These fish-hooks represented to the Ellice Islanders of past generations their most valued treasures. Apart from their intrinsic worth they acquired, as conveying a maximum of wealth in a minimum of space, an artificial value approximating to the coins of more advanced civilisations. Instances have been given of their presentation to the gods (p. 47), of their burial with the owners (p. 53), and of their transmission from atoll to atoll by page 267Frigate-birds (p. 59). In Tonga the hook of the god Tangaloa was an heirloom preserved for many generations.

In this Archipelago their value was heightened by the rarity and inaccessibility of the shell, (Avicula cumingii) from which they are manufactured.; hardly any are found at Funafuti, and the Group is principally supplied from a bed in the Lagoon of Nukulailai, whence they are procured by expert divers.

This type of hook is universal throughout the Pacific, being used alike by Melanesians, Polynesians, and Micronesians. Besides those collected by the Expedition, the Australian Museum contains instances from Manihiki and Mortlock Islands, and the Gilbert and Hawaiian Groups. Among Edge-Partington's sketches may be recognised further instances from Danger Island, Strong's Island, Tahiti, Tonga, and the Solomons.^* In addition, Finsch quotes this type from the Carolines, the Marshalls, and the Marquesas.^† In New Zealand, where the substance of which it is usually manufactured does not exist, the Maories found in the shell of the "pawa" (Haliotis iris), a substitute for the flashing nacre of the Avicula. But this shell being too brittle to stand alone, is supported by a backing of "totara" wood (Podo-carpus totara). It is used, according to Hutton,^‡ for catching the "kahawai" (Arripis salar). The barb is itself single or double recalling the Tongan pattern. Specimens of this interesting variation lie before me in the Museum collection, and correspond fairly to the instances figured by Brough Smyth^§ and Edge-Partington.^|

The habits of the Bonito (Thynnus pelamys), for whom these hooks are intended, resemble those of its near relation the European mackerel; they eagerly rush at and swallow any attractive object, guided apparently by sight, not scent.

This sport is thus vividly described from another island by W. T. Pritchard^*:—"Bonita fishing is, perhaps, the most risky of all Samoan adventures. The natives start off at the dawn of day, and paddle far out to sea in the calm of the morning, and there trail their hooks behind the canoes, heedless of the brewing storm, and trusting to the strength of their arms and the fleetness of their skiffs, to reach the shore before its full force overtakes them. The bonita are found in ' shoals,' with birds hovering over them; and when these birds are still further out to sea, the fishermen bend to their paddles, and the canoes skim over the waves until in the midst of the ' igafo,' as the shoal is called. There the hook, still trailing from a long bamboo rod over the stern, is played to and fro, and as the bonita bites at it with a spring and a splash, he is tossed up with a jerk, and landed in the canoe with a shout and a cheer."

The bamboo does not grow in Funafuti, where the fishing-rods are chosen from the "miro," Thespesia populnea (p. 37). In Tahiti, the rod has bunches of feathers to imitate birds,^† In action the rod butt fits into a rope eye slung from the aftermost thwart (like a sprit-yard when it is shipped in an eye slung from the mast), it reclines in a raised rest carved on the after-decking of a Funafuti canoe (Plate xv.) At Simbo, in the Solomons, Mr. Hardy tells me that a bamboo scoop is drawn through the water to attract the bonito.

The shank "ba," of the hook is carved from an Avicula valve, so that a slice from the thinner part of the valve is attached to a thicker ridge from the hinge. A valve of A. cumingii, from which a hook had been cut, or rather I presume sawn along the sides and snapped off at the tail, which I procured on Nukulailai is figured (fig. 33) to illustrate the mode of manufacture. In one hook from Funafuti (fig. 34) the shank

Fig. 33.

^* Pritchard—Polynesian Reminiscences, 1866, p. 175; see also Wilkes—op. cit., v., p. 11.

^† Ellis—loc. cit., i, p. 148.

page 269is compound, being lengthened and strengthened by a strip of pearl shell, neatly fitted and lashed to the butt-piece. This is the only instance of such that has come to my notice, and doubtless was the result of economy in the use of a rare and valued substance. This hook is the largest of the series from Funafuti, being three inches and a quarter in length, but it is dwarfed by a specimen from Manihiki, six inches long. In weight it is six drachms nine grains. I did not see the whole process of manufacture, but such as I saw, nearly completed, in Funafuti were fashioned with but one Fig 34. tool, a small hard piece of Montipora coral called "lapa," with which the implement was rasped into the desired shape. The tail end of the shank is either made forked or square. The opposite thicker end of the shank is so designed to bear the perforation necessary for lashing on the fishing-line.

