The Influence of Salts in Rock Weathering. 167 



either a high or a moderate angle from the horizontal; or such 

 approximate verticality may not exist in any portion of the cliffs. 

 The base of the cliffs is, in various localities, undermined into 

 irregular caves and hollows varying both in breadth and height 

 from a foot or two to several feet, but the floors of the caves and 

 hollows are not always coincident with the lake floors, although 

 frequently they are so. The caves and hollows in places increase 

 in height away from the cliff face, that is, the roofs become more 

 dome-like. This doming is helped by the tendency to form resis- 

 tant films on the outside surfaces of the rocks; and in its results 

 resembles to some extent the " pocket rock " and the effects partly 

 due to " shadow-weathering " of other areas, referred to by Hobbs.'^ 



The undermining may V)e more or less continuous along the base 

 of the cliff's for some yards ; and where strong vertical joints or 

 other division planes occur, a roughly rectangular outline may be 

 given to the part attacked. 



The roofs, sides and floors may be damp, and the roofs and sides 

 have a very scaly appearance owing to innumerable thin rock- 

 flakes (of from one-half to two inches in length and breadth, and 

 usually from one-eighth to one-quarter of an inch in thickness) 

 being shed from the parent rock. On account of their thinness 

 and of their decay dui-ing the process of splitting oft', these flakes 

 can usually be broken by the fingers. This flaking is, in the 

 writer's opinion, chiefly due to the process of " exsudation," which 

 has been defined above. 



Meteoric water percolating from the surface downward must 

 assist " exsudation " by acting as a solvent, however slight or slow, 

 thereby weakening the rocks and making them more liable to fur- 

 ther decay. This water also, by passing doAvn joint planes and 

 dripping on to the floors of the caves and hollows, helps to enlarge 

 such hollows, one mode of such enlargement that has been noticed 

 being the scooping out of small lens-shaped hollows in uneven 

 floors along joints. 



The rocks at the base of the cliffs must, by reason of the constant 

 drawing up of moisture by capillary attraction, aided by the 

 downward percolation of surface waters, be mostly in a more or 

 less soaked state, which must undoubtedly tend to make the rocks 

 less coherent. 



In many places the sun's rays never reach the sides and roofs of 

 the cavities, so that temperature variations must be negligible in 



1 Holibs, \V. n. — Earth Fcatiuvs and their Mcanlvrj. New Ynik, 191-2, pp. 201-200. 



