250 ANNUAL REPORT SMITHSONIAN INSTITUTION, 195 2 



up the long gentle slope of the plains country to the west. Its uphill 

 course was made possible by its great thickness. Ultimately it met 

 the piedmont glaciers flowing eastward from the Rocky Mountains 

 and merged with them to form a continuous glacial mantle that 

 stretched from the Labrador Sea to the Gulf of Alaska, broken only 

 by high mountain peaks. The line along which the two ice masses 

 merged shifted its position from time to time, but it was never far 

 east of the Rocky Mountain front. This line lay along the general 

 course of the Mackenzie and Liard Rivers, through the vicinities of 

 Fort Nelson and Dawson Creek, and passed west of Calgary and 

 Lethbridge, to the southern limit of the ice at the International Boun- 

 dary. That rock debris was brought to this line by both glaciers is 

 shown by exposures of overlapping glacial deposits, one layer con- 

 taining Rocky Mountain stones, another containing stones brought 

 from the region west of Hudson Bay, and still another layer contain- 

 ing a mixture of both. 



The transport of stones to western Alberta from the country im- 

 mediately west of Hudson Bay involved not only a journey of many 

 hundreds of miles but also a vertical lift amounting to more than 4,000 

 feet. In order to accomplish the lift, the ice sheet must have had a 

 thickness considerably in excess of this value. It is not probable, 

 however, that the ice sheet, when at its maximum, was thickest at its 

 geographical center and thinner elsewhere. The probability is that 

 the ice was thickest at its eastern and southern marginal areas — the 

 areas that intercepted the largest amounts of atmospheric moisture — 

 and that elsewhere the ice was thinner. The glacial striations and 

 other geologic evidence of direction of flow of the ice are still too 

 scanty to justify definite conclusions, but such facts as we have are 

 consistent with this concept. 



If this was the case, then the flow of the spreading ice was most 

 active in the southern peripheral zone and was least active in the 

 vast interior and northern areas. Furthermore, most of the ice mark- 

 ings left on the bedrock and the localized accumulations of glacial 

 drift deposited on the surface must have been made during the waning 

 of the ice sheet; for the majority of the markings and accumulations 

 made earlier would have been erased or reshaped by later movement. 



Floating shelves of glacier ice, like the shelf off northern Elles- 

 mere Island today, undoubtedly fringed many coasts. Beyond the 

 shelves the sea ice was far more extensive and more nearly continuous 

 than it is today. Pack ice filled not only the Arctic Ocean but also 

 the Bering Sea and the Greenland and Labrador seas, and reached 

 into the North Atlantic beyond the southern coasts of Greenland and 

 Iceland. 



The state of exploration of Arctic North America permits us as yet 

 to sketch only the general outlines of the deglaciation — the shrinkage 



