Geography of Ohio 187 



occupied by the Great Lakes the ice pushed far into the Alle- 

 gheny plateau; and that in the Mississippi valley, where no 

 great heights exist, it expanded so far to the south that its mass 

 must have been great. Men have attempted to estimate the 

 thickness of this sheet in particular places by studying the slope 

 of the ice front. It has been found that where the margin of the 

 ice front was registered, by accumulations of drift against the 

 sides of the valley, for example, it had a certain decline. By carry- 

 ing this same slope northward, one gets a suggestion, at least, of 

 the thickness at any given distance to the north. Such a study has 

 assigned astounding depths to the ice, so great that one concludes 

 that this method is not reliable. The fact that the ice sheet had 

 a certain length in the Mississippi valley, and that it moved out- 

 ward from a given place, would warrant the conclusion of great 

 thickness over part of the area beween these termini. 



How then was so much ice made? This is a question on which 

 much has been written. The simplest explanation assumes a 

 heavy and regular annual snowfall, with little or no melting. If 

 this form of precipitation were to continue a sufficient length of 

 time, and little of it were to be melted, eventually an ice sheet 

 sufficiently great to cover any continent might result. But such 

 an explanation hypothecates a condition which, we all believe, 

 never existed. We have to-day some small areas of continuous 

 ice: one about the south pole, the other covering most of Green- 

 land. In both of these areas there is neither continuous snow- 

 fall nor an absence of melting. Man knows of no altitude so high 

 that the ice which forms there does not suffer from some melting. 

 It wastes even when the temperature is continuoush^ below zero. 

 The essential prerequisite for the development of ice fields is that 

 precipitation, in the form of snow, should exceed the wastage. 

 However small the increment left over from each warm season, if 

 this relationship of snowfall and melting continues long enough, 

 an ice sheet will result. 



A study of the glaciated parts of North America, omitting the 

 Rocky Mountain region, points to two centers in Canada, away 

 from which the ice moved. One center lies east of Hudson bay, 

 in Labrador; the other is west of Hudson baj', in the Keewatin 

 district. In these areas, for perhaps thousands of years, the preva- 

 lent form of precipitation was snow; in both, at first, only slight 

 bodies of snow endured from one vear to another. Gradually the 



