Throughout polynesia, the principal tool was the adze, made from basalt on volcanic islands and from tridacna shell on atolls. Authentic old hafted took are rare, except adzes from central Polynesia, and most of the material available for study consists of the stone heads which have survived the decay of their wooden hafts and fibrous lashing material. Adze heads are distinguished by the cutting edge being formed by one bevel on the back like European adzes and chisels, and axe heads theoretically by two bevels, one on each side like a European axe. Hafted adzes have the cutting edge at right angles to the line of the handle, and axes should have the cutting edge in the same line as the handle.

Adze heads are best described in the upright position with the cutting edge below and the bevel forming it, to the back. The four surfaces of a quadrangular adze are thus front (anterior), back (posterior), and two sides (lateral). The surfaces are defined by longitudinal edges which are characteristic of Polynesian adzes, whereas in Melanesian adzes they are rounded off, forming an elliptical cross-section. Some adzes are triangular in section; those with the base at the back having a back surface with two inclined sides in front and those with the base in front having a front surface with two inclined sides at the back. The sharp median edge forming the apex of the triangular section in both groups may be ground off to form an extra narrow surface or it may be absent because the stone used was not thick enough for the converging sides to meet. Such adzes are termed subtriangular.

The top extremity of an adze head is the poll and the upper part which is lashed to the haft is the butt, the part below being the blade. The front and sides of the butt, or the sides alone, may be reduced to form a better grip for the lashing and such implements are termed tanged adzes. In tanged adzes, the shoulder formed by the reduction defines the division page 180between butt and blade but in untanged adzes, the division is purely arbitrary and was determined by the craftsman who hafted the adze. The lower end of the back surface was ground to the bevel required to meet the front surface in forming the cutting edge. The junction of the cutting edge bevel and the back surface may form a distinct edge or it may be rounded off. Triangular adzes with the base in front have a triangular bevel and a wide cutting edge. Triangular adzes with the base at the back have the front median edge ground down at its lower end to form a front surface so that the posterior bevel may form a cutting edge, which is much narrower than in the other form of triangular adze.

Fig. 35. Terms and hafting.
a, b, terms; c, transverse lashing; d, triple triangle lashing (Rarotonga); e, triple triangle lashing (Mangaia).

The haft consisted of a shaft or handle with a foot set vertically at less than a right angle with the shaft. The shaft was usually selected from a secondary tree branch and a section of the larger branch was cut out with it to form the foot. The foot usually projected above and below the junction with the shaft, the part above being the heel and the part below, the toe (Fig. 35a, b). Some hafts have no heel and others have no toe. It is interesting that a Maori myth states that the form of the haft was derived from the human leg, the leg indicating the handle and the foot indicating the part to which the stone head was lashed.

Throughout marginal Polynesia, the general form of adze was quadrangular and the type of haft had a toe long enough to accommodate the page 181entire butt which was lashed to the toe with simple transverse turns of sennit (Fig. 35c). In such hafts, the heel did not function and in some hafts, it was trimmed off. In central Polynesia (Society, Cook, and Austral Islands), the form of adze in general use at the time of European contact was triangular in section with the base in front. A fair number of old hafted adzes from the Society and Cook Islands, preserved in museums, clearly indicate the form of haft and the method of lashing. The hafts had no toe and the middle part of the foot was grooved to fit the posterior median edge of the triangular adze heads. The adze heads were practically all tanged and the tanged butt was lashed to the foot by figure-of-eight turns which passed around the heel to form an ornamental as well as an efficient lashing (Fig. 35d, e). This technique has been described by me for the Cook Islands (99, p. 160) as the triple triangle lashing and a variation has been described for Mangaia (99, p. 165). Quadrangular adze heads have been collected in central Polynesia but I have never seen an authentic old hafting of one from that area.

The Origin of Polynesian Forms

The ancestors of the Polynesians must have made various forms of adzes before they left Indonesia on their eastward movement through Micronesia. On the volcanic islands such as Truk, Ponape, and Kusaie, they had basalt with which to reproduce the forms they were already acquainted with. However, the Marshalls and the Gilberts at the eastern end of Micronesia were atolls without any basalt. It may be assumed that the Polynesian ancestors spent some time in the eastern atolls before they moved further east. The stone adzes they carried with them from the last volcanic island must have worn out, thus forcing them to make adzes of shell. The shell material made it difficult, if not impossible, to reproduce the various forms of the basaltic adzes which they had previously made. Thus the lack of raw material created a gap in the direct transmission of ancestral forms of basaltic adzes from Indonesia. When the Polynesian ancestors reached Samoa and the Society Islands, the abundant supply of basalt enabled them to discard shell in the manufacture of adzes. The abundance of suitable material, however, could not revive the memory of forms they had never seen, let alone made. It would seem that they had to start all over again with the making of basalt adzes after they reached Polynesia. This may be partly the reason for the divergence of the popular forms of adzes made in the different groups of islands.

