Vul. 50.] IN CANADA. AND ALASKA. 7 



would be confined to the deeper western portion of what is now 

 Bering Sea, forming there a limited gulf, without outlet to the north, 

 from which the region where the ' ground-ice formation ' is now 

 found would be so far removed as to greatly reduce its mean annual 

 temperature. Snow falling upon this nearly level, northern land, 

 and only in part removed during the summer, would naturally tend 

 to accumulate in nevee-like masses in the valleys and lower tracts, 

 and the underlying layers of such accumulations would pass into 

 the condition of ice, though without the necessary slope or head to 

 produce moving glaciers. The evidence does not seem to imply that 

 the Mammoth resorted to this extreme northern portion of the 

 region during the actual time of ice-accumulation, but this animal 

 may be supposed to have passed between Asia and America along 

 the southern parts of the wide land-bridge then existing. 



At a later date, when the land became depressed to about its 

 present level, Bering Sea extended itself far to the eastward, and 

 Bering Straits were opened. The perennial accumulation of snow 

 upon the lowlands ceased, and in the southern parts of Alaska such 

 masses as had been formed may have been entirely removed. 

 Parther to the north and at a greater distance from the Pacific 

 waters, while the total precipitation would probably be increased, 

 a greater proportion would fall as rain, and floods resulting from 

 this and the melting of snow on the higher tracts would be 

 frequent. Thus it may be supposed that deposits of clay and soil 

 from adjacent highlands and from the overflow of rivers covered 

 large parts of the remaining ice of the lowlands, and that wherever 

 so covered it has since remained ; the winter temperature being 

 still sufficiently low to ensure the persistence of a layer of frozen 

 soil between the surface annually thawed and the subjacent ice. 

 Over the new land thus formed the Mammoth and associated animals 

 appear to have roamed and fed, and wherever local areas of decay 

 of the ice may have arisen, bottomless bogs and sink-holes must 

 have been produced which served as veritable traps. 



It will be observed that this hypothesis requires a rather abrupt 

 passage from the conditions under which the ice accumulated to 

 those in which, before it had timo to disappear, it began to be covered 

 up by soil, but the change may nevertheless have extended over a 

 considerable number of years. The association of the Mammoth 

 with an animal so essentially Arctic as the Musk-Ox requires — as 

 lias frequently been pointed out — the admission that the Mammoth 

 was capable of living in a rigorous climate, though it may be that the 

 southern limit of the migration-range of one animal merely over- 

 lapped the northern limit of tho migration-range of the other. 

 The occurrence of the Moose (Alces americanus) implies the existence 

 at that time of woodland, or at least of well-grown thickets. 



In the Cordillcran region generally, the Pliocene and Glacial 

 periods were characterized by several important changes in elevation 

 and depression of land; 1 but it is unsafe to assume that these 



1 Trans. Royal Soc. Canada, vol. viii. (1890) sect, iv. p. 54; BuD. U.S. Geol. 

 Survey, 110. 84, p. 278. 



