February 20, 1914] 



SCIENCE 



269 



to find any other route to tlie Des Moines 

 half as probable, but it has been argued 

 elsewhere* that eastern Iowa was formerly 

 relatively considerably higher than now. 



Could it have been southeast to the 

 Grand River of Missouri ? From the mouth 

 of Platte River to a point in Grand River 

 of equal altitude to the Missouri at Leaven- 

 worth is considerably shorter than to the 

 latter point. It is quite possible that 

 Grand River was not much higher in pre- 

 glacial time than now. The region has not 

 been examined with reference to this prob- 

 lem, but there is nothing in the course of 

 present streams nor anything reported con- 

 cerning the geology of that region which 

 suggests that the Platte ever followed that 

 course. We must therefore leave the prob- 

 lem not definitely solved, though the second 

 hypothesis seems on the whole the more 

 probable. 



8. Kansas River. — The preglacial course 

 of this stream is known to have followed 

 nearly its present course but from 125 to 

 150 feet higher than now. This is shown 

 by stretches of chert gravels with no ad- 

 mixture of northern erratics, such as the 

 glaciers brought into the region. We may 

 conceive that it followed the course of the 

 Missouri from Kansas City to the vicinity 

 of Miami and thence southeast by Salt Pork 

 to the Lamine River and back to the Mis- 

 souri again. If the Platte reached Grand 

 River it would here have joined the Kansas. 



From the distribution of the drift and 

 loess in Missouri I argued some years ago'" 

 that the Osage in early Pleistocene did not 

 join the Gasconade but turned north by the 

 present course of Auxvasse Creek and over 

 into Salt River and so to the Mississippi. 

 It does not seem very probable, though not 

 clearly forbidden by known facts. The up- 

 land between rises to 870, and there is no 



*Kan. JJimi. Sd. Bull., Vol. VI., p. 375. 



5 Mo. Geol. Suri'ey, Vol. X., pp. 200 and 212. 



trace of divide between the Osage and the 

 Gasconade, nor are they different in depth 

 and age, as we should expect, if they had 

 been separated till recently. The other 

 alternative is to suppose that its course was 

 down the present Missouri to its mouth, and 

 that the shallow valley along the line of the 

 Auxvasse and Salt Fork is due to general 

 erosion acting on softer rocks, rather than 

 to the presence of a master stream. 



B. THE MISSOURI DURING THE NEBRASKAJST 

 STAGE 



We now proceed to consider the probable 

 histoiy of the Missouri during the succes- 

 sive stages of the Ice Sheet of the Pleisto- 

 cene. The stages usually recognized are in 

 order : Nebraskan, Kansan, lUinoian, lowan 

 and Wisconsin, with substages of the last 

 Altamont, Gary Antelope, etc., correspond- 

 ing to the principal moraines left. 



We know little of the Nebraskan Stage 

 except that the ice did not then extend so 

 far as the Kansan and that its deposits 

 have been mostly carried away or hidden 

 by the ice of the latter stage. The ice ad- 

 vancing from the north gradually pressed 

 southward up the Red River valley, and 

 dammed the master stream, formed of the 

 streams we have enumerated as flowing 

 northeast from the present James River 

 vallejr, and caused an overflow into the Des 

 Moines valley. Later it closed up their 

 entrance into the Red River valley from 

 the west, and caused the Missouri, Cannon 

 Ball, Grand and Cheyenne to become a lake, 

 and to overflow whatever divide separated 

 them from White River, and eventually to 

 overflow the divide separating them from 

 the Niobrara, outlining for the first time 

 the course of the James River. 



Meanwhile, the ice passed more freely 

 southward in the Des Moines valley, filling 

 it and forcing the Niobrara to overflow 

 what ever divide separated it from the 



