Eskers of the Kansnn Epoch. — Ilershey. 205 
are in every way similar to the otiiers, except that they are 
less strongly developed and more inclined to become discon- 
tinuous. 
There are certain features that are common to all the belts, 
which may be detailed as follows: 
1. They are not confined to any level but cross ridges of 
any hight, in several cases 150 feet or more above the present 
bottom of adjoining valleys. However, they are best devel- 
oped at lower levels, and the most prominent knolls and 
ridges stand in the center of valleys. 
2. They endeavor to follow valleys, if such can be found 
trending in their direction. Their situation is apparently less 
affected by the lower canon valleys than by the upper basin- 
like troughs. They will not bend from a idirect course upon 
encountering a low ridge, but are readily compelled to change 
direction upon meeting one of the high ridges wliich consti- 
tute the remains of the peneplain. 
3. When they are in a narrow valley there is usually but a 
single ridge, and only few secondary belts; but in such a wide 
basin as that of the Pecatonica river and Yellow creek there 
are a number of parallel ridges. 
4. The prominence of a belt and the amount of material 
which it contains are dependent largely on the size of the 
valley or basin in which it occurs. Where it is compelled to 
cross the peneplain without any favorable basins or valleys, 
it is so attenuated as to be often difficult to trace, while the 
same belt passing into a valley becomes conspicuously devel- 
oped. 
5. By far the most important feature of the gravel and 
sand belts of this district is their habit of suddenly passing 
into areas of marked development, at intervals averaging two 
or three miles apart. It is important to determine whether 
these areas are original or have been produced by subsequent 
unequal erosion of the belts. The glacial origin of these "areas 
of special development," which, for convenience in discussion 
I shall denominate special areas', is undoubtedly not merely 
apparent. It is true that commonly they are portions of the 
belt where coarse gravel is most abundant, and this has pro- 
tected the deposit from being as deeply alfected by subsequent 
<M-f)sion as are some other portions; but, on the other hand, 
