272 



SCIENCE 



[N. S. Vol. XXXIX. No. 999 



junction with the Missouri and joined the 

 James several miles north of Yankton." 



At this pomt we may offer a word con- 

 cerning the loess. This problematic deposit 

 covers the Kansan till almost universally. 

 The surface of the latter is generally much 

 eroded, and becauseof thisithas beenargued 

 that a long time elapsed between the depo- 

 sition of the two formations. It was onee 

 claimed that the loess was largely of lowan 

 age, but it can be reasonably questioned 

 whether erosion might not have been 

 rapidly accomplished by the waters closely 

 attending the edge of the ice, while the 

 loess may have been deposited in flooded 

 streams and local lakes remote from the 

 edge. This silty material has doubtless 

 been further modified by wind action and 

 by later erosion and slumping. A strong 

 objection to the flood theory, formerly felt, 

 was the supposition that the channels of that 

 time were as deep as the present, which we 

 have seen was not the case. The drainage 

 level of Kansan time was from 80 to 120 

 feet higher than that of the present, and it 

 would have required scarcely a higher rise 

 of water to have overflowed the uplands of 

 that time than that which we shall find 

 attended the retreat of the "Wisconsin ice 

 sheet. 



E. PROM THE END OP THE KANSAN STAGE TO 

 THE BEGINNING OP THE WISCONSIN 



This was a long time during which the 

 ice occupation of the region under consider- 

 ation was not conspicuous. It corresponds 

 to the lUinoian and lowan advances and 

 secessions of the ice. Elsewhere the ice 

 found a growth into southern Illinois and 

 far enough into southeastern Iowa to push 

 the Mississippi westward nearly fortymiles, 

 then a recession and an advance from the 

 northwest in eastern Iowa till the Missis- 



«BmK. G. S. a., Vol. 23, pp. 463-470. 



sippi was pushed eastward over into Rock 

 River. 



Meanwhile the ice may have lingered in 

 northern Minnesota and Dakota and have 

 caused greater volume to the streams than 

 now. The course of the master stream was 

 doubtless in the James valley approxi- 

 mately along the course of the present 

 James River, and further south along the 

 prftsent Missouri. Erosion deepened all 

 channels, till their bottoms approached the 

 level of the present surfaces of the streams. 

 The valleys were widened considerably, 

 particularly in the soft Cretaceous forma- 

 tions in the Dakotas and northwestern 

 Iowa. 



p. DURING THE ADVANCE OP THE WISCONSIN 

 ICE 



As in earlier stages the Minnesota and 

 Dakota lobes of ice were the only ones 

 advanced southward along the same lines 

 which affected seriously the Missouri. They 

 advanced southward along the same lines 

 as before, and though probably with greater 

 velocity and vigor, yet they did not attain 

 so great extent. This may be explained 

 possibly by a warmer climate than in 

 former advances. The Minnesota lobe 

 reached only to Des Moines, Iowa. The 

 Dakota lobe only to the south line of South 

 Dakota, although it pushed farther west- 

 ward than before up the valleys of the west- 

 ern tributaries. Why this should be is not 

 very evident. Possibl.y because of the 

 greater depth and width of the trough of 

 the Niobrara River where it was transverse 

 to the paths of the ice, and because of the 

 greater volume of water skirting the ice 

 front there on the south. On the west, how- 

 ever, the greater maturity and therefore the 

 greater width of the valleys of the western 

 tributaries, like the White, Cheyenne and 

 Grand, enabled the ice to push westward 

 easier than before. At any rate the de- 



