A Comparison of Eskimo and Palaeolithic Art 9781463221126

Citation preview

Comparison of Eskimo and Palaeolithic Art

A n a l e c t a Gorgiana

237 Series Editor George Kiraz

Analecta Gorgiana is a collection of long essays and

short

monographs which are consistently cited by modern scholars but previously difficult to find because of their original appearance in obscure publications. Carefully selected by a team of scholars based on their relevance to modern scholarship, these essays can now be fully utili2ed by scholars and proudly owned by libraries.

A Comparison of Eskimo and Palaeolithic Art

Frederica deLaguna

l gorgias press 2009

Gorgias Press LLC, 180 Centennial Ave., Piscataway, NJ, 08854, USA www.gorgiaspress.com Copyright © 2009 by Gorgias Press LLC Originally published in All rights reserved under International and Pan-American Copyright Conventions. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning or otherwise without the prior written permission of Gorgias Press LLC. 2009

1

ISBN 978-1-60724-466-0

ISSN 1935-6854

Extract from The A^merican Journal of Archaeology, vol. 36 (1932).

Printed in the LTnited States of America

A COMPARISON OF ESKIMO AND PALAEOLITHIC

ART

PLATES X I X - X X I I I * A COMPARISON of Eskimo and Upper Palaeolithic art has been undertaken in this s t u d y . I t forms p a r t of an investigation of the general thesis t h a t t h e modern Eskimo are t h e descendants by blood and culture of the cave men of Upper Palaeolithic Europe. This assertion is an old one and one which has been continually appearing in t h e literature of E u r o p e a n prehistory f r o m t h e time when it was first uttered. I t has gained a new interest through the researches of D r . P a u l R i v e t of Paris and Professor Sollas of Oxford, while the recent work of D r . Birket-Smith of Copenhagen reformulates t h e old problem in a new and what may prove to be a f r u i t f u l way. F u r t h e r investigation of the question m a y not only lay a troublesome ghost, b u t m a y also lead to related problems of importance. The assertion t h a t t h e Eskimo were descendants of the cave men rested upon evidence drawn f r o m physical anthropology and ethnology. I t would lie outside t h e limits of t h e present study to enter into the controversy over the racial affinities of Chancelade m a n . T o investigate t h e possibility of a Palaeolithic basis of Eskimo c u l t u r e would necessitate both a comparison of all similar types of specimens and a s t u d y of the distribution of those possessed in common, not only to determine t h e closeness of t h e possible relationship, b u t to see if the great gap in time and space might be bridged. Since the similarity between Eskimo and Palaeolithic art is one of t h e arguments most frequently advanced by those who postulate a connection between the two peoples, we shall compare the two arts in some detail. The explanation of the similarities existing between the material culture of t h e Eskimo and t h a t of t h e U p p e r Palaeolithic peoples of E u r o p e constitutes a real problem. W e should be prepared, however, to find certain likenesses due to a somewhat similar environm e n t . I t is a question whether these resemblances are too great to be explained b y fortuitous circumstances and must therefore argue a common origin of culture. T h e aim of the present study will be to find out what would constitute proof of a connection between Palaeolithic and Eskimo culture; and whether the a r t of t h e t w o peoples can furnish such a proof. Our conclusions in the field of a r t can not prove or disprove the theory of a connection until all possible similarities in material cult u r e have been compared. N o m a t t e r what t h e final answer, our studies should cast some light on the more general problem of diffusion versus independent origin, inasm u c h as a large body of pertinent d a t a will be collected. THE HISTORY OF THE PROBLEM T h e early works dealing with the problem fall into several divisions. T h e first discussions were characterized by dogmatic assertions based upon scanty ethnological material alone, t h a t t h e Eskimo of today are the direct descendants of t h e cave men of France. T e s t u t ' s publication on the Chancelade skull changed the n a t u r e of the problem by introducing anatomical material. Following him, there have been several works dealing almost exclusively with this new evidence, as well as *For list of illustrations, see pp. 509-511. 477

THE ARCHAEOLOGICAL INSTITUTE OF AMERICA

478

F R E D E R I C A DE L A G U N A

more comprehensive researches containing all the available material. L a t e r research has tended toward a more cautious statement of hypotheses, with a fuller description and enumeration of actual specimens and sources upon which the proof is based. The earliest supporters of t h e theory of a Magdalenian parentage for Eskimo man and his culture were superficial in their investigations, uncompromising in tone, and their writings scanty in documentation. However, t h e material to which they allude was probably better known to their contemporaries t h a n to modern readers, since there was not at t h e time the vast accumulation of publications which now confronts us. A short sketch of the history of the problem and of the contributions of various writers on the subject will be useful for our study and informing in regard to t h e evolution of a scientific a t t i t u d e toward the material. According to Sollas, 1 P r u n e r Bey was the first scholar who a t t e m p t e d to identify t h e men of t h e Upper Palaeolithic with various Mongoloid races. H e was criticised by H a m y for advancing theories without sufficient proof, although H a m y admitted some t r u t h in his assertions. 2 One of the earliest writers to introduce on an extended scale comparative ethnological material into the study of prehistory was Sir John Lubbock. 3 His analogies were not limited to comparing a particular ancient culture with any particular modern one, and he certainly did not a t t e m p t any such identifications. T h e first to claim the Eskimo as the descendant of Magdalenian m a n was William Boyd Dawkins. 4 His sources consisted of works by L a r t e t and Christy. 5 H e also referred to the journals of the arctic explorers, P a r r y , Ross, and Lyon. I t is interesting to know Dawkins' sources, for m a n y who relied on his work never verified his statements, which are repeated over and over again with all the assurance of wellproven facts in t h e publications immediately following t h e first presentation of his theory. After briefly reviewing the discoveries at Les Éyzies and La Madeleine, Dawkins says: 6 " T h e sum of evidence proves t h a t man, in a h u n t e r state, lived in the south of Gaul, on reindeer, musk-sheep, horses, oxen, and the like, a t a time when the climate was similar to t h a t which those animals now inhabit. To what race did he belong? 7 I n solving this, the zoological evidence is of great importance. T h e reindeer and musk-sheep now inhabit the northern p a r t of the American continent, and are the principal land animals t h a t supply the Eskimo with food." These animals are known to have retreated further and further north during the historic period until t h e musk-ox completely disappeared from Europe. " M a y not the race t h a t fed on these two animals in Southern Gaul have shared also in their northern retreat, and m a y it not be living in company with them still? T h e t r u t h of such an 1

W. J. Sollas, Ancient Hunters and Their Modern Representatives,3 London, 1924, p. 562. E . T . H a m y , Précis de paléontologie humaine, Paris, 1870, p. 355. 3 Prehistoric Times, as Illustrated by Ancient Remains and the Manners and Customs of Modern Savages, London, 1865. 4 "Esquimaux in t h e South of Gaul," in the Saturday Review, London, December 6, 1866. s Christy, Les Cavernes du Perigord, Paris, 1864; Lartet, " L a m e d'ivoire fossile," Les Annales des sciences naturelles, F i f t h Series. IV, Paris, 1866. = Op. cit., p. 71S. 7 This question implies t h a t the cave man must have been related to some living race. 1

ESKIMO AND PALAEOLITHIC ART

479

hypothesis as this is found by an appeal to the weapons, implements, and habits of life of t h e E s q u i m a u x . " T h e elements common to t h e two cultures mentioned by Dawkins are: the fowling spear, harpoon, scraper, and marrow spoon, t h e habit of crushing bones for t h e marrow, and t h e lack of regard for the dead. Both peoples are of small bodily size. 1 T h e only difference between t h e cultures is t h e possession of the dog by t h e Eskimo, b u t a long enough time has elapsed to account for its adoption f r o m some other people or for its domestication. As the climate of southern France became warmer a t the end of the Palaeolithic period, "musk-sheep, reindeer, and Esquimaux would ret r e a t further and f u r t h e r north, until they found a resting place on t h e American shores of t h e great Arctic sea. Possibly, in the case of the Esquimaux t h e immigration of other a n d better armed peoples might be a means of accelerating this movement." This hypothesis was given a more extended treatment in Cave Hunting, in 1874.2 Dawkins augmented his previous list of cultural types common to t h e Eskimo and Palaeolithic man. 3 T h e elements in common are: the harpoon (the Palaeolithic forms differing only in their grooved barbs), heads of fowling and fishing spears, darts and arrows (with the same types of base for hafting), t h e shaft-straightener, scraper of stone in handle of ivory, stone lance head, sewing needle, and realistic art (except in so far as t h e hunting scenes depicted by t h e Palaeolithic artists were not familiar to the Eskimo). 4 I n conclusion he advanced t h e argument already quoted, pointing out t h a t " t h e r e are no two savage tribes now living which use the same set of instruments without being connected by blood. . . . T h e conclusion, therefore, seems inevitable, t h a t so far as we have any evidence of t h e race to which t h e dwellers in the Dordogne belong, t h a t evidence points only in t h e direction of t h e Esquimaux." 5 Practically t h e same argument is m a d e in his later work. 6 Dawkins' point t h a t similarities between two sets of culture elements make a stronger argument for a common origin than a similarity between isolated items is, of course, well taken. There remains, however, the question of fact. F u r t h e r more, it seems rash to a t t e m p t to prove racial affinity by ethnological material. Dawkins' theory did not pass unnoticed. Sir John Lubbock incorporated his views in a later edition of his w o r k 7 and wrote a favorable review of Dawkins' position. 8 Also, Charles E. Ranee 9 summarized some of Dawkins' arguments and added t h e suggestion t h a t t h e European Arctic f a u n a and Palaeolithic m a n m a y have reached the New World during interglacial times by means of a land bridge f r o m Siberia to America. I n 1870, H a m y 1 0 followed Pruner Bey in assigning the races of Cro-Magnon and Furfooz to t h e group of hyperborean peoples. There is, of course, no question as 1 As proved by the finding a t La Madeleine of poignards with handles so small t h a t " t h e y can only be grasped by three fingers of an ordinary-sized hand—a fact that indicates a small race of men." 1 Dawkins, Cave Hunting, London, 1874. a Op. cit., pp. 334-335. 4 Op. cit., pp. 356-357, figs. 99, 107, 114, 121-124. * Op. cit., pp. 357-358. 6 7 Early Man in Britain, London, 1880, p. 233. Prehistoric Times; cf. Dawkins, 1874, p. 359. 8 Introduction to Sven Nilsson's, The Primitive Inhabitants of Scandinavian London, 18C8. 9 Nature, April 22, 1875; X I , p. 493, quoted in the Reliquiae Aquitanicae, p. 284. 10 Op. cit., p. 355.

480

FREDERICA DE LAGT'NA

yet of the Chancelade remains. As soon as they were found, every one except Déchelette admitted t h a t they belonged to a different race from t h a t of Cro-Magnon, and an a t t e m p t was m a d e to prove the affinity of this race with t h a t of t h e Eskimo, while Cro-Magnon man was left out of t h e argument. According to H a m y , on the ethnological side we must t u r n to the Arctic peoples for analogies with Palaeolithic Europe. H e first warns us against being misled by superficial and fortuitous similarities. T h e works of primitive m a n in stone are the same everywhere, and this is true of some simple bone implements. Only in the extreme north, among t h e Lapps, Eskimo, and Chukchi, do we find reproduced, " i n all their details," t h e tools of stone and bone of t h e Upper Palaeolithic. T h e list of similar types includes: t h e "couteau à soie" of t h e Vézère valley (?) which is t h e same implement as t h a t used by t h e Eskimo, Chukchi and Lapps; the flint scraper, t h e awl (but he has already listed this among the simple tools of universal distribution), the smoother (lissoir), " h a r p o o n , " bone point (already listed as of universal distribution), spatula, needle, dart point, etc., and style of art, particularly in t h e carving of animals, though in his opinion Eskimo art was decadent. A similarity of customs was also noted: in regard to the eating of marrow, boiling liquids with hot stones, (which m a y explain the presence of so m a n y stones in t h e French caves!), making fire by friction, as a t Les Êyzies (sic!) or by pyrites as a t Chaleux, and the lack of regard for the dead. H a m y also called attention to t h e similarity between t h e Scandinavian stone age and the French Magdalenian, as proof t h a t Palaeolithic m a n accompanied the reindeer in its migration to the north. Some of the circumpolar tribes are actually the descendants of Palaeolithic man, and " c o n tinue in our own day in t h e circumpolar regions t h e age of the reindeer of France, Belgium, and Switzerland, with all its zoological and ethnographical characteristics. " 1 D u p o n t 1 accepted Boyd Dawkins' conclusions, and added a few items of resemblance between the Eskimo and the cave man, though the lack of weapons suggestive of organized warfare is t h e only significant and valid point mentioned. 3 Although published after the discovery of the skeletal remains so important in any consideration of this problem, the opinions of two French writers, Laloy and Déchelette, and of one American, Hoffman, ought to be mentioned here because they deal only with the material known to Boyd Dawkins and do not consider Testut's arguments. Hoffman 4 andLaloy 5 advanced argumentsagainstBoyd Dawkins'theory. They seem to believe t h a t a crossing from Europe to Greenland, via the British Isles, for which they rightly find no evidence, is necessarily involved in the " folio wing-thereindeer" theory. This was certainly not a p a r t of Boyd Dawkins' thesis. They arbitrarily dismiss the possibility of a crossing via Bering Strait. The argument for a Palaeolithic basis of Eskimo art, resting on t h e similarity of realistic figures, can not be held, since, according to Hoffman, Eskimo pictographic representation is 1

Op. cit., pp. 357-367. 3 M. E. Dupont. Les Temps préhistoriques en Belgique,2 Brussels, 1872. Op. cit., p. 187. 4 Walter James Hoffman, "The Graphie Art of the Eskimos," Report of the U. S. National Museum for 1895, Washington, D. C., 1897, pp. 764-765. 5 L. Laloy, a review of Hoffman's monograph in L'Anthropologie, IX, Paris, 1898, pp. 585-587. 2

E S K I M O AND PALAEOLITHIC A R T

481

modern, attributable to contact with the Russians. 1 Similarities in types of implements or in realistic art are due to the use of the same materials and of the same animals as models. Déchelette 2 also attacked Dawkins' thesis, thereby exhibiting his reluctance to explain cultural phenomena by a movement of peoples. Thus, he denied any proof of migration during the whole course of the European Palaeolithic and Epipalaeolithic periods or even of the direct diffusion of a foreign culture into Europe during that time. 3 Déchelette was unable to understand why Palaeolithic man should have preferred to follow the reindeer to the north when he could have hunted the stag and the wild boar, and when the climate in Europe was becoming so much more agreeable.4 Similarities between Eskimo and Palaeolithic culture he ascribed to a similarity in material life and natural environment. He also mentioned Hoffman's opinion that Eskimo art is recent. Déchelette is open to criticism because of his extreme views, the more so since he made no allusion to the evidence which skeletal remains or the temporal and geographic distribution of culture might give. He also erred in denying a migration because the motives for a migration to the north following the reindeer are not apparent. Hollas ' answered in part, at least, some of Déchelette's objections, and we know from the spread of Epipalaeolithic culture into Scandinavia after the retreat of the ice that movements of some sort did occur. Furthermore, Déchelette nowhere alluded to Testut's attempt to dissociate Chancelade man from that of CroMagnon and to connect him with the Eskimo. Last of all, an article by Lartet should be mentioned6 in which he discussed resemblances between the Eskimo and the Lapps on the one hand and the cave men on the other in the matter of bone needles and their manufacture, the use of sinew thread, love of embroidery,7 and the use of tooth pendants. 8 The studies were carried out in detail and include excellent illustrations and quotations from travelers in Lapland and the American Arctic. These comparisons were intended only as analogies, not to prove a relationship. The writers discussed above made practically no use at all of skeletal material and such material as has been mentioned is not that on which later arguments were based. Yet these writers have attempted to prove an identity of race between the Eskimo and Palaeolithic man because of a claimed identity of culture. The validity of such a proof is not evident. There has also been the assumption that Upper Palaeolithic man must have been related to some living race, and that to point out the race he most closely resembles is to prove an identity. The alternative, that Palaeolithic man may have become extinct without leaving any direct descendants 1 Mathiassen's discovery of realistic art in the ancient Thule phase of Eskimo culture, of course disproves this statement. 1 J . Déchelette, Manuel d'archéologie préhistorique, I, Paris, 1908, pp. 309-312. 3 Déchelette, op. cit., I, pp. 310-311. * Ibid.., p. 312. ' Sollas, op. cit., 1924, pp. 593-594. 6 Eduard Lartet, Henry Christy, Reliq uiae Aqmtanicae, Paris, 1865—75, pp. 130 ff. 7 Lartet's evidence from the Palaeolithic on this point are two darts from La Madeleine, on which are incised " h a n d s " with four "fingers," the " a r m s " decorated by a zig-zag line in one case, and in the other by a row of chevrons, supposed to represent the embroidery on the sleeves (Plate B I X - 1 , and 8 Pp. 42—43. BXVII-6).

488

R R E D E R I C A DE LAGUNA

does not seem to have been entertained. We have already remarked upon this unconscious assumption underlying Boyd Dawkins' first statement of his thesis and we are to meet it again in the writings of Testut and Sollas. There is a similar criticism to be made of the attempt to prove an identity of culture. If we examine the actual material types, Eskimo and Palaeolithic, said to show an absolute identity in every detail, we shall see with what little critical judgment these claims have been made. Dupont, Dawkins, Hainy, and more recent investigators alternate between the notion that Eskimo culture shows analogies useful in explaining Palaeolithic remains, and the notion that these analogies constitute real similarities, proving a common cultural tradition. The confusion of these two points of view constitutes an essential weakness in the treatment of the problem. I t is to be explained by the spirit in which archaeologists have approached their task of interpreting Palaeolithic remains which have survived to us, sadly altered by the vicissitudes of time and isolated from the body of culture which alone could render them intelligible. The comparative method, as proposed by Lubbock, offered the best solution. I t brought to light many analogies between the culture of the Eskimo and those of Upper Palaeolithic Europe, which archaeologists have exploited to the full, exaggerating both the number and closeness of the similarities. Not having to deal with differences, they tended to ignore their existence. Moreover, they have failed to perceive that an analogy sufficiently striking to suggest a plausible interpretation may not be close enough to prove connection. Furthermore, it has become a habit, as we can see from the writings of Breuil, 1 Rivet, 2 SaintPerier, 3 etc., to turn first to the Eskimo when seeking an explanation of a Palaeolithic object. This furnishes plenty of analogies, but at the same time makes for a greater possibility of error than if the problematic objects had been compared impartially to a wide range of corresponding forms. On the other hand, the ethnologists have not had the same interest in establishing affinities between Palaeolithic and modern cultures and have usually dismissed the possibility as fantastic. Testut's monograph cast an entirely new light on the old problem, for in the remains of Chancelade, Testut claimed to have found a new race, differing from that of Cro-Magnon, 4 and exhibiting the anatomical characteristics of the Eskimo. The Chancelade skeleton had been buried in a flexed position, and true to the old tradi1 Breuil, "Notes de voyage paléolithique en Europe centrale," part I I , L'Anthropologie, XXXIV, Paris, 1924, p. 532, compares fig. 532 from Predmost with the "bag-handles" of the Eskimo. In Breuil and Saint-Perier, Les Poissons, etc. dans l'art quaternaire, Mémoire 2, Archives de l'Institut de Paléontologie humaine, Paris, 1927, p. 81, footnote, he compares the little bifid points of the upper Magdalenian with the bird arrows of the Eskimo and other peoples. 2 Rivet, "Interpretation ethnographique de deux objets préhistoriques," Congrès International des Americanist.es, XI Session, Part I I , Goteborg, 1925, pp. 203-206, compares the feline figure from Isturitz with the Eskimo ajagaq. 3 Saint-Perier, " L a Grotte de Gouëris à Lespugue," L'Anthropologie, X X X V I I , Paris, 1927, compares a segment of a bone ring from Lespugue, fig. 17-1, with the head of an Eskimo comb, and perforated ivory plates from Mas d'Azil, fig. 18-2 and -3, with Eskimo belt buckles. In "Engins de Pêche paléolithiques," L'Anthropologie, X X X V I I I , Paris, 1928, he compares, pp. 20-21, Palaeolithic specimens with Eskimo fish needles, and Eskimo salmon decoys. 4 Cro-Magnon man, according to Testut, is a cross between Chancelade man and the invading Neolithic race: L. Testut, Recherches anthropologiques sur le skelette quaternaire de Chancelade (Dordogne), Lyon, 1889, p. 115; however, cf. M . Boule, Les Hommes fossiles? Paris, 1923, pp. 285-293.

ESKIMO AND PALAEOLITHIC ART

483

tion, T e s t u t compared this with Eskimo burials, though he found his analogy, not as did Dawkins and the others, in the lack of proper respect for t h e dead, but, in the m a n n e r of burial. 1 Testut m a d e a complete study of the skeletal remains, at every point comparing t h e measurements and striking characteristics with those of other races, both living a n d extinct. This otherwise careful work was rendered somewhat less valuable because the skull h a d to be reconstructed from fragments, and was incomplete in several important parts. Furthermore, T e s t u t did not tell from what sources he obtained the figures about other races which he used in his comparison. This is an u n f o r t u n a t e omission, for we should like to know whether t h e methods by which they were obtained and compiled are comparable to those employed by T e s t u t on t h e Chancelade remains. T e s t u t states in conclusion: "Parmi les races actuelles, celle qui me parait présenter la plus grande analogie avec l'homme de Chancelade est celle des Esquimaux." They have the same skull, t h e same type of face, t h e same height, t h e same nasal a n d orbital indices, the same degree of torsion of the humerus. H e mentions the " f o l lowing-the-reindeer " theory, a n d adds: "La découverte de Chancelade, en mettant en lumière une analogie frappante entre le squelette de notre troglodyte périgourdin et celui des Esquimaux actuels, apporte à cette opinion, aussi séduisante que naturelle, l'appui de l'anthropologie anatomique, qui, dans l'espèce, a une importance capitale."2 Testut's work, in particular his estimation of t h e cranial capacity of the Chancelade skull, has received considerable criticism. Sollas, 3 however, after comparing it with t h e Greenland skulls published by F u r s t and Hansen 4 heartily indorses Testut's conclusions. Sollas sums up 28 points of resemblance between Chancelade m a n a n d the Eskimo, a n d finds only 5 doubtful or differing characteristics. " I t has been objected—and no doubt with justice—that there is a great danger in forming a j u d g m e n t on the evidence of a single skull. ' U n u s est nullus'—it is said. B u t t h e r e is no universal rule, and in t h e present instance we are dealing with a skull of very distinctive characters, and with t h e Eskimo, a race of equally distinctive characters. 6 . . . I think it will be generally admitted, on reviewing the evidence t h a t t h e assemblage of characters presented on t h e one hand by the Chancelade skull, and on t h e other hand by t h e Eskimo, are in very remarkable agreement, and t h a t t h e onus of discovering a similar assemblage, b u t possessed by some other race, rests with those who refuse to accept what seems to be a very obvious conclusion." H e does not here admit of a third possibility, namely, t h a t the similarities observed, however striking, are not sufficient to prove a relationship and t h a t there m a y be no race living t o d a y which is closely related to t h a t of Chancelade. H e a t t e m p t s to answer objections by pointing out t h a t " t h e bones of Chancelade, however, resemble those of the existing Eskimo, as closely as those of the French musk ox resemble those of t h e Greenland musk ox; and are not separated in space by so wide an interval." 6 Sollas here argues for t h e "foliowing-the-reindeer" t h e o r y . ' 1

3

Op. cit., p p . 3 - 4 .

2 T e s t u t , p . 117.

" T h e Chancelade Skull," Journal of The Royal Anthropological Institute of Great Britain and Ireland, LVII, London, 1927. > Crania Groenlandica, Copenhagen, 1915. 5

Sollas, 1927, p p . 1 0 0 - 1 0 1 .

« Sollas, 1927, p . 119.

' Op. cit.. p . 121.

484

F R E D E R I C A DE LAGUNA

Henri Martin has described the human remains which he discovered in the Vallée du Roc, Charente, France. After comparing the du Roc skulls with those of various races, living and fossil, he concludes that the du Roc group belongs to the same race as Chancelade man, for whom Testut established a striking comparison with the Eskimo. In any case, he felt that one could not deny the Mongolian affinities of this race. 1 However, even with the du Roc skulls, is there sufficient material to establish the racial type of Chancelade? And are the similarities between this type and the Eskimo sufficient to prove a relationship notwithstanding the great separation in space and time? It has often been asserted by Boyd Dawkins and his successors that the distribution of animal species common to the European Pleistocene and the present fauna of Arctic North America offers a strong argument by analogy in support of the " folio wing-the-reindeer " theory. We must call attention to the fact that at best the evidence show's only that man might have followed these animals into the New World, not that he did so. The picturesque consideration that the same animals furnished food and tool materials to both prehistoric man and the modern Eskimo adds nothing. Dawkins and Hamv seem to imply that the musk ox and caribou of the New World are the actual offspring of the same herds which roamed Europe during the Ice Age. Moreover, for Chancelade man and the Eskimo a closer relationship has been claimed than that proved for these animals, i.e., a racial affinity, over and above the participation in a common species. As a matter of fact, not all the American Arctic and European Pleistocene animals mentioned in this connection have as close a relationship as that of the same species.2 The "following-the-reindeer" theory definitely specifies a post-Pleistocene immigration of the original Eskimo into the American continent. However, most of the animals of the American Arctic crossed to the New World before the close of the Pleistocene, in many cases much earlier, so that, if man accompanied them, he must have done so at a much earlier date than that assumed by the theory. 3 In Ancient Hunters and their Modern Representatives, Sollas makes the most complete case in favor of Boyd Dawkins' thesis, though the skeletal evidence is given in more detail in his more recent article on the Chancelade skull. On the ethnological side, Sollas enumerates 11 main points of resemblance: the 1 H. Martin, "Crâne trouvé en Charente (Vallée du Roc)," L'Anthropologie, X X I V , Paris, 1924, and "Caractères des squelettes humaines quaternaires de la Vallée du Roc (Charente)," Bulletins et mémoires de la Société d'Anthropologie de Paris, seventh series, VIII, Paris, 1927, p. 129. Boule, in Les Hommes Fossiles, p. 299, supports Testut's identification of Chancelade man with the Eskimo. 2 The cave bear, ursus spelaeus, which became extinct during the Pleistocene, is related to the American grizzly, and to the brown bear of northern Europe and Asia, though none of these are classed within the same species. (Max Weber, Die Sàugetiere, Jena, 1904, p. 535.) Though there is only one species of reindeer, the European Pleistocene form belongs to the same subgroup as that of Lapland, while the various other sub-groups of American caribou are distinct. (H. F. Osborn, Men of The Old Stone Age,'- New York, 1916, p. 209; Andrew Murray, The Geographical Distribution of Mammals, London, 1806, pp. 1 5 1 - 1 5 9 ) . The musk-ox of the European Pleistocene and the American Arctic are classed within the same species. (Weber, op. cit., pp. 680, 725.) 3 The evidence for Pleistocene man in North America is not yet definite enough to shed any light on the problem of Eskimo origins.