A quadrangular form was established early as indicated by its wide marginal distribution. The haft with a long toe could have been carried through Micronesia, for hafts had to be made for the shell adzes and the toe haft has a wide distribution throughout the Pacific (99, p. 442). The triangular form with the base to the back, termed "hog-backed" or "hoof-page 182shaped" (Fig. 36d), must have been a fairly early form for it is widely distributed. From its narrow curved cutting edge and heavy weight, it evidently served a widely spread need such as cutting out parallel grooves

Fig. 36. Adze subgroups.
a, quadrangular, untanged; b, quadrangular, tanged; c, triangular, untanged; d, triangular, tanged; e, inverted triangular, untanged; f, inverted triangular, tanged.

to be followed with the removal of the intervening wood by the quadrangular adze. Another form of triangular adze (toki kouma) was popular in the Marquesas. The triangular form with the base to the front (Fig. 36e, f) became popular in central Polynesia and evidently replaced page 183the older quadrangular form. The three groups of quadrangular, triangular, and inverted-triangular adzes were present probably in central Polynesia at the time of dispersal. Emigrants to other islands carried away some forms and developed others in their new homes.

New Zealand Adzes

New Zealand provided a rich variety of stone for the adze-makers. In the South Island, andesite, argyllite, greywacke, and a few other substances were utilized in addition to basalt. The prize material was nephrite or greenstone (pounamu) but, the supply being limited, basalt remained the most common material.

Skinner (75), from an intensive study of a large number of adzes from the southern half of the South Island as well as comparative material from Polynesia, classified them into ten types which include a number of varieties. All, with the exception of an axe-like form, he found to occur in various parts of Polynesia. Duff (32), after an additional study of adzes from the northern half of the South Island, verified the existence of all Skinner's types and varieties but for greater simplicity in study, he regrouped Skinner's nine types of adzes into four and kept the axe form as a fifth type. Duff's four types of adzes were based on the cross section of the blade above the cutting edge bevel and they are as follows: tanged-rectangular; rectangular without tang; triangular (with base in front); and inverted-triangular (with base at back). Three of the types have four varieties each and one has three.

Duff's classification by cross section is simple and can be applied equally well to the adzes of the whole Polynesian area. An objection occurs, however, to the use of the term "rectangular" for his first two types because rectangular means that the four angles made by the four sides are right angles. With rare exceptions, the Maori adzes so classified have the front surface wider than the back and consequently the angles cannot be right angles. The better term is quadrangular which means four sides with four angles, the kind of angles not being specified (Fig. 36a, b). Duff's classification is also inconsistent in that quadrangular adzes are divided into two types, tanged and untanged, whereas the two classes of triangular adzes are not so divided, though tanged and untanged forms may occur in each. Furthermore, the term triangular would be applied more aptly to triangular adzes with the base to the back (Fig. 36c, d) for if the adze head is laid on its back, the triangular section will be in its natural position with the base below and the apex above. Hence the inverted-triangular form is the opposite with the base in front and the apex at the back (Fig. 36e, f). In my study of Cook Islands adzes (99, p. 136), the general form of adzes with the base in front was described as "inverted triangular adzes."

The three forms of cross-section are too general to be restricted by the term types. It would be better to regard them as groups and refer to them as quadrangular, triangular (base at back), and inverted-triangular (base in front). Each group may be divided into two subgroups, according to whether the butts are untanged or tanged. The six groups so classified are illustrated by Figure 36 as quadrangular-untanged (a), quadrangulartanged (b), triangular-untanged (c), triangular-tanged (d), inverted triangular-untanged (e), and inverted triangular-tanged (f). Each of the subgroups is illustrated by three views which reading from the left are front, right side, and back and the cross-sections have the back, below. The triangular-untanged speciment (c) is Samoan and though the front median edge has been ground off, it is triangular enough to illustrate the subgroup. The triangular-tanged adze (d) is the "hog-backed" or "hoof-shaped" type so widely spread. The inverted triangular-tanged specimen (f) is from a moa-hunter's deposit which shows that the form was ancestral. The untanged specimen (e) is from the Cook Islands and is ground on all surfaces.

