Eskers of the Kansan Epoch. — Hershey. 199 
to the reputed direction of ice-movement. They are practi- 
cally continuous throughout their entire length. On a hasty 
reconnoissance, this statement might be disputed; for the 
continuity of the deposits is now frequently interrupted by 
gaps. Many of these gaps are occupied by the alluvial planes 
of the present streams, and the ends of the ridges on either 
side correspond, clearly showing that the gaps have been pro- 
duced by stream erosion subsequent to the abandonment of 
the region by the icesheet. There are other apparent gaps 
where subaerial erosion cannot explain the absence of the de- 
posit, but in these places careful observation usually reveals 
some trace of it. 
The main belts of stratified drift are found to follow the 
main lines of drainage wherever these agree with the general 
direction of the ice-movement. Three of these belts cross 
Stephenson county in a generaLeast to west direction, which, 
beginning at the southernmost and most prominent one, may 
be named the Pecatonica, Cedarville, and Orangeville belts. 
The next belt northward seems indicated by a heavy develop- 
ment of stratified sand and gravel in and south of the city of 
Monroe, Wisconsin. This may be referred to as the Monroe 
belt. But the most prominent and sharply defined belt of the 
district is the one next south of the Pecatonica belt, which, 
from its fine development in the vicinity of the village of 
Adeline, in Ogle county, may be denominated the Adeline belt. 
Other less developed belts occur both north and south of those 
already mentioned, notably one in southeastern Carroll 
county, designated by Mr. Frank Leverett the Hazelhurst 
belt; but these have not yet been mapped and will not be 
considered in the present paper. 
The five belts thus especially noted are not all that occur 
within the limits of the district, for there are still a larger 
number of shorter and less marked belts, which, although 
generall}^ approximately parallel to the glacial movement, 
gratlually converge toward the main belts, until finally the de- 
])i)sits of one are merged into those of another. If, instead of 
mapping the deposits by their present areal extent, lines are 
<lrawn tlirough the crests of the ridges, the analogy with an 
ordinary drainage system will be remarkably close. The main 
<^>r primary belts represent the trunk streams, and the second- 
ary belts the tributaries. 