E S K I M O AND PALAEOLITHIC A R T

485

shaft-straightener, bone arrowhead with ownership marks, hairpin, ivory pendant, snow scraper, bone needle (though he lays no stress on this because of its wide distribution), hollow bone needlecase, wound pin and plug,1 ivory smoother, barbed bone spearheads, spear thrower (though he mentions that this is widely distributed and the shape of the Magdalenian form is quite different from that of the Eskimo), and art. 2 Minor resemblances to which Sollas alludes in passing, but does not mention in his conclusion are: the gorge,3 stone implements from Greenland which are upper Palaeolithic in character without belonging to any definite type, 4 the stone lamp, 5 and the bow drill.6 He admits that some of the most important and characteristic implements of the Eskimo are lacking in Magdalenian culture: the sledge,7 the kayak, and the fully developed (toggle-head) harpoon. However these are specialized types that we should hardly expect to find in the Palaeolithic. 8 To this "prima facie case in favor of an alliance by culture" Sollas would add racial identity, indicated by Testut's work. The Chancelade skeleton is that of " a veritable Eskimo, who lived in southern Gaul during the Magdalenian age." 9 The Magdalenian peoples emigrated towards the north owing to pressure exerted by Neolithic peoples and to the withdrawal of the sub-arctic fauna upon which the Magdalenian hunters depended. Proof of a shifting of population is afforded by the discovery of several Magdalenian stations in regions formerly glaciated. 10 Sollas here states that a connection between Magdalenian Man and the Eskimo does not necessarily involve the " foliowing-the-reindeer" theory, and though attractive it is not supported by facts. 11 The hypothesis he here advances is as follows: "During the Magdalenian age at least two races of dolichocephalic Leiotrichi, differing greatly in stature, extended from Western Europe to the east, across the entire breadth of Asia, occupying a zone which included much of the tundra and the steppes. They possessed a common Magdalenian culture, and resembled in their mode of life the Algonkians and Athapascans of the tundra as they existed before the advent of the white man, feeding on reindeer and the mammouth, horse, and bison, together with various kinds of fish." The taller race lived to the south, while the shorter race was forced to the north and so came to develop a special kind of culture. "As the climate became warmer, the pressure of the rapidly increasing Neolithic 1 Sollas, 1924, p. 583, figs. 293 and 294. He compares the ivory peg from Brassempouy with a modern central Eskimo wound plug, fig. 295, a not very fortunate comparison, since the wound plug is recent (T. Mathiassen, "Archaeology of the Central Eskimos," Report of the Fifth Thule Expedition 1921-2b, IV, Copenhagen, 1927, I I , p. 41), and the Brassempouy specimen appears to have had an ornamental significance, rather than a utilitarian function. 2 Op. cit., pp. 304, 529, 531, 583-585; figs. 284-285, 293-294, 301, 305, 327-329. 3 Op. cit., p. 586, no 4 Op. cit., p. 582, no figures. figures. 5 Op. cit., p. 541, fig. 303, the lamp of La Mouthe and an Eskimo lamp from Kodiak I., southwestern Alaska. sOp. cit., p. 365, fig. 302-g, Eskimo, and p. 540, fig. 302-d and -j, Magdalenian. (-d is called a spear thrower by Breuil.) I t is not generally accepted, moreover, that the bow-drill was invented before the Neolithic. 7 He thinks the Magdalenians may have had the sledge, basing this opinion on the interpretation of a design on a slate pendant, fig. 318-2, but does not stress the point. 6 Op. cit., pp. 582-583. » Op. cit., pp. 586-591. « Op. cit, pp. 593-594. 11 Op. cit., p. 563, footnote 3. I t will be remembered that he adopted this hypothesis in his paper of 1927.

480

F R E D E R I C A DE LAGUNA

peoples began to make itself felt, acting probably from a region between the Caspian and India. A movement of the Leiotrichi was thus set up towards the north; but as there was no room for expansion in that direction, it was diverted towards the only possible egress, and an outflow took place into America over Bering Strait or the Aleutian Islands. The primitive Eskimo, already accustomed to a boreal life, extended along the coast." The other peoples came in behind and gradually spread over the whole of the New World. 1 I t is difficult to criticise this theory because it does not seem to follow directly from the evidence and arguments which Solías has previously presented in his work. Though Upper Palaeolithic sites have been discovered in Russia and Siberia, the culture of these eastern stations apparently differs considerably from the European Upper Palaeolithic cultures. 2 Unfortunately, in this country we do not yet have access to the results of recent excavations in Siberia, though a comparison of Siberian Upper Palaeolithic types with those of the Eskimo is of the highest importance for our problem. According to Solías, comparison should be made between the Upper Palaeolithic of Siberia and the Algonkian and Athabaskan Indian, not between the Palaeolithic of Western Europe and the Eskimo. This theory approaches those of Hatt, Hallowell, and Birket-Smith, except that Solías believes that the differentiation of the Eskimo culture out of the common preEskimo-Indian background took place in the Old World rather than in the New. Solías' hypothesis is of value because he shows that what we have here is only the fragment of a much larger problem, involving the culture history of Siberia and North America and the peopling of the New World. Dr. Paul Rivet has also considered this problem, but he has published little more than a statement of a hypothesis to explain similarities between the Upper Palaeolithic cultures and those of Arctic Europe, Asia, and the Eskimo, and a short paper comparing some Palaeolithic objects with the Eskimo ajagaq. Among the cultural types he considers common to the Eskimo and the Palaeolithic are: the thro wingstick, shaft-straightener, barbed "harpoon," snow knife,3 ajagaq (or cup-and-ball game), and art. 1 The hypothesis which he advances to explain these similarities is as follows: "On peut supposer que les peuples ouralo-altaïques ont eu pour centre de dispersion quelque région située à l'est de la Caspienne et que de là est parti un premier essaim vers le nord de l'Europe occidentale dès le quaternaire, un second essaim vers le nord, representé par les Finnois, les Lapons, et les Samoyedes, et enfin un troisième essaim vers l'est, par les régions circumpolaires, dont les Eskimos sont les représentants les plus Solías, 1924, p. 596. I t has hitherto been believed (MacCurdy, Human Origins, New York, 1924, I, p. 439, and p. 206, and Von Merhart, " T h e Paleolithic Period in Siberia, etc." trans, by G. G. MacCurdy, American Anthropologist, N. S. X X V , 1923, p. 21-25) that the Siberian Upper Palaeolithic was lacking in art. In the last few years, however, remarkable discoveries of art objects have been made, and I only regret that I am not able to make use of that material in this paper. 3 The presence of the snow knife in the Palaeolithic seems to me very doubtful. Some of the large spatulas of Le Placard, for example, might have been suitable for such a purpose, but there is nothing to indicate that they were intended for that rather than for scraping skins, etc. 4 Rivet, "Interpretation Ethnographique, etc." p. 265, and " L e s Origines de l'homme Américain," L'Anthropologie, X X X V , Paris, 1925, p. 297. 1

1

E S K I M O AND PALAEOLITHIC A R T

487

orientaux." This last migration could not have taken place before the end of the Palaeolithic. 1 This hypothesis has the virtue of bringing all the circumpolar peoples into the problem, but so definite a theory seems premature since we know so little about the archaeology of Siberia. A REFOBMULATION OF THE P R O B L E M

Up to this point the writers considered have not used Eskimo archaeological material. This is because until very recently there has been no available body of material to consult, with only at best isolated collections of Eskimo antiquities of uncertain provenience, inadequate for any systematic conception of the development of Eskimo culture. On the other hand, there have been several attempts to discover the origin of Eskimo culture by rationalistic deduction, either through an analysis of the culture itself, as in the work of Steensby, or through a study of the distribution of Eskimo traits outside the Eskimo area as in the writings of Hatt and Thalbitzer. The influence of these authors has left its mark upon their successors, Mathiassen and Birket-Smith. The attitude of the average ethnologist to Boyd Dawkins' thesis is illustrated by Steensby's criticism. 2 What similarities exist between the two cultures he ascribes to common tool-materials and similar environment, pointing out that the toggle harpoon heads, typical of the Eskimo, have never been found in a Palaeolithic deposit.3 Sollas had also mentioned this fact, but for him it did not present a fatal obstacle to his theory. This is because the two have something quite different in mind when writing of "Eskimo culture." While Sollas would not deny the recent origin of many Eskimo traits, he, like other archaeologists, has rather tended to ignore these latter phases. Steensby, however, is interested in the finer distinctions which set Eskimo culture apart as uniquely " E s k i m o " and which differentiate it from every other, that of the adjacent Indians in particular. European archaeologists have not been interested in these distinctions but have sought to emphasize those points of resemblance by which the Eskimo and the Upper Palaeolithic might be included in a common scheme. They have erred in not distinguishing between what the Eskimo share with other peoples and what appears to be exclusively their own. Yet this distinction is of the highest importance, since it is only for those widely distributed traits that we can expect to find Palaeolithic origin. The purely Eskimo traits, on the other hand, must be considered of local origin. I t is only in respect to the "non-Eskimo" traits that the Eskimo can be Palaeolithic. I t is hardly necessary to say much about Steensby's theory since it has been recently summarized by both Mathiassen and Birket-Smith, whose own theories have superseded it, 4 Steensby believed that Eskimo culture developed out of an Indian background. The pre-Eskimo, speaking the mother tongue of the present Eskimo language, lived in an area between the Barren Grounds and the prairies. Even Op. cit., 1925, p. 2.96. H. P . Steensby, " A n Anthropogeographical Study of the Origin of Eskimo Culture," Meddelelser orn Grönland, L I I I , Copenhagen, 1916. » Op. cit., p. 50. 4 Mathiassen, 1927, I I , pp. 196-198, and K. Birket-Smith, " T h e Caribou Eskimos," Report of the Fifth Thüle Expedition, 1921-2b, 1929, Copenhage n, I I , pp. 221-222. 1

2

F R E D E R I C A DE LAGUNA

488

then there must have been Asiatic influences, though we can not say to what extent this proto-Eskimo culture was American or Asiatic. 1 The most " t y p i c a l " Eskimo culture is the Arctic form, which is, therefore, the oldest, the sub-Arctic forms being derivations. 2 The region most "suitable" for the development of this Arctic Eskimo culture from the original Indian mode of life is the Arctic Archipelago of Canada, and in particular the region about Coronation Bay, so that it is there that we must look for the home of Eskimo culture.3 This Arctic or "Palae-Eskimo" phase of culture was really an inland culture, the sub-Arctic or coastal aspects did not develop until the culture spread into other regions and felt the influence of other cultures, particularly the Asiatic cultures at Bering Strait, where a later "NeoEskimo " phase developed, characterized by borrowings from Asia. 4 It is this culture sequence which we meet again in Birket-Smith's thesis. Steensby's use of the criteria of "typicalness" and "suitableness" is very dangerous, since they depend so much upon individual judgment, and the possibility of error, where there is no archaeological material as a check, is very great. Steensby has attempted to discover the origin of various elements of Eskimo culture. Unfortunately except for the throwing-board, which he believes to be a local Eskimo invention,5 and the lamp which he thinks was not though he fails to state its place of origin,6 he does not discuss those objects capable of being preserved, and of which we could consequently hope to find traces in the Palaeolithic. A long list of elements common to the Eskimo and the tribes of Arctic and subArctic Siberia and Europe has been published by Thalbitzer. 7 This study is in the nature of a bibliography rather than a finished work; the only conclusion drawn is that the resemblances may point to an original geographic community or even a common origin of circumpolar cultures. 8 The Danish scholar Hatt in his monograph on moccasins 9 has distinguished two types of clothing, each associated with a corresponding type of culture, the younger of which originated in the area of the Tungusian tribes, or a little to the south, and now enjoys the wider distribution. With it was associated the snow-shoe. "If this theory be correct, we shall then have to reckon with two large culture waves, which in prehistoric times swept over the northern regions. The oldest of these, now most fully developed in the culture of the Eskimo tribes, did not have snow-shoes and therefore could not conquer the vast inland areas; it must have followed and taken possession of the rivers and coasts, and we would call this first great wave the coast culture. The younger culture wave is found fullest and unmixed in the culture of the Tungusians, although its influence is felt from Lapland to Labrador: it still has the character of an inland culture and must have originated as such. Its most valuable possession is the snow-shoe which has carried it over the greater part of the arctic." 10 1

4

7

S t e e n s b y , p. 20a. Op. ciL, p . 166.

a Op. cit., p . 168. « Op. cit., p . 202.

W. Thalbitzer, "Parallels within the Culture of the Arctic Peoples," Annaes do XX Congresso

Internacional de Americanistas, 9

' Op. cit., p . 170. s Op. cit., p . 163.

R i o de J a n i e r o , 1924.

s Op. cit., p. 283.

G. Hatt, "Moccasins and their Relation to Arctic Footwear," Memoirs of the American Anthropo-

logical Association,

I I I , 1916.

T h i s position was previously developed in Arktiske Skindragter i Eura-

sien og America, Copenhagen, 1914. See also the critical analyses of his work in Mathiassen, 1927, II, p. 198, and Birket-Smith, 1929, II, pp. 212-214. " Hatt, 1916, pp. 247-249.

E S K I M O AND PALAEOLITHIC A R T

489

In another article 1 he explains the origin and evolution of Eskimo culture. As Mathiassen has pointed out,2 Hatt, by placing the coast culture as the earliest phase, inverts the order Steensby tried to establish. Hatt believes that this original coast culture was related to the Palae-Asiatic cultures of northeastern Asia, and possessed the umiak, fish-net, gut-skin shirt, urine tanning, square house, etc., and also a part of the elements found among all Eskimo tribes. The wave of inland culture brought the kayak, and because of this valuable invention was able to sweep over the whole Eskimo area, which accounts for the uniformity of the Eskimo language. Mathiassen considers this theory in the main correct, since it is supported by the discovery of the markedly coastal Thule culture as the oldest culture in the Central regions. However, the kayak was not introduced by the modern, inland culture, but belonged to the Thule phase.3 Systematically excavated archaeological material which has long been needed is provided for us in Mathiassen's 4 classic monograph on the archaeology of the Central Eskimo. No future study of the problem can afford to neglect this contribution. The Thule culture, i.e. the original " c o a s t a l " culture in the Central regions, is the oldest in northeastern Canada and in Greenland of which we have any archaeological record. Mathiassen believes that it originated in Alaska or in northeastern Siberia, because the types or prototypes of certain characteristic implements of the Thule culture are found in the west, and because anthropogeographical considerations point to Alaska as a suitable place of origin since it possesses ice for the development of ice-hunting, wood for the building of umiaks, and large sea mammals which were incentives to whaling.5 Unfortunately the archaeology of Alaska has not been sufficiently explored, especially on the north coast, for us to be certain about the place of origin of the Thule culture. Hence it is impossible to say what relationship the Thule culture of the east bears to the archaeological Alaskan cultures, the so-called Old Bering Sea and the Punuk, and how these are all related to the original Eskimo culture. I t is hoped that the present investigations conducted at Point Barrow by the U. S. National Museum will solve some of these problems. In the meantime it is certain that some of the Thule types now found in Alaska are recent there, at least in the region about Bering Strait,—and were introduced from the east by a backwash of Thule culture into Alaska in very recent times.6 However, it is also fairly certain that at an early date there was a Thule or a related phase at Point Barrow, represented by the Van Valin 7 and Hobson collections in the University Museum, Philadelphia, but only further excavation can determine the exact position of this culture. Birket-Smith carries on the traditions of Steensby and Hatt, although his studies " K y s t - og Inlandskultur i det arktiske," Geografisk Tidsskrift, X X I I I , Copenhagen, 1916. s Ibid,., p. 200. Mathiassen, 1927, II, p. 198. T. Mathiassen, 1927. » Mathiassen, 1927, II, p. 182. 6 H. B. Collins, Prehistoric Art of the Alaskan Eskimo, Smithsonian Miscellaneous Collections, v. 81, no. 14, Washington, D. C „ 1929, p. 44. 7 A preliminary report has been published on the Van Valin collection by J . Alden Mason, " E x cavations of Eskimo Thule Culture Sites at Point Barrow, Alaska," Proceedings of the XXIII International Congress of Americanists, New York, 1930. Mathiassen, "Archaeological Collections from the Western Eskimos," Report of the Fifth Thule Expedition 1021-24, X , No. 1, Copenhagen, 1930, made use of the material. 1

2

4

490

FREDERICA DE LAGÜNA

have resulted in a reformulation of the problem. He follows Steensby's general scheme of the history of Eskimo culture. The Palae-Eskimo, or original inland culture, is represented today by the Caribou Eskimo, the Neo-Eskimo phase by the Thüle culture. Birket-Smith accepts Hatt's theory of two great culture waves in northern Eurasia and North America, but finds Hatt's terminology misleading and would substitute "ice-hunting" and "snow-shoe" for the terms " c o a s t " and " inl a n d " cultures. 1 The Caribou Eskimo should be considered the most primitive and conservative survivors of the ice-hunting culture. The second volume of Birket-Smith's work is largely devoted to the elements of this culture, and to the position of the Caribou Eskimo within it. Perhaps the fullest picture of the "snowshoe" phase is to be found in Hallo well's study of the bear cult in northern Eurasia and North America. B y the expression "ice-hunting culture" Birket-Smith does not want to imply that it was absolutely homogeneous throughout its entire area. The term " i s merely an abstraction, an expression meaning that a certain common basis must be assumed to lie under the building up of the culture everywhere in the circumpolar region." 2 Unfortunately Mathiassen and Birket-Smith have left us two very different pictures, and as yet no way has been found to reconcile their opinions. 3 However, they agree on a point of the greatest importance for our problem, viz., that the original Eskimo culture contained many Asiatic elements. Birket-Smith has suggested a way in which the problem of a Palaeolithic origin of Eskimo culture can be restated. The comparison should be made between the Palaeolithic and the "ice-hunting" culture as a whole, not with the Eskimo alone. He also believes in the possibility of such a connection, suggesting that the connecting links are to be sought in the Scandinavian Neolithic. The stone ages of Scandinavia, and even the Lapp Iron Age carry on traditions established in the Palaeolithic. 4 The modern culture of the Lapps, as Thalbitzer, Birket-Smith, and others have shown, contains a substratum which makes it one among the related circumpolar cultures. I n this way the enormous gaps of space and time may be filled to a certain extent, and a connection between the Upper Palaeolithic and the Eskimo can no longer be argued on similarities between those two cultures alone. A related chain of similarities must be exhibited, with a continuity of tradition. B u t at present such an ideal argument is impossible. However, it is not at all certain that we shall be able to establish a connection in the way Eirkct-Smith has outlined. I t is not easy to trace Palaeolithic traditions in the Scandinavian Neolithic.' At first glance the series: Magdalenian—Maglemosian (or Mullerup)—Arctic Stone Age—Lapp Iron Age, appears as a chain of Birket-Smith, 1929, I I , p. 214. 2 Ibid.. p. 216. See the discussion between Mathiassen and Birket-Smith in the American Anthropologist on the 4 Birket-Smith, 1929, I I , pp. 216 ff. origin of Eskimo culture, n.s. X X X I I , 1930. 5 A discussion of this problem had to be omitted from this paper on account of its length, but cf. C. A. Nordman, " S o m e Baltic Problems," Journal of the Royal Anthropological Institute cf Great Britain and Ireland, I , I I , London, 1922; H. Shetelig, Préhistoire delà A'oreè^InstitutetforSammenlignende Kulturforskning. Oslo, 1926; C. D. Forde, " E a r l y Cultures of Atlantic Europe," American Anthropologist, n.s. X X X I I I , 1930; H. Breuil, " L e s Subdivisions du paléolithique supérieur et leur signification," Congrès International d'Anthropologie et d'Archéologie préhistoriques, XIV Session, Geneva, 1913, p. 236. 1

3

ESKIMO AND PALAEOLITHIC ART

491

continuous development, but if we attempt to trace this development step by step we note the difficulties of the problem. Even though the chain of Palaeolithic influences were unbroken, we should still have to satisfy ourselves that they were embodied in those cultural types shown to have a wide distribution from Lapland to the American Arctic, if we are to prove that Eskimo culture can be traced back to the Palaeolithic over this route. Unfortunately the resemblances to Eskimo culture do not increase as we advance along this suggested sequence; in some respects the Magdalenian seems closest to the Eskimo, in others the Maglemosian, and in still others the Lapp Iron Age. Yet this last culture is too late to represent a direct link in the chain; at best it can be a stage parallel to that which saw the invention of the toggle-head harpoon in the original Eskimo or proto-Eskimo culture, with an art style, which though apparently local, suggests some similarities with that of the Punuk culture in Alaska. However, if, as I believe, it is only in Siberia, if anywhere, that we can hope to find a Palaeolithic-Neolithic culture sequence from which Eskimo culture is derived, the Scandinavian stone age can not tell us much of that development. It would reflect only faint traces of those culture waves which might reach the western limit of the circumpolar area. As a matter of fact, the Neolithic culture which bears the closest resemblance to the Eskimo is that of Japan, 1 and since no Palaeolithic has been found there, it is difficult to see what this may mean for our problem. At best the theory of the inheritance by the Eskimo of Palaeolithic culture via the Scandinavian Neolithic can be proved only by a study of particular cultural types, for most of which Birket-Smith has traced a wide distribution throughout northern Eurasia and elsewhere, and to these, others might possibly be added. The list would include: the stone lamp, sewing-needle, bone needle-case, barbed bone points with tang (not with socket), both fixed and detachable, chipped flint endscraper, gorge, fish-lure, fish-needle, shaft-straightener, bone spoon, marrow extractor, bone awl, throwing-stick, pendant, bead, button with perforated shank, etc. To use these elements in any proof it would first be necessary to show that the identification of the Palaeolithic specimens involved is correct. To these, the Maglemosian or an analogous culture stage in the east may have contributed: the bone comb, barbed fish hook, etc. However, these are all extremely simple types of wide distribution, so that to be sure of dealing with the same element and not with similar objects of independent origins, it would be necessary to demonstrate the constancy of some distinguishing stylistic trait in each case, or perhaps the associated occurrence of a complex of elements. It seems doubtful if any connection between the Palaeolithic and the Eskimo can be limited to even a circumpolar culture area. Again and again Birket-Smith finds a trait common to the "ice-hunting" cultures in South America, too, or even in Oceania; and this, I suspect, will also hold for any of the Palaeolithic elements we may include in the comparison. Some traits which Birket-Smith has discussed appear to be so simple and satisfy such inevitable human wants that we wonder if they are significant in a distribution study, especially when its scope precludes the taking into account of stylistic features. This difficulty would also apply to many 'G. N. Munro, Prehistoric Japan, Yokohama, 1911.

492

F R E D E R I C A DE LAGTINA

of the elements common to Eskimo and Upper Palaeolithic cultures. These considerations make us doubt whether a connection between the Eskimo and the Upper Palaeolithic would depend on anything stronger than those bonds of common human culture which unite all peoples, due either to an inheritance from remote antiquity, or to such simple parallelisms as may arise anywhere under similar conditions. T h e resemblances would then be explained by the fact that the Eskimo remained in a simple stage of culture where these elements are comparatively conspicuous, while a similar natural environment has favored accidental and superficial similarities. Or, it may be that the Eskimo have actually retained an important share of the Palaeolithic substructure which sets them apart among all other peoples. In any case, we have passed far beyond the original thesis of Boyd Dawkins. REPRESENTATIVE

ART

In support of a Palaeolithic origin for Eskimo culture a similarity in art, particularly in the case of animal representations, has been cited as one of the strongest arguments. If this similarity is due to nothing more than an attempt on the part of both peoples to portray men and animals realistically we may dismiss the argument as of little worth. I t will be necessary to show that the style of representation is the same, and that there is some archaeological or ethnological evidence to suggest the possibility of a connection. T h e age of Eskimo realistic art is, of course, of primary importance. Though Hoffman's theory that all Eskimo realistic incised art was due to Russian influence has been disproved, the age of incised representations of men and animals in Alaska is still unknown, our archaeological examples being of uncertain age. Mathiassen ascribes realistic etchings 1 to the hypothetical first stage of the Thule culture in Alaska, but according to Collins this form of art originated in the Canadian Thule culture, being carried to Alaska by a late wave of migration, since such etchings have not been found in the sites of the Old Bering Sea or the Punuk cultures. 2 Representations of animals in the round, however, are known from all the archaeological Eskimo cultures,—the Old Bering Sea, Punuk, and Canadian Thule cultures,— and from archaeological sites in northern Alaska. T h e human figure in the round, on the other hand, is not known from the Old Bering Sea culture, though its absence, in view of the scantiness of our collections can not be taken as very significant. Though the realistic art of the two peoples is really very different, as I think I shall be able to prove, an argument might be based upon purely geometric decorations, motifs composed of straight lines for the most part. T h e geometric art of the Upper Palaeolithic has received very little attention because it is less conspicuous than the representations of animals, and also because too often the patterns have been explained as degenerate animal or human figures. In comparing the non-representative art of the Eskimo and Upper Palaeolithic cultures, we must proceed with caution, for the elements are so simple that they might possibly be bound up with the technique of working bone with stone tools; that is, they might arise spontaneously from any attempt to decorate bone objects with straight-line patterns. T o prove a genetic connection, will it be enough to show that the same decorative elements 1

Mathiassen, 1930, p. 82.

2 Collins, 1929, p. 44.

ESKIMO AND PALAEOLITHIC ART

493

occur in both arts, or must we also show them to be executed by the same technical means? Do they occur too often to be fortuitous? Are they used in the same way, for the same decorative effects, entering into the same types of composition and filling relatively the same place in the two art styles? Certainly the argument would gain weight if these designs, though simple, were really limited in distribution, say, to the circumpolar area. However, I think it is impossible to prove this; the design elements appear sporadically in every part of the world wherever there is a style of incised decoration, unsophisticated enough for simple straight-line patterns to find a place. But to investigate the distribution of these simple patterns would be a task far beyond the limits of the present study. No argument, therefore, which relied upon these simple elements alone could attempt to prove more than the participation in a common aesthetic background, with the alternative theory of parallelism unanswerable. We shall content ourselves with an analysis of the two arts, trying to discover what evidence might be found in a comparison of the complete repertory of the simple design elements, and in their stylistic functions. In this study I have tried to use the archaeological Eskimo material as far as possible, and thus have perhaps appeared to slight modern Eskimo art. Unfortunately the archaeological specimens most exactly datable come from what we must consider a peripheral region when discussing connections with Asia; they belong to the Thule culture of Canada and Greenland. The oldest of these finds dates back perhaps a thousand years. Mathiassen has argued that the art of the Thule culture of Canada appears to be a degenerate form of an older art in the west; but from Arctic Alaska and northeastern Siberia, where presumably the Thule culture and its art originated, we lack properly excavated specimens, and must content ourselves with purely modern or with archaeological material about the age of which we can say very little. Possibly some of these, perhaps a considerable portion, are the result of that western spread of the Thule culture in recent times which I have already mentioned. In the region about Bering Strait, the oldest culture is the so-called Old Bering Sea culture, but its art is so highly evolved and so very individual in character that we shall omit it from our comparison. Collins believes that future excavation will prove the Old Bering Sea culture to be the parent of the Thule culture, 1 but in spite of its great antiquity in Alaska, in my opinion it is too specialized to be the original Eskimo culture which must be much more generalized, at least in art style. On St. Lawrence Island Collins has found a succession of sites showing the development of the Old Bering Sea culture into its most typical forms and its subsequent change into something intermediate between the Old Bering Sea and the modern Alaskan culture, a phase he has called the Punuk. The art of this period is characterized by rigid metal-cut lines and all-over patterns, of a style so local and individual that it is difficult to include it in our comparison. In the Punuk finds there also occur specimens with ornamentation like that of the Thule or the modern Alaskan art, which we are to interpret, perhaps, as forms introduced from a contemporary Thule or a proto-Thule culture in the north. The typical Punuk style with mechanically drawn dot-and-circle, metal-cut lines, round bored dots at the ends of short lines and spurs, with red or black paint rubbed into the incisions, is, I 1

Collins, 1929, p. 47.

494

FREDERICA DE LAGUNA

believe, t h e result of the introduction of design elements and techniques from Siberia, associated with t h e use of metal engraving tools, which produced a revolution in t h e art of t h e Old Bering Sea period. However, some of the designs are also characteristic of t h e Thule culture and, long before their introduction into P u n u k a r t , m a y have existed on the Arctic coast of Alaska where they formed p a r t of what was in m a n y ways a simpler style, lacking the elaborate, all-over, connected p a t terns of the Old Bering Sea and P u n u k art. T h e history of Eskimo art in Alaska and t h e chronological position of the P u n u k phase are naturally of vital importance t o our problem, b u t they can not be understood without more excavations, particularly on t h e north coast. Among t h e Caribou Eskimo, who represent, according to Birket-Smith, t h e original stage of Eskimo culture, carved or decorated bone work is virtually nonexistent. 1 However, t h e original Eskimo could not have been without a r t ; its lack among the Caribou Eskimo is an extreme case of degeneration among the modern Central Eskimo. THE HUMAN FIGURE

T h e h u m a n figures in t h e round of the Upper Palaeolithic might be compared t o t h e dolls and figurines of the Eskimo. However, t h e h u m a n figure is not carved in the round in t h e Magdalenian, b u t is confined to t h e Aurignacian and the Solutrean, with the exception of one example about which opinions differ, a n d this is in itself an argument against a direct connection with the Eskimo. Almost all t h e Eskimo carvings of t h e h u m a n figure appear to be children's toys, except for a few p u p p e t s used by shamans. T h e Eskimo dolls represent both males and females, a n d for most the sex is not indicated at all. T h e carvings of the Palaeolithic, on t h e other hand, appear to be works of art, not toys, and with t h e exception of t h e m a n of Briinn, Moravia, and a somewhat uncertain fragment from Brassempouy, 2 all represent women with exaggerated sexual characteristics. T h e most typical of t h e Aurignacian figures are the so-called " V e n u s e s " of Willendorf in Austria, and of the G r o t t e des Rideaux at Lespugue in France. (Plate X I X A-2 and 7.) * Four female statuettes in crystalline talc from the caves of Grimaldi, near Mentone, are in t h e same style, 3 as is also a modeled figure of a woman from Unterwisternitz, Moravia. 4 T h e female torsos from t h e Grotte du P a p e at Brassempouy have a much greater naturalism, which makes t h e m stand a little a p a r t from t h e rigid stylization of most of the Aurignacian statuettes. (Plate X I X A-l.) T h e Aurignacian t y p e is characterized by t h e bowed head without features, t h e long pendent breasts, t h e spindling arms, folded above the breasts, which hang so low t h a t the arms are usually a t the level of t h e waist, t h e prominent belly, t h e fleshy hips, and the feetless legs which taper below t h e knees. The figures are quite naked, except t h e Lespugue Venus, who wears a little apron or fringed skirt below t h e buttocks. T h e hair, or a headdress, is often shown. T h e full front of a woman carved in low relief from t h e Aurignacian rock shelter of Laussel, Dordogne, should 1

Birket-Smith, 1929, I, p. 253. E. Piette, L'Art pendant I'age du renne, Paris, 1907, pi. LXX-2. Boule, fig. 195-1, 2, 3, 5. (The last is published as a masculine figure.) 4 O. Menghin, Weltgeschichte der Steinzeit, Vienna, 1931, pi. XII-3. *For List of Illustrations, see pp. 509-511. !