By using the group and subgroup arrangement, the confusion caused by the varying type numbers used by different authorities would be lessened. The group would correspond to genera in natural history and the subgroup to subgenera. As it would be difficult to apply further descriptive names, the system of numbering types and lettering varieties followed by Skinner and Duff seems the most practicable.

Types and Varieties

In reproducing ancestral forms in New Zealand or elsewhere, departures from the pattern must have occurred owing to different stone, shape of the spalls, and the varying skill of the craftsmen. The distinction between types and variations must be decided by the study of large numbers of specimens. Skinner (75, p. 147) has defined "type" as "a group of adzes which exhibit a general somatic resemblance comparable to the somatic resemblance between members of a single biological species…. Some of the adze types are further divided into varieties comparable with biological varieties within the species."

In the development of new forms, it may be assumed that variations which proved acceptable were copied by succeeding generations and so became types. Forms which are unique or few in number may be regarded as varieties. Most New Zealand and Polynesian adzes may be placed in the six subgroups already suggested. Doubt may be created by specimens with rounded edges and triangular specimens with the median edge rounded off or replaced by a narrow surface. Subtriangular forms may be treated as types or variations according to their numerical occurrence in a large series.

A variety in one island group may be a type in another group. The quadrangular form, with the front narrower than the back, is rare in New Zealand whereas it is the common type in Samoa. The occurrence of all the New Zealand types and varieties in various parts of Polynesia does not mean that all these forms were brought from central Polynesia. Some were introduced and others were developed independently. For more detail about Maori types and variations, the reader is referred to the works of Skinner and Duff already quoted.

The Hafting of Adzes

The term adze should indicate the complete tool composed of adze head, haft, and lashing. Complete tools are rare, however, and the studies have

Fig. 37. Hafted adzes.
a, toe haft (Best, 12); b, transverse lashing (U.S. National Mus., no. 21229, adapted); c, served lashing (Best, 12).

been made mainly on adze heads. It is convenient, therefore, to refer to the stone adze heads as adzes and to the complete tools as hafted adzes.

The wooden haft, formed of a shaft and a foot, was used throughout Polynesia and the Pacific area. Variations of the foot were made by discarding the heel or the toe. New Zealand followed the usage of marginal Polynesia in making the toe long enough to support the full length of the adze butt. The shaft was termed kakau and the toe, kauae. The toe usually had a transverse groove (tokari) to allow the lashing turns to be countersunk (Fig. 37a). The flat back of quadrangular adze butts was page 186laid against the flat anterior surface of the toe and the braid lashing (taka) was made in transverse turns around the adze butt and the tokari groove of the toe (Fig. 37b). Thin wooden wedges (matia) were sometimes driven between the face of the adze butt and the lashing to tighten the lashing.

A variation in the form of a single composite lashing is described by Best (12, p. 109). A running loop of braid was slipped over the shaft, run down close to the foot, and drawn taut. A number of turns were passed around the butt and the toe groove, and around the shaft. The braid was then twined around the turns in front of the butt to serve them and passed vertically between the butt and foot to tighten the transverse turns by passing around them. The braid was carried to the back of the toe where it was tied to the turns passing between the toe and the shaft (Fig. 37c).

In Polynesia, the lashing element was usually a three-ply braid of coconut-husk fibre termed sennit; and in Micronesia, a two-ply twisted cord of coconut fibre was used. In New Zealand, the substitute for sennit was a three-ply braid of scutched flax fibre and the persistence in the use of braid as against cord is illustrated by the following story told to me by Bishop Herbert L. Williams.

A dealer showed the Bishop a hafted Maori adze which he considered a perfect old specimen. Bishop Williams felt that there was some flaw in its alleged perfection but it kept eluding him.

"Well," argued the dealer, "you must admit that the stone is old."

"Yes," said the Bishop, "the stone part is old."

"The handle is old," persisted the dealer, "because it formed part of an old collection and the original lashing rotted away."

"Yes," agreed the Bishop, "the wooden part appears old."

"Well," triumphantly concluded the dealer, "the lashing material is old because it is part of an old Maori fishing line made of flax fibre."

"Ah," murmured the Bishop with a smile of relief, "now I remember what is wrong. Yes, the fishing line is old as a fishing line but a twisted cord was not used to lash adzes. The lashing material should have been a three-ply braid."

However, I saw two hafted adzes in the British Museum which were lashed with cord. They were catalogued as old but they may have been lashed after European contact. A ceremonial jade adze in the Otago Museum (Fig. 38d) is also lashed with a fine cord. Though braid is the correct material, rare exceptions evidently occur.

