- 11

[See illustrations at the end of previous article]

THE problems which centre round the earliest human migration into New Zealand are of great interest, and their solution is essential before much progress can be made in working out the history of Maori culture. Data concerning these earliest inhabitants can be drawn only from two sources, one traditional, the other archæological.

Traditional evidence relating to the first migration was small in volume and vague in character until the publication of the traditions placed on record by the late T. Whatahoro. 1 The part of these traditions which concerns the first immigrants is considerable in bulk and precise in character, though whether its accuracy is equal to its precision may be doubted. It describes a people whose strange physical characters and culture cannot be affiliated to any area of the Pacific, least of all to Polynesia, and are, indeed, unique.

Turning to the archæological evidence, we find that it is smaller in volume than could be desired, and for large areas of New Zealand does not as yet exist at all. In these Islands stratified sites, which in the Old World supply the most reliable evidence of the succession of ancient cultures, are extremely rare; thus far only a single one, the Moa-bone Point Cave, near Christchurch, has been excavated. There is available, however, another class of archæological or geological evidence, namely the association of Moa bones with the implements of the men who hunted them. It is known that the final extermination of some of the smaller species of Moa came at the hands of man, and there is good evidence for believing that the process of extermination was not long drawn out. As the men who hunted the Moa in any district may thus be regarded as among the earliest inhabitants of that district, the careful exploration of every Moa-hunter camp becomes a matter of great importance.

1. Note.—The numbers in the text refer to the bibliography at the end.

- 12

Moa-bone Point Cave, near Christchurch, is in some respects the most important site yet explored in New Zealand, for it combines association of implements and Moa bones with a considerable depth of stratification. 2 The Moahunter beds in this cave averaged a foot in thickness, whilst the average thickness of the beds above them was seven feet. There is some evidence indicating that these upper beds did not accumulate faster than the lower, and hence it may be inferred that the period since the extermination of the Moa in that part of Canterbury is at least seven times as long as the Moa-hunter period.

The artifacts found in Monck's Cave 3 are of great interest and importance, but the evidence that they are of Moa-hunter date is not conclusive.

The Shag River site is probably the richest that has ever been excavated in New Zealand, its only possible rival being Murdering Beach, Otago, the artifacts from which, however, are now irretrievably scattered among the collectors and museums of Europe and America. If we include the Moa-bones which have been taken from the Shag River site, it is indeed incomparably the richest site of any kind yet worked in New Zealand. Unfortunately the site has suffered much from careless digging and the failure to keep records of excavation. The first recorded excavations here were carried out under Von Haast's direction in 1872. 4 In 1874 B. S. Booth, under Hutton's direction, dug on the site for three months, and was able to send Hutton, then curator of the Otago University Museum, ten boxes full of the material collected. 5 One must suppose that Booth found a considerable amount of ethnographic material in addition to the Moa bones, which doubtless filled most of the space in the ten boxes, but where this ethnographic material went remains an unsolved mystery. It certainly did net remain in Otago. The next investigator of the site was interested in it from a purely commercial point of view. He dug up Moa bones, conveyed them on a punt constructed for the purpose, up the Shag River to the railway, and filled several - 13 railway trucks with them. These were dispatched to Dunedin bone-mills which, however refused them, and all the bones were destroyed. Worse still, not a scrap of the ethnographic material which must have been found at this time can be traced. In the late eighties the site was excavated by Augustus Hamilton, in whose note-books are preserved careful plans of the numerous trenches dug, with notes as to the nature of the deposits. I have to thank Mr. Harold Hamilton for permission to consult these records. It is greatly to be regretted that Hamilton kept no record whatever of the implements, tools and weapons found. All the objects recovered from the site passed into his general collection, and it is now impossible to separate them out from the rest, far less to assign them to definite strata. About the same time Mr. (now Sir Frederick) Chapman did a little digging and made the earliest and one of the most important contributions to our knowledge of the culture-history of the site. In addition to these collectors there were a great many occasional visitors who must in the aggregate have taken away a large amount of material. none of which is now available for scientific study.