3

ESKIMO AND PALAEOLITHIC ART

495

be compared to the Willendorf Venus seen from the same angle. (Plate X I X A-3.) T h e principal stylistic features are the same, though the head is bent to the side, not bowed to the front, probably because of the difficulty of representing the traditional pose. MacCurdy considers the work unfinished, the artist intending to show the profile of the face, 1 but I think the features were purposely omitted in conformity with the tradition. T h e steatopygous female figure from Mentone, with erect, pointed head, and narrow hips, is very different from the other Aurignacian figures. 2 A steatite statuette from Savignano sul Parano, Italy (Plate X I X A-4), also has the narrow hips, with a tendency towards steatopygy, and the shoulders terminate above in a pointed cone, suggestive of the pointed head of the Mentone figure. T h e other proportions of the body, especially the little arms folded over the breasts, place it in the ordinary Aurignacian tradition. T h e only Eskimo figure I can compare with those just described, is the ivory carving of a woman from Punuk, in the St. Lawrence Island group, Alaska. (Plate X I X A-6.) T h e resemblance is not close, however. As in the Aurignacian statuettes, the neck is bent forward, and the head, now missing, was probably bowed. T h e legs are slender and are together, though they do not appear to taper. T h e arms are too short, and are held close to the side and bent, as on the Aurignacian statuettes. This same attitude is also found on a shaman's doll (archaeological), from Point Hope, Alaska, probably representing a masculine figure. (Plate X I X A-5.) T h e belly of the Punuk woman is prominent, though not nearly as exaggerated as on the Aurignacian figures. T h e vulva is shown, which is not usual in Aurignacian statuettes, 3 though it is found on incised representations. I t is in the bodily proportions, however, that the Punuk figure shows the most striking differences from the Aurignacian. T h e breasts are flat, and though pendent, do not fall too low; in consequence the arms are bent below, not above them. T h e buttocks are too small, and the hips are ridiculously narrow. In the narrowness and in the roll of fat over the hips there is a similarity t o the Italian figure and the Laussel bas-relief, though the hips of the latter are very wide. T h e Magdalenian engraving of the " f e m m e au renne," of Laugerie-Basse (Plate X I X A-8), has a few points of similarity with the Punuk statue. T h e vulva is shown, though the artist had to place it on the left side in order to get it into the picture. T h e woman appears to lack breasts, but this may be because they are flat, like the old woman's of Punuk. T h e arm is too short and too thin in proportion to the body. T h e hand ends at the knuckles as if closed or cut off there, as on the Punuk woman and the Point Hope man. T h e hands of the Aurignacian figures are usually open, with the fingers sketched in by a few lines. Beginning with the wrist of the Laugerie-Basse woman, and extending up to the elbow, are rows of parallel transverse lines, which I interpret as tattooing. T h e Punuk figure is also tattooed with transverse lines on the upper arms. Such tattooing is characteristic of Eskimo women, though the pattern may be more complex. 1

3

MacCurdy, I, explanation to fig. 162.

2 Boule, fig. 195-6.

C'f., however, one of the Mentonefigures,H. KUhn, Kunst und Kultur der Vorzeit Europas, I,

Berlin, Leipzig, 1929, pi. 10-a.

496

F R E D E R I C A DE LAGTINA

Other Aurignacian statuettes seem to form a second group, because they are without prominent sex characteristics, or are clumsily made. The crude figures carved out of mammoth metacarpals from the Solutrean station of Predmost, Moravia,1 which suggest seated, shrouded figures, and the Aurignacian figure from Pont-à-Lesse, Belgium, carved from a tip of antler,2 are of this clumsy type. A little better is the " fillette " of Brassempouy (Plate X I X B-3), which might be compared to the Eskimo dolls from the Thüle culture (Plate X I X B-6, 8, 11), though the "fillette" has very prominent buttocks, and probably also had breasts, now broken off, while the Thüle culture dolls from the central regions not only lack sex characters, but arms, and feet, and usually features. Mathiassen considers these dolls male, simply because the sex is not indicated.3 In the Thüle culture in Greenland sex is clearly shown, as on a doll from Comer's Midden, North Greenland (Plate X I X B - l l ) , but this is a later development in Greenland. I suggest that the similarity between the Brassempouy girl and the Eskimo dolls is due only to a sketchy naturalism and a crudity of representation rather than to any similarity in style. The best of the Aurignacian figures of this second group is the " Venus impudica " of Laugerie-Basse (Plate X I X B-l), which Sollas has compared to a modern Alaskan needle case in human form,4 and two ivory figures from Brassempouy,5 of which unfortunately only the lower part of the body remains. These three female figures form a striking contrast to the fat, exaggerated women of the first group, and were perhaps intended to represent young girls. We have nothing quite like them in Eskimo art. The naturalism of Inugsuk, a late Thüle site in Greenland, offers parallels,6 but these are the product of a later and local Greenland development, as already explained. The figures from the Thüle culture of Canada are stiffer, like those of modern Alaska, though there are some similarities between the " Venus impudica" and a modern Alaskan female figure (Plate X I X B-4). Although the Aurignacian figures we have discussed lack features, the artists of the time were able to portray them if they chose. The girl's head from the Grotte du Pape, Brassempouy (Plate X I X B-5) and the head of the man from Brünn, Moravia, have features indicated by the cutting out of the ivory below the brows and about the nose. The mouth and eyes are not shown on the girl's head, while on the man's the mouth is indicated only by a faint line. The girl's hair, or headdress, is shown, somewhat conventionally, by a pattern of crossing vertical and horizontal lines, forming little squares. This is not the Eskimo method. The figure of a woman carved from a horse's tooth from Mas d'Azil, of uncertain age (Plate X I X B-2), is more realistic and has eyes, ears, nose, and mouth. The Eskimo dolls of the Thüle culture and of later periods in the east almost always lack the features. The face is usually cut off flat, and the head may be rounded in the back to suggest the appearance of a head in a fur hood. However, when the features are shown, as they G. H . Luquet, " L e s Origines de l'art figuré," IPEK, Leipzig, 1926, fig. 12. M . Hoernes, Urgeschichte der bildenden Kunst in Europa,3 in collaboration with Mengbin, Vienna, 1925, p. 119, fig. 3. s Mathiassen, 1 9 2 7 , 1 , p. 187. 4 Sollas, 1924, fig. 306, needle case after Nelson. ' Op. cit., fig. 239-b and c. 6 Mathiassen, "Inugsuk, a Mediaeval Eskimo Settlement in Upernivik District, West Greenland," Meidelelser om Grönland, L X X V I I , Copenhagen, 1930, pi. 18. 1

2

E S K I M O AND PALAEOLITHIC

ART

497

often are on Alaskan dolls, both archaeological and modern, the eyes and mouth always figure, even when reduced to a minimum of two dots and a line. The Aurignacian method of cutting out the material below the brows and around the nose appears in the very few representations of the human figure which we know from the Scandinavian stone ages. The human head on the bone comb from Gotland, Arctic Stone Age (Plate X I X B-7) is treated in the Aurignacian way, as are the amber charms, in the shape of human heads, from Carelia,1 and the Kurische Nehrung, East Prussia,2 of approximately the same age. On the latter specimen, however, a line is added for the mouth. We hesitate to infer any connection between these Neolithic figures and the Aurignacian, especially because of the vast gap in time, during which human representations are unknown. The human figure is rare in the stone age of Scandinavia. Among the minor but curious resemblances between Palaeolithic and Eskimo art, we must cite the Solutrean figure from the Vallee du Roc, Charente, France (Plate X I X B-9), carved from a block of limestone, with apparently natural rings, to which a face has been added at one end by incised lines for the eyes and mouth, in the Eskimo way, and on the other hand, a modern Alaskan ivory carving, representing the "man-worm," a mythical figure, with the segmented body of a worm, and the upper part of a man. (Plate X I X B-10.) It is impossible to determine the significance of this similarity, especially since the carvings are unique in their respective cultures. Lastly, the phallic representations of the Palaeolithic are practically absent from Eskimo art. The carved bone penis from the Sukkertoppen District, West Greenland, is a unique exception belonging only to the latter half of the seventeenth century. 3 Mathiassen believes that the flat, featureless face and the lack of arms characteristic of the Thule dolls of Canada are due to the degeneration of Thule art in the east as compared to the earlier Thule stage in Alaska, from which he believes the Canadian Thule culture was derived. The dolls of this hypothetical earlier stage, like the modern Alaskan dolls, had arms and features, 4 and it is with these, of which unfortunately we lack properly excavated examples, that we should compare the Aurignacian statuettes. Therefore, some of the resemblances between the Canadian Thule dolls and the Palaeolithic figures which might be cited—the lack of features, the tendency to omit the arms or to reduce them in size, and to show the legs pressed together and feet lacking—must be the result of chance parallelism, since, if Mathiassen is right, these are later developments, peculiar to Canada. Moreover, another difference between Eskimo dolls in general and those of the Aurignacian must be cited. With the exception of the Lespugue "Venus," and two figures from Brassempouy, one with a " b e l t , " the other with a "cape," both very doubtful,6 the Aurignacian statuettes are naked, whereas the Eskimo dolls are frequently shown with the amulet strap, the natit, or genital covering worn by both sexes, the belt, boots, or even the whole jacket. In fact, the head of the Canadian Thule dolls, Pictures from, the Collections, Helsingfors, p. 4-2. » Hoernes, 1925, p. 243, fig. 1. Mathiassen, "Ancicnt Eskimo Settlements in the Kang&miut Area," Meddelelser orn Grtfnland X C I I I , Copenhagen, 1931, fig. 35, and plate 5-9. 4 Mathiassen, 1930, p. 83. t Piette, pi. L X X - 2 and 3. 1

3

498

FREDERICA DE LAGUNA

rounded in back and flat in front, suggests the head in a fur hood, even though the rest of the jacket is not expressly indicated. The similarities between the Aurignacian figures from Europe and the Eskimo dolls are therefore not sufficient to prove any connection between the art of the two peoples. However, within the last few years, Palaeolithic statuettes have been discovered in central Russia and in the Lake Baikal district, Siberia 1 which may lead us to modify this conclusion. In Central Russia (Kostienki and Gagarino), over 19 female figures were found. Some of these statuettes resemble the conventionalized European Aurignacian form, illustrated by the "Venuses" of Willendorf and Lespugue, others show an exaggerated female type, but less rigidly conventionalized,2 while others, again, are more generalized human figures. The possibility that these statuettes may be related to the Eskimo carvings can not be dismissed. Their great numbers suggest that they may have been dolls, like the Eskimo figures, and the absence in many cases of the rigid conventionalization characteristic of the European statuettes, means a closer resemblance to the Eskimo dolls. Many of these Siberian statuettes have features, like the Alaskan dolls, and one has a rounded head with flat face, like the Thule dolls. However, we must beware of considering an imperfect naturalism as an indication of a true similarity, based on a common origin, for it can equally well result from degeneration of several distinct conventionalized styles. Engravings of human figures are more difficult to find in the Upper Palaeolithic than are the representations of animals, and as a rule they are less skillfully executed. Luquet 3 finds that the human form is usually represented with animal-like aspects : hair on the body, muzzles instead of faces, protruding stomachs, stooping posture, etc. This, he explains, is because the Palaeolithic artists first portrayed animals, and when they turned to human beings gave them a bestial character simply because it belonged to the technique and art style they had already perfected. However, not all the peculiarities of the Palaeolithic human representations can be thus explained. The incised female figure from the cavern of La Colombière,4 which is either final Aurignacian or early Magdalenian, is like an outline of the profile of one of the Aurignacian statuettes. Other female figures are the "femme au renne," already mentioned, and the woman with pendent breasts on the walls of the cave of Combarelles. 5 I exclude from this discussion the painted figures in the Spanish caves, for the style seems to be local, and though analogies to Bushman art have been pointed out, there was, in any case, no spreading of this type of painting to the north. The male figures are more numerous. I have figured (Plate X X A, B) the man stalking a bison from Laugerie-Basse, the man with a stick from La Madeleine, the sorcerer of Lourdes, the anthropomorphic figure from Mas d'Azil, and the sorcerer engraved and painted on the wall of the cave of Trois-Frères in Ariège. Other and »Information and photographs brought from Leningrad by Eugene Golomstock, University Museum, Philadelphia. s Kuhn, pl. 10-c. 3 G. H. Luquet, "Les Caractères des figures humaines dans l'art paléolithique,"L'Anthropologie, XXI, 4 Sollas, 1924, fig. 212. 5 Hoernes, 1925, p. 149, fig. 1. Paris, 1910.

499

ESKIMO AND PALAEOLITHIC ART

cruder figures are, for example, the little men at a bison feast, engraved on a pendant from the cave of Raymondin, at Chancelade, 1 who are drawn with nothing but a body and a head, with a line running up the middle of the back, presumably to indicate the division of the legs. Among the more fantastic figures, we have the grinning faces at Marsoulas, of Magdalenian age; 2 the Aurignacian animal-headed monsters of Altamira, Spain; 3 and the grotesque figures of Combarelles, Dordogne,4 which include a curious creature with the head of a girl, the body of a pregnant woman, and a man's sexual organ. There are also human figures with animal heads and other monstrous forms difficult to interpret. Besides the human figures with animal heads (Altamira, Combarelles, Mas d'Azil), there are men with remarkably high pointed heads (Lourdes, LaugerieBasse, La Madeleine, Marsoulas, La Colombiere, the last superimposed on the figure of the woman, and one figure, though not very human in the features, from Combarelles). This high pointed head is hardly an animal characteristic; neither are the full beards (Trois-Freres and Lourdes). Another striking feature of the style is a long, rather heavy body, which the sorcerer of Trois-Freres illustrates quite clearly. I t is also found in the figures of Lourdes, Laugerie-Basse, La Madeleine, Mas d'Azil, etc. This heavy, disproportioned body, however, is not found on the Aurignacian figure of a man in low relief on a limestone block from Laussel, 5 and it may be a Magdalenian feature. There is also a tendency towards small feet (TroisFreres) . We have seen that on the Aurignacian statuettes there was a tendency to slight the feet for stylistic reasons, which appears also with the Magdalenian male figures from Laugerie-Basse and La Madeleine. The body is often bowed, as in the case of the figures of Trois-Freres and Mas d'Azil,—though here the body of the latter is probably bent to fit the round area of the bone disk,—the figures at Altamira, and especially those at Combarelles, where the men go almost on all-fours. Representations of the human figure in Eskimo graphic art are not closely comparable with those just discussed. They are executed in the silhouetting technique, (Plate X X B-3), which I shall describe in detail in connection with the representations of animals. The Eskimo men are usually reduced to their simplest forms, a dot for the head, and lines for the body and for the limbs. (Plate X X B , 3,5,6.) They are much smaller in scale than the Magdalenian figures, and only in a few cases do we find them large enough to make possible any real comparison. The human figures on two archaeological combs from Alaska have the same long heavy body, with a tendency to a protruding paunch, which Luquet called an animal trait in Palaeolithic art, and the same bent attitude, though for the Point Hope figure it is hardly a stylistic feature. (Plate X X B-3 and 4.) The engravings on Central Eskimo combs (Plate X X B-10) seem to be quite modern and were probably made under white influence. The figures are clothed, not naked like the Palaeolithic representations. Some of the Palaeolithic figures have been taken for men disguised as animals. The sorcerers of Lourdes and Trois-Freres, with their tails and horns, certainly impersonate animals. The little imp-like figures on a Magdalenian "baton" from 1 4

MacCurdy, I, fig. 137. Hoernes, 1925, p. 149, fig. 1.

» Sollas, 1924, fig. 213. ' MacCurdy, I, fig. 165.

= Op. cit. fig. 213.

500

F R E D E R I C A DE LAGUNA

Mége, a t T e y j a t (Plate X X B - 9 ) , M a c C u r d y considers to be dancers disguised as chamois. 1 T h e head is t h a t of an animal, b u t the erect a t t i t u d e and the foot on one of t h e figures are h u m a n . The arms are not shown, b u t t h e figures appear to be enveloped in shaggy f u r robes, which is perhaps the only example in Palaeolithic art of complete clothing. T h e monstrous animal-headed creatures of Altamira m a y also be men disguised as animals. M a c C u r d y mentions the representation of a h u m a n figure with a bear's head from Mas d'Azil, and another with a horse's head from the cave of Espélugues a t Lourdes. 2 A male and female figure, wearing masks representing animal heads, have been found incised on pebbles a t L a Madeleine. 3 Masks and animal disguise for festivals are found among t h e Eskimo, particularly in Alaska, where the masked ceremonial has been greatly stimulated b y contact with Northwest Coast Indians. T h e mask, however, is a common Eskimo trait, presumably very old. 4 I t has a wide, though sporadic distribution in N o r t h a n d South America, as well as in Siberia. 5 I t is not improbable t h a t the mask, or animal impersonation, originated in t h e Palaeolithic, but we must not suppose t h a t t h e connections are any closer between t h e cave men and the Eskimo, t h a n between t h e former and almost any other modern primitive people. Birket-Smith's distribution appears confined to t h e circumpolar area only because of the limits he set for himself. W e have a few examples of figures disguised as animals in Eskimo art. (Plate X X B-7 and 8.) T h e shaft-straightener originally figured by Boyd D a w kins is decorated with figures of caribou and men disguised as caribou, dancing with horns on their heads and in their hands. (Plate X X I I I A-2.) A scene taken from a modern Alaskan specimen shows a festival with two men beating drums, and five dancers with animal heads and ears. An eighth figure lifts something in t h e air, (Plate X X B-8.) Although we do find a similarity in subject between these Eskimo engravings and t h e Magdalenian figures, we must not overlook the great difference in technique. T h e Eskimo figures are shown only in silhouette. ANIMAL REPRESENTATIONS

T h e range of animals depicted during t h e Upper Palaeolithic, incised or carved in bone and ivory, or incised and painted in caves, is rather large. T h e horse was the most popular, followed by t h e red deer, t h e reindeer and other deer, the bison, and t h e wild goat, 6 but almost every species of beast, bird, and fish known to Palaeolithic m a n was represented. 7 I n Eskimo art the whale, caribou, seal, bear, dog, and birds are t h e most common, while other animals are rarely represented. T h e repertory is thus not nearly so extensive as in the Upper Palaeolithic. T h e difference can be explained in p a r t by t h e absence from t h e Arctic of some of the species the cave men knew, b u t also by the fact t h a t where Eskimo art was most highly developed the hunting was chiefly of aquatic, not land mammals. One of t h e chief differences between the two arts lies in the lack of cave painting among the Eskimo. T h e Eskimo of Kodiak I. and the Eskimo formerly inhabiting 1

MacCurdy, I, fig. 167. Kiihn, op. cit., pi. 40-a and b. Op. cit., pp. 201-202, and pp. 365-366, tableB 62. mask in general, not the animal mask alone. 6 MacCurdy, I, p. 268. 3

5

2 MacCurdy, I, explanation to fig. 167. Birket-Smith, 1929, II, p. 288, table A 96. Birket-Smith is studying the distribution of the 4

11bid., pp. 272-285, for list of animals.

ESKIMO AND PALAEOLITHIC ART

501

t h e Cook Inlet region of southwestern Alaska made rock paintings. Those which I saw in Cook Inlet were very small figures of men and animals, executed in solid red silhouette. T h e surface of t h e rock showed no preparation and with one exception the figures did not seem to have much relationship to each other. T h e animals were in general better executed t h a n t h e h u m a n or anthropomorphic figures. However, these paintings can hardly be used to prove connection with Palaeolithic Europe, because they have been discovered only in southwestern Alaska, a peripheral region of Eskimo culture, and their occurrence there is possibly due to influence f r o m Indians in the South. 1 T h e Eskimo method of incising animal forms on bone and ivory is quite different f r o m t h a t employed by Palaeolithic artists. T h e latter generally represented t h e animal on so large a scale t h a t they were able to show the body markings. I n t h e finest Palaeolithic examples the coat of the animal is carefully represented by rows of short parallel strokes in the direction of the hair. This method of rendering t h e hair is so characteristic t h a t Hoernes finds in it the prototype of all the simple geometric decorations, 2 a view I consider extreme. I would say, rather, t h a t t h e rows of short incisions a n d notches which are such common elements in non-representative art have been carried over and incorporated into t h e technique of realistic representation. There is a tendency to enhance t h e decorative effect of t h e natural body patterning. (Plate X X I A-l.) T h e same method of indicating the hair is found on animals in the round, as for example, on the famous Magdalenian horse f r o m Lourdes (Plate X X I I B - l ) , where the edge of the shaggy coat is indicated by very short vertical lines forming a zig-zag band. T h e zig-zag lines on t h e flanks m a y have become further stylized into t h e triangles on t h e flanks of animals incised on t h e walls of caves, like the horse at Combarelles (Plate X X I A-2), and f r o m which Shetelig tried to derive the triangles and zig-zag lines on t h e reindeer engraved on rocks in Norway. 3 Besides the arrangement of short vertical lines in zig-zag bands, we find other conventionalizations of the body markings, for example in the t r e a t m e n t of t h e m a n e of t h e bison on the Magdalenian "baton de commandement" from LaugerieBasse (Plate X X I I I A-l) and on t h e feathers of the grouse on a Magdalenian throwing-stick from M a s d'Azil. 4 E v e n the scales on t h e salmon in the famous carving from Lorthet of t h e stags crossing a salmon stream have been made into patterns. (Plate X X I I A-5.) This attention to the details of the animal's markings and the representation of fur, in particular, is quite foreign to Eskimo art, except during very recent times, in carvings made under white supervision for the tourist trade. T h e older method, represented in t h e realistic Thule etchings, in archaeological finds from Alaska, a n d m modern but old-fashioned Alaskan art, invariably shows the animals and other figures in silhouette, the entire space being filled in with single or cross-hatching, depending on the size of the area. I n modern times, black paint, consisting of charcoal from burned grass mixed with oil, is rubbed into t h e cuts. 5 We have no evi1

" E s k i m o Rock Paintings in Cook I n l e t , " by the author, in preparation. Hoernes, op. cit. first edition, 1898, p. 43. a Shetelig, p. 48. " M a c C u r d y , I, fig. 103. 5 Hoffman, p. 790. 2

502

F R E D E R I C A DE LAGUNA

dence of this method from the Thule culture (or in the art of the Old Bering Sea period), but during use, when dirt would inevitably fill the cuts, the general effect would be the same. The rubbing of paint into incisions was introduced during the Punuk period and is applied to non-representative as well as to realistic art. The same methods of scratching out the space is used in representations of men, houses, tents, boats, sledges, etc. We might be tempted to suggest that the Eskimo shows everything in silhouette in this way, because that is how objects and animals most often appear to him, black against the white snow. However, just as the short parallel lines, zig-zag and crossing lines of the Upper Palaeolithic, which are used in representative art as the mechanism of rendering detail, exist independently in non-representative art, so do the single- and the cross-hatching of the Eskimo appear as parts of geometric patterns, not only in Thule and modern art, but also in the Old Bering Sea period. Even the double line with alternating internal spurs, perhaps the most characteristic element of Eskimo art, has entered the domain of animal representation, as on a modern ivory drill bow from Port Clarence, Alaska (Plate X X I A-5), where both this design and cross-hatching fill in the bodies of the walrus. This use of the double line with alternating internal spurs is quite unusual, but shows how closely the non-representative style and technique control the development of realistic art. Eskimo art has nothing comparable to the sketchy and impressionistic, but often spirited outlining of animals, characteristic of the Upper Palaeolithic. Hoffman has drawn attention to this difference in spirit. 1 However, the subject of this Magdalenian technique, usually an animal head (Plate X X I A-'5, 4), is often repeated several times, resembling the decorative use of the walrus head by the Eskimo. (Plate X X I A - 7 . ) In the Eskimo examples the heads are usually bounded by two straight lines, the lower one representing the water, from which the heads emerge. They are not arbitrarily cut off, as are the Magdalenian animals' heads. The latter are not usually repeated as often as the Eskimo walrus heads and are not transverse but are likely to be placed longitudinally on the object. The neatness of the arrangement, due largely to the bordering lines on the Eskimo specimens, is not characteristic of the Palaeolithic. The bottom line, in more elaborate Eskimo specimens, represents the ground or the surface of the water, sometimes for two sets of figures, one of which, of course, is upside down with reference to the other (Plate X X I B-3), or both borders may serve as ground for two different scenes, producing a similar effect. (Plate X X I B-6.) In Palaeolithic art the ground is rarely shown (as an exception see Plate X X I I A - 2 ) and sometimes in Eskimo art the figures are arbitrarily scattered. 2 In general, however, the bordering lines, in both realistic and geometric art, make one of the most striking differences between the style of Eskimo art and Palaeolithic. Besides conventionalizing and repeating the heads of animals, the artists of the Upper Palaeolithic did the same with smaller parts, for example, the horns and the ears, or the horns and the eye.3 Though this is very different from anything in H o f f m a n , p. 765, for examples, cf. Hoernes, 192.5, pp. 36-37. Like the caribou on an archaeological kayak scraper from Point Barrow, Alaska (Copenhagen, P 15-297). 3 Hoernes, 1925, p. 149, and Breuil, 1913, fig. 19-2, and fig. 40-2. 2

ESKIMO AND PALAEOLITHIC ART

503

E s k i m o a r t , I can cite a n isolated case of similarity in tlie use of t h e fish's or whale's tail. T h e E s k i m o are fond of t a k i n g t h e l a t t e r as it sticks u p o u t of t h e sea, a n d r e p e a t i n g it as an element in a decorative b a n d , as t h e y do t h e walrus h e a d . (Plate X X I B - 7 a n d 8.) I t h i n k t h a t t h e whale's tail design a n d t h e Y figures on a line h a v e h a d a m u t u a l influence on each other. T h e similarity of t h e whale's tail t o t h e Y a n d its significance as a whaling a m u l e t h a v e b o t h s t i m u l a t e d its use as a decorative element. On a " b a t o n de commandement" f r o m t h e beginning of t h e final M a g d a l e n i a n , f r o m R a v m o n d i n , Chancelade ( P l a t e X X I B-2) we h a v e t w o vertical whale's tails, connected a r o u n d t h e cylindrical s h a f t b y a c o n t i n u a t i o n of their outlines. H o w e v e r , t h e y do n o t f o r m a real border in t h e E s k i m o way a n d t h e y are combined in a h a p h a z a r d fashion with geometric figures a n d schematized animals of a c h a r a c t e r very different f r o m a n y t h i n g in E s k i m o a r t . A m o n g t h e superficial similarities in subject we should m e n t i o n t h e animal skin, t h o u g h f r o m t h e Palaeolithic I can cite only one example, t h e skin of a fox (?) on a spear point. (Plate X X I B-5.) T h i s can be c o m p a r e d t o t h e figures of skins on m o d e r n Alaskan drill bows, which are, however, blocked o u t b y cross-hatching. ( P l a t e X X I B-4.) On m o s t of t h e E s k i m o examples, t h e skins are r e p e a t e d as t r a n s verse elements of a longitudinal b a n d . 1 T h e simplest t y p e of composition of animal representations is t h e repetition of exactly similar figures, like elements of a f o r m a l design. T h i s is very characteristic of E s k i m o a r t . (Plate X X I A-6, 8.) W e find similar repetitions in t h e final M a g d a l e n i a n . (Plate X X I A-9.) L e s t we be t e m p t e d t o see t h e conventionalized E s k i m o figures as t h e final stage f r o m a n early Palaeolithic n a t u r a l i s m t o w a r d s t h a t greater stylization characteristic of t h e final M a g d a l e n i a n , we m u s t r e m e m b e r t h a t t h e few animal figures of t h e Maglemosian show a r e t u r n t o realism before t h e final extinction of incised animal representation, a n d t h a t t h e conventionalization in t h e final M a g d a l e n i a n , with t h e distortion of t h e bodily proportions, is very different f r o m E s k i m o conventionalization, which is h a r d l y m o r e t h a n a simplification a n d stiffening. W e find some a t t e m p t s a t composition i n t o a p i c t u r e in t h e U p p e r Palaeolithic, especially in t h e l a t t e r half of t h e M a g d a l e n i a n , b u t t h e y are never very successful in grouping more t h a n a few animals, or in t h e h a n d l i n g of perspective. T h e stags crossing a salmon stream, on a piece of a n t l e r f r o m L o r t h e t (Plate X X I I A-5), exe m p l i f y this failure, all t h e more striking because t h e individual animals are so b e a u t i f u l l y executed. W e m u s t note, however, a certain stiffness in t h e a t t i t u d e of t h e last stag, which is looking b a c k w a r d s . T h e M a g d a l e n i a n artists h a v e n o t fully m a s t e r e d t h e t e c h n i q u e of presenting animals in a n y b u t strict profile. H o w e v e r , t h e real difficulty comes f r o m introducing t h e salmon into t h e p i c t u r e a n d is augm e n t e d b y t h e restrictions of space, since the whole composition is incised on t h e cylindrical surface of a n antler. T h e salmon themselves are f a i t h f u l l y represented, b u t t h e y h a v e been, so to speak, t a k e n o u t of t h e water a n d squeezed between t h e legs of t h e stags; t w o are even in t h e air above t h e b a c k s of t h e deer. A s o m e w h a t similar effect is f o u n d on t h e poorer E s k i m o examples, as in t h e h u n t i n g scene f r o m P o i n t B a r r o w , Alaska ( P l a t e X X I B-6), in which a m a n is shooting a t t w o animals, 1

Hoffman, pi. 21-1 and 4, pi. 22-3, and pi. 59-1.

