Ethnology of Tongareva
In the absence of the pig and dog and of cannibalism fish form an even more important flesh food in Tongareva than in the Polynesian islands where there are other flesh alternatives. The supply in the lagoon, in the reef channels, and in the sea outside the reef is varied and abundant. Besides the rockfrequenting fish, flying fish, shark, and porpoises are taken.
Fishing methods are affected by the lack of material for lines and nets The hau (Hibiscus tiliaceous) and the oranga, which provide the best fibers used in most islands, are absent from Tongareva, and coconut sennit fiber is used instead. Self-acting traps and walled fish traps are not used. There is a form of shark noose, but no squid lure. Fish are caught by groping, snaring, sweeping, driving, spearing, and angling.
Groping with hands in the crevices of the rocks is practiced by members of both sexes who are expert divers and can remain under the water for some time. The method used in the daytime when the fish can be seen as well as felt is termed nono (Rarotongan, naonao). At night, when the sense of feeling alone is relied upon, the method is distinguished as haha (Maori, whawha).
Sharks are snared with a slip noose made of a two-ply twisted sennit cord (hau ato). The ruhia shark, which comes into the lagoon after the marau fish when they are plentiful, are caught by diving down and placing the snare over the tail. Sharks termed mango moc awa (sharks that sleep in the channel) that at certain times sleep with their heads in rock crevices are also snared by the tail while they are asleep. The same method of snaring sleeping sharks prevails in Aitutaki. It may be stated that the species of sharks snared are not voracious maneaters and are regarded merely as ordinary fish without any of the fear prevailing among Europeans, who regard all sharks as dangerous. It is quite safe for the expert to dive down and slip a noose over the tail. The snare is termed sele; the method of snaring a shark, sele mango.
The coconut leaf sweep (rau) receives its name from the leaves (rau) of the coconut from which it is made. The leaves are split (sasae); and the half leaves are tied together (sere), twisted round and round to make the leaflets stick out in different directions (viri), and some of them may be braided together (hiri) to get extra thickness. A long sweep is drawn around (taki) so as to inclose the fish and drive them ashore. As the curve decreases the slack ends are doubled in to strengthen and thicken the sweep. The sweep is used by day and by night.
The drive (aroaro) method of fishing consists of driving shoals of flying fish or porpoises onto the reef, or turtles into the shallow water. When a shoal of flying fish is seen outside the reef canoes paddle to the outer side of them, stretch out in a semicircle, and drive them in toward the reef. Paddles are beaten against the canoes, and stones are thrown at the fish to urge them toward the reef. Escape back to the sea is blocked by the line of canoes and by swimmers, or, where canoes are not available, by swimmers only. Such methods have been abandoned, but Lamont (15, pp. 217–218) gives a good description of a drive:
On the third day we sat chatting in the usual quiet way, when a shout at a distance set the whole household in commotion. As Opaka started excitedly to his feet, I asked him, in his own language, what was the matter. “Eia ha?” said I. “No, te maroro,” he replied; and, without waiting to give me further explanation, he seized a “toto,” or bag-net, from the roof, and darted along the beach, calling the rest to follow. Fully as excited as himself, and shouting at the top of their voices, “Maroro! maroro!” each seized a mat-basket of some kind and rushed wildly off in the same direction. I followed them as quickly as the rough ocean shingle, with its burning stones, would permit. With their long hair streaming, and their eyes gleaming with excitement, I saw them diving into the hollow curve of the breakers that raised their white heads aloft, soon to appear again some distance off beyond the force of the waves. Men, women, and children alike fearlessly plunged beneath the foam, seemingly as much at home as on land. The multitudes in the sea, at first scattered over a considerable extent, now began to concentrate towards a point, not only keeping up an incessant noise with the voice, but jumping halfway out of the water, and, as they descended, striking their elbows to their sides, and clapping their hands, producing a report like a pistol-shot. I now observed shoals of flying-fish skimming the water in terror in every direction, often rising beyond the nets of the circle of men, who raised their arms to catch them, and often escaping in their flight the baskets of the outer guard of women and children. When the circle was sufficiently contracted to concentrate the fish in a mass, the men dived amongst them with their nets, which, soon becoming too heavy for them to support, were emptied into the baskets of the women behind, who proceeded with them ashore, riding behind the crest of a breaker that would dash an ordinary swimmer headlong upon the rocks, and returned again after they had emptied them. In about half an hour the shoal was all dispersed or caught, and each family had a bountiful supply of flying-fish, or “maroro.”page 199
The toto hand net and baskets of the tupono type are used to scoop up the fish.
