W. Upham — Diversity of the Glacial Drift. 361 



about 80 miles on the east side of the driftless area in Wis- 

 consin the magnificently developed Kettle moraine, belonging 

 to a late part of the Ice age, forms the extreme border of the 

 drift. Its correlation or continuation in Minnesota, Iowa, and 

 South and North Dakota, is probably the Altamont moraine, 

 the outermost of the series of twelve approximately parallel 

 successive retreatal moraines which I have explored, and 

 mapped, under the direction of Prof. N. H. Winchell, for the 

 Minnesota Geological Survey. Near Des Moines in central 

 Iowa this Altamont or first moraine of the Minnesota series, 

 there forming the southern extremity of an area of the later 

 drift deposited by the Minnesota and Iowa lobe of the ice- 

 sheet, lies 175 miles north of the southern boundary of the 

 drift in its course through Missouri and northeastern Kansas. 

 Within the Mississippi basin the ice sheet forming the Kettle 

 and Altamont moraines occupied less area by 125,000 square 

 miles than was enveloped by the ice forming the earlier drift. 

 But while there had been in general this great decrease in the 

 extent of the ice previous to the accumulation of the outer 

 large moraine, it even advanced at that time farther than ever 

 before upon the east side of the Wisconsin driftless area. 



In the eastern United States, the outer moraine along an 

 extent of about 700 miles, from the Scioto basin in Ohio to 

 Martha's Vineyard and Nantucket, stands upon the boundary 

 of the drift or very near it. At its time of maximum exten- 

 sion the ice-sheet in this region generally reached a short dis- 

 tance, from a few miles up to twenty miles or more, beyond 

 the position of the moraine. Besides this remarkable differ- 

 ence from the drift of the Mississippi basin, it must be con- 

 fessed that we cannot yet be sure that this outer moraine, at 

 least eastward from the angle of the drift boundary in south- 

 western New York, is to be correlated with the Kettle and 

 Altamont belt. Following a suggestion or query of Mr. 

 Leverett, I incline to believe that quite as likely it may belong 

 somewhere within the time of the very large Fergus Falls, 

 Leaf Hills, and Itasca moraines, which are the eighth, ninth 

 and tenth of the series in Minnesota. 



Oscillations of the Boundai'y and Changes in the Thick- 

 ness and Currents of the Ice-sheet during its general Recession. 

 — The great contrast between the glacial retreat from north- 

 eastern Kansas to Des Moines and the contemporaneous 

 encroachment of the ice upon the eastern side of the Wisconsin 

 driftless area seems to be accounted for, in part or wholly, by 

 interdependence of snowfall at the east with rains and ice 

 melting at the west. While the ice was being melted away by 

 rains and sunshine over its 125,000 square miles that had been 

 uncovered south and west of central Wisconsin, the east- 



