THE ORIGIN AND ANTIQUITY OF THE ESKIMO — ^COLLINS 447 



that they may be assumed to have had a common ancestry? And, are 

 the hypothetical ancestors pure-blood Indians? 



As to the first question, the answer seems to be affirmative, as far as 

 the measurements themselves are concerned. The three groups are 

 very much alike in stature, length and breadth of head, and length 

 and breadth of face. It is quite evident, however, that two of these 

 groups — the Caribou and Labrador Eskimos — are by no means pure 

 Eskimo. Stewart (1939) has demonstrated this for the Labrador 

 group, and Birket-Smith's photographs leave no doubt of the consid- 

 erable amount of white blood present among the Caribou Eskimos. 

 The Ammassalik people, on the other hand, show no such evidence 

 of white mixture. If we may judge from the photographs published 

 by Thalbitzer, Holm, and others, the Ammassalik are one of the 

 purest of all Eskimo groups, showing not the slightest resemblance 

 to the Cree or any other Indians. This is one of those instances, not 

 uncommon in anthropology, where metrical comparisons alone are 

 misleading. 



As to the second question, it appears that the Cree Indians at Fort 

 Chipewyan on Lake Athabaska who were measured by Grant (1930) 

 and who, according to Seltzer, represent the type ancestral to the 

 Caribou, Labrador, and Ammassalik Eskimos, are by no means full- 

 bloods. Grant showed that this group of Crees were practically iden- 

 tical, except in stature, with a group he had measured earlier at 

 Oxford House. The latter were clearly mixed-bloods, as they them- 

 selves recognized. They were very close metrically to mixed-blood 

 Sioux, from which Grant (1929, p. 27) concludes: "We surmise that 

 the Oxford House Indians have likely as great an admixture of white 

 blood as have the half-blood Sioux." Considering that the Indians 

 in this part of Canada have lived in contact with whites for many 

 years (Fort Chipewyan was established in 1789 and Oxford House 

 in 1792), it is inevitable that extensive race mixture should have 

 occurred. As the modern mixed-blood Cree do not represent the 

 original physical type of the group it is obvious that their measure- 

 ments cannot be used to trace original relationships. Stewart, in 

 commenting on this same suggested relationship between Eskimo and 

 Cree, has expressed a similar opinion : "I object chiefly to drawing 

 such far-reaching conclusions from such unequal material ... in 

 other words, to concluding from the similarity of a few measurements 

 taken on small samples of widely separated modern groups, speaking 

 different languages (Eskimo, Algonkian Cree) and undergoing dif- 

 ferent stages of acculturation (Whites), that they must have had a 

 common ancestry a little over 1,000 years ago" (Stewart, 1939, p. 120) . 



Birket- Smith remarks on the great physiognomic likeness between 

 the Caribou Eskimos and Chipewyans he had seen on Churchill River, 