504

F R E D E R I C A BE LAGITNA

perhaps caribou, while two whales apparently float through the air over his head. The difficulty here results from an attempt to depict two unrelated scenes in a limited space. A very crude type of composition is illustrated by the pendant from Raymondin, Chancelade, 1 incised with a picture of five men feasting on a dismembered bison, in which the human figures are each viewed from the back, and are grouped about the bison as on a map, not as in a picture in perspective. A similar effect is produced in a modern Eskimo illustration of a shamanistic ceremony, very crudely drawn. 2 On the latter specimen further confusion arises from representing the shaman several times, at different stages of the ceremony. However, the best of the Eskimo realistic pictures surpass those of the Magdalenian in the composition of the figures and in the spirited and natural attitudes of the men and animals. The representation of animals in various attitudes does not seem to present great difficulties. (Plate X X I B-l.) Sollas has already pointed out the great difference between Eskimo and Palaeolithic art in that the Eskimo drawings seem to tell a story, often illustrating actual events, while Palaeolithic art does not.3 The difference in the treatment of animals massed in a herd is significant of the difference in fundamental principles. The two best known Magdalenian examples, both from the latter half of the period, are the herds of wild horses, engraved in broken lines on a stone, from the cave of Chaffaud, Vienne, France (Plate X X I I A-2), and the reindeer engraved on a wing bone of an eagle from the cave of La Mairie, Dordogne. (Plate X X I I A-l.) The leaders of the horses are almost completely outlined, though rather sketchily, especially in the legs. For the rest of the herds, there are only the heads, all in a line on the same level, each somewhat obscured by the one in front, with the outlines of the neck ending abruptly. The bodies are not indicated at all, and the effect of legs is given by a row of short lines, just above the ground. The last animal in each herd is more fully drawn, with the outline of the chest and back indicated. The reindeer exemplify the same impressionistic tendency carried still further. The first three deer and the last one are drawn almost completely, even to the hairs of the coat and the shaggy mane. The rest of the herd is nothing but a line of horns, all too large, and under them a band of short vertical lines, which run into the body hairs on the last of the three leaders and into the legs of the last animal in the herd. On modern Eskimo examples of herds of caribou (Plate X X I I A-3, 4) the bodies are represented by a solid horizontal band, filled in by neat vertical hatching (like a variant of the "ladder" pattern discussed later). Sometimes the heads are lifted above this band, but in other cases the animals stand in the slouching attitude, characteristic of caribou, with only the horns, drawn in neat pairs, sticking up above the band. The legs appear to be in the proper groups of four, corresponding to each pair of antlers, but it is not possible to distinguish the individual animals, and the herd remains merely schematized. The difference between the Eskimo and the Palaeolithic herds seems to result in part from the Eskimo tendency to fit animals into a formal pattern, a tendency perhaps due to the greater relative strength of the purely geometric designs in Eskimo art, and partly from the Magdalenian trend towards a sketchy impressionism. This latter principle is in direct oppo1

MacCurdy, I, fig. 137.

2

Hoffman, fig. 142.

3

Sollas, 1924, pp. 584-585.

ESKIMO AND PALAEOLITHIC ART

505

sition to that other Magdalenian tendency to emphasize the details of the body markings, and both are embodied in the reindeer herd from L a Mairie. The similarities between Eskimo and Upper Palaeolithic carvings of animals in the round have been exaggerated. There is nothing in Palaeolithic art comparable to the little carvings of swimming birds, so characteristic of both ancient and modern Eskimo art. Sometimes the bird ends with the upper part of a human body, rather conventionalized, instead of a bird's head (Plate X X I I I B-13) and such mixing of human and animal parts is quite foreign to Palaeolithic art. The Eskimo, however, have nothing corresponding to the horses' heads cut out of thin bone, with highly stylized hair, and markings about the muzzle, which the Abbé Breuil has interpreted as representations of the teeth under the skin. 1 B u t as these figures are typical of the Middle Magdalenian alone, it is not surprising that they are not found among the Eskimo. I n the Thule culture, however, there are silhouette figures, cut out of thin sheets of baleen, representing a whale, from Point Barrow, Alaska, 2 a bird, from Naujan, Melville Peninsula, Canada, 3 and a bear, from Inugsuk, West Greenland. 4 These figures have no markings, only the outline being given. The chief difference between the carved animal figures of the Eskimo and of the Upper Palaeolithic is that the former are usually quite plain or else are ornamented with non-representative designs, while the animals of the Upper Palaeolithic are finished, like the incised figures,- by lines emphasizing the body markings. However, there are exceptions: some Palaeolithic carvings are plain, and in Eskimo art we may find the eyes, mouth, ears, whiskers of animals, and the fins and gills of fish, etc., indicated, though usually in the simplest manner with dots and lines. I t is not easy to tell whether the dots so common on the swimming bird figures and on other little carvings are intended to represent the patterning of the natural markings or are merely non-representative elements applied to the carvings. (Plate X X I I I A-8 and Plate X X I I I B-4, 8, 13, 15, 17.) However, the most elaborate of the purely realistic Eskimo carvings never approaches the type of work illustrated by the bison from Laugerie-Basse or the horse from Lourdes, nor does the most decorative treatment of hair on Upper Palaeolithic carvings ever produce so geometric a design as the compass-drawn dot-and-circles, the double line with alternating internal spurs, which is not unusual on Eskimo animal figures. (Plate X X I I I B - 1 4 , 1 6 . ) Another difference, though it, too, admits of exceptions, is that the Eskimo carvings of animals, like those of human beings, seem to be toys, while those of the Upper Palaeolithic give the impression of being more serious works of art. There are only certain species of animals represented in the round by both peoples. Carved ivory bears are characteristic of the Thule culture and are common in modern Alaska. Bears are not common in Upper Palaeolithic art, and I have been able to find only one example in the round, a carving in stone, from the cave of Isturitz, Basse-Pyrénées, Erance. (Plate X X I I B-6.) Unfortunately the head is missing, and the hind legs are broken. Moreover the species is very different from the Polar bear, which has a long body and neck. The Eskimo examples, on the Piette, pl. X-4, from Lorthet, pl. L X X V I I I - 3 , from Brassempouy, etc. 1 Mathiassen, 1927,1, fig. 14. Mathiassen, 1930, pi. 11-10. 4 Mathiassen, M.o.O. 1930, pi. 19-15. 1 s

506

FREDERICA DE LAGUNA

whole, seem more conventionalized. T h e neck tends t o run into t h e head without a distinct separation, which is an exaggeration of t h e natural appearance of t h e animal, and the forelegs are underemphasized in comparison with the hind legs. (Plate X X I I B-2.) T h e final stage in this degeneration is found in t h e modern Central Eskimo ajagaq. (Plate X X I I B-3.) This type of conventionalization is certainly foreign to the Palaeolithic. T h e bear from Ponds Inlet is a good example of t h e application of a geometric design to a realistic figure. Over t h e back is a design suggesting a harness, though Mathiassen points out t h a t it is defective, 1 consisting of the line with spurs on one side, the double line with alternating internal spurs, and t h e " l a d d e r " pattern, all elements common in non-representative art. Here, these designs have nothing to do with the coat of t h e animal. I n this connection we might mention a carved bear figure from the Chukchi, decorated with dots and V shaped lines (Plate X X I I B-4) and the figures on an ivory pipe from K a m c h a t k a , ornamented with carvings of two men hunting a walrus and two bears. (Plate X X I I B-5.) Across the backs of the bears and the stomach of t h e walrus is a continuation of t h e same p a t t e r n which covers the lower p a r t of t h e pipe stem. A powder-charge measure from the Ostyak of the Vas-Jugan River, carved to represent a bear, is covered with a p a t t e r n of little gouged triangles. 2 T h a t t h e decoration of bears and other animal figures with conventional patterns may be very old is suggested by the Maglemosian amber bear, from the swamp a t Resen, in Denmark, 3 decorated with the characteristic Maglemosian design of the double line with diagonal cross-hatching and crossing lines forming lozenges. T h e first p a t t e r n is found in both Eskimo and Magdalenian art, t h e second is peculiar to the Maglemosian. Since the specimen is unique in the Scandinavian Neolithic, not much weight can be attached t o it. Fish figures, probably fish lures, are found in both the Upper Palaeolithic and Eskimo cultures. Those of the Eskimo are carved to represent fish, while most of the Palaeolithic examples are m a d e from bear's teeth or are cut from thin plates of bone. On the Eskimo specimens the eyes are indicated by dots, t h e gills are outlined, t h e medial line down t h e side m a y be a simple straight line, or m a y have fine spurs on both sides (like t h a t on one of the salmon in t h e stag picture from Lorthet), or t h e medial line m a y join the outline of the gill. The spines of t h e fins and t h e tail m a y be represented by lines. (Plate X X I I I B-2, 3, 4, 8.) These devices are also found on Palaeolithic representations of fish (Plate X X I I I B - l , 5) but since they are of t h e simplest nature, they can hardly show a real similarity in style. There are resemblances in t h e secondary use of animal forms, especially as decorations on handles, or on the ends of objects. Animal carvings on objects are found in t h e Old Bering Sea, Thule, and modern Eskimo cultures. Buttons, toggles, handles of knives, bodkins, and shaft^straighteners, d a r t foreshafts, finger rests for harpoons, t h e flaring ends of comb handles, of thimble holders, and of reels, t h e knobs on needle cases, etc., are given an animal form, usually the head of a seal, a walrus, or t h e like. T h e Eskimo shaft-straighteners published by Boyd Dawkins 1 3

2 Mathiassen, 19£7, I, p. 186. National Museum, Helsiugfors, 3904.590. S. Miiller, Oldtidens Kunst i Danmark, I, Copenhagen, 1918, fig. 24.

ESKIMO AND PALAEOLITHIC ART

507

and by Hoffman (Plate X X I I I A-2, 5 and 7) show rather striking analogies to the "bâtons de commandement" from the early Magdalenian of Le Placard, figured by Breuil. (Plate X X I I I A-3, 4, 6 , 1 0 . ) The similarity becomes more significant if we accept Boyd Dawkins' suggestion that these are also shaft-straighteners. 1 The placing of two animal heads back to back, illustrated by Boyd Dawkin's shaft-straightener, and by Eskimo combs and thimble holders, 2 also occurs in Palaeolithic art, as Sollas has pointed out, for example on the "bâton de commandement" from LaugerieBasse, which Sollas compares to that figured by Boyd Dawkins. 3 (Plate X X I I I A - l . ) The animal's head on an archaeological hunting dagger from Point Hope, Alaska (Plate X X I I I A - 9 ) , also resembles a head on an implement from Le Placard. (Plate X X I I I A-10.) However, not much weight can be attached to these similarities, which appear, on closer examination, to be due only to a sort of sketchy naturalism, a lack of precise style, rather than to a real similarity in artistic rendering. Animal heads as decorations appear on the slate knives of the Arctic Stone Age, 4 on the famous bone comb from Gotland, with the human face, already described, and animal heads of a similar style are also found in the Finnish Neolithic. 6 However, the decoration of objects with animal heads occurs so widely and is so natural that these specimens can only suggest the possibility that the Eskimo, among many others, share in this common Palaeolithic tradition. However, proof of this must remain uncertain since the animal heads of the Neolithic in Scandinavia and East Prussia 6 have a style of their own, different from both the Eskimo and the Palaeolithic. Hrdlicka 7 has made a comparison between the modern Alaskan Eskimo carvings of a kneeling animal embracing the handle of a knife and a similar Magdalenian carving from the rock shelter of Monastruc at Bruniquel. The specimen from Bruniquel, as Hrdlicka figures it, is incomplete. The object as mended, some years after it was found, does not appear to be a dagger, and there are two reindeer on it, not one, so that some of the similarity to the Eskimo forms is lost. 8 Furthermore, we can not stress the position of the animals, since it is dictated by the requirements of utility and of the restricted space. The Eskimo use the whale tail in the round to decorate objects. Hoffman has illustrated some of the more elaborate forms on modern Alaskan specimens (Plate X X I B-8), some of which have rows of tails, the flukes touching and in some cases 1 It must be remembered that it is the Aurignacian and lower Magdalenian "bâtons"which resemble the Eskimo shaft-straighteners; those of the upper Magdalenian, however, are quite different. As the Abbé Breuil has suggested, it is possible that we have been grouping under the single term, "bâtons de commandement," or shaft-straighteners, implements of several different functions, and that one explanation will not cover them all. As far as the Eskimo are concerned, the shaft-straightener is not found in the Thule culture in Canada, and though found archaeologically in Alaska (Mathiassen, 1930), is probably not very ancient there. 2 Mathiassen, 1927,1, pi. 76-7, and 1930, pi. 11-6. s Sollas, 1924, p. 258. •Montelius, "Sur les souvenirs de l'âge de la pierre des lapons en Suède," Congrès International d'Anthropologie et d'Archéologie préhistoriques, I, Stockholm, 1874, fig. 18. 5 Pictures from the Collections, Helsingfors, figs. 7 and 8. 6 Cf. the amber and clay figures, Hoernes, 1925, p. 243, figs. 4 to 7. ' A . Hrdlicka, "Anthropological Survey in Alaska," Forty-Sixth Annual Report of the Bureau of American Ethnology, Washington, D. C., 1930, pi. 28, and p. 175. 8 MacCurdy, I, fig. 129.

508

FREDERICA

DE

LAGUNA

actually connected. 1 Archaeological specimens from Alaska also show simpler uses of the whale tail as a pendant or as the end link in a chain. (Plate X X I I I B-6.) We might compare these and similar Eskimo carvings, with the fish and fish tails used as pendants (Plate X X I I I B-7), and as handles of spatulas in the Magdalenian. (Plate X X I I I B-9 to 12.) However, in this comparison, differences between the styles are again more apparent than the similarities. (To be continued) FREDERJCA UNIVERSITY M U S E U M , PHILADELPHIA 1

Hoffman, pi. 80.

DE

LAGUNA

LIST OF ILLUSTRATIONS* PLATE

XIX

H U M A N F I G U R E S IN THE R O U N D A

1. 2. 3. 4. 5. 6.

Aurignacian statuette, mammoth ivory, Grotte du Pape, Brassempouy. Scale After Piette, pi. L X X I . Aurignacian " V e n u s , " limestone, Willendorf, Austria. Scale After Hoernes, 1925, p. 121, fig. 1. Aurignacian bas relief, limestone, I.aussei. Scale After M a c C u r d y , I, fig. 162. Aurignacian statuette, steatite, Savignano sul Parano, Italy, Reduced. Menghin, 1930, pi. XII-3. Archaeological shaman's doll, ivory, Point Hope, Alaska. Scale After Mathiassen, 1930, pi. 17-1. Punuk culture statuette, ivory, Punuk Island, Saint Lawrence Island, Alaska. Scale f o • After Collins, 1929, pi. 16. 7. Statuette, mammoth ivory, Grotte des Rideaux, Lespugue, France. Scale After MacCurdy, I, fig. 159. 8. Magdalenian, "Pernme au renne," engraved on antler, Laugerie-Basse. Scale 1 >• After Piette, pi. XXVII-5. B

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11.

Aurignacian, " V e n u s Impudica," ivory, Laugerie-Basse. S c a l e n i - After Piette, pi. VII-1. Aurignacian statuette carved from a horse tooth, M a s d'Azil. Scale ' f . After Piette, pi. XLIII-2. Aurignacian " F i l l e t t e , " ivory, Grotte du Pape, Brassempouy. Scale }•{. After Piette, pi. LXXIV-4. Modern doll, ivory, Alaska. After Nelson, pi. XCIII-1. Aurignacian head, ivory, Grotte du Pape, Brassempouy. Scale ]/{. After Piette, pi. LXX-1. Thüle culture doll, wood, Ponds Inlet, northern Baffinland, Canada. Scale After Mathiassen, 1927, I, pi. 57-18. Arctic Stone Age, bone comb, Gotland, Sweden. Scale After Br^gger, fig. 43. Thüle culture doll with amulet strap, Ponds Inlet. Scale % . After Mathiassen, 1927, pi. 57-15. Solutrean carving, limestone, Vallèe du Roc, Chiarente, France. After Martin, 1927, pi. 12-2. Modern " Man-worm," Alaska. Nelson, fig. 158. Thüle culture doll, wood, Comer's Midden, Northwest Greenland. Scale % . After Mathiassen, 1927, I, pi. 78-16. PLATE

X X

INCISED H U M A N F I G U R E S A

1. Magdalenian sorcerer, engraved on schist, Lourdes. Sollas, 1924, fig. 211. 2. Magdalenian sorcerer, engraved and painted, cave of Trois-Frères, Ariège. Scale Ho- After MacCurdy, I, fig. 151. 3. Aurignacian engraving on mammoth bone, detail, La Colombière, France. After Sollas, 1924, fig. 212. 4. Magdalenian anthropomorphic figure, engraved on bone, M a s d'Azil. Scale Y2. M a c C u r d y , I, fig. 168. 5. Magdalenian engraving, man hunting a bison, Laugerie-Basse. Scale J^. Sollas, 1924, fig. 311-1. B Magdalenian, man carrying a stick, La Madeleine, France. Scale H - Parkyn, fig. 94. Detail from modern bucket handle, Norton Sound, Alaska. After Hoffman, pi. 14-7. Detail from archaeological ivory comb, Point Hope, Alaska. Scale J-f- Nat. Mus., Copenhagen, P 16-1153. Detail from archaeological ivory comb, upper Bering Strait, Alaska. After Hrdlicka, pi. 24-2. Modern, human figure etched on bone or ivory, Alaska. Hoffman, fig. 25. Detail from a modern ivory drill bow, Cape Nome, Alaska. Hoffman, fig. 27-6. Detail from a modern drill bow, Port Clarence, Alaska. Hoffman, fig. 27-8. Modern engraving representing masked dance, Alaska. Hoffman, fig. 140. Dancers disguised as chamois (?) engraved on a baton, Mège, Teyj'at, Dordogne. Scale 14. MacCurdy I fig. 167. 10. Modern comb, Aivilik Eskimo, Canada. Scale lA. Boas, 1907, fig. 215-a. 1. 2. 3. 4. 5. 6. 7. 8. 9.

PLATE

XXI

INCISED ANIMAL F I G U R E S A

1. Upper Magdalenian, ibex, Lourdes.

Scale }{.

Breuil, 1913, fig. 28-3.

2. Magdalenian, horse engraved on the cave wall, Combarelles.

Scale K 6 .

After Sollas, 1924, fig. 199.

* The figures have been somewhat reduced from the scale given above. 509

510

AMERICAN JOURNAL OF ARCHAEOLOGY

3. 4. 5. 6. 7. 8.

Upper Magdalenian semi-cylindrical baguette, T e y j a t , France. Scale Breuil, 1913, fig. 29-7. Semi-cylindrieal baguette, Trou des Forges, Bruniquel, France. Scale Breuil, 1918, fig. 29-14. Detail from modern ivory drill bow, Port Clarence, Alaska. Scaie • Ricks museum, Stockholm, Vega 5054. Detail, modern Eskimo carving, Alaska. A f t e r Hoffman, pi. 60-3. Detail, modern ivory drill bow, Port Clarence. Scale ^-f. Stockholm, Vega 5055. Detail, archaeological ivory handle or drill bow. Bering Strait. S c a l e n i - National Museum, Copenhagen, P 260. 9. Horses engraved on a wedge, Souci, Lalinde, France. Scale V'>- Breuil, 1913, fig. 37-4. B

1. 2. 3. 4. 5. 6. 7. 8.

Detail, modern ivory drill bow, Port Clarence, Alaska. Scale l \. Stockholm, Vega 5057. Upper Magdalenian " B â t o n de commandement," Raymondin, France. Scale 1--j. Breuil, 1913, fig. 28-2. Detail, modern bone box, St. Michael, Alaska. A f t e r Hoffman, pl. 58-5. Detail, modern ivory bag handle, Kotzebue Sound, Alaska. A f t e r Hoffman, pl. 21-1. Detail, Magdalenian spear point, La Madeleine. Scale 3*i- A f t e r Lartet and Christy, pl. B I X - 4 . Modern hunting record, Point Barrow, Alaska. A f t e r Hoffman, pl. 58-1. Detail from modern bag handle, Alaska. Hoffman pl. 40-6. Modern bucket handle, ivory, Sledge Island, Alaska. Much reduced. A f t e r Hoffman, pi. 80-1. P L A T E A.

X X I I

INCISED A N I M A L

FIGURES

1. Upper Magdalenian engraving on a wing bone of an eagle. 2. 3. 4. 5.

Cave of La Mairie, Dordogne.

Curdy, I , fig. 131. Upper Magdalenian engraving on a stone slab, Chaffaud, Vienne, France. Scale Modern engraving, Alaska. Hoffman, pi. 73-3. Modern engraving, Alaska. Hoffman pi. 67-4. Upper Magdalenian engraving on an antler. Lorthet. Scale Piette, fig. 58. B.

Scale j.;;.

Mac-

MacCurdy, I , fig. 132.

A N I M A L S I N THE R O U N D

1. Lower Magdalenian, horse, cave of Espélugues, Lourdes. Scale } { . MacCurdy, I, fig. 172. 2. Thüle culture, bear, ivory, Ponds Inlet, northern Baffinland, Canada. Scale % . A f t e r Mathiassen, 1927, I , pi. 57-9. 3. Modern Central Eskimo ajagaq in shape of bear, Canada. Scale Boas, 1888, fig. 521-a. 4. Modern Chukchi bear, Siberia. Scale J f . Stockholm, Vega 4530. 5. Detail from modern ivory pipe, Kamchatka, Siberia. Scale Copenhagen, K c 164. 6. Final Solutrean, bear, stone, Isturitz, France. Scale A f t e r Passemard, fig. 4. 7. Magdalenian feline figure, Isturitz, France. Scale A f t e r Passemard, fig. 7. P L A T E

X X I I I

A N I M A L S I N THE R O U N D A

1. Magdalenian, " B â t o n de commandement," Laugerie-Basse, France. Scale Vi- A f t e r Piette, pl. VI-1. 2. Modern shaft straightener, Alaska. Much reduced. A f t e r Hoffman, fig. 100. 3. Lower Magdalenian, " B â t o n de commandement," wolf or fox head, L e Placard, France. Scale Breuil, 1913, fig. 23-G. 4. Lower Magdalenian, " B â t o n de commandement," rabbit head, L e Placard, France. Scale Breuil, 1923, fig. 23-2. 5. Modern shaft straightener, Cape Darby, Alaska. Much reduced. A f t e r Hoffman, pi. 8-3 (realistic figs, on handle omitted). 6. Lower Magdalenian, " B â t o n de commandement," feline (?) head, L e Placard, France. Scale % . Breuil, 1913, fig. 23-1. 7. M o d e m shaft straightener, Sledge Island, Alaska. Much reduced. A f t e r Hoffman, pl. 7-3 (realistic figs, on handle omitted). 8. Archaeological, drag handle, heads of a seal and raven, Point Hope, Alaska. Scale A f t e r Mathiassen, 1930, pi. 12-13. 9. Archaeological hunting dagger, ivory, Point Hope, Alaska. Scale >2- A f t e r Mathiassen, 1930, pi. 15-6. 10. Lower Magdalenian, " B â t o n de commandement," L e Placard, France. Scale Breuil, 1913, fig. 23-4. B

1. Magdalenian, pike, engraved on a bear tooth, cave of Duruthy, Sordes, France. Duparc, fig. 37.

Scale } { .

After Lartet and

LIST OF ILLUSTRATIONS

511

2. Thüle culture salmon lure, ivory, Naujan, Melville Peninsula, Canada. Scale Y2. After Mathiassen, 1927, I, pi, 11-10. 3. Thüle culture salmon lure, ivory, Ponds Inlet, northern Baffinland, Canada. Scale 3^2 • After Mathiassen, 1927, pi. 43-6. 4. Modern, flounder, Unalakleet, Alaska. After Hoffman, pi. 57-1. 5. Magdalenian, flounder, Grotte des Boeufs, Lespugue, France. Scale }/[. After Saint-Perier, 1925, fig. 3. 0. Archaeological ornament representing a whale's tail, for shaman's drum, Point Hope, Alaska. Scale After Mathiassen, 1930, pi. 17-8. 7. Magdalenian pendant, Gourdan, France. Scale Breuil and Saint-Perier, fig. 8-8. 8. Modern, grayling, Norton Sound, Alaska. After Hoffman, pi. 57-7. 9. Middle Magdalenian spatula with fish-tail handle, Laugerie-Basse. Scale }/{. Breuil and Saint-Perier, fig. 8-6. 10. Middle Magdalenian spatula with fish-tail handle, Grotte Bout-du-Monde, Les Éyzies. Scale Breuil and Saint-Perier, fig. 8-3. 11. Early Magdalenian handle of spatula, cave of Rey, Les Éyzies. Scale 3^2- After MacCurdy, I, fig. 177. 12. Magdalenian spatula with fish-tail handle, cave of Rey, Les Éyzies. Scale }{. Breuil and Saint-Perier, fig. 8-2. 13. Thüle culture, bird with human head and shoulders, ivory, Naujan, Melville Peninsula, Canada. Scale After Mathiassen, 1927, I, pi. 32-6. 14. Modern, walrus, Nashagak, Alaska. After Hoffman, pi. 56-5. 15. Thüle culture bird, ivory, Naujan, Melville Peninsula, Canada. Scale Y\. After Mathiassen, 1927, I, pi. 32-4. 16. Thüle culture, bird, ivory, Ponds Inlet, northern Baffinland, Canada. Scale After Mathiassen, 1927,1, pi. 57-10. 17. Thüle culture, bird, ivory, Kuk, Southampton Island, Canada. Scale After Mathiassen, 1927, I, pi. 09-13.

A COMPARISON OF ESKIMO AND PALAEOLITHIC ART PLATES PART I I .

