Volume 56 1947 > Volume 56, No. 1 > Notes on Oceanian war clubs, by B. A. De Vere Bailey, p 3-17
                                                                                             Previous | Next   

- 3
NOTES ON OCEANIAN WAR CLUBS.

WAR clubs are common to most areas of Oceania, and as their name implies, are chiefly used for combat purposes. Some club-like implements, however, as the Australian digging-stick, are primarily identified with agriculture, and only occasionally function as offensive weapons. Certain types in addition to being used for warfare are employed in mortuary rites, and a few figure in sympathetic magic. The survey that follows is based on a series of twenty-nine clubs belonging to the University of Michigan collections in the Museum of Anthropology, The specimens in this group are largely without documentation, and conclusions reached, therefore, are based upon an analysis of typological evidence. Nevertheless, careful comparison has been made of characteristics shown by these clubs with those of similar and documented specimens of known origin, and since styles are, in the main, diagnostic, the secondary determination of provenance can in each case be taken as reasonably certain. Different ethnic areas are represented in the collection, and considerable diversity of material is shown. Certain aspects of the study are limited by lack of comparative material within the group, for in several cases there is only a single specimen of some particular variety of club.

The different types may be defined by the designations given in Table 1, below. These suggest in most instances some structural characteristic of the weapon indicating its general shape or purpose; more rarely, the kind of wood normally used in making it. The figures within brackets apply to corresponding numbers on plates 1 and 2, which will help to identify the various types by their respective terms. The different ethnic areas to which the clubs have been referred on typological grounds are listed, the immediate provenience of each specimen being shown in Table 2.

- 4
Type Ethnic Area       Total
  Polynesia Melanesia Micronesia Australia  
1. Bell-head   6     6
2. Billet 5   1   6
3. Carinated 1       1
4. Coconut-stalk 1       1
5. Digging-stick       1 1
6. Forked     1   1
7. Lipped 2       2
8. Missile 2       2
9. Mushroom-head   2     2
10. Paddle 1 1 1     2
11. Pandanus 1       1
12. Parrying       1 1
13. Phalloid   1     1
14. Rootstock 2       2
Totals 15 10 2 2 29
1   10a.
TABLE 1.

Effort has been made in the foregoing table to classify the clubs by types as accurately as possible, but in certain cases opinion as to fitness of the terms used may vary with individual appraisals of salient features. For example, the “carinated” type is here so called because of a keel-like element adown the blade section and a rib carved at the point of greatest breadth. Similar attributes are shown by both the “coconut-stalk” and “paddle” specimens, the latter evidencing faintly the structural presence of a keel. This, however, is not treated as a distinct carving, and in describing the type, therefore, is subordinated to a more prominent feature. The “digging-stick” type resembles an Australian billet of the same name, which is also used as an offensive weapon; while the designation “parrying” refers to a similar variety made in Western Australia. The term “pandanus” applies to the club head, which is thought to resemble a cluster of fruit of that kind of palm.

In order to identify the individual specimens discussed in this paper, they are listed by catalogue number in the table below, together with immediate provenience of each, determined by typological criteria, as before noted. Where a club type seems common to several adjacent localities, the - 5 plural sources are given. Although often included in the Melanesian area, the Fiji islands for convenience are here considered as part of the Polynesian group, since they lie on the perimeter of Polynesia, and show considerable cultural influence of the latter.

Catalogue Number Source Ethnic area Number of specimens
8692, 8693, 8700, 8709, 8710 Fiji, Tonga, Samoa Polynesia 5
8696 Fiji, Tonga Polynesia 1
1326, 2516, 2518, 2520, 2572, 8702 Fiji Polynesia 6
2517 New Zealand Polynesia 1
2528, 8297 Samoa Polynesia 2
8687 Solomon islands Melanesia 1
8706, 8711 New Britain Melanesia 2
8694, 8699, 8703, 8704, 8705, 8707, 8708 New Caledonia Melanesia 7
8697, 8701 Caroline islands Micronesia 2
8695, 8698 Australia Australia 2
    Total 29
TABLE 2.

