- 17

Since the last instalment of these papers was published a number of pieces have been received, of which the following are here figured:

Fig. 143. Soapstone figure of a man, allied in type to fig. 137, from D'Urville island. There is a similar but finer amulet in soapstone from the Marlborough sounds in the Dominion Museum collection, but it is not accessible for figuring.

Fig. 144. Soapstone amulet representing a reptile, closely allied to fig. 63, but having the head on obverse and hind limbs on reverse. The spine seems to have been treated as in figs. 59 and 60.

Fig. 145. Amulet in red argillite from Little Papanui, Otago. I think this represents a spear-head with reversed barbs, a kind not yet recorded from New Zealand.

Fig. 146. An amulet in dark greenstone from Pahia, Southland, is of the same type, but has reversed barbs along one side only.

Fig. 147. Amulet in light-coloured greenstone from Birdlings flat, Canterbury. This is interesting because it represents the barbed point of a composite fish-hook of a type made, so far as is at present known, only in the district about the East cape. Its presence in Canterbury must therefore be due to gift or trade.

I tender my cordial thanks to Miss L. Daff, the accomplished draughtsman to whom these drawings, and all the other figures in this series, are due.

- 18

Illustration

- 19

THE first paper of this series was published in September, 1932. In the period following that date a number of drawings have come to hand representing new varieties of amulet or better examples of varieties already figured. In addition the writer has received a number of valuable suggestions and criticisms which have in some respects modified the conclusions previously reached. It may therefore be profitable to set out the classification of amulets, slightly modified in view of additional data, and to summarize briefly the views that may, in the writer's opinion, be drawn from all the data that has been collected.

Type 1—Amulets of Human Shape. Class 1. The Hei-Tiki. Varieties A, B, and C. Variety A may be regarded as the oldest. Variety B has been definitely affected in shape by the adze. Besides these two varieties there are a good many hei-tiki which may be included in a loosely defined Variety C.

Type 1—Amulets of Human Shape. Class 2. The headless human figure: Variety A, front view. This is a well-established New Zealand form and also occurs widely in Oceania. Variety B, side view. This variety is not so well established as the preceding one, and may in fact belong to the bird-headed man group.

Type 1—Amulets of Human Shape. Class 3. This group is very varied and is more numerous than is generally supposed.

Type 2—Amulets in the Form of a Human Leg. This type is now well established, and is of particular interest since the piece from Waverley, in the Alexander Museum, demonstrates that the motive is religious and that the right leg is probably intended in all cases.

Type 3—Amulets of Bird Shape. Class 1, side view; Class 2, seen from above or below; Class 3, pekapeka, paired birds facing outwards. Variety A, bird heads; Variety B, human heads.

Type 4—Amulets in the form of bird-headed men.

Type 5—Amulets of mermaid form, marakihau.

Type 6—Amulets in the form of reptiles.

- 20

Type 7—Amulets in the form of fish.

Type 8—Amulets in the form of one-piece fish-hooks.

This group is especially variable, hybridizing with many diverse forms.

Type 9—Amulets in the form of implements.

Type 10—Reel-shaped amulets.

Type 11—Teeth as pendants or amulets.

Type 12—The rei puta.

Type 13—The chevroned amulet. Variety A, full-face; Variety B, profile.

The hei-tiki is distributed commonly throughout New Zealand, and it is, therefore, surprising that it has not been reported from the Chathams. The explanation of this lack of representation is probably that the islands are remote and isolated, and Moriori culture has in consequence lost many features of the parent Maori culture. Type 1, Class 2, is rare in New Zealand, being well represented only in the south. It occurs at the Chathams. Class 3 is universal in New Zealand but is not yet recorded at the Chathams. Type 2 is at present recorded only in the South Island. Type 3, Class 1, is universal in New Zealand but is doubtful at the Chathams, the examples reported there being perhaps conventionalized fishes. Type 3, Class 2, is at present represented by a single specimen excavated near Kaikoura (fig. 46). This is supported by a detail in fig. 119, and also by the bird in fig. 124, all from northern or central Canterbury. Type 3, Class 3, has not yet been recorded at the Chathams or from the South Island. It is commonest in the North Auckland peninsula, its most southerly representative being a greenstone piece in private possession found with a burial in South Taranaki. This limited New Zealand distribution constitutes a weakness in the present writer's theory that the pekapeka motive is old and widely spread in the Pacific.