Porpoises are driven into the shallow channels on the reef where men seize them and drag them up out of the water. Much ceremony is observed to ensure success; women are not allowed out of the houses, and children are instructed not to cry, as that would render the operations unsuccessful. Handy (9, p. 176) records a similar method in the Marquesas.
Turtles in the lagoon are driven into shallow water by men in canoes. Men jump overboard and dive down to keep the turtles swimming in the right direction and to prevent their doubling back into deep water. During the drive the men make as much noise as possible by shouting and beating the water with their paddles. As they reach shallow water the noise subsides and the turtles rest on the bottom. Men dive down and, seizing the front flappers from behind, lift up the front of the shells and force the turtles to swim up to the surface.
While examining the marae at Vaiari, we saw a rowboat that had forced a turtle in toward the lagoon reef, but the turtle had stuck on the bottom in fairly deep water and refused to go further in. As the crew of the boat failed to reach the turtle because of the depth, Tupou Isaia of our crew took a hand. In his first straight dive he could just touch the turtle. He came up and noted a high rock on the bottom with a lower one near the turtle. He dived down to the high rock, kicked off from it to the lower rock, and with another kick off from the lower rock he reached the turtle got his hands in position on the front of the shell, and brought the turtle to the surface. He calmly appropriated the turtle as the reward of superior endurance.
Another form of fish driving is termed titoko. Before fish are driven into a channel, loose rocks are placed on stationary rocks and reef projections that are under water. As the fish are driven in, the rocks are kicked off with the soles of the feet by the drivers as they pass. The rocks, as they fall to the bottom, frighten the fish and cause them to go forward (kia soro ki mua). When driven to a confined space, the fish are scooped up with a hand net (ka asu ki te toto).
A method of driving the sikutoto fish is termed toro sikutoto. Thirty or forty men armed with pieces of coconut leaf a span in length and termed usu work around in a semicircle and, by beating on the water, drive the fish into the shallow water. The hand net is used to scoop them up.
Lamont (15, p. 278) saw the people spearing fish in the deep passage between Hakasusa and Vaiari, and he states that it was an exercise at which page 200 the people were expert. Fish spears with metal points are now used exclusively, and no information concerning the spears originally used was available.
The general term for fishing with a line is si or sisi (Maori, hi). Since foreign lines, hooks, and sailing boats have come into use the following native methods of catching fish with the hook have been almost entirely abandoned. Hooks are described on pages 202–211.
1. Fishing from an anchored canoe (tukutuku). In tukutuku (to keep letting down) the baited line was lowered from a canoe which was kept stationary by an anchor resting on the bottom. The length of line was therefore not great. A baited circular hook was used.
2. Diving (hakaruku). The U shaped hook (matau si ruhi) was attached to a very short line, the end of which was tied to the middle finger of the right hand. The hook was baited and held in the midst of a handful of ground bait in the right hand. The mouth also was filled with ground bait. As the line was too short to reach the fish near the bottom, the fisherman dived down with the hook. The Tongarevans maintain that fish are not afraid of anyone under the water. On reaching the fish the fisherman opened his hand and let go the ground bait and the baited hook. He removed the first fish that he caught from the hook and placed it under his left arm. If the first fish was caught quickly there was still time to catch a second one. Seeing that his bait was intact on the hook he blew the ground bait out of his mouth, taking care that the baited hook was in the cloud of ground bait. He often reappeared from the depths with two fish. If he had a canoe he placed the fish in it and repeated the diving as long as he had success. In fishing beyond the outer reef it was not always convenient to launch a canoe, so the fisherman nonchalantly walked over the edge of the reef and swam out to sea with a piece of wood which served as a float or a buoy. The float gave him a certain amount of support when he desired to rest, but its main use was to buoy up the catch of fish and the bait. The same method of fishing was used, but on coming up he would thread the fish through the gills with a tari strip from a coconut leaf midrib and tie it to the float. After a catch was secured the fisherman tied the string of fish in a loop over his shoulder and swam ashore.
