248 
IOWA ACADEMY OF SCIENCE 
or sandy heavy soil is now found on the poorly drained Kansan 
surfaces in Dubuque and Bremer counties, and elsewhere. 
If swampy areas, such as are still common within the Wisconsin 
drift area in Iowa were gradually drained by erosion, a vegetation 
would soon gain a foothold on the exposed parts. This would at 
first be scattered, and would retain but little fine material. Every 
subsequent temporary flooding of the territory during wet seasons 
would result in the covering of the plants and the mingling of 
some of the coarser materials with the fine soil accumulated by 
them. As the drainage became more perfect the flooding would be 
less frequent, and less coarse material would be shifted, until 
finally the permanently exposed surface would be densely covered 
with vegetation and would accumulate a uniformly fine soil, and 
ultimately a loess. 
In such cases the line of demarkation between the drift, sand or 
muck, and the overlying soil would not be sharp, as may fre- 
quently be observed on old sandbars along our streams. The fact 
that gumbo thus frequently grades up into loess is therefore far 
from proof that the two deposits are genetically the same, but sug- 
gests only a gradual change in conditions during the transition 
period. The manner in which vegetation starts on a bare loess 
surface is shown in Plate III, fig. 2, and is of interest here. 
The presence of the several contiguous deposits leads Professor 
Todd to ask: “But how by the aeolian hypothesis can be explained 
the occurrence of different strata of considerable thickness, clearly 
delimited and in close contact ?” Eliminating the gumbo, with 
which, as noted, the aeolian hypothesis has little or nothing to do, 
the question may be readily answered with reference to contiguous 
loesses. Investigation has shown that the contiguous loesses lie 
wholly (as far as observed) outside of the several drift borders, or 
at least near them. The advance of the several ice-sheets was ac- 
companied by a glacial climate the influence of which must have 
been felt far beyond the border in a territory in which large quanti- 
ties of ice and snow remained during the greater part of each year. 
In such territory there would be no evidence of glacial drift, but as 
all vegetation must have been destroyed on the old surface, and as 
new materials were brought down by the later ice-sheet which could 
be re-worked into a newer loess, a decided difference between the 
newer and older loess would be expected, and the line of demarkation 
between them might well be sharp, though not necessarily so, especi- 
ally at points more remote from the ice-border. When the Illinoian 
ice-sheet entered Iowa there was evidently loess on much of the 
