EROSIOXAL HISTORY OF DRIFTLESS AREA 73 



ville limestone, 100 feet of St. Peter sandstone, and 

 many feet of Prairie du Chien dolomite have been removed 

 in such a way as to leave thousands of acres of flat land 

 on the divides, 500 feet above present drainage. The three 

 dolomites are cherty and resistant to mechanical wear, al- 

 though much of the rock is soluble in water. The sandstone 

 and shale are non-resistant physically, but resistant 

 chemically. These same rock formations, in varying 

 amounts, have been removed in making all of the many 

 upland flats of the Driftless Area. On few of these flat 

 surfaces is there any concentration of residual materials 

 such as chert fragments, save those which have been round- 

 ed by stream action, although such residual materials are 

 not entirely lacking everj-where. It seems that the removal 

 of the rocks from positions above the flat upland surfaces 

 must have been accomplished by some agent which was 

 capable of removing products of disintegration and of de- 

 composition, even the coarse material. If these surfaces 

 have always been divides, and this must have been the case 

 if there has been no rejuvenation of streams, all of the 

 originally overlying material could not have been removed 

 by streams, for on many of the flats there are now no 

 streams nor stream channels. It is not conceivable that 

 wind degraded the tops of the divides to make flat summits ; 

 the region has not been glaciated ; waves and currents have' 

 been eliminated. It cannot be that solution by ground 

 water has been the method of degradation of these sur- 

 faces, for much of the material such as shale, the sand- 

 stone, and the chert in the dolomite are practically in- 

 soluble. Even if it be conceived that the shale and sand- 

 stone constituents were removed by the wind and the 

 soluble portions of all the rocks were dissolved and carried 

 away by ground water, there would be left many feet of 

 residual chert. 



(5) It has been made clear that the Dodgeville plain, 

 constructed by joining the various patches of summit plain 

 in each of the three cuestas and the summits of the cuestas 

 across intervening areas, is by no means perfectly flat. 

 But, it is difficult to explain even the rough accordance of 

 summit levels in individual districts and the general slope 