Considering, therefore, the great amount of casual scratching, and even more the vigorous but unsystematic digging that has been indulged in on the site for half a century, it is surprising that anything was left for the systematic investigator. He at length appeared about 1915, in the person of Mr. David Teviotdale, who has described his methods in the paper preceding this. When he began collecting on this site, Mr. Teviotdale had already long been a collector of stone implements. The first adze he found was lying on a hillside east of Hyde at a time when that town hummed with the life of a goldfield. In Hyde he met B. S. Booth, who had fled across the border from Canada, so the story goes, with a price on his head as a result of the part he played in Reil's rebellion. After many years in the deserts, mountains, and alluvial goldfields of California and other Western States he came to Otago, where, digging for gold, he found the great deposit of Moa bones in the Hamilton swamp. His paper on this deposit, published in 1874 6 shows him to have been a man of good education and good intellect. Hutton, as we have seen, employed him on excavations at the Shag River mouth in 1874, and the loss of Booth's report on his work there must be regarded as a - 14 heavy loss to the science of ethnology in New Zealand. A tale of adventure and romance lies in the unrecorded grave of Booth in the deserted goldfield at Naseby.

From the beginning of 1919 up to the present time Mr. Teviotdale has worked carefully and has kept a journal of his excavations. He has generously presented to the Otago University Museum the greater part of the material recovered. It is a most extensive collection, including numbers of knives and saws and cleavers of quartzite brought from inland Otago, scores of drill-points, scores of cutters and polishers, and a great many implements in various other materials, such as bone, basalt, and greenstone.

The Shag River site consists of a small, almost insignificant, range of sandhills separating the estuary of the river from the sea. To south and west lie warm, rolling downlands, admirable country for the flock of Moas which must have swarmed there before man came. It seems probable that the Moa-hunters drove the Moas from the downs into the triangle of sandhills and there killed and cooked them. No one who has read the accounts of the earlier diggers and has seen the bones laid bare by the spade can doubt that many hundreds of Moas were killed there. How long ago that took place we have no means of knowing, but the coast has sunk some feet since that time, for many of the earth ovens are now well below high-water mark.

The principal problem at the Shag River mouth is to decide which implements are of Moa-hunter date and which are later. At Moa-bone Point Cave that problem is simple, for all objects below a clearly marked stratum belong to the Moa-hunters, while all above are more recent. But in the tumbled topography of the Shag River sandhills there is only one test, and that rarely applicable. Articles found below undisturbed beds containing Moa ribs, vertebrae, and other bones not utilisable in industry must be of Moa-hunter date. No doubt many objects which have not been sealed down in this way are of Moa-hunter date also, but their bona fides should be carefully scrutinised before they are admitted as evidence.

I think we are justified in believing that the greater part of the artifacts recovered from the site were made and used by the Moa-hunters. The grounds for thinking so are two. In the first place, the material from the site has a look of its own; no one who has collected from Otago

- i

FIG. 2—A-D. Scale ½

A, Beater resembling Moriori type. O.U M. Coll. and pres. D. Teviotdale., B, Beater. O.U.M. Coll. and pres. D. Teviotdale., C, Slate Scraper. O.U.M. Coll. and pres. D. Teviotdale., D, Cast of Slate Scraper. Original in Dominion Museum. Coll. A. Hamilton.

- ii

FIG. 3—A-F. Scale ½

A, B, C, Unfinished quartzite spear points (?)., D, F, F, Butts of broken spear points., D, chert; rest quartzite., O.U.M. Coll. and pres. D. Teviotdale.

- 15

beaches or has studied Otago material can have the slightest difficulty in “spotting” a group of artifacts from the Shag River mouth. It is true that no unique types are found there, but numerous types which are rare elsewhere are there extremely common. Several classes of work appear, also, to be, on the average, much better executed there than on other Otago sites, e.g., large flaking, the finishing of stone fish-hook shanks, and the shaping of bone into large one-piece hooks. It must be admitted that some part of this general likeness arises from the use of characteristic local material, but when full allowance has been made for this factor there still remains a definite family likeness due, it may be supposed, to a common technical tradition. In this view I have the support of Sir Frederick Chapman and Professor Wood Jones.