All specimens in this collection are made of wood, varying in weight and hardness but generally heavy and fine grained. The weight of each club will be found in a tabulation further on. In various areas of Oceania, including Polynesia, a type of ironwood identified as Casuarina equisetifolia was extensively used in shaping these weapons, also a dark reddish wood called “kauila,” heavy and fine grained, and to a lesser degree the stalk of the coconut palm. Where the wood employed for the present specimens was identifiable, this was taken as a factor in determining the provenience of the club. The timber found in the different types of weapons under consideration appears to have been derived chiefly from the following sources: the simple branch broken from the tree and represented by the billet clubs; the uprooted sapling with side and/or tap-roots partially retained, as in the missile and rootstock types; the limb of the tree at the crotch, which shows a curved configuration as in the “pandanus” and “lipped” varieties; and finally, the plank forms rived from the trunk section, suggested by the broad, paddle-shaped specimens.

- 6

Based on the above, it may be noted that the two clubs ascribed to the Australian area are both of the limb (branch) type; the Micronesian examples include one limb and one plank type; of the ten weapons assigned to Melanesia nine are of the limb variety and one of the plank; while the Polynesian clubs show five items derived from limbs, four from saplings, three from crotches, and three from planks. Since both the Australian and Micronesian areas are represented by only two clubs each, it is difficult to draw any conclusions from this collection as to predominating structural forms of the wood used in these regions. In the case of the Melanesian weapons the figures show the limb type as the usual form, at least for this particular group; while the Polynesian specimens derive with some degree of proportion from all four wood sources.

As in other classifications of similar objects, it may be assumed that in this group, likewise, length of the clubs is determined by use. Those specimens wielded by a single hand would naturally be shorter than the longer and generally heavier two-handed weapons. In most instances the clubs are fashioned from heavy timber, such as ironwood, and their weight, therefore, increases appreciably in ratio to length. Thus, a line dividing the short from the long specimens might be expected to show a limit of length conditioned by expedient handling, and so set up the twin categories of single and two-handed weapons. The maximum weight convenient for one hand in wielding the club, must necessarily be reckoned with reference both to swinging action and ability to exert the force desired at point of contact with the objective. In the case of two-handed weapons, the weight successfully manipulated and the degree of effectiveness in combat are obviously determined by the physical capacity of the wielder.

The length of each club in the present group was measured, and the results, expressed in centimetres, may be conveniently summarized as below. Figures for specimens showing a curved contour are here based on the outer curve only:

- 7
Length Number of specimens
38.9 to 40.6 cm. 2
61.6 to 67.0 cm. 4
70.7 to 78.1 cm. 11
80.0 to 87.0 cm. 4
98.4 to 112.4 cm. 7
124.5 1
Total 29

Of the foregoing, the two items measuring 38.9 and 40.6 cm., and the single club 124.5 cm. in length, may be considered atypical in this group, since the former are “missile” clubs obviously designed for one-handed use, and the latter is an “oversize” paddle club. In attempting to define the other twenty-six weapons on a basis of length as single and two-handed specimens, it will be noted that no examples occur showing measurements between 87.0 and 98.4 cm., thus setting up the greatest differential of length for this group, with a variation of over 11 cm. Assuming that this differential marks the line dividing the short from the long clubs, and that the elements of length and related weight normally determine use of the weapons by one and two hands, respectively, the collection consists of nineteen shorter clubs measuring from 61.6 to 88.6 cm., plus the two short missile types, making twenty-one single-handed specimens in all; and seven longer examples from 98.4 to 112.4 cm. in length, which with the “oversize” paddle club 124.5 cm. long, give a total of eight two-handed weapons. These categories, however, are not set up as arbitrarily determinative, since in the case of some particular club classified as short, because of weight the practicability of its single-handed use may be controversial.

