384 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1940 



Australian, has apparently found few supporters. Some day we may 

 succeed in joining up Eskimo with the Ural-Altaic languages and in 

 proving that the Athapaskan or Dene tongues of North America 

 are genetically related to the Sinitic tongues of eastern Asia. Such 

 relationships, however, even if confirmed, must be exceedingly dis- 

 tant, for we know how little the Greenlandic Eskimo dialects have 

 diverged from the north Alaska ones, despite a separation of 1,000 

 years, and how small is the difference, after an equally long separation, 

 between the Navaho Indian dialect in the Southwest of the United 

 States and the dialects spoken in the Mackenzie River Basin. Prob- 

 ably much of the linguistic diversity among the Indians and Eskimo 

 took place in Asia before their entry into the New World; but the 

 fact that no American tongue is palpably related to any Asiatic one 

 strongly suggests that the inhabitants of the New World, barring the 

 Eskimo to whom I will return later, separated off completely from 

 those of the Old more than 2,000 years ago. 



On etlmological grounds, too, there seems no reason to question 

 this conclusion, because the traits that are common to Asia and 

 America, apart from a few that are concentrated near the bridgehead 

 at Bering Strait, are so widely diffused in both continents that they 

 evidently carry a very respectable antiquity. Even the resemblances 

 between the Palae- Asiatics and the Indians of the northwest coast of 

 America hardly demand a migration in post-Christian times. If 

 there was such a migration it is more likely to have been from Amer- 

 ica to Asia by way of the Aleutians and Kamchatka, as Collins has 

 shown,* than from Asia to America; moreover, it was a relatively 

 insignificant migration that introduced into northeast Asia a few 

 cultural traits such as labrets, certain forms of stone lamps, a certain 

 type of house, and perhaps some folk tales, but failed to effect any 

 far reaching changes. It can hardly account for the much deeper 

 resemblances, e. g., in physical type and clothing, between the Palae- 

 Asiatics and some of the American Indians. 



For the millennia that preceded the Christian Era, the millenia 

 that saw the peopling and subsequent isolation of America, ar- 

 cheology, our safest guide, has afforded us hitherto only one or two 

 uncertain clues. The main props for our theories have come from 

 ethnology, linguistics, and physical anthropology, none of which can 

 furnish more than the vaguest indications of a time sequence. In 

 founding our theories on these disciplines we are building on shift- 

 ing sand, and we need not be surprised if the theories topple over 

 when the spades of the archeologists succeed in uncovering new and 

 possibly unexpected remains. 



* Collins, H. B., op. clt., pp. 875-378. Also Culture migrations and contacts In the 

 Bering Sea region, Amer. Antbrop., vol. 39, No. 3, pp. 375-384, 1937. 



