ILLINOIS GLACIATION 505 



with the more uniform sheet of drift laid down later from a thicker 

 Illinoian ice-border, which could form a low terminal moraine. 



The Keewatin and Labradorian icefields are here regarded as very cer- 

 tainly attaining respectively their greatest southward and southwestward 

 areas in one general stage of prevalent snowfall and ice accumulation, 

 called west of the Mississippi the Kansan stage, and from this river east- 

 ward the Illinoian stage. After the melting of each of these icefields 

 from large southern tracts of their maximum extension, the subsequent 

 more northern drift formations of the mainly receding continental glacier, 

 which yet at many times halted or readvanced, with much complexity of 

 its outlines, drift sheets, and marginal moraines, have been referred to 

 lowan and Wisconsin stages of glaciation. To the last belongs far the 

 greater part of the area of the ice-sheet, including nearly all its surface 

 deposits in Canada; but underlying the Wisconsin drift are seen on 

 many tracts deposits referable to the preceding glacial and interglacial 

 stages. 



Wherever the northern and central icefields may have remained un- 

 melted from the earliest Nebraskan stage onward to the close of the 

 Glacial period in the Champlain stage, only a single sheet of drift re- 

 sulted, such being probably true for much of the interior of Canada and 

 for parts of New York and New England. 



The Sangamon interglacial Stage 



An interval separating the times of formation of the Illinoian and 

 lowan drift, the latter being closely coincident with the chief loess deposi- 

 tion, is shown by an old soil and leached subsoil of the original Illinoian 

 surface, and occasionally by scanty beds of peat, overlain by the loess. 

 It was named by Leverett in 1897 the Sangamon interglacial stage, from 

 the county and river of this name in Illinois where such evidence of it 

 was first observed. 



In a paper before cited, written in 1913, I referred the arid climate 

 by which Lakes Bonneville and Lahontan were dried up, between their 

 times of high water, to this Sangamon stage of interruption in the con- 

 tinental glaciation; but such correlation in the records of these lakes and 

 the Glacial period seems instead to belong surely with the very widely 

 extended Aftonian recession of the ice-sheet. ■ My erroneous study em- 

 bodied in that paper served as a stepping-stone to reach this conclusion, 

 from which the general time relationships in the series of secular climatic 

 changes during the Ice Age are, as I believe, rightly ascertained. 