The actual relation that the weight of these clubs bears to their length in terms of the foregoing division, will now be examined. The following table gives the lengths and weights, in centimetres and kilograms of all clubs in the collection, defined as single and double-handed weapons.

- 8
One-handed clubs
Cat. Number Length (cm.) Weight (kg.)
1326 38.9 .5103
2572 40.6 .4961
8706 61.6 .6379
8707 63.8 .6450
8703 66.7 .6450
8705 67.0 .9639
8699 70.7 .6733
8694 70.8 1.1198
8698 72.0 .9781
8711 74.0 .5316
8709 75.1 .7584
8695 76.2 .5812
8708 76.4 .8789
8697 76.5 .6379
8692 76.7 .7088
8696 77.8 .7867
8693 78.1 .3544
2517 80.0 1.0490
8710 80.8 .6237
8704 83.4 1.1765
8701 87.0 1.1340
Two-handed clubs
Cat. Number Length (cm.) Weight (kg.)
2518 98.4 1.4559
8702 101.3 1.6585
8700 104.3 .8718
2520 107.6 3.3653
2528 108.0 .8789
2516 110.2 1.9620
8297 112.4 1.5380
8687 124.5 .8647
TABLE 3.

From this table it will be seen that the ratio of weight to length for both groups varies considerably. The majority of one-handed clubs show progressive increase of weight in relation to length, but this increase is not always proportionate, nor is it consistently maintained. In several cases the difference between weights of two or more clubs of approximately the same length is so marked that these elements seem unrelated. The two-handed weapons are similarly disparate, and it may be assumed that factors common to both groups are involved. While heavy wood is generally used for making the clubs, certain specimens are fashioned of comparatively light timber, and a few of decidedly open-grained stock. Examples of this light wood are seen in Nos. 8693, 8697, and 8711. These naturally show less increase of weight in ratio to length than weapons made of heavy wood. Lack of comparative specimens of the same type within each group serve to make unusual contrasts between length and weight, more apparent. Likewise, massed weight in the conformation of either head or

- i
PLATE 1
Photograph by courtesy of the University of Michigan., 10: 2517; 8: 2572; 9: 8706; 12: 8695; 13: 8707; 5: 8698; 1: 8694; 2: 8709; 6: 8701.
- ii
PLATE 2
Photograph by courtesy of the University of Michigan., 3: 8297; 7: 8702; 14: 2516; 4: 2528; 10a: 8687; 11: 2320.
- 9

shaft of the club, or both, will effect normal relation of weight to length. Thus, in the single-handed group, No. 1326 with a length of 38.9 cm. weighs .5103 kg., while No. 8711 which is 74.9 cm. long, weighs only .5316 kg. The first is a missile club with large, heavy head and short shaft; the second a mushroom-head club with a relatively small distal end and long, slender shaft. Among the two-handed specimens No. 2520 is an example of unusually heavy bulk massed disproportionately to length; both head and shaft, however, contributing to the weight.

On the other hand, while Nos. 8700 and 2528 (Plate II) are considerably lighter than the rest of the two-handed weapons of similar or even less length, their size and comparative unwieldiness suggest they are best suited to two-handed use. This condition is emphasized in the case of No. 8687 (Plate II), which is at once the longest and lightest specimen of the double-handed group. Here, as in other instances above cited, the density of the wood used seems an important factor in considering relation of weight to length; for it is fairly evident that a club may be long and comparatively light, yet unsuited for use with a single hand, while a relatively short club may be of sufficiently dense wood, or with weight massed in conformation, to warrant it being logically included in the category of two-handed weapons. However, except in cases where controversial elements make the fitness of some classification debatable, there seems little to negative the division of this collection into single and two-handed clubs, as detailed on page 8.