XVIII-XXIV

NON-REPRESENTATIVE ART

Dots DOTS are comparatively rare in Palaeolithic decoration, and only a small number of them are of the round bored type common in Eskimo art. As far as they can be identified, the types of Palaeolithic implements most frequently ornamented with dots are pendants and spatulas. According to Piette, 1 this type of ornamentation is restricted to the upper levels of sculpture and the lower levels of simple engravings, i.e., to the Aurignacian, Solutrean, and lower Magdalenian. Dot ornamentation does not continue to the end of the Magdalenian, though it reappears in a very fully developed form in the Maglemosian; this fact argues against a direct inheritance by the Eskimo of the Palaeolithic dot designs. Piette's stratigraphy, however, needs revision,2 so that perhaps this argument will have to be abandoned. Unfortunately, the material I have gathered is not sufficient to solve this question. The high development of the dot motif among the modern Central Eskimo and Algonkian Indians suggests, according to Speck, that this design originated in the region about Hudson B a y ; 3 but, he admits, it might be an element of remote antiquity, possibly dating from the Aurignacian.4 The prominence of the simple dot design among the modern Central Eskimo is due, I think, to the disappearance of the more complex elements of Thule art, leaving only the simplest patterns in the degenerate modern style. In Palaeolithic art, dots tend to occur without any other decoration. The arrangement is sometimes irregular,5 though usually we can detect a tendency towards longitudinal lines of dots (Plate X V I I I A, 1 to 3). An early Magdalenian prototype of the barbed antler points ("harpoons") of the upper Magdalenian, from Laugerie-Basse, 6 has nine rudimentary barbs, pointing towards the point, just above the tang, and above them, continuing the line, is a single row of round bored dots. However, an Aurignacian bone specimen from the Gorge d'Enfer, Dordogne, has a slightly oblique transverse line of dots.7 The ornamentation is sometimes a little more complicated, when short transverse lines are added to the vertical lines of dots, as on two Magdalenian pendants (Plate X V I I I A, 4 and 7). ' E. Piette, L'Art pendant l'âge du renne, Paris, 1907, explanation of pl. L X X X I , 3. He considered the Solutrean a phase of the upper Magdalenian, instead of a stage between the Aurignacian and the lower Magdalenian. ! F . Speck, "Central Eskimo and Indian Dot Ornamentation, " Indian Notes, Museum of the American Indian, New York, 1925, I I , No. 3, p. 170. * Speck, op. cit., p. 171. 5 As on a spatula from the Abri Blanchard, Sergeac, Dordogne, middle Aurignacian (St Germain 56344), a double-ended paint mortar from Ivulna, Moravia, Aurignacian (Breuil 1925, fig. 11, 7), and a spatula from the Grotte du Pape, Brassempouy, Solutrean (Piette, pl. L X X I , 5). ^8 H. Breuil, " L e s subdivisions du paléolithique supérieur et leur signification," Congrès International d'Anthropologie et d'Archéologie préhistoriques, XIV Session, Geneva, 1913, fig. 30, 4. " S. Reinach, Répertoire d'art quaternaire, Paris, 1913, p. 82, fig. 3. 2

77

78

F R E D E R I C A DE L A G U N A

T h e round bored dot, either alone or in combination with other design elements, is common in t h e Thule, P u n u k , and modern Eskimo cultures. T h e true round dot, however, does not appear in t h e Old Bering Sea culture, or on t h e archaeological specimens in Van Valin's collection from Point Barrow, Alaska, which in this respect shows the influence of t h e Old Bering Sea culture. P e n d a n t s and ornamental plates with dot decoration are wide-spread in Eskimo culture and presumably very old, though precise information is wanting from Alaska. We can cite, however, examples from modern Alaska, from the modern and t h e ancient Central Eskimo, and from t h e Thule culture in Greenland (Plate X V I I I A, 5 and 6). Solías has compared a modern Eskimo drop pendant with one from K u l n a (Plate X V I I I A, 4), and considers the resemblance significant. 1 However, it is problematic how we are to regard t h e occurrence together of two such simple elements as t h e drop pendant and t h e round bored dot in longitudinal lines, when neither of them alone would be of any significance. On the hair ornaments of t h e now extinct people of Southampton Island we find dot decorations consisting of longitudinal lines connected by diverging transverse lines (Plate X V I I I A, 9), similar to the dot design of t h e pendant from Kesslerloch (Plate X V I I I A, 7). A very similar design also occurs in the Danish Maglemosian, where dot decorations are very elaborate (Plate X V I I I A, 8). T h e prevailing tendency towards longitudinal lines of dots with secondary transverse elaborations we can also see on a Thule culture thimble-holder from northern Greenland (Plate X V I I I A, 10), and on an archaeological boot-creaser f r o m West Greenland (Plate X V I I I A, 11). D o t s used as borders form a type of decoration distinguishing Eskimo art from t h a t of the Palaeolithic. D o t s are substituted for more elaborate borders on late and degenerate forms of the " w i n g e d " needlecase from C a n a d a and northern Greenland (Plate X V I I I A, 13, 14 and 16), and serve as borders on other archaeological specimens, for example on a wrist guard from Alaska (Plate X V I I I A, 12). D o t s are a very common ornamentation on t h e backs of small animal figures, particularly t h e little swimming birds. Thirteen of t h e twenty-one bird figures found a t t h e Thule culture site at N a u j a n , Melville Peninsula, Canada, are so decorated. 2 T h e dots m a y be scattered irregularly over the b a c k ; 3 or, more commonly, t h e oval shape of t h e back m a y be outlined by a double row of dots, from which double lines of dots radiate to t h e edge of t h e figure (Part I, P l a t e X X I I I B, 13 and 17). This is, therefore, a simple combination of the decorative principles we have seen on t h e wrist guard and t h e thimble-holder. On one bird figure from K u k , t h e radiating lines form little loops, which Mathiassen has taken as a representation of t h e wings, tail, and feet, 4 b u t a t best this seems only a secondary adaptation of what is primarily a geometric ornamentation. F r o m N a u j a n there is a salmon lure irregularly covered with dots, 5 and from Alaska, as already stated, we have m a n y animal figures covered with dots. A modern figure of a swimming bird from 1

Solías, Ancient Hunters and their Modern Representatives,3 London, 1924, p. 583, and fig. 305. Mathiassen, "Archaeology of t h e Central Eskimos," Report of the Fifth Thule Expedition, 1921-24, IV, Copenhagen, 1927,1, p. 74. 3 4 5 Op. cit., pi. 32, 5. Op. cit., pi. 69, 12, and p. 241. Op. cit., pi. 32, 5. 2

E S K I M O AND PALAEOLITHIC A R T

79

the Pitlikaj Chukchi also has dots scattered on the head, and three chevrons on the back. 1 The line of dots is often enclosed between two straight lines, as on Alaskan needlecases, both archaeological and modern. The decoration often consists of encircling bands of straight lines, sometimes with spurs on one or both sides, with the spaces between filled with evenly arranged dots.2 A similar border is found on modern Central Eskimo combs, and also occurs as an element in the borders of modern Alaskan belt buckles and ear pendants, in which the border is adapted to the circular form.3 I t is a common Aleut decoration (Plate X V I I I A, 15), and occurs frequently in Siberia and northern Europe, from the Chukchi to the Lapps.4 However, to my knowledge, it is not found in Palaeolithic art. Variations of this design, consisting of dots placed between the spaces of the "ladder" pattern (Plate X V I I I A, 17, 18 and 19), are found in Alaskan art, both from the Punuk and modern phases, in Chukchi and in Aleut art. In the art of the Punuk culture, the dot appears enclosed in spaces of various shapes; it is also placed at or near the end of a line, usually a short spur at right angles, or diverging obliquely from the main lines of the design. I t is also characteristic of this art that dots do not appear alone, but are always part of a wellintegrated pattern. The use of dots in Punuk art calls to mind the dot designs of the Lapp Iron Age find from Varanger.Fjord, East Finmark, ' and the various modern Lappish and Siberian examples of the dot combined with zig-zag lines and triangles.6 Rather than regard Punuk art as a purely autonomous growth on Alaskan soil out of the Old Bering Sea culture art, as Collins argues, 7 1 am inclined to see in it strong Siberian influences. In arguing a Palaeolithic origin for the dot designs of northern Eurasia and the American Arctic, we should mention the many examples of dot decoration in the Maglemosian. They are found on amber pendants,8 on hafted bone and antler axes,9 and were much elaborated on bone dart points.10 Müller has pointed out the similarity between these designs and those of the Magdalenian," but we must remember that the latter are much simpler. The dot was also widely used, alone and in conjunction with other elements, on the pottery of the succeeding periods in Scandinavia and elsewhere. The most important argument, however, can be Riks Museum, Stockholm, Vega 5440. As on the Punuk culture needlecase from Punuk (H. Ii. Collins, Prehistoric Art of the Ahskan Eskimo, Smithsonian Miscellaneous Collections, Volume 81, No. 14, Washington, D. C„ 1929, pi. 17, b) and on modern examples from the lower Yukon and Kuskokwim Rivers (op. cit., pl. 18, f and g). 3 W. J . Hoffman, " T h e Graphic Art of the Eskimos," Report of the U. S. National Museum for 1895, Washington, D . C„ 1897, pi. 51, 4; pi. 48, 1; pi. 50, 6. 4 Specimens in the National Museum, Helsingfors. 6 O. Solberg, Eisenzeitfunde aus Ostfinmarken, Videnskabs-Selskabets Skifter, I I , Historisk-Filosofisk Klasse, No. 7, Christiania, 1909. 8 Specimens in the National Museums at Helsingfors and Stockholm. 1 Collins, 1929, p. 14. 8 S. Müller, Oldtidens Kunst i Danmarlc, I , Copenhagen, 1918, figs. 36-38. 9 K . F . Johansen, " U n e Station du plus ancien âge de la pierre dans la tourbière de Svaerdborg," Mémoires de la Société Royale des Antiquaires du Nord, Copenhagen, 1914-1919, figs. 35 and 01. and Broholm, "Nouvelles trouvailles du plus ancien âge de la pierre: Les trouvailles de Holmegaard et de Svaerdborg," Ibid., 1920-1927, fig. 55. i» Müller, 1918, figs. 23, 26-32. n Qp. cit., explanation of figs. 37 and 38. 1

2

F R E D E R I C A DE L A G U N A

80

based on the finding, in the Lake Baikal district, Siberia, of Palaeolithic ornamental plates, carvings suggesting worms or fish, and other objects, decorated with round bored dots.1 The

Dot-and-Circle

The art of the Upper Palaeolithic has only two or three examples of an incised circle with a dot in the center, but a comparison of these with the dot-and-circle designs of the Eskimo may prove interesting. From the lower Magdalenian of Espélugues at Arudy, we have a semi-cylindrical piece of antler with a circle and dot incised on the convex surface near one end (Plate X V I I I B , 1). This design Piette has compared to similar decorations on megalithic monuments, on objects from the Bronze Age and the Gallic period, and on Egyptian objects where it is supposed to represent the sun god. On a Magdalenian bone pendant from SaintMarcel there are three double circles with dots at the lower end of a line with downward slanting spurs on both sides which begins at a circle drawn about the hole for suspension, the whole design being enclosed within a simple border (Plate X V I I I B , 2). The dot-and-circle is sometimes found as the schematization of an eye, according to Breuil (Plate X V I I I B , 5). The circle with central dot appears frequently on the painted pebbles of Mas d'Azil.2 The dot-and-circle is one of the most common decorative elements in fully developed Punuk and modern Eskimo art. I t may stand alone, as on the specimen from Arudy, or it may be associated with the line with spurs on both sides, as on the Saint-Marcel pendant (Plate X I X A, 4 and 5). On Plate 77, Hoffman illustrates various forms which the dot-and-circle takes in modern Eskimo art. We might also be tempted to include the circles and ovals of the Old Bering Sea culture (Plate X I X A, 1 and 2). The style of the Old Bering Sea art, however, is so completely different from anything in Palaeolithic art that such a comparison could not lead very far. The similarity between the Palaeolithic dot-and-circle and that of the modern and Punuk phases of Eskimo culture is really very superficial. Those of the Palaeolithic are incised free hand, and perhaps belong to the same school as the dots, circles, spirals, etc., in raised relief, characteristic of the semi-cylindrical baguettes described by de Saint-Périer. 3 The modern Eskimo dot-and-circle is drawn by means of a compass or pronged fork,4 as can easily be seen from the mechanical regularity of the circle and the depth of the central dot in which the tool pivots. This mechanically drawn dot-and-circle has an interesting history. It is lacking in the Canadian Thule culture, in the Old Bering Sea culture, and in the earliest stages of the Punuk culture.5 I t is not represented in the Van Valin find from Point Barrow. I t first appears in the fully developed Punuk art, in association with the use of metal and metal-cut designs. Collins suggests a considerable antiquity for this period.6 Information and photographs brought from Leningrad by Eugene Golomshtok, University Museum. G. G. MacCurdy, Unman Origins, New York, 1924, I I , p. 218. 3 Saint-Périer, " L e s Baguettes sculptées dans l'art paléolithique," L'Anthropologie, X X X I X , Paris, 1929. 1 Hoffman, pp. 787-789. » Information from Henry B . Collins, U. S. National Museum. 6 Collins, 1929, p. 28. Collins has suggested, arguing from Chinese documents of the third century A.D., which prove that iron was then known to some of the tribes of northeastern Siberia, that iron in 1 2

ESKIMO AND PALAEOLITHIC ART

81

Outside of Alaska, Halin I. Smith has found this design element, apparently drawn by mechanical means, in archaeological deposits in British Columbia, some of which must be very old. I t appears on a bone dagger from Nicola Lake, in the Thomson River region, 1 on a bone plate from Lytton, 2 on an archaeological toggle for a dog halter from the territory of the Thomson Indians, 3 a n d on a pipe, 4 dagger, 3 whetstone, 6 and pestle, 7 from archaeological sites in the Yakima Valley, Washington, though Smith does not believe the design to be old in this last region, b u t to have been recently acquired from the Okonagon. 8 However, it has also been found on a barbed bone dart head 9 and (as eyes) on a stone carving of a h u m a n head 1 0 from shell heaps on t h e lower Fraser River, whose great antiquity can not be questioned. H o w is it possible to reconcile this evidence with t h e theory of a Siberian origin for t h e dot-and-circle? These archaeological cultures of British Columbia have a certain unity of f u n d a m e n t a l traits, as Smith has shown. Moreover, they contain elements which are also common to a phase of the ancient Eskimo culture in Cook Inlet, southwestern Alaska, which I h a v e called the Third Period of t h e Kachemak B a y culture. 11 This phase is characterized by decorations apparently m a d e with metal tools, as are t h e decorations of the related cultures at P o r t Moller, on the north side of the Alaska Peninsula, 12 at Kodiak Island, and in Prince William Sound. Though I did not find any example of the mechanically drawn dot-andcircle in the sites of Kachemak Bay, Cook Inlet, I feel sure t h a t it belongs to the Third Period of t h e Eskimo culture there, and t h a t f u t u r e archaeological investigations will discover this element either in Cook Inlet or somewhere else in southwestern Alaska, where corresponding cultures are represented. 13 I t seems more likely, therefore, t h a t this design element traveled to British Columbia from Bering Strait along with other elements of culture, rather than t h a t t h e dot-and-circle in British Columbia was an independent Indian invention, as Leslie Spier a n d D o r o t h y Smith have argued from its present distribution in northwestern N o r t h America. 14 small amounts m a y have reached St. Lawrence Island at t h a t time. T h e mechanically drawn dot-andcircle may, therefore, be sixteen or seventeen hundred years old in Alaska. (Collins, "Prehistoric Eskimo Culture on St. Lawrence Island," The Geographical Review, X X I I , part 1, New York, 1932, pp. 117-118.) I H. I. Smith, "Archaeology of t h e Thompson River Region, British Columbia," T h e Jessup N o r t h Pacific Expedition, Memoir of the American Museum of Natural History, New York, 1898-1900, I, part VI, figs. 360 and 378. 2 H . I. Smith," T h e Archaeology of L y t t o n , " Ibid., part I I I , figs. 109-110. 3 James Teit, " T h e Thompson Indians of British Columbia," Ibid., part V, fig. 296. * H . I. Smith, "Archaeology of t h e Yakima Valley," Anthropological Papers of the American Museum of Natural History, VI, New York, 1910, fig. 106. 5 Op. cit., fig. 120. « Op. cit., fig. 378. ' Op. cit., fig. 296. « Op. cit., p. 131. 9 H . I. Smith, "Shell Heaps of the Lower Fraser River, British Columbia," T h e Jessup N o r t h Pacific Expedition, etc., 1898-1900, II, part IV, fig. 52, b. Op. cit., fig. 58. II " T h e Archaeology of Cook Inlet and Prince William Sound," in preparation. 12 E . M . Weyer, "Archaeological Material from the Village Site at H o t Springs, Port Moller, Alaska," Anthropological Papers of the American Museum of Natural History, X X X I , part IV, New York, 1930, fig. 17, b (harpoon head), fig. 17, c (barbed dart head), fig. 19 (slender barbed point). 13 Since the above was written, two examples of t h e dot-and-circle (one with double concentric circles) were found in a Third Period site of t h e Kachemak Bay culture, during excavations in the summer of 1932. 11 L. Spier and D . Smith, " T h e Dot-and-Circle Design in Northwestern America," Journal de la Société des Americanistes de Paris, X I X , Paris, 1927, p. 53.

82

F R E D E R I C A DE LAGUNA

I t is quite possible that in the course of its wanderings, the dot-and-circle became divorced from the original metal tools, for native copper would have been a suitable substitute. However, it is not until modern times, when European and American steel have been available in abundance that this decorative design has been able to develop to any great extent. The true compass-drawn dot-and-circle makes its first appearance in the Bronze Age, for example at Susa, Persia, and in Egypt of the First Dynasty, on gambling bones.1 In Western Europe this design is also a part of the Bronze Age culture complex. I t was apparently always made with metal tools. In Africa, where this pattern is rather common, it is associated with the use of metal. Even though it may be possible, or even easy (?) to execute the design with stone tools, as Spier and Smith argue, and though it may, on occasion, have been so engraved, I know of no occurrence of the mechanically drawn dot-and-circle prior to the use of metal. 2 Stone tools were only substitutes. From the appearance of the dot-and-circle at the beginning of the Bronze Age in Scandinavia, as on two antler axe hafts from Denmark, with both the single and the double circle,3 we can trace a continuous distribution across northern Europe and Siberia, to the Aleut and the Eskimo. 4 The dot-and-circle, curiously enough, does not appear in the East Finmark find, described by Solberg, though it was a popular design all over Europe in the Iron Age and among the modern Lapps. However, it does occur on the bowl of an archaeological spoon from Norwegian Lapland.5 The evidence, I suggest, points to a Siberian and ultimately to a Eurasian Bronze Age origin of the Eskimo dot-and-circle. This motif came to Alaska during the Punuk culture stage, along with metal in small amounts, the technique of applying dots to lines and of filling the incisions of the design with paint, — all elements of Siberian art. The discovery of the mechanical dot-and-circle on a needlecase from the Thüle culture site at Inugsuk, West Greenland,6 might at first seem to disprove this theory. Mathiassen, however, very plausibly suggests that the Eskimo at Inugsuk learned the design from the Norsemen, with whom they came in contact. 7 We must remark 'Jacques de Morgan, La Prekistoire Orientale, II, "L'Egypt et l'Afrique du Nord," Paris, 1928, fig. 327. 2 Collins writes: " T h e Punuk art with its deeply and evenly incised lines and mechanically perfect circles could have been produced by none other than metal tools," (1932, pp. 117-118). 3 A. P. Madsen, Affaldninger af Danslcs Oldsager og Mindesmaerker Stenalderen, Copenhagen, 1868, fig. 23, 3, and Müller, up. cit., II, 1921, fig. 22. This specimen Müller originally published as belonging to the Later Stone Age (1,1918, fig. 242), but in the second volume of his work, revised his opinion. 4 Specimens in the University Museum, Philadelphia, in the National Museums in Copenhagen and Helsingfors, in the Riks Museum in Stockholm; from the Finnish peasants, Lapps, Vogul, Ostyak, Yakut, Tungus, Tartars, Samoyed, Kamchatkans. For the Koryak (W. Jochelsen, "Material Culture and Social Organization of the Koryak," Jessup North Pacific Expedition, etc., VI, part I I , New York, 1908, fig. 174, and W. Bogoras, " T h e Chukchi: Material Culture," Ibid., VII, part I, 1904, fig. 197, c). Spier and Smith have argued from the non-occurrence of this design in Chukchi art that the Eskimo design could not have been derived from Siberia, but I think archaeological evidence might reveal it, and its absence today may be due to the recent back-wash into Siberia of the Thüle culture, which 6 National Museum, Oslo, 17869. lacks the design. e T . Mathiassen, "Inugsuk, a Mediaeval Eskimo Settlement in Upernivik District, West Greenland," Meddelelser om Grönland, L X X V I I , Copenhagen, 1930, pi. 14, 1. » Op. cit., p. 299.

E S K I M O AND PALAEOLITHIC A R T

83

in this connection that these Inugsuk Eskimo also obtained steel from the Norsemen, as has been shown by the analysis of the metal blades of some of their harpoons, which shows the same structure and composition as that of Norse iron.1 The occurrence of the mechanical dot-and-circle in pre-Columbian Central America and Peru adds complexity to our problem. I am in agreement with the opinion of Spier and Smith that it is of independent origin in this region, and that the design here has nothing to do with its occurrence among the Eskimo. It is significant, however, that only in an area where metal was used did the dot-andcircle develop independently of the Eurasian Bronze Age culture-complex. Collins believes that the mechanically drawn dot-and-circle of Punuk and modern Alaskan art originated in Alaska out of the free-hand ovals and circles of the Old Bering Sea culture.2 However, not only does the European archaeological evidence and the distribution of this design argue against this theory, but an examination of the circles of the Old Bering Sea art proves them to be quite different in character. The ancient figures are circles and ovals, etched free-hand, often lacking the central dot, or having instead a tiny pit, into which was fitted a plug of baleen or some other substance. The circles and ovals are often double, while those of the Punuk stage are always single. I t is only in more recent times in Alaska that we get the double, triple, or even more elaborate forms of the dot-and-circle. However, there has been some connection with the art of the Old Bering Sea culture. I t is because the free-hand concentric circles and ovals were used so much in the older period that the mechanical dot-and-circle, coming from Siberia, was adopted so readily by the Punuk Eskimo. The neatness and ease with which it was executed must have been largely responsible for its supplanting the free-hand figure. A second factor favoring the adoption of the dot-and-circle is, I suggest, the similarity in the technique of its production with that of boring the round dot.3 In the early stage of the Punuk culture, in which the Old Bering Sea art had already disappeared, a free-hand circle with round bored dot in the middle is occasionally encountered.4 Perhaps this relationship between the round bored dot and the dotand-circle explains the tendency in modern art for the latter element to usurp the place of the former. Boas, in writing of the art of the Southampton Island Eskimo, says that "more recently this design (the simple dot) has been developed into the circle-and-dot design." 5 In modern Alaskan art the dot-and-circle often appears as the eye of an animal, which is natural enough, but it also occurs on the back of the animal, in the same manner as the simple dot (Part 1, Plate X X I I I B , 14). On an archaeological needlecase from West Greenland, dot-and-circles outline the wings and encircle the tube (Plate X V I I I B , 8), as do the dots on the needlecase from North Greenland and the degenerate specimens from the Central Regions. We can see this substitution actually taking place in Punuk art, where dots and dot-andcircles occupy analogous positions on the same specimen (Plate X I X A, 6). The Op. cit., pp. 299-300. 2 Collins, 1929, pp. 38-39. Speck, though inclined to view the dot-and-circle as an Eskimo invention, emphasizes its development as "associated in origin with the intensive use of drill and compasses in working on bone and 4 Information from Henry B. Collins, Jr. ivory" (p. 171). 5 F . Boas, " Second Report on the Eskimo of Baifinland and Hudson B a y , " Bulletin of the American Mnnevm of Natural History, X V , New York, 1907, p. 460. 1 3

F R E D E R I C A DE LAGUNA

84

tradition of spurs on the Old Bering Sea circles and ovals (Plate X I X A, 3), and some of the principles of composition have survived into Punuk art. The spurs are still frequently found on the dot-and-circles of modern Eskimo art (Plate X I X A, 5 and Part 1, Plate X X I I I B, 14), but the beautiful and elaborate composition is gone. Thus, there appears to be no connection between the Eskimo dot-and-circle, or the earlier free hand circle, and the Magdalenian circle with dot in the center. The Spurred, Circle The spurred circle is encountered in Palaeolithic art as well as in that of the Eskimo. On a broken " bâton de commandement" from the lower Magdalenian of (Jourdan, Haute-Garonne, there are two examples of the spurred circle, with a roughly shaped pit in the center of each (Plate X V I I I B, 3). Piette interpreted this design as a sun symbol. He also found at Gourdan a thin bone disk with a hole in the center from which radiated incised lines, and similar specimens were found at LaugerieBasse and Lourdes.1 On a lower Magdalenian fragment of a bone blade, from the Grottes des Fées, Marcamp, Gironde, there are two semicircular elements, consisting of double lines, which are spurred (Plate X V I I I B, 7). This design offers a rather striking analogy to that on an archaeological comb from East Cape, Siberia (Plate X V I I I B, 6). The spurred circle is not uncommon in Alaskan archaeology (Plate X V I I I B, 9 and 10),2 but it has not been found in the Canadian Thule culture, though we have one example of a spurred oval (Plate X V I I I B , 4). It is represented in the Old Bering Sea art, where the spurs are often oblique and seem to frame the circle, though not always (Plate X I X A, 3). This is the form most common in Punuk art (Plate X I X A, 6), where it is associated with the dot-and-circle, and it has even survived into modern Eskimo art (Plate X I X A, 5), though now the dotand-circle with radiating spurs is common. The Line with Spurs The line with spurs in its various forms is one of the most fundamental decorative motifs of both Eskimo and Upper Palaeolithic art. Decorations of this kind may be subdivided into the following types: (1) the line with spurs crossing it, at right angles or oblique, (2) the line with spurs on one side, at right angles or oblique, (3) the line with spurs on both sides, (a) paired and oblique, or (b) alternating, oblique or at right angles. An examination of Eskimo and Palaeolithic specimens shows that these design elements are to a certain extent interchangeable, for not only do we find that a careless artist will often pass from one pattern to another, but even on carefully made specimens we often find two or more variations of the same fundamental design in analogous positions, showing them to be closely related. The line with spurs crossing it is comparatively rare in both Upper Palaeolithic and Eskimo art. From the former I have only four examples, though more could probably be located. On the bone blade from the Grotte des Fées (Plate X V I I I MacCurdy, I I , pp. 217-218; Reinach, p. 89, fig. 5; p. 118, figs. 7 and 8; and p. 136, fig. 13. T . Mathiassen, "Archaeological Collections from the Western Eskimos," Report of the Fifth Thule Expedition 1921-2i, X , no. 1, Copenhagen, 1930, pi. 17, 3, an ornament from Point Hope. 1

