J. G. Goodchild — Glacial Erosion. 327 



all the structural planes near the surface are rapidly widened, by 

 which the removal of the rock is greatly facilitated. 



It will then readily be admitted by most geologists that under 

 purely atmospheric conditions the rock that tends to disappear the 

 fastest is limestone ; next to this shale ; and the slowest of all to 

 weather away is sandstone. 



When subjected to mechanical erosion, as when these rocks are 

 being worn in a river channel, the rates of abrasion are nearly as the 

 relative hardnesses of the three kinds of rock. Shale goes fastest, 

 next to this come the thinner-bedded sandstones, and longest of all 

 in being worn away are the blocky sandstones and the purer kinds 

 of limestone. The last-named rock especially seems able to withstand 

 much of the ordinary wear and tear of even a large stream. In the 

 case of some of the waterfalls, the limestone forming the floor of the 

 ravine is rarely worn down many feet lower at the outer end than 

 it is found beneath the fall. The same remark applies also to many 

 of the harder beds of sandstone. 



Thus far then it seems clear that the rocks that best withstand 

 mechanical erosion are at the same time those that are least able to 

 withstand Subaerial Denudation. Therefore, if Subaerial Denu- 

 dation has really had so much to do with the development of the 

 existing surface characteristics, we ought to find the more prominent 

 features exclusively of sandstone ; while the accompanying limestone, 

 everywhere but near the streams, should be dissolved clean out of 

 sight. But, although there are in the Dale rocks frequent alter- 

 nations of limestones with sandstones and shales, in the majority 

 of cases the more prominent terraces and scars consist solely of 

 limestone. 



If rivers have been concerned in the formation of the features in 

 question, it is difficult to understand how these have retained their 

 regular form in such perfection while the stream that produced them 

 has cut down several hundred feet into the rocks beneath. It is not 

 easy to believe that a river ever extended right across the dale from 

 the highest scar on one side to the corresponding scar on the other ; 

 yet the advocates of the subaerial theory virtually assume that when 

 the scars were formed the rainfall was so much greater than at 

 present that the river filled the dale from side to side. There can be 

 no better proof of the fallacy of this argument than is afforded by the 

 existence of inclined scars that rise towards the lower end of the 

 valley. A good instance is found near Carperby, in Wensleydale, 

 where a limestone scar and terrace, after a rapid descent of nearly 

 four huudred feet in just half a mile, rises again to the same level 

 within twice that distance, in the direction of the mouth of the 

 valley. Yet in this instance the scar and its accompanying terrace 

 are as perfect at the highest point as at the lowest ; and where the 

 bed that forms the scar is faulted, scars of nearly the same character 

 occur at different levels within a few yards of each other on the 

 opposite sides of the dislocation. All such denudation by rivers is 

 limited to the zone between the highest flood-line and the bed of 

 the river — rocks also from points above this are it is true often 