It may be noted that with the exception of one specimen, all the clubs classified as two-handed are referred to Samoa, Tonga, or Fiji in the Polynesian area. The remaining example is attributed to the Solomon islands in Melanesia. Of the single-handed group, specimens from the Fiji region, aside from the short missile clubs, are generally long and relatively heavy, save No. 8693, a billet type, the unusually light weight of which suggests possible use as a ceremonial stave, rather than in combat. The New Caledonian clubs are on the whole shorter than the rest of the group, but show considerable disparity between weight and length, since two examples approximately 70½ cm. long weigh respectively, - 10 .6733 and 1.1198 kg. Two weapons from the Caroline islands, both of which are among the longer clubs, are similarly disproportionate. The Australian “digging-stick” club No. 8698 (Plate I), which is about 4 cm. shorter than the parrying type of weapon from the same area, weighs close to twice as much as the latter, the disparity here being obviously due to character of the wood selected.

With the possible exception of No. 8693 mentioned above, all specimens in this group of twenty-nine seem adapted for combat purposes, the “paddle” clubs being so termed in reference to shape rather than function. In Polynesia similar types were much used for fencing at matches, in which the participants' ability to parry and thrust were developed, thus setting up a fighting technique. As offensive weapons twenty-seven examples in this group may be classed as “contact” types, while two are missile clubs designed to be hurled at the objective. These latter are cast by the half end, but the club capsizes during flight and continues in reversed position, reaching the mark haft first. Penetration is secured by the weight of the heavy, spherical head, which drives the blunt end of the haft into the object aimed at—usually human or animal tissue. The paddle-shaped clubs were swung like bats in the Polynesian area, and weapons pointed at proximal and distal ends alike were gripped around the middle of the shaft to afford dual action and also enable the user to parry a thrust. The paddle-shaped club No. 8687 (Plate II, fig. 10a), is the only double-pointed specimen in this group with terminals of which both were probably at one time sharp. The distal end is now blunted at the tip, apparently due to breakage, while the proximal end shows a complicated, carved design with lateral serrations along the half area. This would provide an inconvenient grip, and gives added reason for handling the club at the middle section of the haft.

Clubs not leaving the hands when wielded, and intended for fighting at close range, are divided into two major classes: those made to produce contusive effect, in terms of bruising or crushing; and those designed to wound by cutting action. In contusive types the element of weight is naturally important, and where weapons have a relatively short shalf, this is secured by using a heavy kind of wood, - 11 or by massing weight in the head, or both. Long shafts will provide certain types of lighter weight cutting clubs with increased impact force, and thus obtain a stunning effect in addition to lacerative action. Other varieties of cutting weapons with shorter shafts are probably equally effective as impact clubs when by reason of thickness in the head area, weight is so massed. With the exception of the missile clubs and two varieties showing obtuse and sharply pointed terminals, respectively, at both ends, the weapons under consideration depend upon the head for effectiveness in combat. This, then, is usually the most important structural unit of the club, and both shaft and grip must be so proportioned as to give the head maximum force, whether contusive or lacerative. With weight massed either in the distal section or in the head itself, balance must be had for proper swinging action, and care taken that the shaft is sufficiently strong to resist without breaking, the impact of a heavy forward section against its objective.

The type of cutting club with a blade-like head showing the two opposite edges strongly serrated, or with head angular in section and marked by rows of spikes, is not represented in this collection. The paddle and carinated specimens, however, belong in the category of cutting weapons, since the edges of the blade-shaped heads are evidently capable of lacerating effect. Three of these are diamond-shaped in section, and one example shows a quadrangular blade banded by prominent carved cross-ribs, which become acutely angular at the opposing edges of the head and form cutting projections. In No. 8702 (Plate II), a lipped club from Fiji, that side of the weapon opposite the lip is roughly triangular in shape, with its apex on the edge of the inner curve. The three edges formed by this triangle are broken by rows of five perforations each. These holes are shallow drillings made from opposite sides of the angles, and suggest use in lashing sharks' teeth or other pointed objects to the head area, from which they would project outward for cutting action.