2

E S K I M O AND P A L A E O L I T H I C A R T

85

B , 7), all the lines, with three exceptions, are crossed by spurs at right angles, and where the lines converge in the lower part of the figure the spurs alternate between one line and the next, producing an effect like that of the characteristic Eskimo pattern of the double line with alternating internal spurs. On this specimen the pattern changes from the line with spurs crossing it to that of the line with alternating spurs on both sides. The total effect of the decoration on this specimen is very Eskimo-like in character. On other specimens from the Magdalenian, the line crossed by spurs is shown to be closely related to the line with paired oblique spurs on both sides (Plate X I X B , 7 and 8). An almost identical pattern is found on the bottom of a toy wooden kayak from the Thule site at Button Point, Ponds Inlet, Canada (Plate X I X B, 9). Transverse lines, slanting down from left to right, occur on almost all the little wooden carvings from this site. On a wooden doll we get the same transformation of the line crossed by spurs at right angles into the line with paired, oblique spurs on both sides (Plate X I X B , 11). I t is only at Button Point, as far as I know, that the line crossed by spurs is used in this way, so that the similarity between these specimens and those of the Upper Palaeolithic appear to be the result of an accidental parallelism. Moreover, Mathiassen has drawn attention to the fact that the Button Point designs are the product of a local "school." 1 Tchekalenko has argued that in the Upper Palaeolithic the transverse lines used for decoration or for roughening the surface are more often slanting from the left downward than in the opposite direction, because this is mechanically easier to make.2 This technical explanation, therefore, weakens the effect of the similarity between the Palaeolithic specimens and those of Button Point, for the transverse lines on these objects may run in the same direction for the same reason. An Aurignacian paint tube from the cave of Les Cottés, Vienne, France, has an encircling decoration at one end, composed of a line of X s and a line crossed by oblique spurs (Plate X I X B , 10). As we shall have occasion to observe many times, the placing of a design about an object is very typical of Eskimo art, but is comparatively rare in that of the Palaeolithic. To my knowledge, the line crossed by spurs does not occur in Maglemosian art, and is very uncommon in the modern art of Northern Europe and Siberia. The only example I can cite from Siberia is a very poorly decorated comb from the Tungus of the Limpusk Tundra. 3 The lack of this design element argues against a relationship between the Palaeolithic and the Eskimo examples. In the Palaeolithic, the line with spurs on one side occurs five times as often with slanting spurs as with spurs at right angles; archaeological specimens from the Alaskan Eskimo show that the two types are of equal occurrence there, while in Canada the line with spurs at right angles predominates. The determining principle in both arts is apparently the same. Whenever the line is used longitudinally the spurs are oblique; when the line runs across the object or encircles it, the spurs are at right angles. In this way, the spurs themselves accentuate the longitudinal direcMathiassen, 1927, I , p. 211. L. Tchekalenko, Étude sur l'évolution de l'ornement géométrique paléolithique, Ukranian University, Prague, 1923, p. 46. a University Museum, Philadelphia, A 1552. 1

2

86

F R E D E R I C A DE L A G U N A

tion and preserve symmetry. The difference in the proportion of oblique spurs and spurs at right angles is due to the fact that the line with spurs is most often used longitudinally in Palaeolithic art; in the Canadian Thüle culture it is most often transverse; while in Alaska both uses are of equal frequency. This same principle is also observable in the art of the Scandinavian stone ages. However, it is such a simple and natural aesthetic device, that it seems rash to advance it as proof of a relationship. In Palaeolithic art the line with spurs on one side appears most often on long objects,—spear points, shafts of an antler, etc., and it is almost invariably in a longitudinal direction. On some objects we hardly know whether we are dealing with the line with oblique spurs, or whether we have only a series of oblique spurs to which a straight line has been added, (Plate X X A, 1); 1 or the place of the line may be taken by a slight ridge, below which the spurs are incised (Plate X X I V B , 4). On the most typical examples of the line with spurs we are apt to find the lines paired, with the spurs pointing towards the same end of the implement (on a spear towards the base) (Plate X X A, 4, 6); or the spurs may point towards opposite ends, producing an end-to-end symmetry (Plate X X A, 5). Sometimes the arrangement, while still longitudinal, may be more irregular (Plate X X A, %).2 Slightly more complex arrangements than the simple pairing are also found in the Palaeolithic (Plate X X A, 3 and 7). These Magdalenian examples may well be compared to those of the Maglemosian. Broholm and Obermaier have called attention to these similarities and believe that the Maglemosian designs are a direct inheritance from the Palaeolithic. 3 The fragment of an axe (?), from Illeb0lle, Langeland, Denmark, 4 is very similar to a fragment of antler from Le Placard, 6 and numerous examples of the line with oblique spurs could be cited from the Danish Maglemosian.6 The motif has also been found at Braband, a station intermediate in culture between the Maglemosian and the Kitchen-Midden. 7 The pairing of lines with spurs, such as we have found in the Magdalenian, is not very common in Eskimo art. Assymetric balance, with the spurs on the same side of both lines, or pointing in opposite directions, does not seem to appeal to Eskimo taste, which shows a preference for patterns with a right-and-left symmetry. This principle we see in the ornamentation of the Alaskan "flanged" needlecases, particularly on the longitudinal lines bordering the concave area between the flanges. There is a tendency to decorate these lines with oblique, downward slanting spurs, an impulse which is probably connected with the conventional ending of the lines Piette, pi. X X I I I , 1. F o r example, an upper Magdalenian antler from Mas d'Azil (Piette, pi. X C I V , 1), an upper Magdalenian pendant from Lorthet (Piette, pi. X X X I X , 6). 3 Broholm, p. 98, and H. Obermaier, Der Mensch der Vorzeit, Berlin, 1912, p. 468. 4 Müller, 1918, fig. 19. * Reinach, p. 171, fig. 5. 6 Bone fragment from Magiemose (G. L . Saraw, " E n Stenalders Boplads i Magiemose ved Mullerup, Sammentholdt med Beslaegtede Funde," Aarb^ger for Nordisk Oldhyndighed og Historie, Det Kongelige Nordiske Oldskift Selskab, second series, X V I I I , Copenhagen, 1903, fig. 44), piece of antler from Svaerdborg (Broholm, fig. 51), a similar antler from Taarbeck, Lyngby (Müller, 1918, fig. 0), a bone point from the baths of Winther at Copenhagen (Müller, 1918, fig. -23), and an antler axe from 7 Müller, 1918, figs. 14, 15, and 31. Refsvindinge, Uinding Herred (Müller, 1918, fig. 18). 1

2

ESKIMO AND PALAEOLITHIC ART

87

in a n inverted Y or a " t r e e " figure (a Y with central stem prolonged). T h e s e lines are usually grouped in pairs on b o t h sides of t h e concave areas (Plate X X A, 9). A similar pairing of lines with spurs 6ccurs on archaeological h a r p o o n heads f r o m Alaska (Plate X X A, 10 a n d I I ) . An example of t h e line with spurs, unusually a s y m m e t r i c for E s k i m o a r t , is on a n archaeological thimble-holder f r o m P o i n t H o p e , on which t h e line with spurs is balanced with t h e double line with little h a t c h e d areas. T h e spurs are on t h e side of t h e line next t h e edge of t h e implement, in itself an u n c o m m o n f e a t u r e (Plate X X I I I B, 6). A m o d e r n Alaskan thimble-holder exhibits a very typical a n d symmetrical design, t h e central element of which is t h e line with spurs on b o t h sides, flanked b y lines with spurs on one side only. All t h e spurs slant downwards (Plate X X I A, 7). N o t all t h e specimens in the U p p e r Palaeolithic h a v e t h e line with spurs applied longitudinally, t h o u g h this is b y f a r t h e m o s t c o m m o n a r r a n g e m e n t . T h e only striking exceptions which I can cite are a p e n d a n t f r o m t h e M a g d a l e n i a n station of Kesslerloch, Switzerland, which is encircled with lines with spurs a t right angles (Plate X I X B, 6), a n d an elaborate p e n d a n t f r o m t h e middle M a g d a l e n i a n of M a s d'Azil, which has encircling lines with spurs a t right angles a b o u t t h e b o t t o m (Plate X I X B, 3). T h e composition of t h e design on this second p e n d a n t is very similar to t h a t f o u n d in E s k i m o a r t . A m o n g t h e Eskimo, decorations like those j u s t discussed are very c o m m o n , a n d a p p e a r on beads, b o t h archaeological a n d m o d e r n (Plate X I X B, 5), 1 on a n archaeological knife-handle f r o m P o i n t Barrow, 2 a " w i n g e d " needlecase f r o m Qilalukan, a T h u l e site a t P o n d s Inlet, n o r t h e r n Baffinland, 3 on t h e little bear figure f r o m t h e same site ( P a r t 1, P l a t e X X I I B, 2), on bird bone needlecases f r o m Alaska, b o t h ancient a n d m o d e r n 4 (Plate X I X B, 2 a n d 4), a n d on m a n y modern Alaskan objects. 5 B a n d s of neatly m a d e transverse lines, one of which has spurs cut precisely a t right angles, are characteristic of a school of a r t f o u n d in P u n u k culture sites, t h o u g h distinct in some ways f r o m t h e typical P u n u k school (Plate X I X B, 1). T h a t this school has Siberian affinities is suggested b y t h e similarity of p a t t e r n s on a K o r y a k reel, 6 a n d t h e ivory pipe f r o m K a m c h a t k a decorated with bears a n d o t h e r figures ( P a r t 1, P l a t e X X I I B, 5). Influence of this style is felt in m o d e r n Alaskan a r t (Plate X I X A, 5). 7 E r o m the ancient M a g d a l e n i a n of L e Placard is a d a r t head (?) decorated n e a r t h e point a n d near t h e b u t t b y two short, converging b u t not meeting, oblique lines, on t h e lower sides of which (i.e., on t h e inside of t h e V) are oblique spurs, pointing d o w n w a r d (Plate X X B, 1). T h e neatness as well as t h e shape a n d arr a n g e m e n t of t h e design remind us of E s k i m o art, t h o u g h t h e r e is no perfectly analogous specimen with which to c o m p a r e it. T h e decoration is given a precise c h a r a c t e r by t h e f a c t t h a t t h e spurs s t a r t f r o m t h e very ends of t h e lines. D e c o r a 1

See also Mathiassen, 1927, I, pi. 31, 8, and National Museum, Copenhagen, P 25, 107. Mathiassen, 1930, pi. 10, 3. 'Mathiassen, 1927, I, pi. 52, 2. Collins, 1929, pi. 18, g, and Hoffman, pi. 52. 5 A bodkin from t h e Yukon (Collins, 1929, pi. 18, d); a snuff box from H o t h a m Inlet (Hoffman, pi. 6 «5, 3); a bone knife from Norton Sound (Hoffman, pi. 15, 3). Jochelsen, fig. 191, a. 7 See also a workbag fastener from t h e Kuskokwim (Collins, 1929, pi. 19, d), and a bodkin from t h e Yukon (Hoffman, pi. 54, 2). 2 4

88

F R E D E R I C A DE LAGUNA

tive units, somewhat similar, appear on Danish amber pendants of the New Stone Age, though here the spurs are much longer in proportion, and a variation is effected by combining spurs of different lengths (Plate X X B , 9). The closest parallels in Eskimo art are found in the decoration of the Thule combs with long, rectangular handles (Plate X X B , 4 and 8). A similar, though simpler design appears on an archaeological thimble-holder from Point Barrow. 1 Though we have noted similarities between Palaeolithic and Eskimo use of the line with spurs, we must remember that the Eskimo, both in ancient and modern times, employ this element as a border, outlining the whole object or a part of it (Plate X X B , 7), and that borders are almost unknown in the Upper Palaeolithic. On an archaeological ulo handle from Point Barrow there is a line like a border along the top of the handle, on the lower side of which are spurs grouped in twos and threes, with a little whale's tail in the middle, like an inverted Y (Plate X X B , 2). This grouping of spurs appears to be characteristic of Alaska, as opposed to the east. It is not common in the Magdalenian, but from Espelugues, Lourdes, there is a blunt point of antler with a line running up the middle, on one side of which are very long spurs in groups of three (Plate X X B , 3). We have already noted that in the Maglemosian the line with spurs on one side is used longitudinally in the same way as in the Magdalenian. In the Megalithic, or Later Stone Age, the line with spurs is almost always transverse, or encircles the object, which is the commonest style in Eskimo art. 2 A triangular plaque of bone from the Danish Neolithic has a border decoration like that on the archaeological combs from Alaska (Plate X X B , 5). The line with spurs does not appear in the Lapp Iron Age find from East Finmark. Though it appears on modern Lappish and Siberian objects, it occupies a very inconspicuous place in the design, usually occurring transversely, and at the edge of the pattern. The decorations employing the line writh spurs on one side show an evolution from the Palaeolithic through the Scandinavian Stone Ages towards the Eskimo uses of this element, and offer us the best argument we have yet encountered for a relationship between Eskimo and Palaeolithic art. The line with alternating spurs on both sides is a characteristic, though not very common Eskimo decoration. I t was used in the Thule culture (Plate X X I V A, 6) and in the Punuk culture, where it was combined to produce bands of the double line with alternating internal spurs 3 (Plate X X I V A, 2). I t is found on a needlecase in the Van Valin collection from Point Barrow, 4 and on modern Alaskan snuff tubes, needlecases,5 and on other articles (Plate X X I V A, 4 and 10). However, this design is extremely rare in the Upper Palaeolithic. I t occurs, but only through accident, on the base of a semi-cylindrical implement made from an antler from the upper Magdalenian of Bruniquel, 6 and on a bone plate from the Grottes des Fees, Marcamp (Plate X V I I I B , 7). The National Museum, Copenhagen, P 15, 19C. Miiller, 1918, figs. 55 and 241; also specimens in the National Museum, Copenhagen. Collins, 1929, pi. 17, h and g. 4 J . A. Mason, "Excavations of Eskimo Thule Culture Sites at Point Barrow, Alaska," Proceedings of the XXIII International Congress of Americanists, 1928, New York, 1930, fig. 1. 5 Hoffman, pi. 52. ' Breuil, 1913, fig. 29, 4. 1

2

3

E S K I M O AND PALAEOLITHIC A R T

89

The line with alternating spurs on both sides is found on a Maglemosian implement made from an antler, resembling a "bâton de commandement," from KleinMachnow, near Berlin. The decoration consists of the double line with alternating internal spurs, with black paint rubbed into the cuts, and has a very Eskimo like character. 1 The line with alternating spurs also occurs in the Later Stone Age of Denmark, where it seems to have developed out of the line with spurs on one side only, with which it is closely associated.2 I t is found on an arrow from East Prussia, contemporaneous with the Arctic Stone Age.3 In the Lapp Iron Age dots applied to the sides of a line produce an analogous effect.4 Thus, at best, we can find evidence for only a Maglemosian origin for this design element. It seems unlikely, however, that this should be the original of the Eskimo pattern, in view of its rarity in Siberian art. The line with oblique spurs on both sides apparently offers close parallels between Palaeolithic and Eskimo art, not only in the forms and variations of the motif, but in the composition. We can recognize various subdivisions of the pattern: (1) the medial line is fairly long with oblique spurs on both sides, pointing in the same direction. and no particular attention is given to the ends, (2) an end-to-end symmetry is obtained because all the spurs on one half of the line point in one direction, while those on the other half point towards the other end, (3) the line is quite short, and the ends are important. In this group should be placed the Y and " t r e e " patterns of the Eskimo and the Palaeolithic " a r r o w " figures. The spurs are not always evenly spaced in these designs, but pleasing variations may be produced by groupings. Under the first head, we can mention the slender point from Lespugue (Plate X I X B , 8), which we compared to the wooden kayak from Button Point, the very fine middle line on the bone plate from the Grottes des Fées (Plate X V I I I B , 7), and two Magdalenian bone disks from Bruniquel and Kesslerloch (Plate X X I B , 2 and 3). From the Magdalenian of Espélugues, Lourdes, this design is applied longitudinally to a bird bone needlecase (?) (Plate X X I B , 1), to a piece of antler,5 and to a flat slab of bone.6 We have already noted the association of this motif with the line crossed by spurs on a stick of antler from Mas d'Azil (Plate X I X B , 7). A more sketchy rendering of this design, in which the lines do not quite touch each other, is found on a barbed point ("harpoon "), from the final Magdalenian (Plate X X I A, 2). A fragment of a spear point (?) from the lower Magdalenian of Le Placard has apparently the same design.7 On a slender bone point from Lorthet the line with oblique spurs on both sides passes over into the line with oblique spurs on one side only, owing to the omission of some of the spurs (Plate X X I A, 6), and a similar error is found on a Chukchi toggle (Plate X X I B , 5). Below the line on the specimen from Lorthet are two detached chevrons, suggesting a continuation of the line with spurs. A similar combination occurs on a slate pendant from the Thule culture (Plate X X I A, 5). On a "bâton de commandement" from the upper Magdalenian of Raymondin, decorated with whale tails (Part I , Plate X X I B , 2), the line 1 3 6

2 Antler hammer from Fredbjerg (Millier, 1918, fig. 55). Obermaier, 1912, fig. 286. Br0gger, fig. 155. * Knife handles (Solberg, figs. 96 and 97). » Piette, pl. X X I V , 9. Op. cit., 6. 7 Breuil, 1913, fig. 18, 4.

90

F R E D E R I C A DE LAGUNA

with spurs is a little irregular because there are more spurs on one side than on t h e other. I t ends in a Y. Examples of t h e line with oblique spurs on both sides are very common in Eskimo art, both on modern and on archaeological specimens. I t is commonly used longitudinally (Plate X X I B, 4), and often, as on the modern Alaskan thimble-holder (Plate X X I A, 7), it forms t h e central element of t h e design. T h e spurs m a y be grouped by twos or threes, in t h e characteristic Alaskan fashion (Plate X X I B, 10). 1 An archaeological comb from Point Atkinson, Canada, shows how closely related t h e line with oblique spurs on both sides is to t h e Y and t h e V figures (Plate X X I I A, 10). A Magdalenian engraving from Laugerie-Basse (Plate X X I B, 11), supposed to represent a plant,—an interpretation which m a y well be questioned,—is comparable to t h e Alaskan specimens with grouped spurs. I t is similar in style to t h e blunt point of an antler from Lourdes, with grouped spurs on one side (Plate X X B, 3). A defective archaeological harpoon head from Southampton Island, Canada (Plate X X I A, 3), resembles the barbed point from La Madeleine (Plate X X I A, 2), in t h a t t h e spurs do not touch t h e central line, and t h a t t h e central line does not run the entire distance. This is also true of the design on the edge of a hook in t h e Van Valin collection from Point Barrow (Plate X X I A, 4). On both Palaeolithic and Eskimo specimens the line with oblique spurs is given an end-to-end symmetry. T h e spurs m a y point towards t h e middle of the line (Plate X X I B, 8), or towards t h e ends (Plate X X I B, 6, 7, 9). In Eskimo art this is closely related to t h e double-ended " t r e e " and Y figures, which we will discuss presently. On the Chukchi buckle, already mentioned (Plate X X I B , 5), there is an a t t e m p t a t this type of decoration. In the Magdalenian we find t h e line with spurs on both sides shortened into " a r r o w " - s h a p e d figures. T h a t sometimes these designs m a y actually symbolize barbed points, is suggested by t h e figures of animals with barbed points on their flanks, painted in the cave of Niaux (Plate X X I I A, 1). On t h e sides of the feline figure from Isturitz (Part I, Plate X X I I B, 7), which Rivet has compared to the Eskimo ajagaq in the shape of a bear, 2 there are similar decorations, with from three to four pairs of barbs. These m a y well symbolize the barbed weapons used in real hunting, whose roll was taken by t h e pin used to pierce t h e figure in the game or ritual. I n an upper Magdalenian burial, in t h e cave of D u r u t h y a t Sordes, Landes, was found a necklace of bear's canine teeth. Saint-Périer suggests t h a t these teeth m a y originally have served as fish lures. 3 On them are a series of " a r r o w " decorations, with from one to four pairs of " b a r b s " (Plate X X I I A, 2 and 3). One of the teeth has a fish engraved on one side (Part I, Plate X X I I I B, 1), on t h e other is a short line with two sets of spurs, as if pointed a t t h e hole for suspension. Another 1 Grouping by twos on a thimble-holder from St. Michaels, Alaska (Hoffman, pi. 35, 3); by twos and by threes on a shaft straightener from Nubiachugalik, Alaska (Ibid., pi. 8, 2). 2 P. Rivet, "Interpretation ethnographique de deux objets préhistoricjues," Congres International des Americanistes, XI Session, P a r t II, Goteborg, 1925. 3 R. de Saint-Périer, "Engins de peche paléolithiques," L'Anthropologic, X X X V I I I , Paris, 1928, p. 18.

ESKIMO AND PALAEOLITHIC ART

91

tooth has the figure of a seal on one side, and an " a r r o w " with three pairs of "barbs " on the other side (Plate X X I I A, 2a). If the teeth were merely pendants, the designs are, no doubt, purely decorative; but if they were used as fish lures, or had some significance as amulets, the figures may be symbols of barbed spears. It is interesting to note that a design of the same kind is incised on a human tooth, with hole for suspension, from the Aurignacian cave of La Combe, Dordogne (Plate X X I I A, 4). The " a r r o w " lacking the shaft, or rather a V with medial prong, appears on upper Magdalenian implements from Bruniquel (Plate X X I I A, 7). On one specimen these elements are used like spurs along a longitudinal line (Plate X X I I A, 6). Ordinarily the line with spurs on both sides is not abbreviated in Eskimo art into the forms which we have just observed in the Magdalenian. However, on a modern Alaskan hair ornament (Plate X X I I A, 8) there is a figure almost identical with that on the human tooth from La Combe, or the tooth with the engraving of a fish from Sordes. On two pendants in a comparatively modern Aleutian necklace, there are encircling lines with groups of three diverging spurs, like those on the semicylindrical implement from Bruniquel (Plate X X I I A, 6). In Eskimo art the line with spurs on both sides is usually shortened into what Hoffman calls the " t r e e " figure.1 In Alaska we find every stage represented, from the long line with many spurs to the simple Y. In the Canadian Thule culture, the simple Y is by far the most common, and we lack the typical " t r e e " figure, or Y with central prong. In fact the only archaeological examples of the line with oblique spurs on both sides from Canada are on the slate pendant from Naujan (Plate X X I A, 5), and the wooden toys from Button Point (Plate X I X B, 9 and 11). Mathiassen has, therefore, argued that the Y figure was derived from the " t r e e " figure, and that this simplification took place in Alaska before the eastward spread of the Thule culture. 2 Alaskan examples of the various forms of the " t r e e " and Y figures show that they were used in the same way. The most common type of decoration is a series of "trees," or Ys, "growing" from a longitudinal, or bordering line, after the fashion of simple spurs (Plates X V I I I B, 10, X X I I A, 11, and X X I I B, 4). This is true of archaeological as well as modern specimens. A bone knife from Norton Sound has the simple Y and the Y with medial spur in exactly analogous positions (Plate X X I I A, 12). The Y figure is quite rare in Punuk art. Collins finds, however, a Y-shaped element of double lines, connected to other lines at the base and at the end of the prongs. 3 This double Y plays such a different role from that of the simple Y that I do not think there can be any close relationship. However, the pure Y does occur sometimes, as on a harpoon head from Cape Kialegak, St. Lawrence Island (Plate X X I I B, 3), and on a knife handle from Punuk. 4 The simple Y rising from a line is not uncommon in the Canadian Thule culture, where it also occurs flanked by, or alternating with, simple spurs (Plates X X I I A, 9 and X X I V A, 6). From a Thule site on Belcher Island on the east coast of Hudson Bay, there are three ornamental ivory plates with borders consisting of lines from 1 3

Hoffman, p. 829. 2 Mathiassen, 1927, II, pp. 123-124. As on an ornamental plate from P u n u k (Collins, 1929, pi. 11, g). < Collins, 1929, pi. 17, a, p. 31.

92

F R E D E R I C A DE L A G U N A

which spring rows of Ys. On the first of these the Ys are detached so t h a t their true character can be recognized, 1 b u t on t h e other two (Plate X X I I B, 5) 2 t h e Ys are so close together t h a t the prongs touch, giving t h e effect of a zig-zag line, connected at intervals with a parallel straight line. T h e Y also occurs free and detached, chiefly as a decoration on harpoon heads, with the prongs pointing downward, just above the line hole. Harpoon heads so decorated are very common in the Canadian Thule culture (Plate X X I I B, 1 and 2), and have also been found a t archaeological sites in northern Alaska and a t E a s t Cape, Siberia. 3 Collins, however, is of t h e opinion t h a t this type of decoration in harpoon heads is fairly recent in Alaska. 1 I think t h a t t h e Y on the P u n u k harpoon head from Cape Kialegak represents the influence of the Thule culture, or of a related culture stage in the north, upon the local P u n u k style of St. Lawrence Island. 5 One of t h e holes on a trace buckle from N a u j a n is decorated with a Y just like the hole on a harpoon head. 6 T h e unattached Y is comparatively rare on objects other t h a n harpoon heads, though it does occur on an archaeological thimbleholder from Point Barrow, 7 and on the archaeological comb from Point Atkinson (Plate X X I I A, 10). T h e double-ended Y appears on t h e b o t t o m of an archaeological f a t scraper from Point Barrow, 8 and on a modern sinew twister from Alaska. 9 T h e double-ended Y with the prongs attached a t t h e top and b o t t o m is very common on modern Central Eskimo needlecases of caribou leg bone. Two short longitudinal lines, parallel to the stem of t h e Y, are often added (Plate X X I I B, 7). T h e unattached doubleended Y with the short parallel lines is repeated several times on a pair of wooden snow goggles in t h e British Museum collection from t h e Coppermine River, and a similar design is found on a modern quiver handle from King William Land (Plate X X I I B, 10). If we have dwelt on the " t r e e " and the Y figures a t such length, it has been to show how very important they are in Eskimo art. T h e almost complete lack of them in Palaeolithic art certainly furnishes a strong argument against the theory of a Palaeolithic origin of t h e Eskimo designs. However, a few parallels can be noted. T h e series of Ys or " t r e e s " on a line is completely lacking in the Palaeolithic as a purely geometric motif, b u t Ys on a line are employed to indicate the markings on one of the salmon in the famous engraving from Lorthet (Part I, Plate X X I I A, 5). 1

Matliiassen, 1927, I, pi. 77, 10. And on t h e unpublished specimen, t h e National Museum, Copenhagen, P 22, 8. 3 F r o m N a u j a n (Mathiassen, 1927, I, pi. 1, 2, 3, 5, 9, and 11); from graves near N a u j a n (Ibid,., pi. 37, 2, 3, 4, 5); from K u k , Southampton Island (Ibid., pi. 69, 3 and 5); from Point Atkinson (Mathiassen, 1930, pi. 1, 1 and 2); from Point Hope (Ibid., pi. 12, 1); from East Cape (Ibid., pi. 18, 4) (cf. C. Wissler, " H a r p o o n s and D a r t s in t h e Steffanson Collection," Anthropological Papers of the American Museum of Natural History, X I V , P a r t I, New York, 1914, figs. 7, 8, and 21, from Point Barrow). 4 Collins, 1929, p. 44. 5 Another example of t h e influence of t h e Thule, or of a related culture, upon t h e P u n u k culture, is the appearance on St. Lawrence Island in t h e second half of t h e P u n u k period of t h e Thule house, built of stones and whale bones (H. B. Collins, "Ancient Culture of St. Lawrence Island, Alaska," Explorations and Field-Work of the Smithsonian Institution in 1930, Washington, 1931, p. 142). 5 T h e National Museum, Copenhagen, P 1, 599. »Ibid., P 15, 515. 8 T h e National Museum, Copenhagen, P 15, 181. 9 Specimen in t h e American Museum of N a t u r a l History, N e w York. 2

ESKIMO A N D PALAEOLITHIC ART

93

T h e double-ended Y occurs on a large Magdalenian "bâton de commandement" f r o m Laugerie-Basse, 1 just below t h e hole, and suggests t h e decoration of t h e hole on Eskimo harpoon heads. T h e double-ended Y with short lines parallel to t h e stem is incised on t h e beveled base of a short weapon point from the early Magdalenian of M a s d'Azil (Plate X X I I B, 6). T h e position of t h e figure suggests t h a t it was not a decoration b u t was used to roughen t h e b u t t for hafting; however, there are other instances of neatly incised figures in t h e same place on similar points. A doubleended Y with only one short parallel line is incised on one of the bear's teeth in t h e necklace from Sordes (Plate X X I I B , 9), and t h e same design is said to be found on a piece of antler from La Madeleine. 2 On an antler chisel from t h e lower Magdalenian of Le Placard two detached spurs a t t h e end of a line give the impression of a Y (Plate X X I I B, 8). T h e " b a c k g r o u n d " of t h e figure of t h e man with a stick, from L a Madeleine (Part I, P l a t e X X I B, 1), might be compared to t h e border on the ivory plates from Belcher Island (Plate X X I I B, 5). We might be tempted to argue t h a t t h e simplifications in Eskimo art of t h e line with oblique spurs on both sides are derived from t h e Palaeolithic motifs ; t h a t for example, t h e radiating lines with spurs on both sides, on the bone disk from Kesslerloch (Plate X X I B , 3), form t h e prototype from which analogous Eskimo decorations have sprung, as on t h e archaeological stone disk from Point Hope (Plate X V I I I B, 10), and the ivory b u t t o n from N a u j a n with radiating Ys (Plate X V I I I B, 4). An intermediate stage would have to be found, and we might cite the Y and t h e line with oblique spurs on both sides in the Megalithic pottery of Sweden, where these two elements appear incised vertically, and horizontally in vertical bands, in analogous positions. 3 However, we are not able to show a gradual evolution, for in t h e L a p p Iron age a stage of simplification more extreme t h a n t h a t of Eskimo art has been reached, since only the simple Y is found. I t occurs free and detached, 4 rising from lines, 6 a n d radiating from the angles of rhomboids. 6 T h e use of t h e Y figure is probably connected with the common pattern of two longitudinal lines which diverge a t the very ends. 7 Though t h e series which we have suggested might be taken as proof of a Palaeolithic origin of the Eskimo " t r e e " and Y figures, t h e rarity of such design elements in Siberian art certainly constitutes a formidable gap. The Double Line with Cross Lines ("Jjadder"

Pattern)

This is one of the few decorative elements of t h e Upper Palaeolithic which enters into combinations and arrangements comparable to those of Eskimo art. I n t h e Palaeolithic the cross bars are usually a t right angles to the double lines. T h e simplest use of this p a t t e r n is longitudinally up t h e side of a long shaft or similar 1 P. Girot and E. Massenat, Les Stations de l'âge du renne dam les vallées de la Vézère et de la Corrèze; Laugerie-Basse, Paris, 1900, pl. XCI1I, 1. -L. Lartet and C. Duparc, " S u r une sépulture des anciens troglodytes des Pyrénées," Matériaux pour l'histoire primitive et naturelle de l'homme, tenth year, second series, V, Paris, 1874, p. 139, but no reference is given. 3 J. Nihlén, Gotlands Stenâldersboplatser, Stockholm, 1927, figs. 11 to 17 and 111. 4 Solberg, figs. 34, 35, 64, 83, and 197. s Ibid., figs. 85 and 148. « Ibid., figs. 112 and 148. • Ibid., fig. 107.

