DEFORMATION OF SHOKE LINES. 513 



might fall considerably short of reaching the extreme margin or even the mean line of the lobate 

 margin. If depression depended on a central ice mass of great thickness and was not influenced 

 by the relatively thin ice of the marginal part with its lobes, it might be expected that the outer 

 boundary or hinge line of depression and of subsequent resilience would lie within the marginal 

 belt of the ice field and would follow a straight or gently curved line substantially parallel with 

 the isobases of the uplifted region to the north and largely if not entirely independent of the mar- 

 ginal lobes and reentrants. Thus, the elevation of the land for probably 100 to 150 miles north 

 of the first hinge line was in all probability due not to depression by the weight of the ice that 

 rested directly upon that land (at the maximum of the ice) but to the greater weight of the much 

 thicker ice farther north, the depression under which died out gradually southward into the mar- 

 ginal area, where the weight of the thinner ice failed to produce depression. 



As the retreat of the ice front progressed from the maximum boundary, the height of the 

 entire ice surface to the north and hence the thickness of the ice and the weight with which it 

 pressed upon the lands beneath it were reduced, and it was presumably this reduction of weight 

 in the central, thickest part of the ice field that caused resilience, and not the uncovering of the 

 marginal belt by the shrinkage of the lobes. In view of the independence of the hinge lines in 

 relation to the irregularities of the ice margin it seems certain that only a small part, if any, of the 

 uplift in the region discussed by Lane (Huron Mountains and Keweenaw Bay, Lake Superior) 

 could have been caused by the removal of the weight of the ice mass of that immediate vicinity, 

 and that the uplift, if due at all to the removal of ice weight, was due mainly or wholly to the 

 removal of the much thicker ice farther north or northeast. 



The area of marked depression under the central mass of the ice sheet may be regarded 

 theoretically as a saucer-shaped basin, its outer margin sloping gently upward and dying out to a 

 hinge line in the marginal belt, where the ice weight was not sufficient to produce depression. 

 Questions arise here as to the probable breadth of the inclined rim of the depressed basin, the dis- 

 tance to which depression may have extended beyond the area of effective ice weight, and the 

 size of the area that might be left bare of ice without markedly modifying the courses of the hinge 

 lines and isobases. 



The boundary of the light and short-lived glaciation around the west end of Lake Superior 

 appears to widen toward the northeast in northeastern Minnesota, as though extending across 

 into Canada. This it may do, and it may include a large part of the high region which lies north 

 of the western arm of Lake Superior and west and northwest of Lake Nipigon. As may be seen 

 on the rebel map of the Dominion of Canada, 1 this area is occupied by a high plateau covering 

 about 100,000 square miles. The widening toward this plateau indicates that it was not a region 

 of strong glacial growth and dispersion and hence of thick ice, but that it was in the marginal 

 belt and was covered, if at all, at the Wisconsin maximum only thinly and that for a relatively 

 short time. If this area extends into Canada, as seems probable, it reaches so far back from the 

 hinge line, as determined by the Lake Agassiz beaches in Minnesota, that it seems to call for too 

 great a width of the propagated marginal belt of the ice-depressed basin. 



Suppose a nunatak, or rather, a flat island in the ice field, like that which existed for a time 

 in the southwestern peninsula of Ontario, surrounded by thick ice at some distance, but with the 

 ice thinning toward it on all sides. If the island were of small area it is easy to believe that it 

 would be carried down along with the depression of the surrounding heavily weighted lands, 

 although it was itself not covered or only thinly covered, and that it would come up again when 

 the ice disappeared and the surrounding lands were affected by resilience. How large such an 

 island must be to be exempt from the influence of conditions directly affecting only the surround- 

 ing lands and not the island itself is a question. Even if the island itself be relatively small, 

 200 or 300 miles of relatively thin ice must be allowed for in the margin around it. The area in 

 Canada exempt from direct depression by ice weight may have been larger than the whole of the 

 high plateau; and, if so, the northernmost boundary of the nonweighted land would be at least 

 400 or 500 miles back from the hinge line at Traverse Lake, Minn. This region of light and short- 



i White, James, Dictionary of altitudes in Canada, Ottawa, 1904, map. 

 34407°- -15 33 