Although the fisherman had no fear of sharks, sometimes a shark did attack, not the man, but the string of fish. Mr. Wilson, resident Government Agent, and other eye witnesses tell of a fisherman who, swimming ashore with his catch of fish looped over his shoulder, had just such a difficulty. He called for assistance, and two men swam out to help him. Supporting him on either side, they enabled him to reach the reef and obtain foothold, when he walked calmly up on to the dry part of the reef. The man had a shark imprisoned horizontally across his back with its head under one arm and its tail under the other; the shark had a grip on the inner side of the upper arm that held its head. The shark, attracted by the fish, had followed the fisherman in and commenced eating the fish strung from his shoulder. The next thing the fisherman knew was that the shark had gripped him by the arm near the fish. The fisherman thereupon promptly closed his arm, pinning the shark's head against his side, and then coolly reached back with his other arm and brought its tail in under his other armpit. He had the shark jammed helpless, while he himself swam with his feet until assistance reached him. The shark let go when released, but it had removed a large piece of flesh from the arm of its captor. The shark was promptly slain by the fisherman's excited relatives. The hero in this fishing adventure made a perfect recovery, but the scar, which was deprecatingly shown to me, remained as a witness of page 201 that same coolness in moments of emergency which enabled the Polynesians to conquer the Pacific.
3. Anchored hook fishing. The large shark hook was used, and also a short line, one end of which was tied to a stone anchor. The hook was baited with the tentacle (mangamanga) of a squid. The fisherman dived down with the hook and anchor, set it on the bottom and covered the line with sand. He then came up and watched from his canoe or float. When the fish took the bait it was prevented from getting very far by the heavy stone anchor. In the struggles of the fish, however, the anchor could be heard bumping on the ground, and the fisherman dived down and secured his line and the fish. Besides shark, maratea, which may weigh as much as 80 pounds, were caught in this manner.
Fish nets were made of two-ply twisted sennit fibre (hau ato), but commercial twine has now completely superseded sennit. A netting needle (ta) and a mesh gage (mata) were said to have been used. As ta is the widespread Polynesian verb meaning to make a net, and as mata is the mesh, the lack of specific terms indicates that special implements were not used. Lamont (15), however, saw the toto hand net in use, so there is no doubt that the netting technique was known. The knot is the same as that in Rarotonga, New Zealand, and Samoa (29, p. 471). The small meshes are termed mata hiohio, and the large ones, mata tua nunui. Four types of net were described:
1. Ordinary hand net (toto). The toto was the commonest net. It had an oval frame made of two thin pieces of wood tied together at either end and with a crossbar tied across a few inches from the thicker end to spread the frame out into oval form. A bag net was made, a circumferential two-ply cord was run through the circumferential meshes, and another cord was run spirally around the circumferential cord and the oval frame to keep them together. The net was used to scoop up fish in sweeps and drives. It was also set in the channels where fish were driven into it.
2. Fine meshed scoop net (sema). The sema had a finer mesh than the toto, but otherwise was similar.
3. Bag net (takeke). The takeke was a bag net without a wooden frame but with ropes attached to it to keep the mouth stretched and open. It was set on the bottom of the lagoon on one side, with the opening in vertical position. Stones were set on the part of the net opening that lay on the bottom.
4. Baited net (taka). The taka was a bag net with a hoop of ngagie around the opening. A line was attached to the hoop, and the net was baited. It was lowered with the line, and fish such as kohiri were caught by drawing the line up quickly when the fish were felt biting at the bait.
A flying fish net with a long handle similar to that used in the Cook Islands (28, p. 288) is now also used in Tongareva. As the aroaro drive method of catching flying fish was recognized, it seems probable that the present form of flying fish scoop net has been introduced in post-European times. The flying fish net seen had two crossbars across the frame at the handle end.
The introduction of foreign fishing lines has led to the abandoning of purely native material and a lack of clarity as regards technical details has resulted.