In this group twenty-one clubs seem designed to produce crushing or bruising effects, while eight specimens suggest chief use as lacerating weapons. Certain specimens, however, appear to have a dual purpose, probably - 12 functioning primarily as contusive implements. Examples of these latter are the bell- and mushroom-head clubs referred to the New Hebrides and New Britain, in which the circular flange forming the base of the head, extends at right angles to the main axis of the shaft and terminates in a rim with sharp edges. In some of these specimens the head begins as an abrupt rectangular outthrust from the shaft, while in others it evolves from a flare in the distal end of the club that gradually increases in diameter and curves outward; added weight being given by thickening of the shaft in the area affected. Bruising as well as cutting action are indicated by both varieties.

Of the twenty-one contusive types of clubs, thirteen may be classed as intrinsically crushing or bruising weapons, and eight defined as “dual-purpose” specimens. Nine of the thirteen in the first category may be referred to Fiji, Tonga, Samoa; two to Australia, and one each to the Caroline islands and New Caledonia. Six of the New Caledonia clubs suggest a dual function, and two from New Britain may be similarly defined. Of the eight lacerative weapons, five come from Fiji, Tonga, Samoa, and one each from the Caroline islands, the Solomons, and New Zealand. From the foregoing it will be noted that the several regions are represented by too few specimens to make this group of much value in determining types prevailing in the areas of origin. However, the ratio for Polynesia of bruising weapons to cutting clubs is nearly two to one, while the totals for Melanesia show eight dual purpose specimens and one each of the two general types. The missile clubs are included among the cutting weapons, as is No. 2520 (Plate II), since the spike with which the latter is provided, is designed to finish off a wounded opponent by piercing the crown of the head without otherwise splitting the skull. This type of club is also used in burial rites and in sympathetic magic. It is probably apparent that the foregoing classifications are broadly based on structural qualities that indicate the major effectiveness of the different clubs when used in combat. Equally obvious, doubtless, is the fact that either a contusion or a laceration, or both, might, under given circumstances, be inflicted by any or all of the clubs in the collection.

- 13

Certain additions made to the conventional structure of the clubs are evidently designed to increase their utility, while others merely serve as ornamental detail. In some specimens these qualities merge, and at times the primary purpose is obscure. Obvious decoration may take the form of material affixed to the substance of the club, as wrappings or ties of sennit, coir, cotton fabric, etc., or it may be developed either as relief carving or as incised design through which certain areas of the club structure are removed to reveal a pattern. In this collection there are no examples of ivory or other inlay, but certain clubs show pigmented detail. Several of the weapons have services of sennit or similar material banding the haft, and evidently intended to facilitate handling the weapon. Some of these utilitarian additions show calculated design, and will be further considered in somewhat more detail. Non-decorative adjuncts include slings of cordage and beckets of plaited fibre, used for convenience in carrying the clubs. Ties of sennit or leaf material occur infrequently in the present group, but are represented by a narrow banding of single strand, plaited fibre on the Solomon islands paddle-club, used decoratively; and by a service of braided fibre with definite pattern of lozenges, seemingly intended to offset a structural crack in a New Caledonian specimen.

Of the five bell-head clubs referred to New Caledonia, four show incised decoration in the head area, extending from a flare in the shaft at the distal end to the edge of a flange forming the head rim. The design consists of diagonal lines in herring-bone pattern, varying in size as length of the lines and spacing differ. The motive is thought by Churchill (4, 132-133) to derive from the pinnate-form leaf of the coconut-palm, and with more advanced technique develops as a meandered zigzag. In the intaglio design on the New Caledonian clubs excision is not always completed, and patterns in such cases is evolved by mere displacement of the wood structure in the minute areas affected, similar to the roughening of an engraving plate by the worker's burin. On four of the clubs this incised detail is coloured white by a lime-like substance, probably “chunam” derived from shells and coral, used to provide contrast with the ground colour of the wood. The fifth specimen of this type of club shows no carved decoration, but in common with - 14 three incised examples, is wrapped at the proximal end with fabric secured by bindings of braided fibre, apparently coir from the coconut husk. This wrapping obviously affords a better grip on the haft section of the weapon, and in three instances is of cotton material, while the fourth is of wool. The plaited bindings are affixed diagonally to form a lozenge pattern on specimens 8694, 8703, and 8708, but on club No. 8704 the cotton wrapping is secured only by a strip of red calico. It is likely that this specimen was originally also equipped with the fibre bindings, which later were lost. No. 8705 has a section of pandanus(?) leaf-sheath wound spirally around the shaft and attached by a tape of red cotton.

