The horizontal bottom beam is called the paetara (lower beam with points). It carries the wooden teeth, or tara, which are about 6 in. long. They are lashed to the under-surface of the beam with fine aka or with fibre of the Phormium tenax, and then a thicker piece of aka is woven in and out in figure-of-eight turns to finish off. The number of teeth are usually about two dozen. The two halves of the paetara are joined on a slant, and aka or fibre lashings passed through holes drilled on either side. The overlap in the paetara I saw in use was 17 in., and its total length 44 in.

Fig. 3.—Dredge-rake frame.

The bend at the sides is called the kauae (jaw). Besides bending upwards, the kauae bends forwards and is continued on into the upper limb, or peke. The two peke do not come close together in the middle line, but are separated by a gap of from 1 in. to 1 ½ in. This point is about 10 in. above the bottom beam. These ends of the peke have holes through them for lashing purposes. From the front, the plane of the kauae and peke forms an angle of about 45° with the plane of the paetara and teeth.

About 7 in. above the lower beam a horizontal rod, called the paepae, is securely lashed at either end to a hole in the ascending limbs. As further support there are two vertical rods, called kume, about 1 ft. to the inner side of the bends. They are made of manuka, with a fork at the lower end. The fork embraces the lower beam from behind, and the rod passes behind the paepae and ascending arms, to each of which they are securely lashed. The two kume and the paepae rods thus brace and strengthen the wooden framework, or kapu.

(b.) Heheki, Or Net.—The net is a bag net with a 1 ¾ in. mesh and about 34 in. long. At the end away from the frame it is wider, if anything, than at its attachment. It is attached to the paetara below, and the upper edge comes up as high as the paepae above, but is not fastened to it. It has a string attached to this upper edge, which is drawn taut and tied to the lower end of the pole or handle. The net has a special name, the heheki. In Museum Bulletin No. 2 Hamilton quotes Best as giving the name of the dredge-rake used at Rotoiti as heki. The Rotorua people were very clear that it is the actual net that is the heheki. In fig. 76 of the above publication a dredge-rake is shown with a punga, or sinker, attached to the end of the net. This is incorrect, as there was no necessity for it in this position, the kakahi weighting the net back as they were dredged up.

(c.) Rou, or Handle.—The handle was called the rou. In order to drag the rake along the bottom the handle had to be from 28 ft. to 30 ft. long. It was not a simple case of getting the longest pole from the adjacent forest, as Newman^* states in his article on Maori dredges. To get a pole of the right length without being too heavy or unwieldy, and yet with sufficient slimness and spring without being too weak, was the problem that faced the neolithic Maori. He solved it by joining four pieces together, thus obtaining length without excessive thickness. Of these four pieces the most important was the lowest, called the matamata. This was carefully sought for in the bush. It had to be a straight piece of toro (Myrsine salicina) or mapou (Myrsine Urvellei) of the right thickness. These woods are very springy, and will not break or snap. In the rou I saw in use the matamata was 12 ft. 4 in. in length and 3 ½ in. in circumference. The thin end was downwards, and near this end a groove was cut round, for a purpose to be detailed later.

The other three pieces were not so important, and the wood was not so carefully selected. The ones I saw were of-tawa (Beilschmiedia. tawa). The piece next to the matamata was named the whakatakapu. It was spliced to the matamata with an overlap of 11 in. and had the thick end down, and was 4 ft. 7 in. in length. The third piece was of the same length, and was called the whakangawari, and had an overlap of 11 in. The last piece was the one which was grasped by the hand, and hence was called the tango-tango (what one lays hold of). It was 9 ft. 6 ½ in. in length, and had the thin end uppermost, being here about 1 in. in diameter. The overlap with the whakangawari was 9 ¾ in.

The various parts of the rou were joined together as shown in the diagram, with a 9 in. to 11 in. overlap, by a double tie. These ties at the joints are called hotohoto.

When the dredge-rake was not in use the handle was untied, taken to pieces, and put in water to preserve it until the next season. In northern France the French farmers, after the pea crop is gathered, place the wooden stakes or pea-props in ponds for a similar reason.

Joining the Rou to the Framework.—The rou, or handle, having been completed, the lower end of the matamata is fastened to the kapu. It is passed down at the back of the two ascending arms (peke), and the groove already mentioned at the lower end is fitted on to the cross-rod (paepae) and securely lashed to it and the peke. The handle, peke, and kauae are now in the same

plane, and, as before mentioned, form an angle of 45° with the plane of the teeth of the rake. This insures the teeth gripping the sand or mud at the bottom when the rake is dragged.

The punga, or sinker, is then attached, not to the end of the net, but to the back of the matamata, between the cross-rod (paepae) and the ascending arms (peke), where in fact it is fastened to all three. Some fern is wrapped round the punga, before fastening, to save the woodwork. Its weight is about 6 lb. Should the weight of the sinker be insufficient, smaller sinkers, called potiki, are attached on either side of the main punga. It will be observed that the function of the punga has nothing to do with the net, but from its position at the lower end of the handle and directly over the middle of the frame it weights down the lower beam and causes the teeth to sink into the soft sand to scrape up the kakahi. The sinker described by Dr. Newman, in his article already quoted, as being flat at the base whilst the other side is rounded, was not so made that the broad flat surface should lie in the lake-mud, but that the flat surface might rest evenly against the back of the framework in the position described above.

Method of Dredging.—As foreshadowed in the proverb already quoted, kakahi dredging required great skill, or, as the Maoris say, He tino mahi tohunga. It was very difficult to get a good quantity, and the kuare, or unskilled dredger, was useless. It is said that skill descended in or was inherited by certain families. The Ngati-Pukaki were a skilled tribe. As there was so much talk about dredging, it is natural that a good deal of show was indulged in. The fisherman going out to the kakahi ground put on his best dress-cloak of dogskin or fine flax. The turuturu was driven in, and the canoe paddled off to the end of the attached rope. The dredge-rake was lowered over the left side of the canoe, and the end of the handle (tango-tango) held in the left hand. After feeling that the rake was on the bottom and that the teeth had gripped, the dredger would work towards the turuturu by successive pulls on the rope with the right hand. In olden days, when conscientious objectors were not even dreamt of, if a Maori held a stick in his hand and started moving it about his fighting-blood was speedily aroused. It is known of many a Maori of the old school, peacefully walking along with a walking-stick in the degenerate post-fighting days, that if he struck once or twice at a tree-branch or a piece of bracken an association of ideas seemed to stir the blood, and it was no uncommon sight to see him leaping about from side to side and going through all the strokes and parries of the ancient pastime. This would happen even with men using the ko in digging. So with the kakahi dredger: as he dredged along he had to move the handle from side to side; gradually the movement would excite him so that anon he was guarding and parrying with the handle of the dredge-rake, quite oblivious of the kakahi below. It was considered good training for war: hence my informants said, He karo rakau tonu te mahi (The method was just like guarding against a weapon). Probably some excitable warrior created a precedent and it became the fashion.

When a larger canoe was used there might be three or four dredgers all facing the same way, and were the angles of the handles of the rakes the same all would be well; but if one were different all would be wrong, and the rake that was out of line would immediately be drawn up, so that the fault might be investigated. The fault might be (a) the tying at the joints (hotohoto) of the handle, (b) the teeth of the kapu loosened or set wrongly, (c) more weight (potiki) needed. When the net became full the weight caused the handle to assume a more vertical position—ka tu te rou.