3/2 TJie A?iicrica?i Geologist. June, isos 
The till covering the earlier modified drift and forming the plain at 
Rush Cit\ appears not to have been deposited until the ice in which it 
had been held was melted away. This ice in its onflow did not erode 
the modified drift, at least in any considerable amount, else its level 
contour would be destroyed; and at the same time it seems impossible 
that a sheet of till having so uniform thickness upon so large an area 
could be formed as a ground-moraine. Obviously the till here was 
brought to this area and spread upon it while it was contained in the 
mass of the onflowing ice, and it seems almost equally sure that it re- 
mained thus in the ice till. this was melted. 
In qtialification of this explanation, it shotild be added 
that I regard the later ice extension, like the earlier, as dtie 
more to snowfall on this area than to ice invasion, although 
all the upper till was brought by the comparatively small pro- 
portion of inflowing ice. In other words, the ice-sheet could 
not advance with a steep border and leave the sand plain so 
unchanged; but it was attenuated in its first advance, being 
formed by local snowfall, and was reinforced soon by some 
inflow from the higher part of the ice-sheet that had remained 
unmelted, bringing its englacial drift. 
So great recession of the ice-sheet, preceding its extension 
in the low^an stage of glaciation, is indicated by wood, peat, 
and fresh water molluscan shells, found in sections of inter- 
glacial beds, underlain and overlain by till, at various localities 
in the southern half of Minnesota, as noted in the first and 
second volumes of the final report on the geology of this 
state. Other localities and general drift features of the same 
region, likewise described in these volumes, show that the 
lowan glaciation, continuing through a far shorter time than 
the earlier and maxiinum Kansan glaciation, was therefore 
insufficient to obliterate some of the larger valleys that were 
channelled by the streams during the long time between these 
principal stages of glaciation. Thus the deep valley of the 
Minnesota river appears to have persisted, without becoming 
generally filled with drift, under the ice-sheet in its lowan re- 
advance; and the valleys of head streams of the interglacial 
Des Moines river are preserved and marked in parts of their 
courses by three somewhat parallel chains of lakes in Martin 
county, Minnesota, adjoining the north line of low'a. Sub- 
glacial transportation of the lowan drift would have caused it 
to fill and thus bridge the transverse Minnesota valley; but its 
englacial transportation would favor the preservation of these 
