Line Fishing

Line fishing (hi ika) was the favourite method of capturing fish and a great number of hooks and fragments have been collected in various localities. Valuable local studies have been made by Skinner, Fisher, Duff, Fairchild, Lockerbie and others. The castor-oil fish (Ruvettus ruvettus) and bonito were not present in New Zealand and the specialized Polynesian methods of catching these fish were thus absent though modifications of the Polynesian bonito hook are present Dried shark was much esteemed by the Maori and this led to the development of more varieties of shark hooks than in central Polynesia. Bone as material for one-piece hooks or the points or point limbs of composite hooks was much in favour and the occurrence of the bones of the extinct moa led to their extensive use in the South Island. Human bone was also used, and the mean custom of insulting a chiefly enemy by using his bones for catching food may have increased the number of human-bone fishhooks. Other sources were whale bone and Fisher (42) has recorded the clever use of the lower jaw bone of dogs in the Thames district.

The fishing lines (aho) were made of dressed flax fibre which was rolled (miro) on the thigh into two-ply twisted cord. Lines varied in thickness according to the size of fish and fine two-ply twisted thread was made for lashing the snood and the points of composite hooks. In Taranaki, where some fishing places were perched high up on cliff girt promontories, the lines were correspondingly long.

Hooks (matau) varied in size, shape, material, and construction. A hook may be described as being composed of a shank limb and a point limb with a bend between them (Fig. 51a). The point limb bears the technical point (mata) which may be straight, incurved, or barbed. The shank limb usually has a knob or projection (koreke) on the outer side of its upper end to prevent the snood lashing from slipping off. In some page 217Maori hooks there are two lateral knobs instead of one outer knob. The shank knob, which may be termed the snood knob from its function, is characteristic of Polynesian hooks and offers a marked contrast to the eye in European metal hooks. Sometimes the snood knob is replaced by grooves to assist in the snood lashing. The Polynesian bonito hooks have

Fig. 51. One-piece hooks.
a, parts; b, c, d, Manihiki pearl-shell hooks; b, U-shaped; c, subcircular; d, circular.

a transverse hole through the head of the shank, not for the snood line itself but for the lashing thread which passes through it and around the snood which is primarily attached to the base of the point piece (Fig. 57). The snood which is also termed the snell, is a short length of line attached to the shank limb all ready to attach to the fishing line. Hooks are conveniently divided into simple hooks made of one piece and composite hooks made of more than one piece.

Fig. 52. Maori one-piece hooks, after Skinner (76). a, subcircular; b, inturned point; c, U-shaped.

The one-piece hooks of central Polynesia, made of pearl shell, are in two main forms, U-shaped and circular. The U-shaped form has straight limbs which are fairly parallel while the circular form has curved limbs which make an even curve with the bend. Intermediate forms occur in which one limb may be more curved than the other to form a subcircular variety or in which fairly straight limbs are not parallel. Usually the points are incurved as in the Manihiki hooks in Figure 51b-d. Variations occur in the curve of the bend, the treatment of the shank knob and of the point, and the respective lengths of the two limbs.

One-piece hooks were common in New Zealand in the U-shaped and subcircular forms. Skinner in his classification of Murihiku hooks (76) placed the one-piece hooks in his type 1 with three varieties; subcircular hooks (Fig. 52a), with an inturned point (Fig. 52b), and U-shaped hooks with a straight point (Fig. 52c). The true circular form was not present in his study material. Most of the Murihiku hooks were made of moa bone and some of human bone. In the North Auckland area, numbers of small hooks were made of paua shell in the U-shaped and subcircular forms. Rectangular blanks for bone hooks were cut out with stone flakes, the outer form shaped with stone rasps, and the inner part removed by drilling and smoothed off again with stone rasps.

The composite hooks, with the exception of the trolling hook for kahawai, were made of two pieces and they have a wide range of form. The shank limb was formed of wood, bone, or stone and the point limb or point was usually of bone and rarely of shell or stone.

Fig. 53. Composite hooks.
a, bone, after Skinner (76); b-d, wooden shank limbs.

