=" 
243 
A steel rod was also driven horizontally into a drilled hole in the even- 
textured but softer rock of the vertical or overhanging walls of the amphi- 
theater-like excavations beneath the falls, and in each case a mark was 
made upon the rod from which from time to time measurements can be 
made as the gradual weathering of the rock continues. 
The excavations beneath the falls are caused chiefly by the saturation 
of the rock by means of spray and mist carried by waterfall breezes or 
winds during winter floods, when frosts follow and complete the process. 
Better results will in all probability be obtained from the measurements 
of the weathering beneath the falls than from those upon the edge of the 
precipice, since the rate of weathering beneath is quite uniform and since 
the wearing away of the softer rock beneath determines largely the 
amount of breaking away and falling of the harder rocks above. _ 
By means of data obtained from such measurements during a period 
of years, a close approximation to the tate of valley erosion can be 
obtained. These data taken in connection with the length of the valleys 
from the Ohio River to the falls, which in the case of Butler ravine or 
valley is approximately 3,100 feet, and in the case of Clifty is 11,000 feet, 
will give us an approximation of the time since the streams began work on 
the valleys. 
The topography of these valleys and of the surrounding region, which 
gives every indication of the youthful stage of erosion, together with evi- 
dence of anothér character, indicates that these valleys have been eroded 
since the retreat of the Kansan or first great ice sheet. This ice sheet not 
only covered all the region referred to, but crossed the Ohio Valley, if the 
valley were there at that time, and advanced at least twelve or fifteen 
miles beyond in Trimble County, Kentucky. All the region was planed 
off to essentially the same level, and there are now within one and a half 
miles of the Ohio River undrained flats produced by glacial action. 
A close approximation to the time required to erode these valleys 
should, then, give a fairly accurate approximation of the number of mil- 
lenniums that have passed since the disappearance of the first great ice 
sheet from the borders of the glaciated region. This period, even approxi- 
mately obtained, will be of great importance in ascertaining the causes of 
glacial periods, as well as being of interest in the discussion of other prob- 
lems of import. 