In the second place, the number of Moas eaten on the site would have gone far towards supporting for some time a village of respectable size, judged by Otago standards; one large enough to have produced the great bulk of the artifacts found on the site. But after the extermination of the local Moas the site would not afford the food requisite for a permanent village. Nor were the geographical conditions at the Shag River mouth such as to compensate for this disadvantage by encouraging trade in food-stuffs from more favoured districts. Thus the place is known to the present Maoris of Otago only as an occasional camping-ground for fishing parties. No tradition exists of the settlement that once flourished there, nor, for that matter, have the Otago Maoris any knowledge, apart from European sources, of the existence of the Moa.

On these grounds I conclude that the great bulk of the articles recovered from the site belong to the Moa-hunter period. It is proposed in this paper to discuss first those culture elements which certainly belong to that period, and afterwards those other elements which probably belong, but about which there is less certainty.

The men who hunted the Moa in North Otago used the Polynesian earth oven. It is to be presumed that they used the Polynesian fire-plough, but, as is to be expected, no examples of this have been preserved. They had dogs which they probably ate, for all the jaw bones found belong to - 16 young dogs. Beneath a stratum of Moa bones Mr. Teviotdale discovered the skull of a dog much larger than any that I have seen from old Maori sites. The adzes found beneath deposits of Moa bones form an important group. The three largest (Fig. 1, A B C) were found by Sir Frederick Chapman. A belongs to the type generally regarded as the most typical Polynesian of all Pacific artifacts. B belongs to the same type, but a wide groove has been pecked across the polished front to hold binding supplementary to that held by the “grip,” which is of normal size. This is not a purely individual feature, for it occurs, though rarely, on other Otago adzes of this type. C is in some respects the most interesting of the group. It belongs to a highly specialised type which was common in the Chatham Islands and in Tahiti, and occurred also at Easter Island. 7 Though it does not appear to have been recorded from other Polynesian groups it probably occurred in them all, since it was well rep resented in the Admiralty Islands. It does not occur, so far as I am aware, in any part of New Guinea or in Melanesia proper. Sir Frederick Chapman was told by old Otago Maoris that this was a type used by the ancient Waitaha people, but not by later tribes. If weight is to be attached to verbal evidence on this class of fact this statement strongly supports the Polynesian affiliation of the Moa-hunters. Mr. Teviotdale found a small unfinished example of the same type beneath Moa-bones, and also another adze of generalised Polynesian form. All five of these adzes are made of local materials. The only other adze of undisputed Moa-hunter date (at present known) is that from the Moa-bone Point Cave, which is entirely Polynesian in character. Not a single adze or fragment of an adze of Papuan or Melanesian type has yet been found in deposits of Moa-hunter age. The evidence of Moriori adze types renders it probable that occasional adzes of Western Pacific types will be found, but it is in the highest degree improbable that these will amount to more than a tiny fraction of the total.

The other artifacts associated with the Moa-hunters are few, but of considerable interest and importance. One is the polished greenstone implement of uncertain use found by Hamilton and Chapman, 8 and the other the fragment of a greenstone adze found by Mr. Teviotdale. As the former

- iii

FIG. 4—A-I. Scale ½.

A, C, D, E, Fine Polishers; A, schist, rest mud-stone. B, schist cutter., F, G, soft drill-points, mud-stone. H, I, hard drill-points quartzite and chert. O.U.M. Coll. and pres. D. Teviotdale.

- iv

FIG. 5—A-F. Scale ⅓

A, C. Quartzite implement, one side untouched, the other retouched., D, quartzite saw. E, large quartzite flake. B, triangular mud-stone cutter or polisher. F, schist polisher, circular in cross-section., O.U.M Coll. and pres. D. Teviotdale.

- 17

cannot now be traced, no judgment can be formed as to which of the two greenstone regions the material is derived from. Nor is the fragment found by Mr. Teviotdale sufficiently characteristic to warrant a definite conclusion. It is unlikely, however, that at this early period the difficult country between Lake Wakatipu and the western coast had been explored. The alternative source—the river and beach-gravels of Westland— is therefore more probable. This would connote considerable maritime activity.