The composite hooks which resemble the simple one-piece hooks in form and classed by Skinner (76, p. 217) as his type 2 have two separate limbs with curved lower ends which complete the bend. The shank limb usually has a vertical groove at its lower end into which a convex face on the point limb fitted to make a more secure join (hononga). The under surface of the bend in both limbs was usually prolonged downward to provide grooves or notches for the transverse lashing turns which bound the two parts securely together (Fig. 53a). In these hooks, both limbs were usually of bone.

In hooks with wooden shanks, particularly shark hooks, the point of junction with the bone part varied considerably. In some, the shank limb (papa-kau-awhi) was prolonged to form the bend (Fig. 53b); and in others, it was further extended to form the point limb to the end of which a short bone point was lashed (Fig. 53c). In the North Auckland area, the page 219bend was produced artificially by tying the stem of a young tanekaha plant in a loose overhand knot so that the natural growth fixed the curve in circular form (Fig. 53d). The snood knob and the bait cleat were sometimes carved. The bait cleat was a projection on the under side of the bend around which turns were made with a bait string (pakaikai) tied to the upper end of the shank limb. Usually it was in the middle line but sometimes there were one or two lateral cleats and more rarely a median and two laterals. Some hooks had a lower median notch (Fig. 52b) for the bait string instead of a cleat.

Fig. 54. Branched hooks.
a, b, Maori, after Fairchild (37a); c, Niue, after Beasley (10, pl. 50); d, long-shanked hook, after Hamilton (46a, fig. 23).

Some wooden hooks resembling the Polynesian Ruvettus hook but smaller and with more nearly parallel limbs were discovered by Fairchild (37a) in a cliff cave near Manukau Heads. He states that the parallel limbs were produced by tying two young branches the required distance apart and allowing them to become fixed by growth. The angular bend formed by the natural junction was strengthened (whakaŭ) by figure-of-eight turns with a thread around each limb above the bends. The upper end of the shank limb was provided with a snood knob and the point was formed of a curved piece of shell (Fig. 54a) which was let into a slot at the upper end of the point limb (Fig. 54b). Beasley (10, Pl. 17a) figured some good specimens in the British Museum. Some Niue hooks closely resemble the Maori form except that they have a median projection below the bend and do not have the strengthening lashing above the bend (Fig. 54c).

A peculiar hook with a long, slender, wooden shank limb, straight or slightly curved, and with a short, barbed bone point was figured by Beasley (10, Pl. 42) and tentatively attributed by him to Tahiti. However, a gummy material over the point lashing caused doubt as to the locality. Skinner (76, p. 256) subsequently examined some similar hooks without gum and, by identifying the lashing thread as New Zealand flax fibre, he established the correct locality and classified them as his type 3. He suggested that the gummy material on the British Museum hooks was probably put on by the collector to preserve the lashing. A specimen with a shorter bone shank limb is shown in Figure 54d.

Fig. 55. Barracuda hooks
a, outer snood knob (Hamilton, 46a, fig. 24); b, inner snood knob (Beasley, 10, pl. 10); c, perforation (Skinner, 74, fig. 52).

A specialized but rather crude hook termed okooko was used for catching barracuda (mangā). The straight and somewhat massive shank was formed of a dark wood such as rimu and the distal end was perforated for an unbarbed or serrated bone point which was sometimes rendered more secure by a small wooden wedge driven into the hole. The upper end of the shank had the usual outer snood knob (Fig. 55a) but rarely the snood knob was on the inner side (Fig. 55b) and some had an antero-posterior hole (Fig. 55c). After European contact, a bent nail usually replaced the bone point. The hooks were used with a short line attached to a stout rod about six feet long. When the hook was threshed about on the surface, the voracious barracouda readily swallowed the lure without any bait.

Some interesting hook points were excavated by Fisher (42) at Oruarangi in the Thames district. Some were made from the lower jaw bones page 221of dogs (Fig. 56a), the horizontal part forming the point limb and the ramus cut on a curve forming the point (Fig. 56b). A number of points with an inner barb were made from the canine teeth (Fig. 56c, d) and a simple hook was made entirely from one of the molars. A form of point piece was made from pieces of the lower jaw with a straight limb, incurved point, and outer barb (Fig. 56e, f). These specialized points were termed the "Oruarangi point" by Fisher and they resemble those of some Hawaiian circular hooks (Fig. 63a). The fishermen of Oruarangi did not waste much of the dog's lower jaw.

