Palu Hooks.

As characteristic an ethnological feature of its especial region as the boomerang of Australia or the bola of South America, is the wooden deep sea fish-hook from the Central Pacific. All authors in dealing with this remarkable type of large wooden hook from Micronesia and Polynesia have termed it a "shark" hook. In the preceding pages, (p. 199) a description by Mr. Louis Becke is given of the "shark," for which this instrument is intended. This excellent account, though not couched in technical language, clearly indicates that the fish in question, the "palu," is no shark, and has suggested to Mr. E, R, Waite the idea of some Macruroid,

"This peculiar fish," writes Becke, "is, as far as I know, only found in the Tokelau, Ellice, and Kingsmill Groups, and at the isolated islands of Pukapuka (Danger Island), Suwarrow, and Manahiki. I do not know for certain, but I have been told by many intelligent natives that the palu is never to be found among the high islands, such as the Fijis, Samoa, New Hebrides, &c. "He also mentions catching palu at Nieue.

Tracing the geographical distribution of this hook, we note it recorded from Nanomea,^* by Brill; from Nukufetau in the Ellice, Nukuor in the Carolines, and Tarowa in the Gilberts, by Dr. Finsch;^† from Nukulailai, Nieue, Tamana, and the Union Group, and possibly an eccentric type from the Louisiades,^‡ by Edge-Partington, and the latter also by Macgillivray; ^§ a drawing of a Penrhyn Island hook, by Wilkes,^‖ may be intended for this type; while a huge form is represented in the Australian Museum from the Mortlock Group, and another variation is pictured from the Trobriands by Finsch.^¶ A specimen resembling the latter, said to come from Milne Bay, B.N. Guinea, was lately procured by Mr. Norman Hardy at Samarai, and will be described shortly in the Proceedings of the Linnean Society of New South Wales.

Lister^** figures a palu hook from Fakaafu, and from Atafu, Dr. Coppinger^‡‡ procured "a large wooden shark-hook, with rope snooding made of coconut fibre." A modification of the usual pattern is shown from Fiji in the Macleay Museum, Sydney, agreeing with a figure by Edge-Partington.^‡‡

The exact shape of the manufactured article depends on the growth of the fork from which it is hewn, and therefore exhibits considerable variation, especially in the angle in which the limbs diverge. I procured on Nukulailai rough forks (fig. 39) such as schoolboys select for making catapults, in the bark, intended for palu hooks. I recognised the bark, and the natives further informed me that the wood was "vala vala," (Premna taitensis). Dr. Finsch supposed that mangrove furnished the material of the Gilbert Island hook he described.

Fig. 39.

Fig. 40.

In Tahiti, Ellis tells us that the wooden shark hooks, a foot or eighteen inches in length, were cut from the roots of the "aito" tree (Casuarina equisetifolia), an exposed growing root of which was sometimes twisted into the shape desired for the future hook.^*

The separate barb is roughly L-shaped, one limb being bevelled to form a scarf-joint with the shank, the other carved into the exact shape of a fowl's spur, to which, when affixed to the shank, its resemblance is increased by occupying the same relative position to the limb of the shank as the spur does to the fowl's leg. The joint is completed by a whipping for its entire length of flat sinnet. The most striking peculiarity of the palu hook is the extent to which the entering barb is carried, almost closing the loop of the hook. As the length of the barb is proportionate to the size of the hook, the size of the aperture is decided, not by the length of the barb but, by the divergence of the limbs of the shank. The specimen figured is extremely narrow, a quarter of an inch only separating the point of the barb from the opposite limb of the shank. Finsch's Tarowa hook exhibits an opposite extreme of width which can be matched in a hook from Nukulailai, where three-quarters of an inch intervene between barb and shank. If the hook is held before the eye so that the shorter limb of the shank appears super-imposed upon the longer, the barb is usually seen to be slightly deflected to the right. When, as in the Mort-lock hooks, this feature is exaggerated, the complete hook is thrown into an ascending spiral. Considerable diversity exists in the method of splicing the barb to the shank. In the El lice Islands the face of the joint is in a plane at right angles to the plane of the hook, but the Funafuti craftsmen attach the barb to the inner face of the shank, whereas the men of Nukulailai fasten it (as is shown in the barbless shank on Finsch's plate, and as Edge-Partington correctly figures it) to the outer side, as do also the fishers of Fakaafu.

Reference has previously been made to a series of hooks from the Mortlock Group^* in the Australian Museum. Compared with the Ellice hooks these are enormous, the largest weighing one pound fifteen and three-quarter ounces, and measuring seventeen and a half inches. Grooves gnawed by captured fish upon the shanks attest their genuineness, and their size suggests that they were intended for a form of palu larger than that taken in mid-Pacific. In all points of construction they conform to the smaller type except in the setting of the barb. Here the scarf-joint is cut in the plane of the hook, that is, at right angles to the Ellice Island joint.

The longer or unarmed limb of the shank terminates in a knob on the outer side, half an inch below which is carved a smaller projection. The cord of attachment is a piece of round plaited coconut rope (oukafakanapoua) about two feet in length; the loop in which it ends is slipped over the smaller projection of the shank, and the cord lashed securely to the inner side of the shank by sinnet passing between the knobs. In the Mortlock hooks the cord of attachment terminates distally in a loop, evidently for "bending on" the fishing-line, in which it agrees with the Gilbert Island type: in the Ellice a knot ends this cord.

One Mortlock specimen has a straight stick, fourteen inches long and half an inch broad, so lashed on to the cord of attachment as almost to hinge to the long limb of the shank. A somewhat similar but not identical method of mounting the palu hook is shown by Edge-Partington^* in an instance from Niue. No Ellice hooks present this feature, but we cannot assert that they may not also be thus prepared for service.

Mr. O'Brien told me that the bait was a whole fish split and laid scale to scale upon either side of the barb. In bolting this the palu, whose jaws are very thin and pliable, gets the barb caught behind the angle of the jaw. Sometimes, when the fish bites, the line is so jerked as to bang its head with the flat stone used as a sinker.

Finsch gives the name of this hook in the Gilberts as "tingia," the name of it on Funafuti is "kou boru."