The artifacts less definitely associated with the Moa-hunters have now to be considered. The adzes which may be studied in the Otago University Museum, are uninteresting, and not as numerous as might be expected. Flax-beaters number only two, though it is probable that some of the elongated water-worn pebbles on the site also served this function. Fig. 2, B represents the University Museum example, which is less complete than one of the same type found in 1872 and now in the Canterbury Museum. A beater of similar shape, but made of whale bone, from Waitangi, Chatham Island Museum, is the only other example of this type known to me from the New Zealand—Chatham Island area.

Some half-dozen pointed blades of quartzite (Fig. 3, A B C) appear to be unfinished spear-heads. No finished and complete spear-heads were found, but there are three chipped butts (Fig. 3, D E F), which seem to belong to such implements. These quartzite blades are easily snapped, and it may be supposed that if used in hunting the Moa they would often be broken in localities away from the settlement. This may account for the absence of perfect completed examples and the presence of broken butts which would be cut loose from the binding after the return from the hunt. The blades are triangular in cross section, and have been detached from a dressed quartzite block in the way widely practised by the Stone Age peoples. They are almost indistinguishable from the quartzite blades of Central and Northern Australia. I am not aware that this method or type of implement has been recorded elsewhere in the Pacific except in the Admiralty Islands. It is remarkable that it has not yet been recorded elsewhere in New Zealand or the Chathams, though this may perhaps be due to the absence of quartzite or large blocks of flint. - 18 If there is any genetic connection with the obsidian spearheads of the Admiralties we should expect to find similar obsidian spear-heads in the Bay of Plenty district of the North Island, where abundant supplies of obsidian could be obtained. This is, of course, if the material culture of the Moa-hunters and of the ancient inhabitants of the Bay of Plenty were identical, a supposition which is as yet unproved.

Several examples of the slate implement usually described as a scraper were found, the most interesting one Fig. 2, D), being perforated for suspension, and decorated with a pair of spirals carved in low relief. 9 This kind of implement is not uncommon in Otago, but no other decorated example is known.

No weapons, apart from the suggested spear-heads, have been found on the site. Wooden objects are limited to a decayed, unfinished bowl, with lip and fiat base, and several fragments of canoes and rough unidentified pieces.

The large supplies of Moa-bone resulted in the extensive manufacture of articles from that material. Numbers of drill-points (Fig. 4, H and I) made of quartzite, chert and chalcedony were found, and these were doubtless used principally in the drilling out of one-piece fish-hooks. It is remarkable that though the ancient inhabitants of Otago were capable of executing the most delicate retouch along the edges of flakes, they seem to have lacked any conception of long, slender drill-points, such as were common in aboriginal America. Our present evidence indicates the same weakness throughout the Pacific. In spite of the large amount of drilling done on the Shag River site, even the best drill-points are crude.

Polishers and finishers were very numerous, and varied greatly in size, shape, and material. Sandstones, mudstones, and micre-schists were all well represented, the former two rocks being found locally, and the last-named coming, doubtless, from Central Otago. Some of these polishers are large and weighty, and are marked by dishshaped hollows in which the implements were rubbed. Others are small (Fig. 4, A C D E), and are capable of the extremely delicate work called for in finishing some classes of fish-hook and pendant.

- v

FIG. 7—A-H. Scale ⅓

A, Pickers, needles, and threaders., B, Stout Moa-bone points., C, Pickers, prickers, and threaders., D, Wood-working implements., E, Prickers., F, and G, Wood-workers (?)., H, Needles., O.U.M. Coll. and pres. D. Teviotdale.

- vi

FIG. 8—A-F. Scale ⅓.

A, Broken one-piece bone hooks of large size., B, One-piece bone hooks., C, E, F, Method of manufacture., D, Broken parts., O.U.M. Teviotdale Collection.

- 19

All of the very numerous knives, saws, scrapers and cleavers are made of Central Otago quartzite, and the best of them show a high degree of skill in knapping. Occasional scrapers show very delicate retouch. A number of varities of scraper and plane appear to be represented, and the whole class would well repay careful study, which no one has yet given it (Fig. 5, A C D E).