Narrow fibre binding intended either as a decorative accessory or as a practical means of securing the haft wrapping, is likewise common to certain clubs of this group from Fiji, New Britain, and the Solomon islands. It is made of three bundles of fibre comprising from two to five strands each, usually brown in colour and braided in flat, tape-like form. The root-stock club from Fiji, No. 2516, illustrated on Plate II, shows nearly complete coverage of the shaft with this binding, applied horizontally to the main axis of the club and extending from the distal edge of the grip to the beginning of the head area. The tie is again found encircling the proximal terminal of the haft, and throughout the coverage a regular pattern is developed by rows of brown, black, and mixed brown and black fibre in alternate units of each color and mixture, the distal end of this binding giving way to several rows of cord made by twisting two strands of bundle fibres. On the paddle-club, No. 8687 (Plate II), this tie occurs as an isolated band braided of several strands and located in mid-section of the shaft.

Incised decoration with the zigzag or pattern of diagonal lines, characteristic of Polynesian culture, occurs on five clubs of this group; all of which are here assigned to that area. The simpliest expression of this motive occurs on the root-stock club No. 2516, where the zigzag design encircles the haft in opposing parallel lines placed lengthwise to the main axis of the shaft and separated by a dividing band; this perhaps being a more developed and stylized version of the pinnate leaf design, referred to above. A similar handling of the same motive is shown on the two missile clubs, - 15 Nos. 1326 and 2572 (Plate I), with the zigzag lines forming less acute angles, and the ornamental panel crossed at both ends by lines enclosing a single elongated zigzag carved across the shaft. This design is repeated with slight alterations of detail on the haft area of the Fijian lipped club, No. 8702 (Plate II), where the units of longitudinal bands and double zigzag lines are interrupted about midway of the grip by an encircling panel that extends at right angles to them and is composed of the same design elements. The field area of zigzags is also broken by sporadic units of short bands enclosing diagonal lines which extend obliquely across the shaft, and in two instances form triangles unrelated to the general pattern.

The carinated club No. 8297 (Plate II), shows incised decoration covering the entire head area and part of the shaft. The pattern is developed by a succession of panels marked by wide, transverse bands enclosing single, double, and triple zigzags running across the blade and separated into smaller panels by longitudinal bands. The whole series of panels is divided by the carina on the blade, and this keel-like carving which extends along the middle of the distal area and on both opposite sides, is emphasized in the design by a strip of zigzags paralleling its course and framed by lateral bands. The blade portion of the club is diamond-shaped in section, and the keel on either side, therefore, forms the apex of a theoretical triangle. The decorative patterns on opposite sides of the blade area are nearly identical, and design is continued around the distal end of the circular shaft from the point where the head or blade merges into the former. The distal end of the blade comes obliquely to a rounded point, and the two uppermost panels of decoration form a banded lozenge. It may be noted that the main pattern of zigzags and bands within framed panels is broken by a triangle pointing proximally and placed at the right and left, respectively, of each alternating panel. The design is also interrupted by a projecting cross-rib extending the width of the blade and on both sides, but continues on to juncture of the blade with the shaft proper. The decoration then encircles the rounded structure of the shaft, and appears as bandings of diagonal lines; a herring-bone pattern with separating band; and triangles in simple outline and with fields crossed by oblique incisions.

