Drift ef Northwester ?i lozva. — Bahu 175 
regions, the active erosion by which the stained and leached 
material has been carried away as fast as formed, is probably 
the main factor in the explanation. 
In ver}' many instances where the more weathered portion 
of the Kansan is absent the field evidence shows that its ab- 
sence is probably due to local intensity of erosion. The fact 
that erosion over a- general surface presents the widest varia- 
tion in intensity and efifects from point to point, is not perhaps 
always appreciated as thoroughly as it should be. A differ- 
ence of five to ten feet in the amount of erosion on neighboring 
swells and divides is not unusual but where the uppermost 
stratum happens to be so strongly marked as is the ferretto 
zone the effect of this difference becomes very striking. On 
high land much cut up by streams, erosion is very active and 
it seems reasonable to believe that these minor differences 
from point to point would be correspondingly magnified. 
These very conditions prevail now over most of the territory 
in question and the stream history of the region, so far as it 
can be read, indicates thatconditions were not greatly different 
before the lo€ss was laid down. 
The high divide in Madison and Union counties between 
Clanton creek and Grand river shows a corresponding local 
variation in the amount of erosion. At numerous points the 
yellow unleached drift of the Kansan is exposed at the surface 
in the heart of a region where leached drift and ferretto are 
widespread. In this case the erosion, since the exposures are 
loess-covered, seems to be recent. In the case of the Carroll 
county outcrops the erosion would be mainly pre-loessial. 
No attempt can be made here to fix the age of the cxtra- 
morainic and fresh-looking drift found in the counties to the 
north. The work of the present field season has shown that 
the reference of this drift to the lowan is probably wrong. 
The work in Carroll county has shown that there is danger 
of confusing certain phases of the Kansan with the later drifts. 
It is possible that a limited extent of northwestern part of 
Carroll county is covered by the later drift. The exposures 
are few and such as are found are of the equivocal type. 
There are. however, no marked border phenomena such as 
elsewhere denote the limits of a drift sheet and the possibility 
'of a later extra-morainic drift within the limits of the county 