Type 4 is not recorded at the Chathams. It is not yet recorded from the North Island, but, in my view, is common as a motive in wood-carving.

- 21

Type 5, the marakihau, appears in wood-carving in the Bay of Plenty region, none of it old; if it does occur in old carving it has not yet been recognized. The theory that the motive is old in Maori art at present depends on fig. 9, a greenstone piece in the Auckland Museum, described by its former Maori owner as marakihau. I am inclined to agree, but Mr. Archey takes the opposing view. If my view is to stand it will require additional data in support.

Type 6, reptiles, are generally distributed in New Zealand but are nowhere common. They have not yet been reported from the Chathams.

Type 7, fish, occur at the Chathams and in the South Island but are, as yet, reported only from Taranaki in the North Island.

Type 8, one-piece fish-hooks, are found at the Chathams and throughout New Zealand.

Type 9, implements, are not known at the Chathams but occur throughout New Zealand.

Type 10, reel-shaped, have a similar distribution.

Type 11, teeth, occur universally.

Type 12, rei-puta, are not reported from the Chathams but occur throughout New Zealand.

Type 13, chevroned amulet, is not recorded from the Chathams. Variety A occurs throughout New Zealand. Variety B occurs only on Otago peninsula.

It is obvious that many of the gaps in distribution recorded above will be filled when all local collections have been examined, and when careful archaeological work has been carried out in all districts. The only area in which the gaps are probably real is the Chatham islands, where isolation and the hurried process of colonization must almost certainly have resulted in the loss from the beginning of many culture-elements.

- 22

Type 1, Class 1. The human figure worn as a breast-pendant has been recorded from the Marquesas (human bone) and Easter island (wood). The Marquesan rendering seems to be closely allied to what may be regarded as the oldest variety of Maori hei-tiki. The Easter island male figure rendered in wood and used as a breast-pendant appears to be closely allied to the Moriori wooden male figure. That these two similarities are not accidental but indicate close relationship is rendered more probable by the fact that the cultures of the Marquesas and Easter island are in general closely related to Maori-Moriori culture. The connecting link between New Zealand and these two distant spots—as also with the Tuamotus and Hawaii—must be the Tahitian group. But this group does not afford a single amulet of this or, indeed, of any other type. The same is true of a number of other elements of material culture, including the whole of decorative art. I suggest that where such gaps in distribution do occur in the Tahitian group the culture-element once existed there but has since been lost.

Type 1, Class 2. The headless human figure is recorded from Easter island, the Marquesas, the Australs, the Cooks, New Zealand-Chathams, Tonga, and the Solomons. The occurrences at Easter island are in the script cut on wooden tablets. Laura Maud Thomson 1 has recorded numerous headless human figures from the Marianas in a context which strongly suggests writing.

Type 1, Class 3. The third group of amulets in human form is at present so ill-defined that it is not yet profitable to discuss its overseas relationships.

Type 2. The human-leg amulet is at present recorded beyond New Zealand only from the Santa Cruz group, and there depends upon a single piece. In this case, also, it represents a right leg. The human leg occurs as an amulet in Europe, but I have not seen any record of it in south-east Asia or in Indonesia.

- 23

Type 3, Class 1, birds rendered in side-view. This is a very common form in New Zealand, but I do not know of it elsewhere in Polynesia. Class 2, birds seen from above or from beneath. This appears in the Easter island script and in Tonga. In Tonga the motive is present as an amulet and also carved as an ideograph on clubs. It is present in amulets of Santa Cruz and the Solomons. Class 3, paired birds facing outward, occur in New Zealand, Easter island (script), and the Solomons. This motive appears sometimes to have been applied to the half-moon breast-pendant. This pendant is seen unaltered in New Zealand, Easter island, Tahiti, the Cook islands, Tonga, Fiji, and the Solomons. The influence of the bird-motive on the half-moon breast-pendant seems apparent in the pekapeka of northern New Zealand and in amulets of Easter island and the Solomons.