Axial Hafting

A form of axial hafting, or chisel hafting, for felling large trees has been described in detail by Best (12, p. 129). A double-bevelled implement page 187termed a toki ngao tu or poki, about five or six pounds in weight, was attached to the end of a spar which had been cut to form a flat upper surface with a shoulder for the butt of the tool. The lashing consisted of transverse turns and the cutting edge of the implement was held horizontally. For details of the cross bow arrangement and supporting scaffold, see Best. The implement cut across the grain of the wood and was used to make two deep grooves. The wood between the grooves was cut out by a heavy implement with one bevel termed a toki ngao pae which was hafted to a spar in a similar manner to the axe implement. The implement was wielded without a scaffold by two or three men, the axis of the cutting edge being held vertical. Fire was used to aid the deepening of the scarf. This technique is unique for Polynesia and was probably inspired by the larger forest trees of New Zealand. A special form of haft and lashing was devised for certain nephrite adzes.

Nephrite Adzes

The pounamu of the Maori, popularly called greenstone from its colour, is geologically termed nephrite and is a form of jade. It was discovered in rock boulders in the Arahura River on the Poutini or west coast of the South Island by the Maori in their testing of rocks for making tools. A myth was created regarding its discovery. Grey's account (44, p. 132) states that Ngahue, a contemporary of Kupe, brought his fish named Poutini from Hawaiki because of trouble with one Hinetuahoanga, who may be recognized as a female personification of the sandstone (hoanga) used to grind stone. Ngahue arrived at Tuhua (Mayor Is.) but Hinetuahoanga followed him up. Ngahue then fled to Arahura which he made a permanent abiding place for his fish. He took a portion of the fish with him on his return to Hawaiki. He killed a moa at Wairere and sailed from Cape Runaway. The portion of the fish was made into two adzes named Tutauru and Hauhauterangi, a heitiki neck ornament, and an ear drop named Kaukaumatua. Grey's informant was evidently a member of the Arawa tribe, for the succeeding story (44, p. 134) states that the Arawa and six other canoes of the Fleet were made with the greenstone adzes (toki pounamu) named Tutauru and Hauhauterangi made from the fish of Ngahue, who is further credited by the informant as the discoverer of New Zealand.

Greenstone was procured by expeditions to the Poutini coast, by barter, and through war. From its rarity and beauty, it was made into valued ornaments and from its taking a keen edge, it was worked into adzes, chisels, and short clubs. The adzes were mostly working tools but one form was purely ceremonial.

The working adzes in form were variations of the quadrangular group but Skinner (75a, p. 158, Figs. 43, 44) described two fine specimens of page 188tanged triangular form. In his study to test the influence of material on design, Skinner (75a, p. 161) found that, though the craftsman was able to impose old shapes on nephrite and greywacke, both these media imposed limitations on the craftsmen. Thus a full kit of greywacke or nephrite tools would be appreciably less varied than a full kit of basalt tools.

The ceremonial adze termed toki pou tangata was formed of a nephrite adze lashed to a carved haft. The adze was quadrangular, long, thin, and of medium width. It was untanged but some had notches cut on the sides of the butt and others had one or two holes drilled through the upper end of the butt. An interesting feature was the fine carving of the enlarged heel of the haft. Some were carved as single human figures, the head forming the apex of the heel, but the most attractive design consisted of a human figure with elongated lips and tongue forming a series of interlocking curves with the beaklike lips of an opposing manaia motif. The toe was usually shortened and grooved, forming a raised lower rim. In some, the adjacent parts of the shaft and the foot below the carved figures were cut down and roughened like the toe groove for the lashing turns (Fig. 38a). The end of the shaft had a knob (koreke) which was carved into a human head.

Owing to the shortening of the toe, the adze butt was fitted higher up on the foot. In hafts with the lashing groove confined to the short toe, the lashing consisted of transverse turns around the butt and the toe groove (Fig. 38b). The notches on the sides of the butt helped the lashing turns to prevent the adze head from slipping down and further security was provided by passing the braid through the holes in the butt and around some part of the heel. In hafts grooved on the shaft and the foot, it was possible to make figure-of-eight turns in the lashing which in some, resulted in the triple-triangle pattern (Fig. 38c) so characteristic of the Cook and Society Islands technique with toeless hafts (Fig. 35d). Some hafts departed from the general pattern by retaining a long toe which allowed the use of two main sets of transverse turns as well as subsidiary turns. In the specimen figured (Fig. 38b), the lashing is made with cord and the turns which pass through extra holes above, evidently pass through holes in the butt of the adze.

