446 ON THE ACTION OF RUNNING WATERS. 



We often see furrows scooped out on the walls 

 that bound the narrows of rivers ; we also see rocks 

 rounded, and entirely destitute of moss. But let the 

 facts be examined with attention, and we shall find 

 that this erosion always takes place in the parts of their 

 course, where, on account of the nature of the neighbour- 

 ing soil, the torrents carry with them, in their risings (or 

 floods), debris and detached stones from their banks ; 

 and it is by means of these stones that they wear the 

 rocks which are in their bed. 



It is very easy to appreciate these circumstances. It 

 is remarked, that this erosion has never taken place 

 at the sources of powerful springs. All the pebbles 

 which had to be carried off have been so long ago, 

 and the mosses which grow abundantly on the rocks 

 at the level of the water, and in the bed of these torrents, 

 have nothing more to fear from the destructive action of 

 these solid bodies. The case is the same with the parts 

 which immediately succeed a lake, or a great excavation, 

 capable of arresting all the hard bodies carried off by the 

 river* There the mosses appear in abundance ; because 

 they are not subjected to the action of any other sub- 

 stance than of the water alone. 



The present rivers do not therefore appear to have 

 any erosive power upon the rocks which are completely 

 aggregated, when they act by themselves, and when no 

 other cause, such as frost^ decomposition, &c. has disin- 

 tegrated the rock. The absence of these foreign circum- 

 stances is proved by the vegetation or the enamel which 

 then cover the rocks exposed to the action of the water. 

 These rivers, in proportion as they remove from the 



