IOWA ACADEMY OP SCIENCE 
63 
among which they live, with that of the fossils in the loess, shows a striking 
similarity of habitats, and suggests, as the writer has repeatedly stated, that 
the conditions which prevailed when the loess was formed were not materially 
different from those now existing in the same region. The abundance of ter- 
restrial mollusks and plants precludes the possibility of the presence of large 
bodies of water, or of glacial ice in a glacial climate. They render the glacio- 
fluviatile hypothesis wholly untenable. 
IV. Inter-loessial periods. 
In many cases where two distinct loesses are in contact, the lower is gray 
(post-Kansan) and contains numerous iron tubules which have been manifestly 
formed by roots, and these tubules often terminate abruptly in the oxidized 
band which so frequently separates these loesses. Less frequently they terminate 
similarly in the surface of contact of the loesses, the oxidized line or band 
being absent. (See Plate VII, fig. 1.) 
This suggests that when the bluish-gray loess was at the surface it was cov- 
ered with an abundant vegetation which has left evidence of its existence in 
these tubules. When the succeeding ice-sheet overwhelmed the territory this 
vegetation was destroyed, and the loess surface, just beyond the ice-border 
and on the elevations which had not been denuded by the thin ice-sheet, and 
from which the ice first disappeared, remained bare of plants for some time, 
and finally gradually developed a scant flora. During this partial or complete 
denudation the oxidization of the iron-stained surface probably took place, 
though it is possible that the iron was in part washed down to its present level 
after some of the superimposed newer loess was formed. (See Plate VII, fig. 2.) 
The infiuence of the second ice-sheet would have been felt for some distance 
beyond its border, and the same sharp division of loesses would have resulted. 
It is probably for this reason that the sharp division of the post-Kansan and 
later loesses may be traced for a considerable distance beyond the borders of 
the later drifts, especially the Iowan. 
A subsequent restoration of favorable conditions resulted in the develop- 
ment of another fiora which built up the upper (newer) loess. 
In addition to the foregoing important considerations there are other ques- 
tions of interest which may be solved by due reference to the behavior of mod- 
ern plants. The peculiar loose lobular structure of the upper three to five feet 
of much of the loess is evidently due in large part to the action of roots; the 
vertical cleavage of loess is probably due to the same cause; the fact that roots 
will decay in loess and completely disappear without leaving a trace of car- 
bonaceous material is not without interest; this, together with the fact that 
exposed plant materials completely decay long before they could be covered by 
dust in the manner suggested by the aeolian hypothesis, accounts for the ab- 
sence of plant remains in the loess; the effect which roots might have by dis- 
solving shells of the fossil mollusks and thus making some of the loess ap- 
parently non-fo'ssiliferous is also noteworthy; and perhaps other minor prob- 
lems may be suggested. 
It is also worthy of note that sometimes the study of the local fiora makes 
it possible to determine the limits of loess and drift formations without making 
sections. Thus along the timbered portions of the Iowan drift border in John- 
son and Iowa counties in Iowa it is possible in some cases to accurately limit 
the Iowan sands and the loess by the species of trees making up the forest 
covering, the dominant species on the sands being Quercus velutina, the yellow 