The occurrence of the triple-triangle lashing in connection with a practically toeless haft raises the question as to whether or not they were introduced from central Polynesia. The discovery of the inverted-triangular form of adze (Fig. 36f) in moa-hunter deposits adds further interest to the subject. However, it is not known how the inverted-triangular adze was formerly hafted in New Zealand. Both in central Polynesia and in New Zealand, the early inverted-triangular adzes could have been lashed to a long toe with a groove to fit the adze butt. I believe page 189that the development of the toeless haft and the triple-triangle lashing for the inverted-triangular form of adze took place in central Polynesia after the Maori ancestors had left that area. The toki pou tangata with its specialized haft could not have originated until after nephrite came into vogue and after the great development in carving. The carvers in fashioning a haft fit to carry a chiefly emblem, enlarged the heel in order to

Fig. 38. Ceremonial adze lashing.
a, carved foot (from field photo); b, transverse lashing (Best, 12, pl. 36); c, triple triangle lashing (Oldman coll., no. 491); d, complex lashing (Otago Univ. Mus.).

effect their full design; and to balance matters, they shortened the toe. In spite of the shortened toe, some craftsmen adhered to the older technique of transverse lashing but for better security, they notched the sides of the butt and also bored suspension holes through the upper end of the butt. Others sought better security by grooving the foot and the shaft and using figure-of-eight turns in the lashing. In making an orderly sequence page 190of turns, the triple-triangle pattern was developed independently but its comparatively recent development is indicated by the fact that the perfect pattern shown in Figure 38c is somewhat rare. It had not had time to supplant the older technique of transverse lashing. It would appear that the triple-triangle lashing was developed independently in central Polynesia and New Zealand through convergence in the type of haft.

The toki pou tangata was never meant to be used by a craftsman in adzing wood. It formed an exclusive article in the property of a chiefly family, to be borne on ceremonial occasions, to accompany the gestures of the family orator, and to lie in state on the breast of the chiefly dead. I, too, have been told that its name was derived from its occasional use to indicate prisoners selected for the oven but I felt that my informant was trying to convince himself as well as me. As the verb pou has the meaning "to establish firmly," a rationalization more in keeping with the prestige of the toki pou tangata would be "the adze which establishes man in authority."

Rotatory Adzes

Rotatory adzes, sometimes termed socketted adzes, are a specialized form in which the cutting edge could be rotated on a vertical axis to acquire the advantages of an axe for working in narrow spaces such as the bow part of a canoe hold. Rotation was obtained by introducing a short rounded staff as an intermediate vertical axis between the adze head and a toeless haft. The lower end of the axis was shaped to fit the adze butt to which it was lashed with transverse turns. The front of the foot was grooved to receive the upper part of the axis which was lashed to it by a separate lashing. The groove and the lashing formed a "sleeve", which, while supporting the axis, allowed it to rotate the cutting edge of the adze into the angle desired with the line of the shaft. The part of the axis carrying the adze head projected below the foot and took the place of the toe. The axis was thus a rotating toe. The term "socket" has been applied to this intermediate piece but as it is a misnomer as regards function, the term "axis" seems preferable.

Hafted rotatory adzes have been recorded from Hawaii and Tahiti. The Tahitian specimen in Figure 39a has a lower transverse lashing for the adze head and an upper triple-triangular lashing to complete the sleeve for the axis. The pointed upper end of the axis projects well above the heel of the haft.

Stone Axes

Stone implements with a cutting edge bevel on two sides have raised the question as to whether they were hafted and used as axes or were simply a variety of adze. Best (12, p. 136) conducted an exhaustive inquiry among old Maori but as the use of stone tools had ended long before the page 191informants were born, the evidence was partly late hearsay and partly rationalization. The best evidence was afforded by the remembered classification of the toki implements into two groups termed toki aronui or toki hangai and toki titaha. The words aronui and hangai mean full face or face to the front and they were applied to the toki hafted with the face to the front with the axis of the cutting edge at right angles to the axis of the shaft. The word titaha means to turn sideways or turn the face to the side and hence toki titaha implies that the face or aronui was turned to the side to bring the axis of the cutting edge into line with the shaft. Furthermore, the shoulder formed by the upper border of the cutting-edge bevel with the back surface was termed uma (breast) and informants stated that the toki aronui had one uma whereas the toki titaha had two uma. Thus the toki aronui had one cutting-edge bevel

Fig. 39. Uncommon hafting.
a, Rotatory (La Rochelle Mus., no. H526); b, lateral hafting (Best, 12, pl. 42).

characteristic of adzes and the toki titaha had two cutting-edge bevels characteristic of axes. Best pertinently asks why there should be distinctive qualifying terms such as aronui and titaha unless there were tools corresponding to the terms. When the Maori first saw the European axe, they called it a toki titaha. On the evidence, the implements with two bevels were termed toki titaha and were hafted as axes.

