452 ON THE ACTION OF RUNNING- WATERS. 



corals and sponges are fished, are more sheltered from 

 currents than others ; nor that these places, after violent 

 tempests, have been deprived, and consequently, as it 

 were, despoiled of those productions, which, by covering 

 the rocks, demonstrate that they preserve the integrity 

 of their surface. Many of these bodies, however, as 

 sponges, fuci and confervse, contract but a feeble adhe- 

 rence to the bodies upon which th^y are placed. 



It therefore appears, if not completely proved, at least 

 extremely probable, from the facts and reasonings which 

 we have related, 



1. That the presently existing waters, that is to say, 

 in the state of purity in which we are acquainted with 

 . them, have no erosive action upon rocks, whatever be 

 the nature of these rocks, when, 1st, the rocks are com- 

 pletely solid, and when they are neither friable nor disin- 

 . tegrated ; 2d, When these waters act by themselves, that 

 is to say, when their action is not complicated with the 

 really erosive action of solid bodies, such as pebbles, 

 sand, and perhaps even pieces of ice. 



. That water, sometimes acquiring, on account of its 



quality and velocity, a great transporting power, may 



remove masses, already detached, and of great size, ac- 



- cording to its degree of velocity, and the bulk of its mass, 



and so far as it preserves this same power. 



3. That the presently existing waters may have at- 

 tacked, undermined, and caused to fall down, portions 

 of solid and steep rocks, by mixing with beds of clay, 

 marl, and sand, interposed between their solid strata ; 

 that they may also, in their rapid falls, have scooped 

 pretty deep ravines in very inclined deposites, consisting 

 of disintegrated rocks ; but that these waters could not 



