364 W. Ujpham — Diversity of the Glacial Drift. 



tions of the ice-front recorded by the forest beds appear to be 

 ample for the attendant stream channelling. 



Farther northward, along the 700 miles of length of the 

 glacial lake Agassiz the retreat of the ice-sheet, with the forma- 

 tion of numerous large moraines, was demonstrably very rapid, 

 occupying apparently no more than one thousand years.* In 

 all, I believe that a duration of five thousand years is sufficient 

 to account for the records of the waning and closing stages of 

 the Ice age, from the time of the maximum area of the ice- 

 sheet depositing the early drift to the time when Lake Agassiz 

 was drawn off into Hudson Bay and the northern United 

 States and Canada were freed from their glacial mantle and 

 occupied by hunting and fishing tribes of the red race, who 

 have abode here with many intertribal wars and migrations 

 during the 5,000 years or more since the departure of the ice. 

 A very large part of my belief that these oscillations and im- 

 portant changes of the currents of the ice-sheet during its 

 recession were comprised within a time geologically so short, 

 comes from my studies of the shore erosion and beach accumu- 

 lations of Lake Agassiz, and of great changes in the relations 

 of the northeastern and northwestern convergent ice-currents 

 in Minnesota during the deposition of the later drift.f These, 

 I am convinced, took place fast, in a geological sense, and I 

 am persuaded also that the ice oscillations and varying condi- 

 tions causing the diversity of the drift near its boundaries 

 were not of great duration and are referable to a continuous 

 and brief Ice age, not divided by interglacial epochs. 



Loess deposition mainly continuous from the time of maxi- 

 mum ice extension to the time of formation of the Moraines 

 of the Later Drift. — On the drift border in southern Illinois 

 and Indiana, Prof. R. D. Salisbury finds that the deposition of 

 the loess ensued immediately after that of the early till and 

 was in part contemporaneous with the till. As soon as the 

 ice-sheet retired from its farthest limit the glacial drift was 

 covered by this fine silt of the modified drift supplied by 

 streams that flowed from the melting and retreating ice.;}: 



In the northeastern part of Iowa Mr. W J McGee similarly 

 finds the loess to have been deposited while the ice-sheet that 

 spread the upper portion of the early till was melting away. 

 The very remarkable paha of that district, which are eskers of 

 loess, were accumulated while the waning ice-sheet walled 

 them in on each side.§ 



*Geol. and Nat. Hist. Survey of Canada, An. Rep., new series, vol. iv, for 

 1888-'89, pp. 50, 51E. 



t Proc. A. A. A. S., vol. xxxii, for 1883, pp. 231-234. 



X Geol. Survey of Arkansas, Ann. Rep. for 1889, vol. ii, "The Geology of 

 Crowlev's Ridge" (1891), pp. 228, 229. 



§ U. S. Geol. Survey, Eleventh An. Rep., for 1839-90, pp. 435-471. 