The material for fishing lines was restricted to coconut husk fiber, owing to the absence of the more suitable plants used in other parts of Polynesia. The lines were twisted (miro) on the bare thigh into two-ply twisted cords, but fairly thick three-ply braid was used to form attachment cords for the large shark hooks. Gudger (7, p. 230), in describing two shark hooks obtained by the Wilkes expedition, states that a long fibrous material was used in addition to sennit in the lashings. My informants did not mention anything but sennit. Wilkes (31, vol. 4, p. 287), in figuring some hooks from Tongareva, shows one attached to a long line which is wound in longitudinal lengths and then with transverse turns, leaving the longitudinal turns projecting at either end. The point of the attached hook is then evidently stuck in under one of the transverse turns at one end. This corresponds to the Samoan method of winding the pa ala line and hook (29, pl. 47, B), except that the middle part is not covered by the transverse turns. In the figure by Wilkes the transverse turns are continuous over the middle of the hank. The method of winding thus weighs against affinity with the Samoan line, but the method depicted is used in Tahiti, from which island the hook probably came.
Parts of the Hook
The hook is conveniently divided into the shank, the bend, and the point. Beasley (1) and Gudger (7), the most recent writer on Polynesian fishhooks, refer to the point as the “barb,” but as the term “barb” is specifically applied to a projection near the point in trade metal hooks and some forms of Polynesian hooks, its use as a general term to include hooks without the special barb projection is apt to convey an erroneous impression. The barb is a distinct invention added to the point to prevent the hook from working free during the struggles of the hooked fish. Most Polynesian hooks obtained a like result from the extra inward curve of the point toward the shank, and form a marked contrast in shape to hooks of foreign make which are more open. In the open point of the bonito hook neither the inward bend nor the barb were desired, as quickness in detaching the hook and retrolling it was important when the fisherman was on a school of bonito. The skilled fisherman could either flick the fish off into the canoe page 203 with a jerk of the rod, or as he brought it in, strike the fish with his hand to jerk the body upward so that it fell off the hook. Neither of these methods of quick detachment could be so readily carried out if the hook had an inward bend or a true barb.
The bend of the hook is either curved or rounded, but in some hooks it forms an angle. It is then convenient to distinguish the two limbs formed as the shank limb and the point limb. In simple hooks (fig. 58) made from one piece of material, the shank, bend, and point are quite clear. In composite hooks made of two pieces the form of the point element affects the terms used. The upper end of the shank is provided with a projection on the outer side or part remote from the point. The projection is triangular, with the lower side at right angles to the shaft and the upper side sloping upwards and inwards, sometimes with a distinct, concave curve. The Tongarevan term for the projection is reke (knob), and as it is constant in most types of hook, it may be conveniently termed the shank knob. Its use is to prevent the lashing which binds the snood to the shank from slipping. The bend is termed kopu (belly) and the point, mata (point).
One-Piece Shell Hooks
The two types of baited hook described as native to Tongareva are simple hooks with the shank, bend, and point shaped from one piece of pearl shell. Both types are provided with shank knobs, and the points are without barbs. They may be classified as U shaped and circular.
Figure 55. One-piece pearl shell hooks: a, U shaped ruhi hook in course of manufacture; b, completed U shaped hook; c, circular hook. a, U shaped hook in course of manufacture: 1, part of shell to become shank knob; 2, part to become point; 3, bend. Shell 63 mm. long by 47 mm. wide shaped to oval form; gap cut out at smaller end to form inner curve 29 mm. long by 15 mm. at widest part; dotted lines show shell in completed hook, though part of left gap is broken off and knob might actually be higher. b, c, completed hooks: 1, shank limb; 2, point; 3, bend; 4, point limb; 5, shank knob. Shank knob has same form in both hooks.
The circular hook is used for catching fish less than 9 inches in length, such as the marau and the kokiri.
No Tongarevan baited hooks with lashings were seen but, judging from similar Rakahangan hooks, they must have been attached to three-ply twisted snoods of sennit fibre. The lower thicker end of the snood was unravelled, and each ply separately bound round the shank and knob in a particular technique. A fine two-ply twisted thread of sennit was then bound over the snood attachment by making figure-of-eight turns around the shank on one side and alternately above and below the shank knob on the other. The end of the thread was finished off with a number of close transverse turns around the snood ending in half hitches and stoppered with an overhand knot.