Trolling hooks.—Some types of hook indicate a memory of the period when the Maori ancestors trolled for bonito in the seas of central

Fig. 56. Fishhook points (after Fisher, 42).
a, dog's jaw; b, point from jaw; c, d, points, canine teeth; e, f, points with outer barbs.

Polynesia. One form figured by Teviotdale and classed by Skinner as his type 4 was excavated from a moa-hunter camp at the Shag River (Fig. 57a). The stone shank with a proximal pointed head, triangular in section with a median front edge and a transverse hole through the head, closely follows the form of the Polynesian pearl-shell shanks of bonito hooks. However, the Polynesian shank head is formed by the thick hinge of the pearl shell, but the part beyond the hinge formed by the inner surface of the shell is flat. In the stone shank, the triangular section of the head with its front median edge is continued to the distal end except that the median edge is levelled off at the far end to fit the base of the bone point. The base of the bone point is evenly expanded to provide sufficient width for two perforations. Two grooves are present on the under surface of the shank, corresponding with the holes in the base of the point, and it is evident that the point was lashed to the shank by transverse turns passing through the holes and around the shank grooves, page 222

Fig. 57. Trolling hooks.
a, Maori, after Skinner; b, Fanning Is.; c, Hawaii; d, Maori, after Duff; e, Samoa; f, Manihiki; g, Maori, after Skinner; h, Tahiti.

page 223thus closely following the present Polynesian technique. The evenly expanded point base with two holes is similar to that of an old bonito hook dug up in Fanning Island (Fig. 57b). A modified form has survived in Hawaii in which the base width is sufficient for only one hole (Fig. 57c) through which both the shank lashing and the snood pass.

Another trolling hook, discovered in the moa-hunter deposits at Wairau and figured by Duff (30, Pl. 9, g), raises the subject of the treatment of the point base for attachment to the shank. In addition to the evenly expanded base, the base may be prolonged on the proximal side towards the shank head or on the distal side. The Wairau hook (Fig. 57d) has a bone point which is prolonged on the proximal side to provide sufficient width for two holes and it thus resembles the western Polynesian technique as exemplified by the Samoan bonito hook (Fig. 57e) in which the snood is attached to a hole in the point base. This form of point base was also present in Manihiki but the snood was looped around the point base instead of being attached to a hole (Fig. 57f). The stone shank of the Wairau hook is similar to that of the Shag River hook but the pointed head had been broken off, probably at the original transverse hole, and the owner had drilled another transverse hole to keep the hook in use. As the Polynesian bonito hooks had a hackle of pig's bristles or fibre attached to the shank by the distal lashing of the point, it is probable that the Maori hooks also had hackles, probably of feathers.

Perforated points were rare in Skinner's study material but he figured some other bone points (76, Fig. 49) which were obtained from the same excavations as the perforated point. These have a distal prolongation of the base and both Teviotdale and Skinner associated them with flat bone shanks found in the same excavation (Fig. 57g). The mode of attachment to the shank probably resembled that of the Tahitian bonito hooks except that the Tahitian points have one perforation in addition to the distal extension (Fig. 57h). From the archaeological finds, it is evident that all three forms of point were in use during the moa-hunter period.

A number of stone shanks have been found in the North Island, mostly on the west coast, but they differ in being circular in cross-section. Attention has been drawn to their resemblance to stone shanks from the Gilbert Islands and Tikopia. Some differ again in having the hole through the head bored from before back instead of from side to side. So far as I know, none of these stone shanks have been definitely associated with points, let alone a complete hook with the lashing intact.

Bone shanks, elliptical in section, have also been found in both islands, with the head hole bored from before back and the distal end bevelled and grooved for the point base. Some have a knob on each side of the head (Fig. 58a) which would seem to indicate that the snood was primarily attached to the head. Fortunately, some specimens have been page 224preserved with points and snoods complete. Three are shown in Figure 58 and though all have the two lateral knobs, the snood in b and d is looped around the end of the point lashing; and in c, some threads of the point lashing have been twisted together to form the snood line. The snood is carried along the front of the shank to the head perforation but it is not clear from the illustrations of b and c what exact technique was used. In specimen d, a knotted line appears to have been passed through the head hole from the back and the line from the point lashing attached to it in front. All three points have an outward projection at the distal end of the base to prevent the lashing from slipping over the end. Parts of a feather hackle remain attached by the point lashing which proves that the hooks were trolled as lures. The technique of the snood loop around the point base is similar to that of the Manihiki bonito hook (Fig. 57f).

