21Q THE GEOLOGY OF MINNESOTA. 



[Topography. 



lating or rolling character except along the immediate river bluffs, where 

 the rocks frequently appear in craggy bluffs and cause precipitous or steep 

 hillsides. 



The valleys excavated by the streams are remarkable and instruct- 

 ive. Not only have the large streams cut gorges of enormous depth 

 in the rocky floors on which they run, but every little creek and tribu- 

 tary runs in a gorge which shows the same rock-sculpture. Even the 

 ireshet creeks, and the rivulets born of every summer shower, dry entirely 

 the greater part of the year, find their way to the main valleys through 

 rock-bound, canon-like valleys. This makes the county present the usual 

 characters of southern latitudes where the northern drift sheet has not been 

 spread. There is nothing more evident than that these valleys antedate the 

 great ice age. In other portions of the Northwest where the drift does pre- 

 vail, larger streams than those found in Houston county have generally 

 worn their channels only through the drift sheet. The Mississippi river 

 itself, above the falls of St. Anthony, has no rocky bluffs. It very rarely 

 strikes the rock. It is occupied still in dissolving and removing the mate- 

 rials of the drift which covers that portion of the state. It would require a 

 great many inter-glacial periods, or pre-glacial periods, to excavate it as 

 deeply as the same valley is wrought in the southeastern portion of the 

 state. In the limestone areas the valleys are narrow and more generally 

 rock-bound ; they widen out so as to inclose good farm lands on the bottoms 

 in the sandstone areas. This distinction, however, is less evident than in 

 Fillmore and Winona counties, where the St. Peter sandstone plays a more 

 important part in bringing about the present topography. It is, however, 

 well illustrated in the upper portion of many of the tributaries of Root river. 

 In descending one of these valleys from the upland the first descent is rocky 

 and very impracticable. This is caused at first by the cut through the 

 Shakopee limestone. The Jordan sandstone that underlies the Shakopee 

 sometimes relieves this ruggedness a little, but its thickness is so small 

 compared to that of the whole series of strata involved that it is barely 

 observable in this way. Through the underlying St. Lawrence limestone 

 the descent is also rough and the valley narrow, with little or no arable 

 land in the valley. On reaching the horizon of the top of the St. Croix 

 sandstone the change introduced into the aspect of the valley is very notice- 