94

FREDERICA DE LAGUNA

object (Plate X X I I I A, l). 1 On a "baton de commandement" from the cave of Les Espelugues, Lourdes, the cross bars on the " l a d d e r " pattern are grouped, though irregularly (Plate X X I I I B , 2), a device very common in Eskimo art, though somewhat unusual in the Palaeolithic. On other Palaeolithic specimens, the transverse lines may be oblique (Plate X X I I I A, 2). On an implement made from an antler, from the middle Magdalenian of Espelugues, Lourdes, the double lines are so far apart that the " l a d d e r " covers one whole side, and I am not sure whether the incisions are intended to be decorative or are to roughen the surface for some useful purpose (Plate X X I I I A, 3). On a slender Magdalenian shaft from LaugerieBasse, there is the same spread-out form of the " l a d d e r " design, except that instead of the double lines, a wide groove has been incised, in which the cross bars are cut. 2 The combination of a simple engraved pattern with elements in low relief is very characteristic of the Magdalenian, especially in the final stages, and is one of the ways in which this art differs greatly from that of the Eskimo. I n the Palaeolithic, the " l a d d e r " is also employed transversely or encircling a cylindrical object, where it may be arranged in bands to form a pattern of squares (Plate X X I I I A, 8 and 11). A " l a d d e r " with irregularly spaced cross bars encircles the shaft of an upper Magdalenian "baton de commandement" from Mas d'Azil (Plate X X I I I B , 3). T h e design also appears on a late Magdalenian "baton" from Kesslerloch, in a frame about the holes (Plate X X I I I B , 1). T h e cylindrical pendant from Mas d'Azil (Plate X I X B , 3), and a somewhat similar object from the lower Magdalenian of Gourdan (Plate X X I I I A, G), are decorated with the " l a d d e r " pattern in an arrangement which we have already mentioned as being very Eskimolike in style. On an ornamental ivory plate from the lower Magdalenian of Lorthet, the " l a d d e r " element appears as the degeneration of the double line with paired internal spurs (Plate X X I I I A, 7). On the pendant from Mas d'Azil there is the same merging of these two decorative elements. T h e simple " l a d d e r " is found in the Canadian Thule culture, from which, however, I have only one good example (Plate X X I I I A, 9). I t is found in the art of the Old Bering Sea culture, 3 on archaeological specimens from northern Alaska (Plate X X I I I A, 12), and in modern Eskimo art (Plate X I X B , 4), though I know of no example from the Punuk culture. On a modern Alaskan snuff tube or needlecase on which several elements of the ladder pattern coincide, the cross lines are alternately placed (Plate X X I I I A, 10). This specimen should be compared to the ivory cylinder from Brassempouy (Plate X X I I I A, 8), on which, however, the cross lines have no regular arrangement with reference to each other. As an example of oblique cross lines in Alaskan art, we may cite a modern bow drill from Point Barrow (Plate X X I I I A, 4). In Eskimo art there is a common tendency towards grouping the cross bars, as on 1 See also the design on a composite dart from the cave of Tuc d'Audubert, Ariege (Comte Begouin, " L ' A r t mobilier dans la caverne du Tuc d'Audubert (Ariege)," IPEK, Leipzig, 1926, pi. 4, 9); and on a Magdalenian throwing stick from Kesslerloch, Switzerland (Conrad Merk, Excavations at the Kesslerloch near Thayngen, Switzerland, trans, by J . E . Lee, London, 1876, fig. 42). 2 Lartet and H. Christy, Reliquiae Aquitanicae, etc., Paris, 1865-1875, pi. B X X I I I , 2. 3 With both vertical and oblique cross bars; information from Henry B . Collins, J r .

ESKIMO AND PALAEOLITHIC ART

95

1

an archaeological brow band from Point Hope, and a modern kantag handle from Iskaktolik, Alaska (Plate X X I I I B, 4). Apparently the only form of the " l a d d e r " design in the P u n u k culture is found on objects typical of what we have called the second "school" (see page 87), in which the lines are double, and are connected by double cross lines (Plate X X I I I B, 7).2 An archaeological thimble-holder from Point Hope (Plate X X I I I B, 6), and another from Kitikarjuit, the Mackenzie Delta, Canada, 3 have double lines with cross lines, but the cross lines are so close together that we hardly know whether we are dealing with the " l a d d e r " pattern, or with the related motif of the double line with alternate blank and hatched spaces. The same difficulty applies to the border on an archaeological comb from Point Hope. 4 On most of the specimens from the central regions on which this design occurs there is no doubt but t h a t we have the double line with hatched and blank fields (Plates X X I I I B, 5 and X X I V A, 6). T h e hatching may be plain or cross hatching, and it is the same as that used to fill in the silhouetted figures of men, animals, etc. I t also occurs, though rarely, in the art of the Old Bering Sea period. 5 The double line with alternate blank and hatched areas is not found in Palaeolithic art, and the only parallel for it in European archaeology is found in the hatched checker-board figures of the Maglemosian. The simple " l a d d e r " pattern, with oblique cross lines, also occurs in the same period. 6 The " l a d d e r , " both with evenly spaced and grouped cross bars, is common in the ceramics of the later periods in Scandinavia, and elsewhere. The " l a d d e r " is not very important in modern Siberian art, though it has a wide and somewhat sporadic distribution in northern Eurasia. Among the Vogul and Ostyak, the birchbark baskets are often decorated with bands of this design element, the cross bars grouped by twos and threes. 7 The Double Line with Alternating

Internal

Spurs

This decorative motif, called by Hoffman the "seal tooth or fish trap ornament," is perhaps the most important single element in Eskimo art. I t occurs in the Thule culture of Greenland and Canada; it is represented in archaeological collections from northern Alaska, and is found in Punuk art, even in its earliest stages, together with the freehand circle with round bored central dot, 8 (though it is not typical for that period). I t is very common on modern Alaskan objects. I t is lacking, however, in the art of the Old Bering Sea culture. This design element has been employed to decorate almost every type of object except the harpoon head. We find it employed in longitudinal lines (Plate X X I V A, 1); it may encircle objects (Plates X X I V A, 1, 8 and X X A, 9); it forms borders (Plates X X I I B, 4 and X X I V A, 3); 1

T h e National Museum, Copenhagen, P 1G, 817. See also a somewhat similar object from t h e upper Bering Sea region (A. Hrdlicka, "Anthropological Survey in Alaska," Forty-Sixth Annual Report of the Bureau of American Ethnology, 1928-29, Washington, D. C., 1930, pi. 25, 3). = T h e National Museum, Copenhagen, P 17, 1498. • Ibid., P 10, 1415. s Information from Henry B. Collins, Jr. «See a spear point, Danish Maglemosian (Madsen, 1868, fig. 40, 1), and a bone implement from Stensby, Denmark (Miiller, 1918, fig. 22). 7 Specimens in the National Museum, Helsingfors, the Riks Museum, Stockholm, etc. 8 Information from Henry B. Collins, Jr. 2

96

F R E D E R I C A DE LAG UNA

it occurs in transverse lines (Plates X X I I A, 11 and X X I V A, 4); or as transverse elements in a longitudinal band or border (Plate X X I V A, 7); it may form circles or arcs (Plate X V I I I B, 6, 10) ; it frequently serves as the base for a series of Ys or " t r e e " figures, or for simple spurs (Plates X V I I I B, 10, X X I I A, 11 and B, 4, and P a r t I, Plate X X I I I B, 16). Occasionally in Eskimo art we find examples of the double line with paired, not alternating spurs (Plate X X I V A, 5). I t is apparently a degenerate form of the double line with internal alternating spurs. We have already mentioned a few examples of this same pattern in Palaeolithic art, where it is associated with the " l a d d e r " pattern (Plate X X I I I A, 7). However, this element is rare in the art of both peoples. While we can hardly imagine Eskimo art without the double line with alternating internal spurs, we find this element completely lacking in Palaeolithic art, unless we accept as examples the few instances where it occurs by accident, as when spurs along the edge of a figure are brought together, and alternate by accident. 1 T h e only intentional use of a design similar to that of the Eskimo is on the ivory " p e g " from Brassempouy, which is decorated about the head with a band of two raised ridges, the edges of which are alternately nicked (Plate X X I V A, 9). Vertical ridges, similarly nicked, cover the head. The resemblance to anything in Eskimo art is thus not very close. The earliest example of this decorative motif is on the antler implement, resembling a Magdalenian "bâton de commandement," from a Maglemosian site at KleinMachnow near Berlin. The incisions are filled with soot-blackened pitch. This design, however, is not characteristic of the Maglemosian. If we a t t e m p t to follow its distribution, we shall find it on an antler hammer from the Danish Megalithic period, 2 an arrow head from East Prussia of a type related to those of the Arctic Stone Age, and on Megalithic pottery in Denmark. 3 This design is lacking in the L a p p Iron Age find from East Finmark, except on one of the newer specimens, on which, however, dots are substituted for the usual spurs. 4 A comb for making ribbons from a late Iron Age find (tenth century A.D.) from the Island of Björkö, Stockholm, is decorated with the double line with alternating internal spurs, as well as compass-drawn circles, so that the design looks strikingly Eskimo in character. 5 The double line with alternating internal spurs has a wide distribution all over northern Eurasia, from the Lapps to the Chukchi. I t appears to play the same rôle in Lappish and Siberian art t h a t it does in t h a t of the Eskimo, the only difference being t h a t in some cases, under the influence of the gouged triangle ornamentation so popular in Siberia, the spurs are enlarged until they almost become little triangles. However, even though we may be able to prove a very early Eurasian origin for this Eskimo design, we can not claim for it a Palaeolithic beginning. 1

As on an incised figure of a seal, upper Magdalenian, Laugerie-Basse (Breuil, 1913, fig. 29, 6). Müller, 1918, fig. 55. 3 H . Rydh, " O n Symbolism in M o r t u a r y Ceramics," Bulletin I, The Museum, of Far Eastern Antiquities, Stockholm, 1929, pl. I l l , 3. ' Solberg, fig. 190. 5 H . Stolpe, " S u r les découvertes faites dans l'île de Björkö," Congrès International d'Anthropologie et d'Archéologie Préhistoriques, Stockholm, 187II, Stockholm, 1876, fig. 12. 2

ESKIMO AND PALAEOLITHIC ART

97

Zig-zags and Chevrons Chevrons or Vs a n d zig-zag lines play a large rôle in Palaeolithic a r t , b u t t h e y a r e of little i m p o r t a n c e in E s k i m o decoration. I n t h e Palaeolithic, chevrons m a y be incised in longitudinal rows, one a b o v e t h e other, or side b y side, m a k i n g broken zig-zag lines; or t h e y m a y be combined in b o t h ways a t once, producing all-over p a t t e r n s . T h e various short strokes m a y n o t be connected, a n d this is typical of t h e s k e t c h y c h a r a c t e r of Palaeolithic a r t . Longitudinal lines of Vs are very c o m m o n (Plate X X I V B, 4). O n e of t h e bear t e e t h f r o m Sordes is decorated with t h r e e b a n d s of Vs, enclosed in double lines. 1 Chevrons, with or w i t h o u t t h e bordering lines, are used t o indicate t h e t e e t h on t h e c u t - o u t horses' heads, typical of t h e middle Magdalenian. 2 T h e chevron is very r a r e in E s k i m o a r t . I t occurs on t h e stone p e n d a n t f r o m N a u j a n (Plate X X I A, 5), on t h e archaeological c o m b f r o m P o i n t Atkinson (Plate X X I I A, 10), a n d occasionally appears in m o d e r n Alaskan art. 3 Closely related t o t h e chevron in Palaeolithic a r t is t h e zig-zag line. Single zigzags are a c o m m o n longitudinal decoration (Plate X X I V B, 1), a n d m a y also a p p e a r enclosed between two bordering lines (Plate X X I V B , 6). Like t h e chevron, t h e zig-zag m a y indicate t h e t e e t h on t h e horses' heads of t h e middle Magdalenian. 4 T h e simple zig-zag is extremely r a r e in E s k i m o a r t , a n d to m y knowledge is n o t f o u n d on archaeological specimèns. 6 T h e zig-zag between two bordering lines is c o m m o n , however, on needlecases a n d bone i m p l e m e n t s for m a r k i n g skins f r o m K i n g William Land. 6 I n t h e Palaeolithic, double, triple a n d multiple zig-zag lines a n d b a n d s are comm o n ( P l a t e X X I V B, 5). T w o lines of zig-zags m a y combine to f o r m lozenges ( P l a t e X X I V B, 2). N e s t s of chevrons, placed side by side, combine to f o r m b a n d s ( P l a t e X X I V B, S). F r o m t h e Solutrean station of P r e d m o s t we h a v e several examples of objects covered b y a n all over p a t t e r n of zig-zags a n d chevrons, f o r m e d b y disconnected lines. 7 T h e r e are few E s k i m o examples c o m p a r a b l e to these. T h e double zig-zag line between borders is f o u n d in t h e P u n u k c u l t u r e ( P l a t e X X I V B, 7), b u t t h e border lines c u t off t h e edges of t h e zig-zag, so t o speak, so t h a t only t h e lines on t h e insides of t h e angles m e e t . A m o d e r n needlecase f r o m K i n g William L a n d ( P l a t e X X I V B , 8) is decorated with a c o m b i n a t i o n of spurs a n d zig-zags, very similar to analogous decorations in Palaeolithic a r t (Plate X X I V B, 9). On t w o archaeological needlecases, one f r o m P o i n t Harrison, L a b r a d o r (Plate X X I V A, 6), a n d t h e o t h e r in t h e Van Valin collection f r o m P o i n t Barrow, 8 there are encircling lines a b o u t t h e t o p a n d b o t t o m of t h e tubes, with triangles, or zig-zag lines on t h e m . T h e row of triangles on a line also occurs on a border a b o u t t h e hole in t h e h a n d l e of a n archaeological c o m b f r o m P o i n t Hope. 9 i Lartet and D u p a r t , fig. 35. = Piette, pl. X, 4, and fig. 52. Hoffman, pi. 34, 2. 4 Piette, fig. 52. ' Except from Kaehemak Bay, Cook Inlet, where t h e y appear to be derived f r o m an Indian source e (Southern British Columbia). In National Museum, Oslo, 15699, 15708, 15703. ! H. Breuil, " Notes d'un voyage paléolithique en Europe Centrale," p a r t II, L'Anthropologie, X X X I V , 1924, fig. 16, 3, 9, 10, and fig. 22. « Mason, 1930, fig. 1, b. Mathiassen, 1930, pi. 16, 8. 3

FREDERICA DE LAGUNA

98

Hatched

Figures

These are not very common in Palaeolithic art, and are confined, as far as I know, to the Solutrean and first half of the Magdalenian. Some of the mammoth ribs from Predmost are decorated with triangular areas filled with hatching in one direction only. 1 Both simple (Plate X X I V C, 3, 6, 9) and cross hatching (Plate X X I V C, 7 and 8) appear in the Magdalenian. However, only in one of the examples cited (Plate X X I V C, 6) does the shape of the hatched figure correspond to those found in Eskimo art. Besides the double line with alternate hatched and plain areas, hatched figures are found in Eskimo art, though they are not common. Sometimes the space between the prongs of a Y is filled with hatching, as on the Thule harpoon heads,2 or on the hook in the Van Valin collection from Point Barrow (Plate X X I A, 4). In the Old Bering Sea art, small triangles, attached to circles, and other small areas (Plate X I X A, 1) were filled with hatching, usually simple.3 Triangles on a needlecase in the Van Valin collection are filled with single hatching. 4 As a modern example we might cite the bands encircling the snuff tubes and needlecases from Alaska (Plate X X I V C, 4). A few more figures, common to both Eskimo and Palaeolithic art, might perhaps be found, but this survey has exhausted all the more important decorative elements. CONCLUSION

I n the field of representative art, whatever similarities we have been able to cite between that of the Eskimo and the Upper Palaeolithic were due only to a sketchy naturalism, or lack of style; and whenever conventionalization occurred, the styles of the conventionalization were different. The question remains how we are to interpret the similarities and differences in non-representative art. I f we set aside all the decorative elements, which we found to be post-Palaeolithic in origin, or which are precluded from the possibility of a relationship, by restriction to a limited locality or a particular period, or those whose occurrence in the art of either people is so rare as to be fortuitous, we have only a small list of decorative elements left. These are: the dot, spurred circle, the line with spurs on one side, the line with oblique, paired spurs on both sides (but not all of its abbreviations) and the " l a d d e r " pattern. I t is questionable whether the possession of these few decorative elements is enough to prove a closer degree of relationship between Eskimo and Upper Palaeolithic art than might be shown to exist between the latter and the art of almost any people who incise comparatively simple figures. When I began this study I expected to find that the repertory of elements common to Upper Palaeolithic and Eskimo art would be much greater, and that it would be possible to show that these elements had a limited distribution, possibly within the circumpolar area, which would in itself strengthen the theory of a common origin. However, after attempting to trace the distribution of these elements, I was forced to abandon the over' Breuil, 1924, fig. 21. * M a t h i a s s e n , 1927, I , pi. 1, 4. 3 E x a m p l e s of single hatching (Collins, 1929, pi. 1, d, and pi. 6. a ) ; of cross hatching (op. cit., fig. 1). 4 M a s o n , 1930, fig. 1, a.

ESKIMO AND PALAEOLITHIC ART

99

ambitious attempt, for they seem to occur sporadically in every part of the world. Thus, the conclusion has forced itself upon me that the mere possession in common of simple decorative motifs is not sufficient to prove a Palaeolithic origin for Eskimo art. The survey which we have just made included practically every element in the art of the Thule culture and of the modern Eskimo. I t seemed easier to take Eskimo art as a starting point, because its motifs are not only limited in number, but have precise and well-established forms. I t will be remembered that the dotand-circle, the double line with alternating internal spurs, and most of the forms of the Y and " t r e e " figures,—the three most important elements in Eskimo art,—were either entirely lacking in the Upper Palaeolithic, or, as in the case of the last motif were so rare that they must be disregarded. However, our survey has far from exhausted the repertory of the Upper Palaeolithic. We have mentioned only a few of the many developments of the V and the zig-zag, of the short crossing lines, etc.—motifs composed of simple combinations of straight lines, which might easily have found a place in Eskimo art. Some of the Palaeolithic patterns are of such wide distribution that their absence from Eskimo art is all the more striking. For example, one of these patterns is a band of triangles fitted together, with hatching running in alternate directions (Plate X X I I I A, 1). This is not only characteristic of the Palaeolithic, but appears in the Neolithic of various parts of Europe including Scandinavia, in the Bronze and Iron Ages, among modern northern tribes such as the Lapps, in southern Asia, in Pre-Columbian America, including the Northwest Coast of North America, and yet it has never, to my knowledge, appeared in Eskimo art. Besides decorative motifs such as these, Palaeolithic art is very rich in curved lines and spirals, in combinations of incised figures and areas in low relief, in decorative elements suggestive of highly conventionalized zoomorphic forms. All of these are very characteristic of the Palaeolithic, yet are utterly foreign to Eskimo art. Although so far all the evidence seems to be against a Palaeolithic basis for Eskimo art, it may well be argued that we should not expect too great a similarity. We should expect Eskimo art to have lost some of the Palaeolithic elements and to have adopted others. However, the degree to which this must have taken place is very great and presents a difficulty which the theory of a relationship must explain. Eskimo art contains such a small part of the whole Palaeolithic repertory, and that part is of such limited importance, that the difference between the two arts is very significant. It remains to be investigated whether both arts have the same underlving principles of style, that is, the same principles of combining decorative motifs and applying them to objects. We are struck, even from the first, however, by very great differences. Palaeolithic style is fluid: the creation and arrangement of motifs appears to be a matter of individual taste, and we often seem to catch a particular pattern in the process of formation. The cases in which the decoration appears to be imperfectly achieved and but poorly adapted to the shape of the object, are due, I feel, to the fact that the artist has not really thought out his design. He has not thoroughly mastered his technique; he is learning through experimentation. Tchek-

100

F R E D E R I C A DE LAGUNA

alenko, in his study of the art of the Aurignacian station of Mezine, Russia, has shown how this process has been active in the formation of meanders out of combinations of chevrons. I believe, however, that he has overemphasized the part played by blind trial and error and by mechanical techniques. Eskimo art, however, is different. One feels that the patterns and the principles governing their arrangement, and even the choice of objects to which they are applied, were developed and fixed long ago. The clumsiness which we so often encounter is the result only of slovenly workmanship. In some cases the aesthetic appreciation of the design has atrophied to such an extent that it is apparently no longer necessary to execute the design neatly; it is enough if it is incised in the conventional manner. This, however, is an extreme case of the degeneration which has taken place in Eskimo art. The chief differences between Eskimo and Palaeolithic art are those between age and youth, and are perhaps what we should have expected, were Eskimo art derived from that of the Palaeolithic. There are other differences in style, however. The greater part of the Eskimo decorative elements are border patterns, and are most often employed to outline the edge or a part of the object. Even the dot-and-circle and the dot, which from their nature might have achieved a greater degree of independence from this general principle, very often form borders, as we have shown. The Y and " t r e e " figures, also, are utilized as spurs on a line and so are made part of a border. The aesthetic principles underlying the emphasis of the edge of an object by means of a border are so simple and appear to be so universal that we are astonished at the rarity of the border in Palaeolithic art. Among the few examples we can cite, the pendant from Saint-Marcel (Plate X V I I I B , 2) has a line around the edge. The holes in the " batons de commandement" of the final Magdalenian were enclosed in borders of straight lines,1 or other simple elements (Plate X X I I I B , 1). Partial borders are found on the decorated bone plates from the middle Magdalenian layers of Marsoulas (Plate X X I V C, 5 to 7). On two of these, there is a border separating the decorated portion from the undecorated area, while the spurs along the edge suggest that it, also, is to be seen as a border. Another of these plates (Plate X X I V C, 5), with the reentering angle decoration, suggests the handle of some of the Thule combs (Plates X X B , 4 and 8, and X X I V A, 3). In fact, this similarity is so striking, that were the example from Marsoulas not unique in Palaeolithic art, its resemblance to the Eskimo combs might be argued as significant. Closely related to the border around a flat object is the encircling line about a cylindrical one. As we have seen, the encircling line is very common in Eskimo art, particularly in the decoration of needlecases. In spite of the great number of cylindrical objects in the Palaeolithic, this type of composition appears but infrequently. We find it, however, in its simplest form on the bird bone needlecases of the upper Solutrean of Le Placard, in the form of short lines or notches (Plate X X I V C, 2). 2 The same type of decoration, in a slightly more complex form, apBreuil, 1913, fig. 39. The specimen figured is polished smooth on the side with the hole, probably where it rubbed against the clothing. I t is lightly smeared with red ochre. Similar needlecases are: The National Museum, St. Germain, 54967, 54931, 48210, and 47429; Lartet and Christy, pi. B X V I I , 1; and Breuil, 1925, fig. 4, 8. Some of these cases are closed at the lower end. Some Eskimo cases are plugged at the bottom. 1

2

E S K I M O AND PALAEOLITHIC A R T

101

pears on an Aurignacian paint-tube, which has a border consisting of a line of Xs, a line crossed by spurs, and a line of larger X s (Plate X I X B , 10). Bone needlecases, very similar to those of Le Placard, have been found among the Eskimo. From Point Atkinson, Mathiassen reports an archaeological "cylindrical bone tube, 7.6 cm. long, 1.0 cm. diameter, . . . decorated with five pairs of rings (encircling lines?) and three longitudinal rows of short transverse lines," which might have served as a needlecase.1 From King William Land he figures a needlecase made of an animal leg bone with three encircling lines at the top and at the bottom. 2 Two modern needlecases from the same locality have a decoration of partially encircling lines (Plate X X I V C, l). 3 These last two specimens bear a further resemblance to those of the Palaeolithic because they have a suspension hole at the top. In the Thule culture, Mathiassen mentions tubes of swan's bone, which may well have been needlecases, and from the Punuk culture we have bird bone needlecases, on which, however, the decoration is slightly more complex (Plate X I X B , 2). I think it might well be argued that the original needlecase was a natural bone, with a hole at the top for suspension, and that a decoration of encircling lines became associated with it very early in its history. The Alaskan bird bone needlecases, and those of King William Land with suspension hole, would represent the modern survivals of the original form. How the elaborations of the bone case, such as the "winged" or the "flanged" types (or their prototypes), originated, we cannot discover without more archaeological information. The profile of the wooden needlecases of the modern Central Eskimo, as well as the fact that they are plugged at the bottom and suspended by a cord from the top, suggest that they were derived from the caribou leg bone case. Comparatively simple needlecase forms, but little elaborated, are the usual type in Siberia and Lapland. 4 However, it might well be argued that the naturally hollow bone and the decoration of transverse notches and encircling lines are too simple to be of great significance. The very nature of the needlecase prescribes a tubular form, and the hollow bone is the natural, almost inevitable form, to which any people who used the needlecase would turn, whenever, and for whatever reason (degeneration of style, convenience, etc.), the elaborate forms were discarded. Like the naturally hollow stone lamps of the Caribou Eskimo, it may just as plausibly be called degenerate as primitive. 8 Besides the hollow bone tube, the Palaeolithic offers a few other examples of encircling decorations. Pendants from Kesslerloch (Plate X I X B , 6) and Mas d'Azil Mathiassen, 1930, p. 16. 2 Mathiassen, 1927, I, pi. 8,3, 18. And a specimen of caribou leg bone (The National Museum, Oslo, 1568). 4 Thalbitzer reports the needlecase among the Lapps, Samoyed, Tungus, Gilyak, and Ainu, and argues that it is probably a "cultural relic of great antiquity" (W. Thalbitzer, "Parallels within the Culture of the Arctic Peoples," Animes do XX Congresso Internacional de Americanistas, Rio de Janeiro, 1924, p. 284). O. Menghin, Weltgeschichte der Steinzeit, Vienna, 1931, pi. X X I , 17, figures a bird bone needlecase with bone needles in it from a Neolithic site on the banks of the Angara River, Irkutsk, Siberia. I see no reason, in view of this wide distribution, and of the discovery of the needlecase in Palaeolithic deposits, contemporaneous with the first sewing needles, for the assertion that the needle-cushion is older than the needlecase (K. Birket-Smith, " T h e Caribou Eskimos," Report of the Fifth Thvle Expedition 1921-2Í, V, Copenhagen, 1929, II, pi. 199). 5 T. Mathiassen, " T h e Question of the Origin of Eskimo Culture," American Anthropologist, N.S. X X X I I , 1930, pp. 596-598. 1 3