A bait string (nape) is formed of a fine two-ply twisted sennit cord, which is tied with a slip knot below the angle made by the point with the point limb. It is used for tying the bait on to the hook.
1. The pearl shell bonito hook (matau uhi). Whether introduced or not, the bonito hook deserves description, as it is now made locally.
The hook is composite, consisting of a shank and a separate point. (See pl. 8.) The shank is made of strips of shell averaging 18 mm. in width and cut so as to include a part of the thick hinge at one end. Four shanks in Bernice P. Bishop Museum range in length from 113 to 124 mm. The hinge end is termed the head, and the other thinner end, the tail. The nacreous inner surface of the shell forms the front of the hook. The rough outer surface forming the back of the hook is ground down to remove the rough dull material and form a clear shiny surface which has irridescent colors toward the tail. (See fig. 56, a, b.)
The point of the composite hook is made of pearl shell taken from near the edge of the shell away from the hinge. The point piece is cut out on the flat, so that one side shows the natural inner surface of the shell, and the other shows the black outer surface which is smoothed down but not polished. The point piece is fitted against the front of the shank, and the part which fits against the shank may be termed the point base. The problem is to extend this base to make it long enough to support two lashings and so to render the attachment firm. The Tongarevan point follows the form of the Samoan hook, in which the base is prolonged proximally or on the same page 205 side as the functioning point. As the point has to be cleared, the part of the shell between the point and the base has to be hollowed out. The point piece thus carries the technical “bend” of the hook as well as the functional point and the base for lashing. For lashing purposes two holes are drilled through the base, which must be deep enough to hold the holes. (See fig. 56, c, d, e.) In three out of four hooks in Bernice P. Bishop Museum the proximal perforations broke through the upper edge, forming grooves. In three hooks the point was everted.
The Tongarevan point is characterized by the proximal prolongation of the base with two holes for lashing and the material of shell. Beasley (1, pi. 14) figures a point with the typical proximal prolongation of the base but with three lashing holes, giving bone as the material. No mention is made in the text of the hook to which the point belonged. Because of the difference in material and number of holes the locality needs confirmation.
Supporting the theory that the bonito trolling hook was introduced, Solomon of Tautua stated that the drill with the crossbar and two cords was unknown in ancient times.
Figure 56. Bonito hooks: a, b, shank; c, d, e, points. a, front; b, side of shank: 1, thick hinge part ground on both sides in curve to proximal point; 2, point; 3, median line at meeting of sides ground inward at slant, forming triangular head, maximum thickness about 14 mm.; 4, surface of hinge; 5, transverse hole drilled through horizontally below median ridge at about 18 mm. from point; 6, shank beyond head, gradually diminishes in width to about 12 mm. and in thickness to 4 or 5 mm. at tail end, where it is cut off square or with slight convexity from side to side; 7, two shallow grooves cut vertically on both sides at tail end. Longitudinal natural curve of back and corresponding front concavity apparent. c, d, e, points, side view: 1, base in c, 22 mm. wide and 6 mm. high in clear proximal part, in d, 19 mm. wide but only 5 mm. high in clear proximal part, causing proximal hole (5) to break through upper edge, and in e, 23 mm. wide; 2, point, slightly everted (c, d) and straight (e); 3, bend cut out between base and point; 4, distal hole bored through base; 5, proximal hole or hole broken through to upper edge (d, e) because of high placement, and leaving groove (e) effective in lashing; 6, distance between point and shank (c) 21 mm., and corresponding clearance (d, e) between point and shank 17 mm.