Type 4, bird-headed men, is seen in the Easter island script, in Easter island rock-carvings on the flat, in rock-carvings in Hawaii, and in wood-carvings in the Solomons. The bird-headed man occurs in the iconography of Egypt, Mesopotamia, Persia, India, and Indonesia. It would appear rash to deny the probability of connection between, the Indonesian forms and those of Melanesia and Polynesia.

Type 5, marakihau, may perhaps be based on the European mermaid, which has undoubtedly influenced some of the modern renderings in wood. The authenticity of the marakihau motive depends, at present, on a single greenstone amulet in the Auckland Museum. I do not know of the motive elsewhere in Polynesia or Indonesia.

Type 6, reptiles. The lizard-amulet occurs at Easter island, carved in wood in the round, and in the New Hebrides in pearl-shell. The lizard occurs frequently in the plaitwork and carving of the Marquesas and in the wood-carving of northern Melanesia. The snake seems to occur as an amulet only in New Zealand. It occurs in wood-carving of New Ireland, the Tami islands, and north-east New Guinea, in the latter two localities in a rendering almost identical with the New Zealand amulet. 2

- 24

Type 7, fish. This occurs, carved in the round in wood, at Easter island, in turtle-shell at the Santa Cruz islands, and in turtle-shell and tridacna-shell at the Solomons. The stone fish of the Society islands (puna) and of the Marquesas, though no larger than the Easter island fish, are definitely to be classed as religious rather than amuletic.

Type 8, amuletic one-piece fish-hooks. These occur throughout New Zealand and the Chathams, but are not recorded elsewhere in Polynesia. It seems probable, however, that the one-piece stone fish-hooks of Easter island, Pitcairn island, Mauke, and probably of other parts of Polynesia, were more often worn as pendants or amulets than used in actual fishing.

Type 9, amulets in the shape of implements. These have not been recorded in Polynesia outside New Zealand.

Type 10, reels. Related forms are present at the Marquesas and on archaeological sites in Cambodia.

Type 11, teeth, or imitations of teeth. Outside the New Zealand-Chathams area this interesting group is at present known only in the Marquesas and the Hawaiian islands.

Type 12, rei puta. If, as I have suggested, these have developed from the pearl-shell shank of the Polynesian bonito-hook, it is to be supposed that related amulets were present in most Polynesian islands, for favourite shanks seem to have been worn from time to time with something approaching amuletic significance in most of the groups. But so far as is at present known these shanks developed into true amulets only in Western Polynesia.

Type 13, the chevroned amulet. Variety B has been shown to be a local Otago variety. Variety A seems allied to forms in the Marquesas. Related forms are widely spread in central and northern Melanesia.

A study of the distributions here set out seems to indicate that this aspect of the material culture of New Zealand finds its closest relationships with the material cultures of Easter island and of the Marquesas. In respect of the motives used, relationship with central and northern Melanesia also seems fairly close. Whatever Marquesan culture has in common with New Zealand culture must surely have been drawn from the culture of the common - 25 parent, the Society islands. Why is it then that Society islands culture seems to be extraordinarily lacking in related motives? The writer's suggestion is that Society islands culture once had these motives but that it lost them as the interest of the islanders turned some centuries ago from the material side of culture to religious ceremonial.

A further point of considerable interest to students of Polynesian culture is that certain units of the Easter island script are to be found as amulets in other parts of Polynesia. Are they to be regarded as broken remnants of the script? Or was the script elaborated on Easter island by some genius who incidentally made use of some old Polynesian amulet motives?

Fig. 143. Human amulet. Soapstone. Scale 2/3. D'Urville island. Hornsey collection.

Fig. 144. Reptilian amulet. Soapstone. Scale 2/3. Goose bay, North Canterbury. McTaggart collection.

Fig. 145. Spear-point (?) amulet. Red argillite. Scale 2/3. Little Papanui. Otago University Museum. D. 21.5.

Fig. 146. Spear-point (?) amulet. Dark greenstone. Scale 2/3. Pahia, Southland. Sorenson collection.

Fig. 147. Fish-hook point amulet. Light greenstone. Scale 2/3. Birdlings flat, Canterbury. Sorenson collection.