The most interesting objects from the site are probably those made of bone. They include needles, pricers, threaders, wood-carvers, tatooing chisels, bird-spear points, fish-hooks (both composite and in one piece), harpoon points, and pendants. Fig. 6, represents two delicate

FIG. 6—Two bone needles and the case in which they were found.

Scale ½ O.U.M. Coll. and pres. D. Teviotdale.

needles and the bone needle-case in which they were found. The case was probably not designed as such, though it resembles a type of bamboo case for small fish-hooks and minute objects occurring in the East Indies and, I believe, in Northern Melanesia. Fig. 7 includes needles (H), wood working implements (F G, some of D, and probably B), prickers (some of A C and E), threaders for threading up fish (some of A), shell-fish pickers (A and E).

One-piece fish-hooks, most of them in a fragmentary condition, are very numerous. The most interesting variety is of very large size (Fig. 8, A), only two or three examples of this variety have yet been recorded elsewhere in New Zealand. The remaining pieces in Fig. 8 illustrate other varieties of the type, and indicate the method of manufacture by drilling.

Barbed bone points of the spears used in securing pigeons and smaller birds are numerous on most Otago beaches, but are here represented by four specimens only (Fig. 9, I). The common implement usually known as a fish-gorge is here represented by only two examples (Fig 9, E). Three interesting fragments (Fig. 9, G) probably belong to harpoon-leads working on the toggle - 20 principle, which become loosened from the wooden shaft when the prey is struck. This form is found, though rarely, on most Otago sites, and is represented from Chatham Islands by a fine example in the Otago University Museum.

The rest of the objects represented in Fig. 9, with the exception of the three notched pieces in J, which are probably shanks, are the bone points of composite fish-hooks. A and C are the few poor representatives of a variety which is found in large numbers and often beautifully executed on the other Otago beaches. B represents two points made from seal's teeth. F and D are the unbarbed points of barracouta hooks, all made of Moa-bone, except one in F, which is made from the lower jaw of a dog, and has the canine tooth as a part of it. The rest are stout points made of Moa-bone, and representing several distinct varieties of hook.

Perhaps the most interesting class of object found on the site is the composite fish-hook with stone shank and bone point, represented in Fig. 10, L M N. Stone shanks of this type are recorded from all parts of New Zealand, but the Shag River examples are notable in number and in excellence of finish. The fact that no example has ever been secured with bone point attached may legitimately be advanced as evidence that the type belongs to an early ethnic wave. 10 As no complete examples of this variety occur, the kind of bone point attached to the stone shank can be determined only by examples in which point and shank are found in association. Only two examples with shank and point in association have thus far been recorded, one of which is represented by Fig. 10, L, and Fig. 11. This was found by Mr. Teviotdale lying as shown, Fig. 11. That the line was attached as in the Polynesian example (Fig. 12), is rendered certain by the fact that the perforation is too small to take a line, and can have been used only for a fine bridle. It is believed that the unbarbed bone

- vii

FIG. 9—A-K. Scale ⅓

A, C, Small barbed bone points of composite fish-hooks., B, ditto, of seal-tooth., E, “Fish-gorges.”, D, F, Unbarbed bone points of barracouta hooks., G, Parts of bone harpoon toggle-points., H, Butts of large bone points., I, Barbed bone points of bird spears., J, Points of large fish-hooks and shanks., K, Large points of composite fish-hooks., O.U.M. Teviotdale Collection.

- viii

FIG. 10—A-N. Scale ⅓

A, Dentalium fossils., B, Greenstone pendant in form of tattooing chisel., C, Unfinished bone tattooing chisel., D, Rings cut from dentalia, and ornamented with horizontal grooves,, E, Finished bone tattooing chisel., F, Seal teeth perforated as pendants., G, Human ribs, shaped and perforated as pendants., H-K, Bone pendants in form of shanks of fish-hooks., L, Composite fish-hooks., M. Stone shanks of composite fish-hooks., N. Bone points to be attached to stone shanks (?). O.U.M. Teviotdale Collection.