Figure 57. Snood loop and lashing of proximal hole of bonito hook. a, snood, matai (1), is passed around base of point (2) with its end (3) overlapping by about 15 mm. b, two-ply twisted thread commences distally to lash overlap together by burying end of thread under subsequent transverse turns which cross it; transverse turns are continued for length of overlap; end (4) is turned back under three loose turns which are subsequently tightened; slack is removed by drawing on end; end is cut off short. c, overlapping ends of some loops are spliced (3) for distance of about 10 mm., a method probably recent. d, side view: point (2) is placed in position on tail end of shank (3) with back edge projecting a little beyond end of shank; grooves (6, 7) are cut on sides of shank opposite holes (4, 5) in point; loop (1) of snood is placed around base of point and loop end bent down under projection formed by protruding back of point base. e-h, back of the shank tail. e, with point and snood loop in position, lashing thread (1) is laid on back of shank (4) between pair of proximal grooves (7, 7); end is bent down at angle and held in position by left thumb while left forefinger keeps point in position on other side of shank; thread is passed transversely to left by right hand to lateral groove (7) up over snood loop and through proximal hole of point base. f, thread descends on the right side over other limb of snood loop and right groove (7) of shank to reappear on back, where second turn (2) in passing transversely across to opposite groove crosses obliquely bent end of first turn (1). g, oblique end (1), crossed by the second turn (2), is bent over second turn, h, third turn (3), after passing through point hole in making transverse turn around shank, crosses and doubly fixes end (1); five or six lashing turns are made through the one hole. i-l, side view, i, six lashing turns (9) pass through hole (5) of point (2), over snood (1), and around shank (3). j, thread (8) is brought up on right side and passed from proximal side under lashing (9) in space between shank (3) and lower part of snood (1). k, the thread (8) is brought back over lashing (9) and looped through under its standing part to form overhand knot (10). l, two other loops with overhand knots are made and thread (8) is continued on through hole (5) to other side, where it makes similar set of loops and knots over lashing; thread end is cut and proximal lashing of base is completed.
Figure 58. Bonito hook lashing, distal hole and pig's hair hackle: a, lashing thread is fixed on under side of shank (1) in exactly same technique as proximal hole, and hackle (4) is laid transversely along over lashing on under surface of shank; b, the thread (3) is brought around on right and crosses hackle (4) obliquely (5) to the left, on distal side, to continue lashing turn through hole in point base; c, thread (3) is brought around on right on distal side of hackle (4) which is bent forward to approximate limbs of hackle, crosses bend of hackle obliquely to left on way to lateral groove to continue lashing turns—second turn (6) over hackle also crosses first turn (5) in middle line; d, thread is brought around on right proximally to hackle and makes turn (7), keeping on far side of the first turn (5), passes on through point hole and, coming back on right or distal side of hackle, makes crossing (8), keeping to far side of previous turn (6) in same direction. Second pair of turns (7, 8) lashes limbs of hackle and maintains them in bent position, and thread carries on with finishing turns around lashing on either side as in figure 57, j-l.
Figure 59. Bonito hook, completed lashing of point, snood loop, and hackle: 1, shank; 2, head; 3, point; 4, snood drawn over mesial ridge of head; 5, snood loop fixed under lashings of point; 6, first lashing through proximal hole of point; 7, second lashing through distal hole of point; 8, hackle fixed on back of shank by second lashing; 9, hole through head.
Figure 60. Bonito hook, lashing of snood to shank head: a-d, front view, e, f, side view. a, snood (1) rests on median ridge of shank head (2) with transverse head hole (3) showing below snood on either side; lashing thread (4) is passed through hole, and short end (5) on right is bent upward. b, lashing thread is brought over snood (1) from left and down through hole on right, passing over bent short end; short end (5) is bent upward over first turn, and two more turns are made through hole and over snood; short end (5) is thus fixed in position. c, next turn (7) passes from left over snood (1), makes complete turn around it, crosses itself on middle of snood, and passes down on right through hole to prevent snood from slipping off median ridge—three similar turns were made on the hook examined and then straight turn without looping around snood. d, eight more turns with loop around the snood are made, first crossing (8) being distal and subsequent ones following on near side, so as to develop neat ornamental pattern. e, two to four circumferential turns (9) are made by passing thread (4) between snood and head and over both limbs of lashing (6). f, thread is continued on in close spiral for about 8 turns (10) around snood and fixed with a couple of half hitches (11).
2. The Ruvettus hook. The wide distribution of the composite wooden hook for catching the castor-oil fish (Ruvettus pretiosa) has been shown by Gudger (7) and Beasley (1). Hedley (11) calls attention to the manufacture of the hook from a forked branch, Kennedy (13, pp. 12–27) gives details of the hook in the Ellice Islands, and Nordhoff (19, pp. 221–232) discusses its distribution and its diffusion in the Society Islands, Cook Islands, and elsewhere. The typical hook (kau) is now used in Tongareva, but the inhabitants think it was introduced. None was seen, but they are said to be made of ngangie wood (Pemphis acidula or a closely allied species) and shaped like those used in Rakahanga and Manihiki.
