhyn canoe under sail. Its mode of propulsion is, I should think, the most original to be seen in any part of the world. The sail is as simple in construction as it is primitive in appearance. For the purpose three long palmboughs are cut from the nearest tree, and, after a few strips of bark have been torn from them, they are conveyed to the canoe. The lower or thick end of a bough is placed at the bottom of the canoe, with its long slender leaves standing perpendicularly to the height of about ten feet, and made fast to the cross-bar of the outrigger, which runs across the little vessel. A bough is then placed on either side of this, attached to it at the bottom, but inclined outward, and also fastened to the cross-bar. At the top the slender ends of the latter are bowed over to the centre end, the mingling leaves of all being interlaced a little to present further resistance to the wind. This, when completed, forms a broad sail. Strips of bark are fastened to the most extended part of the outer boughs, which are again secured to the stern outrigger: and thus the cocoa-nut tree supplies more of their necessaries—sails, masts, spars, and rigging being all constructed in a few minutes from its boughs.
As there is no step for the mast to rest in, this deficiency is supplied by a little boy, who sits in the bottom of the oaka with his feet against it. By trimming the lee side of the sail a little aft, the boat will keep her course by the help of the paddle when the wind is on the quarter, but will not sail on a wind; or even with the wind abeam, when, having little hold on the water, she drifts to leeward. When the wind shifts thus the sail is taken in in as primitive a manner as it is set. The lashing being cut, and the backstays cast off, away goes the whole ship's rigging overboard, the work of refitting being very speedy and easy.
Towards the end of our voyage we had the wind on the quarter, blowing pretty fresh, and I had to remain with my feet at the bottom of the mast, whilst the small boy sat on the out-rigger to keep her from capsizing, moving out on it or in again as the wind increased or fell.
In about an hour and a half we landed at Haka Shusha, a distance of eight miles, my boatmen cutting away their sail, and paddling off with all their might, in fear of being caught by their enemies.
The “few strips of bark” torn from the palm boughs were the tari strips from the butt end of the leaf midrib, which formed the common material for tying. It is possible that the sail which Kotzebue saw in the distance was of the coconut leaf type, and that the lauhala matting sail was not used. The term ra was equally applicable to the coconut leaf sail.
The canoes could not sail to windward but had to sail down on the wind, hakasekeseke ki raro. This corresponds to Lamont's statement, “the wind on the quarter.”
The state of war that so frequently existed between different islands necessitated the use of much of the available timber in making war canoes. It is thus probable that the smaller fishing canoes were not common, because page 197 of shortage of timber. This may have reacted on fishing methods and led to the more extensive use of wooden floats and swimming, for a wooden float must primarily have been a makeshift for a canoe.