Though several quadrangular implements with two bevels have been collected, only one hafted specimen has been recorded. It was figured by Best (12, Pl 42) but is known to have been hafted after European occupation. The cutting edge is formed by equal bevels on the two wide surfaces, the foot of the haft is flattened on one side, and the adze butt is lashed to the toe by a braid of partly dressed fibre in transverse turns (Fig. 39b). The implement, as figured, is a laterally hafted axe and owing to its position on the side of the foot, the axis of its cutting edge is not page 192in the direct axis of the shaft but is parallel to it. The Mangarevan axes (97, p. 219), bevelled on two sides, were hafted to the front of the foot so that the cutting edge was in direct line with the axis of the shaft and the lashing was apparently effective (Fig. 41d). As it would be an advantage to have the cutting edge in direct line with the shaft, some doubt arises as to the age of the lateral technique supported by one specimen which was lashed long after such implements had gone out of use.

Side-hafted Adzes

A peculiar form of implement found mostly with moa-hunter remains in the South Island, has been described by Skinner (75, p. 166) and Duff (32, p. 158) as a laterally hafted or side-hafted adze. It is triangular in cross-section, the butt is markedly reduced to form a tang, and the cutting edge is formed by one bevel. The two specimens in Figure 40 show that

Fig. 40. Side-hafted implements.
a, after Skinner (75, p. 166); b, after Duff (33, p. 158).

the tang was formed by reducing the sharp apical edge of the butt and the sharp edge was continued from the butt shoulder to the lower end of the implement. The narrow surface opposite the apical edge corresponds to the base of the triangular section and forms the back or posterior surface. Instead of the back being bevelled to form a cutting edge as in "hog-backed" triangular adzes, the posterior surface is continued for the full length of the implement thus creating wide lateral surfaces. The lower end of one of these sides is bevelled to form the cutting edge.

The term "side-hafted" implies that the side of the implement was hafted either to the front of the foot, when the axis of the cutting edge would be that of an adze, or hafted by its side to the side of the foot, when its cutting edge axis would be that of an axe. If hafted as an adze, there would have been no reason for a tang on one side of the butt and making one side border thin and the other thick. If hafted as an axe, it is doubtful that the side of the implement was hafted to the side of the foot like the doubtful technique of the toki titaha shown in Figure 39b.

A comparison is made in Figure 41 with the Mangarevan axe and the Tahitian "hog-backed" triangular adze. All three implements have a narrow posterior surface and are convex longitudinally in front. The Mangarevan axe (a) is not tanged but the upper end of the butt is roughly chipped for the same purpose as the tang in the other two. Though its front is rounded, the cross-section approaches the triangular form of the other two (b, c). In spite of its clumsy appearance, the Mangarevan axe was securely lashed with its narrow posterior surface against the front of the foot (d). The "hog-backed" triangular adze which almost approached

Fig. 41. a, Mangarevan axe; b, New Zealand implement; c, Tahitian hog-backed adze; d, hafted Mangarevan axe; e, hafted Tahitian adze.

an axe in function was also hafted with its narrow posterior surface against the front of the foot (e). It is apparent that there would have been no difficulty in securing the moa-hunter implement (b) with its narrow back against the front of the foot or the toe of the haft. If this was done, the term "side-hafted" is inappropriate.

The question arises as to whether the number of cutting edge bevels or the direction of the axis of the cutting edge is the more important in deciding whether an implement is an adze or an axe. Personally, I think that the axis of the cutting edge is the more important. Unfortunately, however doubtful implements are not hafted and the direction of the page 194cutting edge remains conjectural. On the evidence, I think the moa-hunter implement was tanged to function as an axe. This form of axe with one bevel was old and was evidently succeeded by the toki titaha type with two bevels.

The Polynesians were devoted to the adze form of implement and the need for a longitudinal cutting edge was met in different ways. In New Zealand and Mangareva, implements with two bevels were hafted as axes but in Hawaii and the Society Islands, adze heads were hafted to a rotatory axis to perform the funcdons of an axe.