The Ruvettus is a deep sea fish, and a considerable length of line is required. The sinkers are attached in such a way that on reaching the bottom they automatically detach and so relieve the fisherman from drawing up the page 209 heavy sinker in addition to the fish. The Tongarevan method of attaching the sinker to the hook by a forked branch differs from the usual method, the use of the detachable loop. The local names for the parts of the hook and the sinker attachments as described by Nordhoff (18, pp. 42, 43) are shown in figure 61.
Figure 61. Ruvettus hook: a, parts of hook; b, c, sinker attachments from Tongareva and Cook Islands (after Nordhoff). a, parts of hook: 1, shank limb; 2, point limb; 3, bend (kopu); 4, knob (reke) to support lashing; 5, lashing of snood and shank; 6, snood (matai); 7, point (reinga); 8, lashing, point limb to vertical end of point. b, Tongarevan attachment: strip of hala (1) is knotted at one end (2), which is passed through split made in strip to form running noose around coral sinker (3); other end is tied to straight end of forked stick (4); fork (5) is hooked over bend of hook (6); baited hook is lowered and, when sinker touches bottom, forked stick falls away from slackened line, and sinker is detached. c, Cook Islands attachment: longer strip of hala (1) in similar method attaches sinker (3), but upper end of strip (1) is passed over hook bend, brought around shank limb, and end (5) is passed under loop; weight of stone presses short end against hook and prevents it from slipping; on release of pressure when sinker touches bottom, loop slackens and frees sinker.
3. Shark hooks. Composite wooden shark hooks are made of ngangie wood and differ from Ruvettus hooks in that the shank leg and point leg meet at a curve rather than at an angular bend. The separate point, also of ngangie wood, is curved so as to carry the point of the hook toward the shank limb.
The wood is hard and tough and does not break easily. Gudger (7, p. 230, 232) infers that the shark hooks examined by him were made of the roots, but it is unnecessary to suppose that the growing roots were trained to the required curve. The ngangie is a stunted plant with very crooked branches, and naturally curved branches were selected. The curve could be exaggerated in young growing branches by bending and tying them in position, and the subsequent growth would make the curve permanent. In New Zealand the young growing branches of straight branched trees such as the tanekaha (Phyllocladus trichomanoides), were formed into page 210 a loose overhand knot and allowed to grow into a permanent curve for wooden fishhooks.
Two shark hooks collected in Tongareva by the Wilkes Expedition are now in the United States National Museum. If the hook illustrated by Wilkes (31, vol. 4, p. 287) is intended to represent either of the two in the National Museum the artist, according to Gudger (7, pp. 230–231), has not made the bend sharp enough. (See fig. 62, a.)
Figure 62. Shark hooks collected by Wilkes Expedition: a, hook figured by Wilkes, no description, probably meant for hook no. 3674 in National Museum, but bend (1) made too rounded; b, hook figured by Hough, National Museum no. 3674; c, hook after Gudger (7, p. 230, fig. 19), described as largest hook in National Museum.
Hough (12, pl. 26, no. 1) pictures a large hook, similar to the one represented by Wilkes in the lashing of the snood, and a sharper bend. (See fig. 62, b.) In Hough's description the hook is wrongly labelled “Fiji,” and Gudger (7, p. 231) states that the same hook was labelled “Tahiti.” H. W. Krieger, Curator of Ethnology of the U. S. National Museum, says in correspondence that it is the larger of the two Tongarevan hooks in the Museum, and gives an additional description:
The shank leg (straight line measurement on outside) is 9 inches, the barb leg 4 inches, and the barb (point) measures from lashing tip 2.5 inches. The tip of the barb (point) approaches to 15–16 of an inch of the shank leg. The attached cord of sennit is 44 inches in length. It is a flat braid of three-ply twisted fiber of coconut palm which tapers toward the knotted end. At the shank end the cord is separated into its component parts which are again subdivided into two-ply strands lashed about the shank end where they are partly embedded in an encircling groove. Loose strands are used to cover 7 inches of the proximal end of the attachment cord where it is attached to the shank. This ferrule is loosely plaited. The point of attachment of barb [point] to barb leg is ferruled with a two-ply braid of sennit like that of the lower end of the attachment cord. The form of the lower end of the hook made by shank, barb leg, and barb [point] is roughly that of an acute triangle with slightly rounded angles.