- 16

Zoömorphs and human effigies seem to be relatively common in the incised designs on Polynesian weapons, but neither type appears as intaglio decoration on clubs of this group. Curvilinear design, employed by the Maori but foreign to the rest of Polynesia, is shown on paddle-club No. 2517 (Plate I), which has been referred to New Zealand. 1 Four rows of meandered spirals extend along the paddle area of the weapon, and are bounded at the distal end by a panel of smaller, interlocking spirals that run at right angles to the first. The terminal of the head describes an upward arc and is undecorated. At the proximal end of the design there is a single row of zigzags running across the blade, from which depend eight arrow-shaped incisions pointing toward the butt. Bisecting the four rows of spirals is a series of elliptical figures enclosing a straight line, and extending along the surface of the blade parallel to its main axis. The haft terminates in a dome-shaped carving which roughly outlines the head and features of what seems intended as human semblance. The entire club was apparently coloured with a black stain before it was carved, and the incised areas, therefore, give contrast by showing a light-coloured wood of considerable density.

The haft end of paddle club No. 8687 (Plate II), is acuminate at the terminal and diamond-shaped in section. Each of the two opposite sides is carved to represent a human face, surmounted by radiants probably depicting hair. The exaggeratedly large eye-forms, the mouth, and part of the hair are coloured black, the rest of the face is painted red. Twin spiral forms and dentelle depending from the carved head suggest a simulated necklace, and below these in the usual grip area are two sets of concentric squares incised and placed lengthwise to the main axis of the shaft. The sides of the carved portion of the haft are marked by - 17 deep serrations, suggesting, in conjunction with the shape of that terminal, use of this end of the weapon as a lacerating element. The head of the blade is roughly diamond-shaped in section, becoming elongated proximally as it merges into the shaft. The tip of the distal end, now obtuse, seems originally to have been pointed, and the lateral edges of the blade area present fairly sharp edges. The shaft and most of the blade surface are painted red, but the uppermost portion of the head shows a triangle reserved in black pigment, with the terminal forming its apex.

BIBLIOGRAPHY.
  • BISHOP, MARCIA BROWN—Hawaiian Life of the Pre-European Period. Peabody Museum of Salem (1940).
  • BURROWS, EDWIN G.—Western Polynesia, A Study in Cultural Differentiation. 45-48. Göteborg, 1938.
  • BUSCHAN, GEORG.—Illustrierte Völkerkunde, II, pt. 1: 1-272.
  • CHURCHILL, WILLIAM—Club Types of Nuclear Polynesia. The Carnegie Institution of Washington, 1917.
  • DODGE, EDWIN STANLEY—The Marquesas Islands Collection in the Peabody Museum of Salem. 27-28, Salem, 1939.
  • EDGE-PARTINGTON, JAMES—An Album of the Weapons, Tools, Ornaments, Articles of Dress, etc., of the Natives of the Pacific Islands, I: 12-14, II. 10. (Manchester, 1890-1898.)
  • KRIEGER, HERBERT W.—Design Areas in Oceania. Smithsonian Institution of Washington. Publication No. 2896, 1932.
  • LINTON, RALPH—Ethnology of Polynesia and Micronesia. 108-128. Field Museum of Chicago, 1926.
  • MCCARTHY, FREDERICK D.—Australian Aboriginal Art. 24-25, 34-35, 44, 46. Sydney, 1938. (Australian Museum.)
  • MONTAGUE, L. A. D.—Weapons and Implements of Savage Races. 6-137. London, 1921.
  • RATZEL, FRIEDRICH—Vökerkunde, II: 3-362. Leipzig, 1887.
  • WARNER, W. LLOYD—A Black Civilization. 167, 489-90. New York, 1937.

(Checked against original sources.)

1   It seems most doubtful that this weapon has been correctly assigned to New Zealand. From the illustration in Fig. 1 and the description in the text it is almost certainly from the Trobriand islands in the Massim area in Melanesia. Compare the examples of Trobriand weapons figured by Skinner in J.P.S., Vol. 40, facing page 183. Although the surface patterns on the present specimen are found in various forms in Maori carving, there is no evidence, so far as we are aware, that such patterns have been used on the blades of Maori weapons of this type.—Eds.