Volume 37 1928 > Volume 37, No. 146 > Stone implements of Pitcairn Island, by Kenneth P. Emory, p 125-135
                                                                                             Previous | Next   

- 125

[See illustrations at the end of the previous article]


PITCAIRN ISLAND, at the extremity of the south-east Polynesian chain of islands and on the track of wanderers to and from Easter Island, promises much to Pacific archaeologists in the study of its varied stone implements. From the time of the settlement of the island by the mutineers of the “Bounty,” adzes, pounders, and other implements have been dug from the plateau in surprising quantity. Many of them have been scattered and lost, but happily many specimens have found their way into the British Museum and the Pitt-Rivers Museum.

The prospect that the Pitcairn implements in British institutions will be studied by Henry Balfour has encouraged me to describe those at present available for study in America. The material consists of seventeen adzes, one chisel, and one cleaving implement. Except for three of the adzes, these specimens were collected in Pitcairn in 1920 by Captain V. A. Brisson, who very kindly submitted them to the Bernice P. Bishop Museum for examination.


The adzes are diversified beyond a range which might be expected from the various uses to which they are put. Differences in the cultures reaching out to Pitcairn alone will account for certain of the wide distinctions. Ten of the seventeen adzes, or more than half, are of one unvarying type: a thin, chipped, and partially ground adze of quadrangular cross-section, with a tang, or grip. The tang in the larger specimens is set at an angle to the blade. The other six adzes belong to almost as many types; one adze is characteristic of Samoa, others of the Cook Islands and Austral Islands. Except for specimens from Rapa, the - 126 large collections of Cook Island and Austral Island adzes in Bernice P. Bishop Museum include no example of a quadrangular tanged adze, common at Pitcairn. However, this type of adze appears exclusively in marginal Polynesia (Hawaii, Marquesas, Society Islands, Tuamotu Archipelago, New Zealand), where it is more widely distributed than any other form. It is the sole type in Hawaii; elsewhere it is more or less uncommon. In the Society Islands there is evidence for supposing its disuse during the historic period. Of some thirty hafted adzes which I have seen from the Society Islands, collected during the voyages of Cook, Vancouver, and others of that time, all the adze-heads are of quite another type—triangular in cross-section, with a tang finished by pecking. 2

The absence in existing collections from outside of marginal Polynesia of the quadrangular adze with a tang at an angle with the blade indicates that this form was developed in the marginal area. Its wide distribution in this area, compared with other adzes, points to a very early appearance, though later, I believe, than the tangless adze of the same cross-section and workmanship. This tangless adze may reasonably be considered the prototype, and the tanged adze a modification of it. The tangless, quadrangular adze appears almost everywhere in Polynesia except in Hawaii, where it is represented only by a few very small adzes, which in comparative study should not have the same weight as the larger. On such small adzes normal or striking characteristics are often obscured or omitted. I think the tang has been omitted on these Hawaiian adzes because of the difficulty of making a tang by chipping on such a small object. Moreover, there is less need of such a feature, as the binding covers a larger proportion of the adze-head.

The absence in Hawaii of the tangless adze, the apparent prototype of the adze with a tang at an angle to the blade, gives grounds for the assumption that the development of the angled adze took place outside of Hawaii. The Society Islands, located in the centre of the marginal area, and possessing both forms in numbers, suggests itself as the most likely disseminating point. Although perhaps the - 127 adze employed by Tahitians during an early expansion to Pitcairn and Hawaii, it apparently remained in use only in Hawaii, being superseded elsewhere by derived, and by entirely different forms.

The presence of the unpecked, quadrangular, angled adze on Pitcairn extends the immense range of this form into an isolated corner of Eastern Polynesia, rendering more tenable the theory of its early distribution throughout marginal Polynesia. Likewise some of the other adzes of Pitcairn, because of their heterogeneous though definite affiliations, contribute to the understanding of the complex movement of culture in Polynesia.


In describing the Pitcairn adzes I have followed the terminology used by Skinner, 3 with these deviations: I have kept the often used term “tang” instead of adopting his new term “grip,” as “tang” perfectly describes this feature, which is defined in Websters' New International Dictionary as “the piece forming an extension from the blade…to connect with the handle.” For Polynesian adzes I am considering the tang as that part of the adze modified by special shaping of the front, front and sides, or sides to receive the binding by which the adze is attached to the haft. When no special shaping has taken place, the adze is referred to as tangless. I have retained the term “blade” as necessary in describing that part of the adze, or adze-head, exclusive of the tang. Though every adzehead has a blade, not every one is a blade.