Fig. 58. Trolling hooks.
a, bone shank, after Fisher (42); b, c, after Beasley (10, pl. 11); d, after Hamilton (46a, fig. 22).

A different type of trolling hook continued in active use long after European contact and it bore the widely spread Polynesian name of pa which is still applied to bonito hooks. It was made for catching kahawai and was distinguished by the specific name of pa kahawai and also paua from its inlay of paua shell. The stone shanks, which copied the pearl-shell shanks, were lacking in the iridescence which formed an important feature of the pearl-shell shanks in attracting fish. The bone shanks lacked the same quality. It is apparent that some fishermen, in seeking a local shell substitute, selected paua shell but as it was too thin to form strong enough shanks, they backed the shell with wood and sometimes bone to form a composite shank which would stand the strain (Fig. 59). The page 225Tongans followed a similar technique in backing large bone shanks with pearl shell on the outer side but the Maori hooks have the shell on the inner or front side.

Fig. 59. Pa kahawai trolling hook.

The bone point of the pa kahawai is barbed and its base, though widened, is not perforated but is notched on its curved distal side for the lashing, as shown in Figure 60.

In Figure 60a, the lashing thread is fixed as shown on the proximal side of the distal knob of the shank and the turns also fix the end of the shell inlay. The shank is turned sideways in b and the thread carried over the top notch. In c, the turns are carried around the shank and over the other two notches in turn and then commence a figure-of-eight turn around the shank and point. The figure-of-eight turns cross on the proximal side of the point d, continue in sequence e and working upwards finish as in f.

Fig. 60. Point lashing.
a-f, stages of technique.

This technique differs markedly from that of the bonito hooks, owing to the notches used instead of perforations through the point base.

The snood of the hook I unravelled consisted of a two-ply cord, each ply being formed of a smaller two-ply cord. The end had been untwisted page 226into four plies, of which two had been cut short and two left long. The two long plies were wound around the proximal end of the shank below the two lateral knobs as shown in Figure 61a-d.

In Figure 61, a shows the front of the proximal end of the shank with two lateral knobs and a mesial notch. In b, a thread has been tied around the snood and the unravelled four plies below it are placed in position. One of the long plies is wound obliquely below the right knob and passed transversely across the back. In c, the long ply ascends obliquely from the left to pass around the snood in the notch, descend obliquely below the right knob, around the back and across the front, thus binding down the two short plies and the second long ply. In d, the second long ply is turned to the left, makes a loop around the snood on the back of the shank, and a transverse turn around the front.

Fig. 61. Snood lashing.
a-d, with unravelled snood; e-h, with lashing thread.

The unravelled plies of the snood have anchored the snood to the shank but the fixation is completed by the thread attached to the snood as in Figure 61 e-h, in which the turns shown in a-d are left out for the sake of clarity.

In e, the free end of the binding thread is laid vertically against the snood and shank and commencing from the tied end, the thread is taken obliquely down under the right knob, across the back, and upwards from below the left knob to cross its first downward turn. Both oblique turns cross the free end of the thread and fix it in position. All turns are made with a long loop, the short shank being easily turned in and out of the loop to prevent tangling. In f, the thread passes obliquely down on the page 227back from the position in e to below the left knob, crosses transversely in front to below the right knob, and obliquely upwards on the back to the left of the snood. The thread is then in the commencement position, as in e, and it repeats the turns made in e, close below the preceding ones. The oblique turns in front cross over the vertical end of thread but the horizontal turns do not. The oblique and transverse turns in front and back are continued in sequence, as in g; the upper parts of the turns covering the upper part of the shank knobs and the lower parts of the turns passing below the knobs. With the lashing turns completed, the rest of the thread forms a slack loop. In h, the upper free end of the thread is pulled to remove the slack and the thread end is seized around the snood.