Chisels and Gouges

Wood carving was so well developed in New Zealand that the great variety of chisels (whao) was a natural result of the craft. Some of the finer carving was probably done with the smaller adzes, of which there is also a rich variety. The chisels, according to Best (12, p. 280), ranged in length from one to eight inches and in weight from one-quarter of an ounce to one pound. They were made of basalt and nephrite. Some were quadrangular in section and others rounded. Those with a straight cutting edge are definitely chisels and those with a rounded edge are regarded as gouges. Some implements used as gravers probably were not hafted. Evidence supports the use of a mallet with at least some of the chisels.

The type of hafted chisel is represented by an old chisel in the Dominion Museum described by Best (12, p. 287, Pl. xx, No. 124). The straight handle is six inches long and seven-eighths of an inch in diameter. One end is cut down to a flat surface an inch long with a shoulder for the butt end of the chisel. The part of the circumference below the flat surface is reduced but leaving a raised flange at the free end. The lashing with transverse turns of a vine is thus countersunk (Fig. 42a). The blade of the chisel projected for an inch beyond the lashing.

Some nephrite chisels were perforated at the butt end and thus worn by craftsmen as ear ornaments betweentimes. This was also a safe way of keeping a valuable tool.

The Cord Drill

The cord drill (tuwiri, porotiti) was formerly used to drill holes in stone, bone, and shell. Though said to have been used also on wood, the elongated or four-sided holes in old woodwork indicate clearly that the general tool for making holes in wood was the stone chisel.

The drill was a composite implement composed of a wooden shaft, balance weight, stone point, and pulling cords. The shaft or spindle (pou, turuki) was between one and two feet long and three-quarters of an inch thick. Its lower end was cut to a shoulder against which the point (mata) of a flake of kiripaka or other stone was lashed with a flax-fibre cord. A page 195balance weight (huapae) of two stones or two pieces of heavy wood was tied to the shaft some inches above the point (Fig. 42b). A fly wheel sometimes formed the balance weight. It was made by tying four rods by their middle to the shaft and bending out the resultant eight spokes to tie their ends at even intervals to a circular hoop of vine or supplejack (Fig. 42c). A twisted cord (aha) was tied by its middle to the top end of the shaft, thus providing two pulling cords.

In using the tool, the point was placed on the object or in a depression pecked on it and the shaft was rotated by hand to wind up the two cords around the upper half of the shaft. Holding a knotted end in each hand, the cords were pulled outwards in opposite directions with just

Fig. 42. Chisel and drills.
a, hafted chisel (Best, 12, pl. 15); b, c, cord drills (Best, 16, vol. 2, p. 196); d, pump drill (Hamilton, 46, p. 267).

enough strength to unwind the cords by rotating the shaft and causing the momentum to rewind the cords. The cords were pulled again which caused the shaft to rotate in the opposite direction and rewind the cords. The strength of the outward pull and the inward slackening of the cords to allow them to rewind had to be timed to keep the shaft rotating in alternate directions with each pull. Best (12, p. 66) quotes an authority that first one string was pulled and then the other but I do not see how such a process could succeed. It is the simultaneous pull of the two cords in opposite directions that maintains the equilibrium of the shaft whereas a pull on one cord would draw the shaft over in the direction of the single pull.

The evidence collected by Best proves that the cord drill was the only form known in New Zealand previous to the introduction of the pump drill by Europeans. The pump drill is an improvement on the cord drill through the addition of a cross-bar (kurupae), to the ends of which the cords were tied. A pump drill used in the Urewera district and figured by Hamilton (46, p. 267), has a middle perforation through the cross-bar which moves freely up and down on the shaft. A circular wooden disc with a central hole formed a fly wheel (Fig. 42d). On winding up the cords, the attached cross-bar moved upwards on the shaft. Downward pressure on it caused the shaft to rotate and removal of the pressure allowed the cords to rewind on the overspin. By timing the pressure and release, the shaft was kept rotating in a manner exactly similar to that of the cord drill.

The pump drill has been described for Samoa but with the cross-bar free of the shaft. I saw one in Tutuila in 1927 being used to drill holes in the pearl-shell shanks of bonito hooks. After some practice, I could make the shaft rotate continuously and keep it upright by applying the intermittent pressure with the forefinger and middle finger on either side of the shaft. It was so easy to a novice that the Maori experts must have experienced little difficulty in keeping the cord drill vertical without a cap or other contrivance though such were apparently used in some districts. The pump drill was most probably introduced into Samoa by Europeans as it was in New Zealand.

Owing to the nature of the drill point, a hole bored from one side through the entire thickness of the material would have entailed much labour and made an unsightly large opening on the side of entrance. An easier and a better job was made by drilling from the two opposite sides so that the holes met in the middle to complete the perforation. The stone flake which increased in diameter from its point made a funnel-shaped hole and the opening was slightly increased or even rendered irregular by the unavoidable wobbling of the staff during drilling.