102

FREDERICA DE LAGUNA

(Plate X I X B , 3) are encircled with lines with spurs; a "báton de commandement" from Mas d'Azil, (Plate X X I I B , 3), and an ivory cylinder from Brassempouy (Plate X X I I I A, 8) by the " l a d d e r " pattern. We have already commented on the Eskimo-like ornamentation on the pendant from Mas d'Azil; it might be compared to a Punuk culture needlecase from Cape Kialegak, St. Lawrence Island (Plate X X I V A, 2). However, most cylindrical objects in the Palaeolithic are decorated as if they were long, flat surfaces, and they must be seen from a certain angle in order for the decoration to appear symmetrical. The most common form of decoration, as we have pointed out, is the simple longitudinal arrangement. T h e type of decoration of the Palaeolithic shafts and spear points indicates that these objects are supposed to be viewed from one end or the other. The development of end-to-end symmetry, and of the right-to-left symmetry in which one end is different from the other, indicate this very clearly. A class of Eskimo objects with a consistently longitudinal arrangement are the drill bows and handles, yet even here we find that the object is not to be viewed from the end. The bow and the handle are held in a horizontal position, with the eye above the middle, as is reflected by the realistic decorations. Both halves may be symmetrical with respect to each other, but this is not an endto-end symmetry, like that of the Palaeolithic, but a right-to-left symmetry of a particular kind. There are, of course, exceptions to this generalization. In Palaeolithic art it is permissible to take units which are not really the same, but which produce a similar effect. T h e line with spurs on both sides may be matched with the line crossed by spurs (Plate X I X B, 7), or a line of Vs matched by a zigzag line (Plate X X A, 7). Even when the units of the pattern are really identical, an asymmetry may result because the spurs on the longitudinal lines are all on the same side, or those on the middle line are on one side only (Plate X X A, 7). T h e typical Eskimo arrangement produces a perfectly symmetrical pattern, of which the modern Alaskan thimble-holder is a good example (Plate X X I A, 7). Cases of poorly balanced patterns occur also in Eskimo art (Plate X X I I I B , 6), but they seem to be due to degeneration of technique, and are not, like the Palaeolithic examples, the result of a somewhat sophisticated stylistic preference. T h e art of the Thule culture or that of the modern Alaskan Eskimo does not show as highly developed a spatial sense as that of the Old Bering Sea culture, which was distinguished for its sophisticated use of balanced asymmetry (perfect symmetry is also common), and for the way in which the pattern fills the entire space, enhancing the aesthetic value of the shape of the object and of its parts, without recourse to the simple expedient of outlining. Modern Eskimo art, on the other hand, can fill space only by the clumsy method of setting one border within another. In Palaeolithic art, on the whole (with the exception of the semi-cylindrical objects described by Saint-Périer, ornamented with scrolls in low relief), the spatial sense is even less developed. Cylindrical objects, with few exceptions, are treated as if they were flat. Very rarely is their roundness emphasized, and on most objects the longitudinal direction alone is heightened by the decoration; the shape of the object as a whole is ignored. These considerations seem to show that a fundamental difference separates the

E S K I M O

A N D

P A L A E O L I T H I C

103

A R T

art of the Eskimo from that of Upper Palaeolithic Europe. On the basis of the material at present available it is impossible to prove that Eskimo art is more closely related to that of the Palaeolithic than are other arts of comparatively simple content. Yet I am not sure that this negative conclusion is final. The differences in style upon which we have laid so much emphasis, may be only those which we should expect to find between an old and a young art; they are not differences in fundamental principles. Indeed, we may be demanding too great a uniformity and stability of tradition over such an enormous lapse of time. In view of the great changes in Alaskan art style from the Old Bering Sea period into modern times, is it surprising that Eskimo and Palaeolithic art have so little in common? If we knew the whole history of Eskimo art and of its pre-Eskimo origins, we might see how it grew from a Palaeolithic beginning and how the original Palaeolithic traits have been sloughed off, one by one. The archaeology of Siberia still holds the key to this mystery. A simple comparison of Eskimo and European Palaeolithic art, to which our present inadequate knowledge limits us, does not reveal, I think, similarity in style sufficiently striking to prove a Palaeolithic origin of Eskimo art. I doubt that even for the art of the "ice-hunting" cultures, we can prove a Palaeolithic origin more intimate and direct than for the art of any primitive group. Perhaps a study of all the possible similarities in cultural traits, as suggested in Part I, may reverse this opinion, or future archaeological discoveries in Siberia and Alaska bring new material and a clearer understanding. FREDERIOA UNIVERSITY MUSEUM,

DE

LAGUNA

PHILADELPHIA

LIST OF ILLUSTRATIONS * PLATE

XVIII

A. T H E DOT

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 1,5. 16. 17,

Solutrean rib, Grotte du Pape, Brassempouy, France. Scale After Piette, pi. L X X X I , 3. Magdalenian ivory plate, Wierzchovie, Poland. Scale Sollas, 1924. fig. 301, 2. Lower Magdalenian rib, Saint-Michel, Arudy, France. Scale K - After Piette, pi. L X X X Y I I I , 3. Ivory pendant, Kulna, Moravia. Scale % . Sollas, 1924, fig. 30S, B . Archaeological ivory pendant, Inugsuk, Upernivik District, West Greenland. Scale % . After Mathiassen, M . o . G . 1930, pi. 18, 4. Thüle culture, drop pendant for hair ornament like Figure 9, Southampton Island, Canada. Scale J ^ . After Mathiassen, 1927, I, pi. 76, 10. Magdalenian pendant, coal, Kesslerloch, Switzerland. Scale }/y. After Merk, fig. 85. Maglemosian amber pendant, Denmark. Scale }•, After Miiller, 1918, fig. 37. Thüle culture hair ornament for woman, ivory, Southampton Island, Canada. Scale After Mathiassen, 1927, I, pi. 76, 6. Thüle culture ivory thimble-holder, Comer's Midden, Thüle, Northwest Greenland. Scale % . After Mathiassen, 1927, I, pi. 78, 12. Archaeological ivory baot-sole creaser, West Greenland. Scale N a t . Mus., Copenhagen, L 8074. Archaeological wrist guard, Alaska. Scale l ^ . N a t . Mus., Copenhagen, P 365. Archaeological ivory needlecase, Southampton Island, Canada. Scale After Matthiassen, 1927, I, pi. 76, 5. Modern needlecase, Smith Sound, Northwest Greenland. Reduced. Am. Mus. of Natural History, N . Y . Modern Aleut ivory pendant, Alaska. Scale > { . R i k s Mus., Stockholm, 06. 21. 21. Archaeological miniature ivory needlecase, Ponds Inlet, northern Baffinland, Canada. Scale After Mathiassen, 1927, I, pi. 66, 1. 18, 19. Modern Alaskan motifs. * T h e figures have been slightly reduced from the scale indicated.

FREDERICA DE LAGUNA

104

B . T H E D O T - A N D - C I R C L E AND S P U R R E D C I R C L E

1. Lower M a g d a l e n i a n serai-cylindrical object, antler, cave of Les Espélugues, A r u d y . Scale Yl- A f t e r Piette, pl. L X X X V , 7. 2. M a g d a l e n i a n bone p e n d a n t , Saint-Marcel, I n d r e , T r a n c e . Scale } {. lìreuil, 1902, fig. 4. 3. " B â t o n de commandementlower Magdalenian, G o u r d a n , F r a n c e ; a n d detail of reverse. Scale YAfter M a c C u r d y , I I , fig. 353. 4. T h ü l e culture ivory b u t t o n , K u k , S o u t h a m p t o n Island, C a n a d a . Scale A f t e r Mathiassen, 1937, I, pi. 72, 17. 5. Magdalenian schematizations of eye a n d horn. R e d u c e d . A f t e r Hoernes, 1925, p. 149. 6. Archaeological comb, E a s t Cape, Siberia. Scale Y - N a t . Mus., Copenhagen, P 25, 62. 7. Lower M a g d a l e n i a n bone blade, G r o t t e des Fées, M a r c a m p s , Gironde, France. Scale Breuil, 1913, fig. 25, 6. 8. Archaeological " w i n g e d " needlecase, with compass-drawn dot-and-circles, West Greenland. Scale YN a t . Mus., Copenhagen. 9. Archaeological comb, P o i n t Barrow, Alaska. Scale Y . N a t . Mus., Copenhagen, P 15, 486. 10. Archaeological slate disk, P o i n t H o p e , Alaska. Scale YN a t . Mus., Copenhagen, P 16, 1561. PLATE

XIX

A . T H E D O T - A N D - C I R C L E AND S P U R R E D C I R C L E

1. Old Bering Sea culture, whaling charm?, P o i n t H o p e , Alaska. Scale Yl- University Mus., Philadelphia. 2. Old Bering Sea culture, h a r p o o n h e a d , Alaska. Scale Yl- N a t . Mus., Oslo, 30122. 3. D e t a i l f r o m object similar t o Figure 1, Old Bering Sea culture, E a s t Cape, Siberia. Scale Yl- N a t . Mus., Copenhagen, P 350. 4. Detail, m o d e r n b u c k e t handle, mechanically d r a w n dot-and-circles, N o r t o n Sound, Alaska. A f t e r H o f f m a n , pi. 38, 1. 5. D e t a i l , m o d e r n h a t o r n a m e n t , P r e m o r s k a , Y u k o n River, Alaska. A f t e r H o f f m a n , pi. 54, 2. 6. P u n u k culture drill rest, mechanically d r a w n dot-and-circles, P u n u k Island, S t . Lawrence Island, Alaska. Scale Y - U. S. N a t . Mus., 343427. B.

T H E L I N E C R O S S E D BY S P U R S AND T H E L I N E W I T H S P U R S ON O N E S I D E

1. Detail, P u n u k culture ivory plate, P u n u k Island, S t . Lawrence Island, Alaska. Scale Y - U. S. N a t . M u s . , 343502. 2. P u n u k culture, bird b o n e needlecase, P u n u k Island, Alaska. Scale Yi- After Collins, 1929, pi. 17, c. 3. M i d d l e M a g d a l e n i a n ivory p e n d a n t , M a s d'Azil, F r a n c e . Scale YA f t e r Piette, (ig. 72. 4. M o d e r n bird bone needlecase or snuff tube, Alaska. After H o f f m a n , pi. 52, 5. 5. T h ü l e culture ivory bead, N a u j a n , Melville Peninsula, C a n a d a . Scale Y . A f t e r M a t h i a s s e n , 1927,1, pi. 31, 8. 6. M a g d a l e n i a n p e n d a n t , Kesslerloch, Switzerland. R e d u c e d . A f t e r M a c C u r d y , I, fig. 128, 12. 7. U p p e r M a g d a l e n i a n semi-cylindrical implement, antler, M a s d'Azil. Scale YAfter Piette, pl. X C V I I , 15. 8. B o n e point, G r o t t e de Gouërris, Lespugue, F r a n c e . Scale }4. Saint-Périer, 1927, fig. 5. 9. T h ü l e culture toy k a y a k , wood, B u t t o n Point, n o r t h e r n Baffinland, C a n a d a . Scale YtA f t e r Mathiassen, 1927, I, pi. 62, 11. " . 10. Aurignacian p a i n t t u b e , reindeer leg bone, cave of Les Cottés, Vienne, France. Scale Y . M a c C u r d y , I, fig. 143. 11. T h ü l e culture doll, wood, B u t t o n Point, n o r t h e r n BafBnland, C a n a d a . Scale Yi- A f t e r M a t h i a s s e n , 1927, I, pi. 62, 7. PLATE

X X

T H E L I N E W I T H S P U R S ON O N E S I D E A

1. Detail, M a g d a l e n i a n antler implement, Espélugues, Lourdes. A f t e r P i e t t e , pl. X X I I I , 1. 2. Lower M a g d a l e n i a n throwing stick, Saint-Michel, A r u d v . Scale YCf. Pl. X X I V B, 1. A f t e r Piette pi L X X X 1 X , 6. 3. M a g d a l e n i a n spear point?, Kesslerloch, Switzerland. Scale Yl- A f t e r M e r k , fig. 92. 4. M a g d a l e n i a n semi-cylindrical object, Espélugues, Lourdes. Scale Yl- A f t e r Piette, pl. X X X V I I , 6. 5. Upper M a g d a l e n i a n semi-cylindrical implement, Bruniquel, F r a n c e . Scale A f t e r Cartailhac, 1903, fig. 90. 6. M i d d l e M a g d a l e n i a n semi-cylindrical implement, Espélugues, Lourdes. Scale i f . A f t e r Piette, pl. X X I . 6. 7. M a g d a l e n i a n b o n e plate, Laugerie-Basse, F r a n c e . Scale YReinach, p. 117, fig. 2. 8. M a d g a l e n i a n spear point, Kesslerloch, Switzerland. Scale } 3 '. A f t e r M e r k , fig, 96. 9. M o d e r n flanged needlecase, N o r t o n Sound, Alaska. Scale YBoas, 1908, pl. X X I V , 2. 10.-Archaeological h a r p o o n h e a d , P o i n t Barrow, Alaska. Scale YN a t . Mus., Copenhagen, P 15, 169. 11. Archaeological h a r p o o n head, Point H o p e , Alaska. Scale Yl- N a t . M u s . , Copenhagen, P 16, 810.

ESKIMO AND PALAEOLITHIC ART

105

B 1. L o w e r M a g d a l e n i a n d a r t h e a d , L e P l a c a r d , F r a n c e . S c a l e B r e u i l , 1 9 1 3 , fig. 18, 9 . 2. Archaeological ulo ( w o m a n ' s knife) handle, ivory, P o i n t B a r r o w , A l a s k a . S c a l e N a t . Mus., Copenhagen, P 15, 178. 3 . M a g d a l e n i a n b l u n t a n t l e r p o i n t , Espélugues, L o u r d e s . S c a l e } - f . A f t e r P i e t t e , pi. X X X V I I , 4. 4 . T h ü l e c u l t u r e c o m b , N a u j a n , M e l v i l l e Peninsula, C a n a d a . S c a l e A f t e r M a t h i a s s e n , 1927, I , pi. 29, 2. 5. N e o l i t h i c b o n e p l a q u e , D e n m a r k . Scale M ü l l e r , 1895, fig. 2 4 8 . 6 . Archaeological c o m b , B a r t e r I s l a n d , A l a s k a . S c a l e ^-f. N a t . M u s . , Copenhagen, P 13, 2 7 4 . 7 . Archaeological c o m b , P o i n t H o p e , A l a s k a . S c a l e A f t e r M a t h i a s s e n , 1930, pi. 16, 5 . 8 . Archaeological c o m b , S o u t h a m p t o n I s l a n d , C a n a d a . S c a l e Yò- M a t h i a s s e n , 1927, I , fig. 8 9 . 9 . M o t i f s from a m b e r p e n d a n t , M e g a l i t h i c S t o n e Age, D e n m a r k . S c a l e A f t e r M ü l l e r , 1 9 1 8 , fig. 4 8 . P L A T E

X X I

T H E L I N E W I T H P A I R E D S P U R S ON B O T H S I D E S A

1. T h ü l e c u l t u r e h a r p o o n head, from a grave, N a u j a n , Melville Peninsula, C a n a d a . S c a l e After Mathiassen, 1 9 2 7 , I , pi. 3 7 , 9 . 2. F i n a l M a g d a l e n i a n b a r b e d a n t l e r point, L a M a d e l e i n e , F r a n c e . S c a l e A f t e r L a r t e t a n d C h r i s t y , pi. B X X X I X ,

3.

3 . T h ü l e c u l t u r e harpoon head, S o u t h a m p t o n I s l a n d , C a n a d a . S c a l e N a t . M u s . , Copenhagen, P 11, 5 . 4 . E d g e of archaeological ivory h o o k , P o i n t B a r r o w , A l a s k a . S c a l e U n i v e r s i t y M u s . , V a n Valin Coll. 5 . T h ü l e c u l t u r e slate p e n d a n t , N a u j a n , M e l v i l l e Peninsula, C a n a d a . S c a l e 3a- A f t e r M a t h i a s s e n , 1927, I , pi. 31, 2 . 6. F i n a l M a g d a l e n i a n b o n e i m p l e m e n t , L o r t h e t , F r a n c e . S c a l e 3-f • B r e u i l , 1 9 1 3 , fig. 4 0 , 10. 7 . M o d e r n thimble-holder, A l a s k a . S c a l e A f t e r Nelson, pi. X L I V , IG. B

1. 2. 3. 4. 5. «. 7. 8.

Upper M a g d a l e n i a n bird b o n e needlecase?, Espélugues, L o u r d e s . S c a l e K . A f t e r P i e t t e , pi. X X V , 4 . M a g d a l e n i a n b o n e disk, B r u n i q u e l , F r a n c e . R e i n a c h , p. 4 0 , fig. 9 . F r a g m e n t of M a g d a l e n i a n b o n e disk, K e s s l e r l o c h , Switzerland. S c a l e j . A f t e r M e r k , fig. 7 7 . Archaeological sinew-twister, P o i n t H o p e , A l a s k a . S c a l e A f t e r M a t h i a s s e n , 1930, pi. 15, 4. M o d e r n buckle, P i t l i k a j C h u k c h i , S i b e r i a . S c a l e J ^ . R i k s M u s . , S t o c k h o l m , Vega 4 9 3 5 . M o d e r n b a g - h a n d l e , A l a s k a . M u c h reduced. x\fter H o f f m a n , pi. 3 1 , 1. F i n a l M a g d a l e n i a n b o n e rod, G o u r d a n , F r a n c e . S c a l e } B r e u i l , 1913, fig. 40, 9. D e t a i l from lower M a g d a l e n i a n ivory c a r v i n g of g o a t s in b a s relief, M a s d'Azil, F r a n c e . S c a l e 1 { . After P i e t t e , pi. X L I V . 9 . Archaeological needlecase, E a s t C a p e , S i b e r i a . S c a l e ^-f. N a t . M u s . , C o p e n h a g e n , P 25, 5 8 . 10. D e t a i l , modern b a g - h a n d l e , N o r t o n Sound, A l a s k a . R e d u c e d . A f t e r H o f f m a n , pi. 3 9 , 11. M a g d a l e n i a n b o n e blade, L a u g e r i e - B a s s e , F r a n c e . S c a l e 1 { . P a r k y n , fig. 9 0 , 4.

P L A T E

X X I I

" A R R O W , " " T R E E , " AND " Y "

FIGURES

A

1. D e t a i l , arrows in t h e side of a bison, M a g d a l e n i a n painting, c a v e of N i a u x , Ariège, F r a n c e . M u c h reduced. A f t e r Sollas, fig. 1 9 5 . 2 . D e t a i l s of decorations on upper M a g d a l e n i a n b e a r t e e t h , c a v e of D u r u t h y , Sordes, F r a n c e . S c a l e J-f. After L a r t e t a n d D u p a r c , figs. 3 4 (5, 10, 11) a n d 3 8 . 3 . U p p e r M a g d a l e n i a n b e a r t o o t h , c a v e of D u r u t h y , Sordes, F r a n c e . A f t e r L a r t e t a n d D u p a r c , fig. 3 6 . 4. Aurignacian human tooth, L a Combe, Dordogne, I'rance. Scale A f t e r M a c C u r d y , I , fig. 72. 5 . M o d e r n Aleut p e n d a n t , A l a s k a . S c a l e } 4 - R i k s M u s . , S t o c k h o l m , 02, 2 1 , 21. 6 . U p p e r M a g d a l e n i a n semi-cylindrical a n t l e r implement, B r u n i q u e l , F r a n c e . S c a l e % . A f t e r C a r t a i l h a c , fig. 9 8 . 7. U p p e r M a g d a l e n i a n b o n e p o i n t , B r u n i q u e l , F r a n c e . S c a l e % . A f t e r C a r t a i l h a c , fig. 5 8 . 8 . M o d e r n i v o r y hair o r n a m e n t , A g i a k c h u g u m u t , south of N e l s o n Island, A l a s k a . S c a l e Collins, 1 9 2 9 , pi. 18, b . 9 . D e t a i l , T h ü l e culture arrow h e a d , N a u j a n , Melville Peninsula, C a n a d a . S c a l e 3>f. N a t . M u s . , C o p e n h a g e n , P 1, 1098. 10. Archaeological c o m b , antler, P o i n t A t k i n s o n , C a n a d a . S c a l e N a t . M u s . , C o p e n h a g e n , P 14, 2 6 1 . 11. D e t a i l , modern i v o r y snuff box, H o t h a m I n l e t , A l a s k a . R e d u c e d . A f t e r H o f f m a n , pi. 65, 3 . 12. D e t a i l , modern b o n e knife, N o r t o n Sound, A l a s k a . R e d u c e d . A f t e r H o f f m a n , pi. 15, 3.

FREDERICA DE LAGUNA

106

B 1. Thüle culture harpoon head, Naujan, Melville Peninsula, Canada. Scale After Mathiassen, 1927, I, pi. 1, 11. 2. Thüle culture harpoon head, Naujan, Melville Peninsula, Canada. Scale After Mathiassen. 1927, I, pi. 1, 9. 3. Punuk culture harpoon head, Cape Kialegak, St. Lawrence Island, Alaska. Scale M> TT- s. N a t . Mus. 4. Thüle culture comb, Kuk, Southampton Island, Canada. Scale N a t . Mus., Copenhagen, P 4, 29. .5. Thüle culture ornamental ivory plate, Belcher Island, east coast of Hudson 13ay, Canada. Scale }/{. N a t . Mus., Copenhagen, P 22, 7. 6. Lower Magdalenian weapon point, Mas d'Azil, France. Scale Breuil, 1913, fig. 16, 6. 7. Modern needlecase, caribou leg bone, King William Land, Canada. Scale V2. After Boas 1907, fig. 259, 2. 8. Detail, lower Magdalenian antler chisel, Le Placard, France. Scale }'{. N a t . Mus., St. Germain, 55065. 9. Detail, Magdalenian bear tooth, cave of Duruthy, Sordes, France. After Lartet and Duparc, fig. 34, 8. 10. Modern quiver handle, King William Land, Canada. Much reduced. N a t . Mus., Oslo, Amundsen Coll. PLATE

XXIII

THE "LADDER"

PATTERN

A

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.

Magdalenian double-ended chisel, Saint-Marcel, Indre, France. Scale Breuil, 1902, fig. 5. Magdalenian antler shaft, Espclugues, Arudy, France. Scale Vi. After Piette, pl. L X X X V , 1. Detail, Magdalenian antler implement, Espélugues, Lourdes, France. Scale 1 •>'. After Piette, pl. X X I I , 2. Detail, modern bow drill, Point Barrow, Alaska. Reduced. After Hoffman, pi. 37, 5. Magdalenian pendant, Jancovics, Hungary. Scale After Breuil, 1923, fig. 17, 16. Lower Magdalenian ivory pendant?, Gourdan, France. Scale After Piette, pl. VII, 4. Lower Magdalenian ivory plate, Lorthet, France. Scale 1 ( . After Piette, pl. VII, 3. tapper Aurignacian cylindrical object, Grotte du Pape, Brassempouy, France. Scale }[. After Piette, pi. LXXIV, 2. Thüle culture comb, Ponds Inlet, northern Baffinland, Canada. Scale After Mathiassen, 1927,1, pi. 52, 13. Detail, end of modern needlecase or snuff tube, Alaska. Reduced. After Hoffman, pi. 52, 6. Solutrean? cylindrical object, Gourdan, France. Scale / f . After Piette, pl. L X X X I I I , 3. Archaeological comb, East Cape, Siberia. Scale N a t . Mus., Copenhagen, P 25, 61. B

1. Magdalenian "bâton de commandement,'' Kesslerloch, Switzerland. Scale }-{. After Merk, fig. 41. 2. Upper Magdalenian "bâton de commandement," Espélugues, Lourdes, France. Scale M- After Piette, pl. X X I I , 1. After Piette, pl. LV, 7. 3. Upper Magdalenian "bâton de commandement," Mas d'Azil, France. Scale 4. Modern bucket handle, Ishaktolik, Alaska. Much reduced. After Hoffman, pl. 36, 1. 5. Edge of Thüle culture ivory plate, Kuk, Southampton Island, Canada. Scale }-{. N a t . Mus., Copenhagen, P 4, 850. 6. Archaeological thimble-holder, Point Hope, Alaska. Scale H . N a t . Mus., Copenhagen, P 16, 1496. 7. Punuk culture ivory plate, Punuk Island, St. Lawrence Island, Alaska. Scale U. S. N a t . Mus., 343613. P L A T E A.

X X I V

L I N E S WITH ALTERNATING

SPURS

1. 2. 3. 4. 5. 6. 7. 8. 9.

Modern snuff tube, St. Michaels, Norton Sound, Alaska. Scale Trocadero Mus., Paris, 15045. Punuk culture needlecase, Cape Kialegak, St. Lawrence Island, Alaska. Scale J-f- ! S. N a t . Mus., 346719. Thüle culture comb, Button Point, northern Baffinland, Canada. Scale K - N a t . Mus., Copenhagen, P 8, 169. Modern antler bucket handle, Norton Sound, Alaska. Reduced. After Hoffman, pi. 14, 4. Detail, modern bucket handle, Cape Darby, Alaska, Reduced. After Hoffman, pi. 36, 2. Thüle culture needlecase, Point Harrison, Labrador, Canada. Scale J4- N a t . Mus., Copenhagen, P 21, 14. Detail, border of Punuk culture wrist guard, St. Lawrence Island, Alaska. Scale }{. U. S. N a t . Mus. Modern ivory beads, Port Clarence, Alaska. Scale }{. Riks Mus., Stockholm, Vega 5152. Detail middle Aurignacian ivory "peg," Grotte du Pape, Brassempouy, France. Scale K . After Piette, pl. LXXV, 3. 10. Detail, modern bag handle, north of Norton Sound, Alaska. Reduced. After Hoffman, pi. 31, 4. B.

T H E ZIG-ZAG AND

CHEVRON

1. Upper Magdalenian throwing stick, Saint-Michel, Arudy, France. Scale K - Cf. also plate X X A, 2. Piette, pl. L X X X I X , 6. 2. Upper Magdalenian barbed point, Lorthet, France. Scale . After Piette, pl. LX, 11.

After

ESKIMO AND PALAEOLITHIC ART

107

3. Lower M a g d a l e n i a n bird b o n e needlecase, Le P l a c a r d . Scale N a t . Mus., St. G e r m a i n , 55151. 4 a n d 5. Details, M a g d a l e n i a n throwing stick, Saint-Michel, A r u d y , France. Scale }/{. A f t e r P i e t t e , pi. L X X X I X , 2. 6. D e t a i l , middle M a g d a l e n i a n semi-cylindrical implement, Espélugues, Lourdes, F r a n c e . Scale After Piette, pi. X V , 3. T 7. Detail, border of P u n u k culture wrist guard, St. Lawrence Island, Alaska. Scale }••{. l . S. N a t . M u s . 8. Detail, transverse design on modern caribou leg bone needlecase, K i n g William L a n d , C a n a d a . Scale After Boas, 1907, fig. 259, b. 9. Design on edge of final M a g d a l e n i a n " b a t o n de Commandernent" La Madeleine, France. Scale After Breuil, 1913, fig. 39, 3. C.

1. 2. 3. 4. 5. 6. 7. 8. 9.

N E E D L E C A S E S ANO B O N E P L A T E S

M o d e r n bird b o n e needlecase, K i n g William Land, C a n a d a . Scale N a t . Mus., Oslo, 15702. Solutrean bird bone needlecase, Le Placard, F r a n c e . Scale N a t . Mus., Saint-Germain, 54967. M i d d l e Magdalenian f r a g m e n t of antler, L e Placard, France. Breuil, 1913, fig. 18, 13. Detail, m o d e r n needlecase or snuff t u b e , Alaska. A f t e r H o f f m a n , pi. 52, 3. M i d d l e M a g d a l e n i a n b o n e plate, Marsoulas, F r a n c e . Scale Breuil, 1913, fig. 27, 6. M i d d l e Magdalenian bone plate, Marsoulas, F r a n c e . Scale Breuil, 1913, fig. 27, 2. M i d d l e M a g d a l e n i a n bone plate, Marsoulas, France. Scale Breuil, 1913, fig. 27, 3. M i d d l e M a g d a l e n i a n b o n e plate, Marsoulas, F r a n c e . Scale Breuil, 1913, fig. 27, 4. M i d d l e M a g d a l e n i a n b o n e plate, Marsoulas, F r a n c e . Scale Breuil, 1913, fig. 27, 1.

w It

¿

B

U

T

'

,

) ^ m

/ 1 ) > > > > > > » »

1 1

) ) » > • ) • > )

t

>

ì

i

i -

ß »

j

/ N

>