The Pitcairn adzes are made of compact basalt, and their colour is black, or blue-grey patinated olive-grey. The shaping has been done by chipping and more or less grinding. Adze B, 3612 (Fig. 20), a Cook Islands form in every particular, and adze 5983 (Fig. 18), are the only ones bearing the pit-marks of pecking. 4 All adzes possess a tang, except the small adze, L 2181 (Fig. 16).

- 128

Hawaiian type.—Seven of the Pitcairn adzes are closely alike in form, and because of their remarkable similarity to Hawaiian adzes I refer to them as the Hawaiian type. (See Figs. 1, 2, 3, 6, 7, 8, 9, and Table I). They are relatively thin and broad, with sides diverging from the poll. In quadrangular cross-section, the width of the front exceeds that of the back. The tang is at an angle with the blade in the three largest and finest specimens; it is formed by chipping away of the front part of the adze, near the poll. The front of the adze is convex longitudinally and transversely; the back is flat or concave; the sides are flat or convex. The poll is flat, in some specimens sloping inward from front to back. The adzes are ground on the bit (the biting part of the adze), and more or less elsewhere. The cutting edge is straight in most specimens, but in L. 2169 and L. 2170 (Figs. 1, 2) is slightly convex in front view, and upwardly convex in face view. There is no pronounced bevel shoulder. Adze L. 2183 (Fig. 19) tapers into a nondescript, pointed tang. I believe the tang may be a re-shaping of the broken butt end of an adze of the ordinary type; however it resembles the carelessly-worked Mangarevan adzes.

These Pitcairn adzes are distinguishable from the Hawaiian thin, broad adze (Fig. 22), which they most closely resemble, by their more rapidly tapering tang and by the inward sloping of their sides from front to back.

Notes and sketches made by H. D. Skinner in May, 1927, and put into my hands for study, show that of the 36 Pitcairn adzes in the British Museum, 18 belong to the Hawaiian type, and it is my impression that 20 or 30 of approximately 50 Pitcairn adzes in the Pitt-Rivers Museum at Oxford, are also of this type. It is the Pitcairn adze, and the larger specimens should be recognizable in unidentified collections.


Five Pitcairn adzes may be described as relatively narrow and, in proportion to their width, thick, with the longitudinal edges of the blade nearly parallel to each other. These features are about all they have in common. (See Figs. 4, 5, 10, 13, 14.)

- i
Pitcairn adzes and a chisel: A, L, 2173; B, L. 2172; C, L. 2174; D, L. 2175; E, 2178; F, Chisel. Bernice P. Bishop Museum. (Uniform reduction.)
- ii
Pitcairn splitting implement and adzes: A, splitting implement; B, 26025; C, L. 2179; D, L. 2180; E, L. 2181. A. C. D. E., Bernice P. Bishop Museum; B, Peabody Museum, Cambridge, Mass.
- 129

Adze L. 2176 (Fig. 4), olive-grey, and L. 2178 (Fig. 10), black. These two adzes might be included in the Hawaiian type, but corresponding to the narrow, thick Hawaiian adze. However, the tang is far more tapering than the Hawaiian, and adze L. 2176 (Fig. 4) exhibits the peculiarity of a lateral tang pointed out to me among the Pitcairn adzes at Oxford by Henry Balfour. This one-sidedness is also a feature of three adzes in the British Museum, and one adze figured by Seurat. 5 These five adzes have the same size, outline, and workmanship, and the tang throughout on the same side (the left side when the adze rests on the poll with the front towards the eyes). An adze in the British Museum, described by Brown, 6 differing from these in having a flaring butt and lateral shoulders where the blade meets the tang, has its tang skewed to the same side.

Of Society Island adzes with a lateral tang I have seen only two. These have the tang on this same side (see Plate 5), but they are quite different in type. (See Figs. 24, 25.) Perhaps this is due to some greater ease in reducing this side by people who are right-handed.

Adze L. 2177 (Fig. 5), black, triangular in cross-section, the base of the triangle forming the front. It narrows towards the cutting edge, which is convex. Its longitudinal edges have been bevelled and partly rounded by grinding.