The Passing of the Stone Age

Stone tools as symbols of the stone age were the first material elements of native culture to be discarded after European contact. At first the Polynesians were not particularly interested in the metal articles that early white navigators offered as presents to establish good will or as exchanges for native goods. However, when the superiority of metal over stone for working implements became apparent, a feverish desire to acquire the new material spread like a pandemic. The most desirable articles of European culture were the old-fashioned square spike nails and pieces of hoop iron cut off from the spare hoops of water casks which all ships carried. The spike nails were fashioned into chisels for use in wood carv-page 197ing and making square holes in wooden shafts and planks. The hoop iron termed paraharaha by the Maori from its flat form was made into adzes termed toki paraharaha. These metal heads sharpened with one bevel were lashed to the Polynesian form of haft with the stone adze technique. With increased contact, more manufactured articles such as hatchets, axes, and steel adzes became the more desirable objects of barter and exchange. The warlike Maori sought trade in firearms and they converted trade tomahawks into weapons of war.

In the earlier days of white contact in New Zealand, the supply of manufactured tools never met the demand and some of the sales of land in the north Auckland area were known as the hoko paraharaha, the hoop iron sales. In the records of the early land sales, a variety of other goods such as prints, blankets, tin pannikins, and even umbrellas testify to the manner in which the white buyers hypnotised the Maori owners into parting with their heritage. However, the value of surplus lands in those days was vastly different from what it is now and one is reluctantly forced to admit that the Maori owners accepted what to them was an adequate return in the strange new articles that their hearts desired.

In spite of innovations, the Maori, besides being a fighter, was a craftsman and the adze remained his favourite tool. He continued to carve his houses and work was rendered easier by the substitution of steel adzes and steel chisels for the stone implements with which his ancestors had founded the craft. He could handle the steel adze with its curved blade as well as, or better than, any white craftsman. However, the trade adzes were limited in supply and again he sought a substitute. The duller edge of the hoop iron of former days was not acceptable to a craftsman who had become acquainted with the superiority of steel. The steel blades of carpenter's planes met his requirements perfectly. The fact that they lacked a perforated socket for the handle did not bother him in the least for to him the hafting technique of the trade adze was unnecessary. He simply hafted the plane blade in the same way as the hoop iron adzes. The use of plane blades for adzes was also adopted in other parts of Polynesia and to-day, the Samoan carpenter's kit of tools contains a number of plane blades lashed to short handled native hafts with sennit braid.

The use of hafted plane blades has had one curious result. A Samoan carpenter hafted a stone adze for me with the bevel in front. He supported his action by showing me his kit of tools with ten plane blades all hafted in the Samoan technique to short hafts with sennit and all with the bevel in front. He argued that if the bevel were to the back, the steep front would cause the steel edge to sink too deeply into the wood, whereas with the sloping bevel in front such a danger would be avoided. As he was a master carpenter of life-long practice, I bowed to his experience. In contradiction to the expert's opinion, the photographs of over 50 specimens page 198of authentic old hafted adzes from central Polynesia showed that the bevel was to the back. Reconsideration of the problem revealed that in the European adze with the bevel at the back, the front of the blade is longitudinally convex so that the curve prevents the blade from digging too deeply into the wood. In the Bishop Museum collection of hafted plane blades, some blades had been bent, probably by a blacksmith, and such specimens resembled the European adze not only in the curve of the blade but also in having the bevel to the back. In the collection, there were also some with straight blades, and in these, the bevel was in front Thus it was evident that when the native craftsmen hafted a straight plane blade, they compensated for the lack of curve in the blade by turning the bevel to the front. The danger of digging in too deeply with a steel edge was not present with the less keen stone edge. Hence, the stone adzes hafted with the bevel in front have been hafted as curios by native craftsmen who had never used a stone adze in practice. They were accustomed to the use of straight plane blades reversed with the bevel in front and naturally concluded that stone adzes had been hafted in a similar way.

With the use of steel tools, carving was rendered easier but it also became over elaborate and too ornate. In the ceremonial nephrite adzes, the shaft, which was plain except for a carved knob at the end, was covered with carving which did not improve the artistic effect. Paddles and weapons also had extra carving which may have improved them for ordinary trade purposes but spoiled them for actual use. However, the metal age is now in full operation throughout the land. Yet, though the stone implements ceased to function long ago, they remain as imperishable relics of the skilled craftsmen, who, by their aid, were able to produce masterpieces of art which steel tools have never equalled let alone excelled.