Gudger also describes the two hooks in detail. As he states (7, p. 231) that the smaller of the two hooks is more like Wilkes' and Hough's figures and states that Hough's figure is taken from the larger hook, some confusion is evident. Gudger states that a long-fibered material was used to seize the lashing in both hooks. The textile-like stipule (kaka) from the page 211 base of the coconut leaves was used in Manihiki to protect the lashing and snood, and as this is the only fibrous material besides sennit available in Tongareva the material mentioned by Gudger is probably the same.
A metal hook with two long limbs meeting at an acute angle with the point bent in toward the shank leg is said to have been introduced from the islands to the south.
Characteristic of Tongarevan fishing is the short line, the common use of which was rendered possible by the cultivation of diving abilities, and may have been necessitated by the lack of fiber. The hooks used, however, are not peculiar to Tongareva. The wide U shaped pearl shell hook is found in Manihiki, Rakahanga, Pukapuka, the Tuamotus and Tahiti, and the composite wooden shark hook has a wide distribution. The third form of circular hook used with an ordinary length line from a canoe at anchor also has a fair distribution. The use of but two simple pearl shell hooks in Tongareva is in contrast with the greater number of forms in use in other atolls such as Manihiki, Rakahanga, Pukapuka, and the Tuamotus, where pearl shell material was also abundant. The method of driving flying fish is known elsewhere. Kennedy (13, pp. 61–63) describes its use in the Ellice Islands. Handy (8, p. 176) describes the method of driving porpoises in to the reef in the Marquesas.
That diffusion of the Ruvettus hook into the Society Islands, Tubuai in the Austral Islands, and Anaa in the Tuamotus is comparatively recent has been shown by Nordhoff (19, pp. 224, 225). The contention of the Tongarevans that it is a recent introduction to their islands is also probably correct. In most localities Ruvettus fishing requires a great length of line. Tongareva lacked suitable material for long lines and, though sennit fiber was available, the only line fishing from the surface was from an anchored canoe. Even for this use, the Tongarevan fishing line was comparatively short, the same length as the anchor rope. Nordhoff associates Ruvettus fishing with cannabalism, which, again, seems to have been absent in Tongareva.
The recency of diffusion of the bonito hook is more doubtful. The chief objection to the age of the bonito hook raised by the Tautuan people was that they had no drill. Though the proximal hole in the point base may be displaced by a groove, the distal hole and the hole through the head both require drills. If the bonito hook came by fairly recent diffusion it did not come from Tahiti, as the second lashing of the point is made over a distal prolongation of the point base toward the tail. The bend of the Tahitian hook is equally formed by the shank and the point, whereas in the page 212 Tongare van hook the bend is formed on the point piece by the proximal prolongation of the base. The Tongarevan point has affinity with the Samoan point, but Tongareva and Samoa differ in the principles of the lashing technique. In Samoa (29, p. 501) the snood is tied to the proximal hole of the point base, and an extra lashing element which extends from the proximal hole of the base point to the head hole is introduced. The method of attaching the hackle is also much different. The Tongarevan technique is similar to that of Manihiki and Rakahanga.
On the other hand, Kotzebue (14, p. 219) says, “We obtained from them some fishing hooks, which were two pieces of mother-of-pearl joined together, and wrought in the most tasteful manner, perfectly resembling those of the Sandwich Islands.” The Hawaiian islands (Sandwich Islands) bonito hook has a simple point without prolongation of the base, and the points themselves are of bone. Though the resemblance cannot be perfect, it may have seemed so to Kotzebue, enough at all events to form evidence for the presence of the bonito hook in Tongareva if Kotzebue's hooks can be located and examined. Wilkes (31, vol. 4, p. 287) shows a group of hooks, one of which may be a bonito hook, but Kreiger says in correspondence that the hooks shown are apparently from several sources.