ij4 ^^i(^ American Geologist, March, vm> 
36.18.* In the northwest it varies from 15 to 20 inches and 
in the southeast from 20 to- 25 with areas running from 25 
to 30. The run-ofT also varies widely. There are no data 
relative to Iowa streams but it is well known that run-ofif is- 
proportional to the character of the surface, the slope, and 
the time distribution of the rainfall. It is greater on a non- 
absorbent surface, on greater slopes, and when the rain fall 
is bunched. In the present case the surface of the Kansan 
drift seems not to vary in any systematic way with relation to 
its capacity to absorb water. The rainfall in Carroll county 
is probably as evenly distributed as in other parts of the state. 
There are, however, considerable dififerences in slope. It is 
to be remembered that the region represents the high upland 
between the Mississippi and the Missouri. The railw^ay grades 
across the county are heavy and the stream grades are even 
greater. The drainage consists of the headwater portions of 
the streams only, and the water reaches them by running over 
the steep surface slopes rather than through the ground. 
Springs, in the southwestern part of the county, are almost 
unknown and seepage is rare. The whole of the evidence 
indicates that the water passes over rather than through the 
drift and hence that solution is relatively slight. This seems 
to be one of the important factors in the failure of the leaching 
test. 
Where the grades are high and the amount of surface 
water notably in excess it must be obvious that erosion will 
be very active. This will be true of the slower and less easily 
noticed surface erosion of the inter-stream areas as well as 
the direct corrasion of the streams. It has already been sug- 
gested that stream action in the region is intense. It is also 
true that the erosion of the general surface is much greater 
than on the low wide divides further south. The relations 
of the loess to the river valleys indicate that the latter occupied 
approximately the same position in the interval between the 
Kansan and the loess tliat they do now. sO' that erosion was 
probably at least as active then as now. This is probably 
the main explanation of the absence of the ferretto and leached 
drift in the region under discussion. Aside from the fact that 
there may have been less leaching here than in lower lying 
*Rept. Iowa Weather and Crop Service, 1897. 