Fig. 34.

Fig. 35.

In the article (fig. 35) taken half-finished from the workshop, the perforation has not yet been made. This hole is drilled with a tool just like that figured by Fig. 35. Wilkes^* from Fakaafu, in the Union Group. No specimens of this existed on Funafuti when we were there, though they were described to me as having formerly been used tipped with Terebra maculata or Mitra episcopalis. Critical examination reveals that these perforations were not drilled from one side through to the other, but half through from one side to meet half through from the other. The face of the shank corresponding to the exterior surface of the valve was ground till the dull dark surface disappeared, the convex surface of the finished hook always presenting the most brilliant lustre. It is asserted by fishermen that a particular color of the nacre is preferred by the fish, and a hook is tried, polished, and re-polished till the exact play of light is obtained.

Fig. 36.

Fig. 37.

Fig. 38.

To the shank the barb is securely lashed by twine threaded through the perforation, the distal of the two lashings also serves to hold the beard • in the specimen figured (fig. 38) this latter is of European cotton thread, but usually it is of native fibre. The hook is made more secure by wedging on either side of it under the lashing, a piece of wood, which, in the examples at my disposal, is invariably from the mid rib of a coconut frond pinnule. Finsch^* describes such wedges as of bone or fish-bone splinters.

A hook which differs from the usual type is represented in the Australian Museum from Mortlock Island. This pattern has been noted from Strong's Island by Edge-Partington,^‡ and has been well figured from Mortlock by Finsch.^† It differs markedly by the shape of the barb, the angle at which it is set, and especially by its mode of attachment to the shank and severance from the fishing-line. The tail end of the shank is deeply cut by two pair of notches to which the barb is fastened by a species of "cross-seizing." The hinge of the Avicula is cut lengthwise to form the shank of this hook, not as usual across.

The hook with which the great god Tangaloa dragged up Tonga from the bottom of the sea, was described as "made of tortoise-shell, strengthened by a piece of the bone of a whale; in size and shape it was just like a large albacore hook, measuring six or seven inches long, from the curve to the part where the line was attached, and an inch and a half between the barb and the stem."^*

The fishing-lines attached to these hooks are always sold together with them; being required to endure tremendous strain, they are fastened to the hooks inseparably. In the Ellice, as in the Gilbert and Manihiki specimens, these are composed of Broussonetia, and are fine, white, three-ply cord, two to three mm. in diameter, of immense strength. In the words of Captain Cook,^† the Polynesians "make the best fishing-lines in the world: with these they hold the strongest and most active fish, such as bonetas and albacores, which would snap our strongest silk lines in a minute, though they are twice as thick." Dr. Finsch informs us that in the Carolines the fishing-lines were constructed of Hibiscus fibre, and that the Archipelago was chiefly supplied with this article from Nukuor.

Synopsis.—This kind of fish-hook may, on the model of systematic biology, be classified as follows:—-

Genus Trailed Pearl Shell Hooks.

Description.—Of two pieces, pearl shell shank and attached hook of the same or other substance, large, bearded, trailed on the surface without bait, principally employed for bonito; extends throughout the Pacific.

Type.—Fig. 38, p. 270.

Species A.—Type species.

Description.- -Shank mother-of-pearl, bored at thick end to attach fishing-line, which is then carried along the face of the shank and made fast to the barb, tail not serrated; beard and barb of various substances.

Locality—Pacific.

Sub-species A.

Description.—Metal barb, shank flat and notched to fasten fishing-line.

Locality.—Ellice Group.

Type.—Fig. 36, p. 270.

Species B.

Description.—Fishing-line not carried to barb, barb lashed to serrations on the tail of the shank, shank perforated for fishing-line.

Locality.—Mortlock and Strong's Islands.

Type.—Finsch, Ann. K.K. Naturhist. Hofmus., viii., 1893, pl. iii., fig. 1.

Sub-species B.

Description.—Shank notched or toothed, not perforated, for reception of fishing-line.

Locality.—Solomon Islands.

Type.—Edge-Partington, Ethnol. Album, ii., pl. ccix., fig. 5.

Species C.

Description.—Kahawai hook, shank of pawa face and wood backing, barb bone and double barbed at tip.

Locality.—New Zealand.

Type.— Smyth, Aborigines of Victoria, i, 1878, p. 392, fig. 230.

Species D.

Description.—Shank round, barb shaped like a scythe blade, no beard.

Locality.—New Guinea.

Type.—Finsch, Ethnol. Atlas, pl. ix., fig. 3.