282 DENUDATION TO LAND Chap. YI. 



of every prominence at nearly the same level, 

 and would indent the turf between them ; and 

 such intermediate indentations would again 

 arrest the castings. An irregular ledge when 

 once formed would also tend to become more 

 regular and horizontal by some of the castings 

 rolling laterally from the higher to the lower 

 parts, which would thus be raised. Any pro- 

 jection beneath a ledge would not afterwards 

 receive disintegrated matter from above, 

 and would tend to be obliterated by rain and 

 other atmospheric agencies. There is some 

 analog} 7 between the formation, as here sup- 

 posed, of these ledges, and that of the ripples 

 of wind-drifted sand as described by Lyell.* 



The steep, grass-covered sides of a 

 mountainous valley in Westmoreland, called 

 Grrisedale, was marked in many places with 

 innumerable, almost horizontal, little ledges, 

 or rather lines of miniature cliffs. Their 

 formation was in no way connected with the 

 action of worms, for castings could not any- 

 where be seen (and their absence is an 

 inexplicable fact) although the turf lay in 

 many places over a considerable thickness of 



* < Elements of Geology,' 18G5, p. 20 



