218 
IOWA ACADEMY OF SCIENCES 
the sandy borders so prevalent in the second stage of the stream valleys, 
both signifying extensive ’wmter action, that we would expect the streams 
to have been greatly overloaded with detritus, which would then naturally 
be deposited in extensive terraces and valley trains.* 
It would seem plain that the melting of the ice sheet must have been 
very slow or else the streams would have been greatly swollen in size and 
have left extensive deposits high up along the sides of their valleys. The 
outwash plains of sand and the hills and ridges partaking of the nature 
of kames and eskers, might be formed without a great flow of water at 
any one time. 
The relations of the Iowan and Kansan drift sheets along their con- 
tact line are intricate and puzzling. One finds Kansan-like contours of 
hills and valleys, dotted with Iowan borders, or again on Iowan topog- 
raphy exposures of materials are apparently Kansan in the gravel beds 
and roadside sections. A considerable section, beginning one mile south 
of West Union on the east road to Fayette, appears like Kansan in topog- 
raphy, but has no mantle of loess. It might be expected at the Iowan 
border, where the deposit of that ice sheet thins out, that there would 
be frequent outcroppings of the Kansan. Where the deposit seems to be 
a mixture of the Iowan and Kansan, it may be believed that the Kansan 
deposits were worked over and somewhat intermingled with Iowa drift 
during the Iowan ice invasion. 
*A. G. Wilson, in an article a few years ago, suggested the interesting 
hypothesis, to account for the preservation of such valleys, that the waters suc- 
cessively overflowing and freezing, gradually fllled full with solid ice up to the 
top of the hills, thus protecting the valleys from glacial action. 