Adze 26025 (Fig. 13), according to data kindly supplied by H. D. Skinner, is triangular in cross-section, but, unlike L. 2177, the base of the triangle forms the back. The longitudinal edges have been rounded by grinding.

Adze L. 2179 (Fig. 14). The cross-section of the tang is quadrangular, and the cross-section of the blade before grinding was probably quadrangular, the width of the back throughout not exceeding that of the front. However, the blade has been ground so convexly on the front and sides as to give an almost elliptical cross-section. The outline of the cutting edge has been destroyed by a series of nicks.

- 130

The adzes numbered L. 2177 (Fig. 5), 26025, and 2179 (Figs. 13, 14) may possibly indicate a period of Pitcairn culture in which cross-section was unstable and all longitudinal edges subject to bevelling and rounding.

Adze with Lateral Shoulders.—Adze L. 2180 (Fig. 15), olive-grey, a thin, wide adze, the blade diverging towards the cutting edge. The cross-section is quadrangular, the width of the back not exceeding that of the front. The tang is formed by the extra chipping and grinding of the sides only. All surfaces have received grinding. The poll is a blunt, transverse edge. Its colour is olive-grey.

Adzes approximating this form were found on Tubuai, but in all specimens seen the tang has been shaped by pecking, which has been further extended in forming the front. (See Fig. 23.)

Tangless Adze.—Adze L. 2181 (Fig. 16) is black in colour. Strictly speaking, this adze is tangless, in that no special or extra shaping appears on the end towards the poll. However, the roughness on the face of this part makes unnecessary further shaping to retain the lashings, especially on such a small specimen. Hence it cannot be said that tangless adzes were at any time a form of Pitcairn culture, unless adzes are found which more clearly are without any special feature, intentional or accidental, to facilitate the binding. The rectangular cross-section of adze L. 2181 is unique in having the width of the back greater than that of the front. All longitudinal edges have been bevelled.

Adzes of this outline, cross-section, and workmanship (save for the amount of grinding) are characteristic of Samoa.

Foreign Adzes found on Pitcairn.—Adzes B. 3612 (Fig. 20), B. 5983 (Fig. 18), which on fairly good authority came from Pitcairn, have been reduced to their final form by pecking, a method not employed on any of the other Pitcairn implements described in this paper. These two adzes are distinctly different in form, and resemble none of the others. Their colour, also, does not match that of the other Pitcairn adzes. However, B. 3612 has all the characteristics of an average Cook Islands adze, and B. 5983 has a close parallel in a Tubuai adze (B. 4601) of the Bishop - 131 Museum. If the collectors have not made an error in location, I am inclined to attribute these adzes to the H.M.S. “Bounty,” unless more of them are discovered. After the mutiny the “Bounty” touched at Rarotonga and Tubuai. It is barely possible that the Tahitian men took stone adzes from these two islands. In Hawaii, as late as the last half of last century, Brigham 7 observed steel adzes laid aside and stone adzes being used for the final dressing of canoes.

Adze B. 3612 (Fig. 20), blue grey, found at Pitcairn in 1844. It was entirely pecked before grinding. The blade is triangular in cross-section, the tang is rounded and tapering. On the front the tang and blade rise to form a slight ridge. All these features are typical of adzes from Cook Islands (see Fig. 21). The bit has been seriously battered.

Adze 5983 (Fig. 18), olive grey, found by digging in 1866. It is quadrangular in cross-section, the width of the back not exceeding that of the front, and it shows an extreme thickness of bit. It may be a broken adze reshaped. In the Bishop Museum collection the Tubuai adze most closely resembling 5983 shows the same processes of shaping, and has the same outline and proportions, except that the maximum thickness occurs midway between the cutting edge and the poll. However, the Tubuai adze is undoubtedly a broken blade re-sharpened, and should not be taken as typical.

The Adzes used for comparison.—Adze 5984, blue-grey basalt, found on Oeno Island by one of the crew of the “Wildwave,” wrecked there in 1858 (Fig. 17). It is interesting to note that although Oeno is the nearest island neighbour of Pitcairn, lying a hundred miles to the north-west of it, the only Pitcairn adze which resembles in any way the Oeno adze is L. 2181 (Fig. 16). Both are tangless and have a quadrangular cross-section, with the width of the back greater than that of the front. The sides of the Oeno adze converge more rapidly towards the poll, and slope in towards the front at a smaller angle. The back of the - 132 Oeno adze is ground full length, while the back of the Pitcairn adze is ground only on the bit. The Oeno specimen is Mangarevan, and must have been brought from that island rather than from Pitcairn.