- 21

FIG. 11—Stone shank and associated bone point. Scale ½

O.U.M. Coll. and pres. D. Teviotdale

FIG. 12—Composite Fish-hook showing method of attachment. Tahiti (?). Scale 1/1

O.U.M. Pres. Sir F. Chapman.

points (Fig. 10, N) were designed for attachment to this type of shank, for they are common on this site, where the shanks are common, and rare on other beaches where the shanks are rare. This kind of point is used with stone shanks at Ocean Island. The large shank (Fig: 10, M, right) is a beautiful piece of work, probably the finest in any collection.

The upper end of a very large shank (Fig. 10, M, left) is without history, but has been in the museum for a long time. The material from which it was made is found in situ at the head-waters of the Waitaki not far distant from the Shag River. It is thought that it may be the sole remaining piece of the ethnographic material which Booth must have sent to the museum in 1874.

That the stone-shanked composite hook was an important culture element in Maori eyes is indicated by their use of pendants imitating its shape. Fig. 10, H I J K represents four pendants of this shape made of Moa-bone. It is possible, however, that J may have been an actual shank. K departs from the normal in having a hole at the distal end.

- 22

Pendants and other personal ornaments were poorly represented on the site. Fig. 10, A are fossil dentalia from the Waitaki valley. D are what appear to be sections of dentalia ornamented with grooves arranged circumferentially. Numbers of beads of this kind have been found in North Otago. F are two perforated seal's teeth, while two human ribs ground into shape and perforated are represented by G. C is an unfinished tatooing chisel, with six teeth, and E a finished one from which the teeth have been broken. B is a greenstone pendant unique in form, representing a tatooing chisel.

The objects illustrated form but a small fraction of the artifacts found at the Shag River mouth, but it is believed that they include the most important pieces from the Hamilton, Chapman, and Teviotdale collections. Von Haast's collection included only one important piece—a beater or pounder similar to Fig. 2, B. Of the rest of the material collected we know nothing, except that it was probably considerable in amount, and that it included shark teeth from a tuatinisimilar to the well-known Gilbert Island weapons.

The articles not figured include obsidian, which must have been imported from the North Island, red ochre, hammers, two or three rough grooved sinkers, a small unfinished wooden bowl, a number of stone adzes and chisels, and a great number of flakes of clert and quartzite. There were also a number of greenstone chisels, drill-points and adzes.

[While this paper was in the press Mr. Teviotdale sent to the University Museum the small argillite adze, Fig. 13. It was found in a shallow deposit under about four inches of sandy loam. The deposit consisted of moa and other bones lying on sand, the adze lying on the sand beside a moa rib-bone. The adze is made of clean grey argillite with black strain lines, which is found in the neighbourhood of St. Andrews, South Canterbury.]

- 23

FIG. 13—Argillite adze, associated with Moa bones. Scale 1/1

O.U.M. Coll. and pres. D. Teviotdale.

1. Most of the articles from the site are probably of Moa-hunter date; hence the collections, particularly that of Mr. Teviotdale, may be used, though with reservations, as evidence of the material culture of the Moa-hunters.

2. The Moa-hunters were probably the first inhabitants of the district.

3. The objects found in deposits of undoubted Moa-hunter age prove the material culture of the hunters to have included finely finished and highly specialised adzes of Polynesian types. The hunters also possessed the dog and the earth-oven. Their possession of nephrite indicates a close acquaintance with the geography of the South Island, coupled, probably with considerable coastal canoe travel, for it is unlikely that nephite was brought to the Shag River mouth across the Southern Alps at that early date.

4. The culture of the Moa-hunters did not differ in any of its elements from that of the Otago Maoris in times immediately preceding the arrival of Europeans. But on the whole, craftsmanship appears to have attained a higher standard than in more recent times, as is indicated by: (a) The frequent use of attractively coloured stone for the shanks of fish-hooks and the excellent work displayed by these shanks. (b) The excellence of work in bone, as illustrated by design, and execution of one-piece hooks. (c)

- 24

The decoration of articles usualy left plain, e.g., the slate scraper.

5. There is no definite evidence as to the date of the Moa-hunter occupation of the site, though the subsidence of the land since that time and the general appearance of the excavations suggest considerable antiquity. The most definite evidence as to Moa-hunter chronology is that afforded by the Moa-bone Point Cave, Christchurch.

6. Only some of the smaller species of Moa are represented in the middens.