Adze B. 3485, blue-grey, Cook Islands (Fig. 21). This adze is proportionately shorter in blade and tang than the Pitcairn adze, B. 3612, and than some Cook Islands adzes, but resembles the Pitcairn adze more closely than any other specimen in the Bishop Museum collection.

Adze B. 4575, dark grey, Tubuai (Fig. 23). The tang is rounded and entirely pecked, but in the diverging sides of the blade towards the cutting edge, the continuation of the front of the adze in one plane, the tendency towards lateral shoulders, and in thinness, this adze is somewhat like the Pitcairn adze, L. 2180.

Adze B. 8790, light-grey, Kauai, Hawaii (Fig. 22). About eight per cent. of Hawaiian adzes are of this thinness and have this convergence of the sides towards the poll. The sides of most of them are flat, but in a few specimens converge slightly towards the back. The adze illustrated is remarkably patinated, an indication of great age.

Adze 9264, brown, Raiatea, Society Islands (Fig. 24). Because of its round cross-section, an adze exceptional for Raiatea. It is quite likely that the lateral tang was formed by subsequent pecking, and therefore that the adze was originally tangless.


Among the artifacts from Pitcairn Island is the anomalous implement, broken, but preserving the upper part and tang, shown in Fig. 12. A breadfruit splitter is suggested by some resemblance to the form of the Tahitian and Marquesas wooden splitter. 8 This tool, however, was obviously designed for hafting. Though far more finely finished, it calls to mind the hafted knives of Easter Island and Chatham Islands described by Skinner. 9

- iii
Adzes: D, B. 3612, Pitcairn Island; C, L. 2183, Pitcairn Island; B, 5983, Pitcairn Island; A, 5984, Oeno Island; E, B. 3485, Cook Islands. Bernice P. Bishop Museum. (Uniform reduction.)
- iv
Adzes for comparison with Pitcairn adzes: A, B. 8790, Kauai, Hawaii; B, B. 4575, Tubuai, Austral Islands; C, B. 9264, Raiatea, Society Islands; D, Leeward Group, Society Islands. A, B, C, D, Bernice P. Bishop Museum; D, Papeete Museum, Tahiti. (Uniform reduction.)
- 133

Weight, Measurements, and Proportions of the Adzes. 10

Fig. No. Weight in oz. Max. Length in mm. Max. Thickness Cutting Edge. L-C 11 L-T 12
1 L.2169 18 187 28 64 34.2 14.9
2 L. 2170 17 188 25 63 33.5 13.4
3 L. 2171 10.5 163 21 62 38 12.8
7 L. 2172 2.8 82 15 46 56 18.3
6 L. 2173 1.75 78 11 27 34.5 14.2
8 L. 2174 2.8 88 17 37 42 19
9 L. 2175 2.8 123 18 40 32.5 14.6
19 L. 2183 10 143 33 49 34 2 23
4 L. 2176 24.5 226 34 45 19.9 15
5 L. 2177 21 237 38 29 12.2 16.9
10 L. 2178 3.5 111 19 24 21.6 17.1
14 L. 2179 8.75 149 29 34 22.8 12.7
13 26025 6.75 152 27 20 13.5 17.8
15 L. 2180 7 142 18 61 42.9 12.6
16 L. 2181 3.8 93 18 35 37.6 19.4
20 B. 3612 13.75 142 39 50 35.2 27.9
18 5983 4 80 24 39 48.7 30
Fig. No. Weight in oz. Max. Length in mm. Max. Thickness Cutting Edge. L-C† Index. L-T‡ Index.
17 Oeno (Tuamotu Is.) 5984 3.75 73 19 45 61.5 26
21 Cook I. B. 3485 10.5 115 38 48 41.6 33
22 Kauai, H.T. B. 8790 24.5 230 25 75 32.6 10.8
24 Raiatea B. 9264 20.25 161 36 54 33.5 22.9
23 Tubuai B. 4575 7 128 17 77 60.1 13.2
- 134

The blade is 95 mm. wide; 16 mm. thick at the middle, and 2 mm. thick at the sides, which are ground flat. The back and front are equally convex, transversely. In the saggital plane the thickness of the implement increases from 14.5 mm. where the grip joins the body, to 16 mm. three inches from the poll, and then decreases to 15 mm. at the broken edge.

The implement has been ground smooth on both faces, but rough chipping is left on each face and at each side of the junction of the tang with the blade. Undoubtedly this chipping has been done to facilitate the keeping of the lashings of a haft. The whole proximal end of the implement has a sharp edge, except at the very tip of the poll, which is 2.5 mm. thick.


The Pitcairn Island chisel shown in Fig. 11 is of basalt, blue-grey, patinated olive-grey in colour. It weighs 14 ounces, and measures 338 mm. in length. The cross-section is angular and oval, the point of the oval being towards the face. A thin, sharp ridge runs the whole length of the face; the back is almost flat; the sides are very convex, with a bevel along their lower edges. The width of the cutting edge is 8 mm. The maximum width of the tool, which is at the middle, is 25 mm., and the thickness at this point is 28 mm. The chisel tapers equally towards the cutting end and towards the butt end. The poll is blunt and rough. The under bevel of the cutting edge is flat, transversely, but the front of the bit is rounded, causing a cutting-edge which is convex from the face view. The cutting edge has been ground flat preparatory to regrinding, and is about 1.5 mm. thick. The nick which caused the native to set about preparing a new edge is traceable on the bevel.

[The writer of the above paper employs the term “pecking” in describing the unground surface of a stone implement. Is this usage a justifiable one? It has been used by several New Zealand writers, albeit they were not field workers, but we know of no justification for such use. The Maori tells us that, when an implement was chipped into form, a rounded, usually waterworn, stone was used “to reduce the asperities” of the chipped surface, as Sir John Evans puts it. This was a hammering or bruising process - 135 that much lessened the labour of the next process, that of grinding, and the little pits resulted naturally from the hammering; they were not produced by a pointed instrument. Is the pecking process proven for the Polynesian area? The hammering process referred to is described, as an Australian usage, at pp. 548-549 of Sir Baldwin Spencer's work, The Arunta, Vol. 2, and Fig. 213 shows a native engaged in the task. The term “pecking” seems to imply the use of a pointed instrument; such a tool would be an unsuitable one for the purpose, and unless it can be shown that a pointed tool was used the term “pecking” should be discontinued.—Editors.]

1   Published by permission of the Director, Bernice P. Bishop Museum.
2   On this term “pecking,” see note by Editors at end.
3   Skinner, H. D., “The Morioris of Chatham Islands” B.P. Bishop Mus. Mem., vol. 9, no. 1, p. 90, 1923.
4   See note by Editors at end.
5   Seurat, L. G., “Sur les anciens habitants de I'ile Pitcairn”; L'Anthropologie, vol. 15, pp. 369-372, fig. 14, 1904.
6   Brown, J. A., “Stone Implements from Pitcairn Island”; Jour. Anthrop. Inst., new ser. vol. 3, p. 88, plate IV, fig. 5, 1900.
7   Brigham, W. T., “Stone Implements and Stone Work of the Ancient Hawaiians”; B.P. Bishop Mus. Mem., vol. 1, no. 4, p. 77, 1902.
8   Breadfruit was growing on the island when the mutineers of the “Bounty” arrived: Moerenhout, J. A., Voyage aux Iles du Grand Ocean, p. 45, 1837.
9   Skinner, H. D., “The Morioris of Chatham Islands”: B.P. Bishop Mus. Mem., vol. 9, no. 1, p. 99, 1923.
10   The data is arranged to conform to the method of presentation being used by H. D. Skinner in his adze notes, which he has kindly put in my hands.
11   The L-C index is the width of the cutting edge divided by the maximum length, which gives approximately the relative width of the adze. If the cutting edge is more than 25 per cent. of the length, the adze is relatively broad, as opposed to narrow (see Emory, K. P., “Island of Lanai”; B.P. Bishop Museum Bull. 12, p. 78, 1924). Index.
12   The L-T index is the maximum length divided by the maximum thickness, which gives the relative thickness of the adze. If the thickness is more than 20 per cent. of the length the adze is relatively thick or heavy, as opposed to thin. Index.