63 



of the particular stratum, but for that of all those 

 which may lie above it in succession, as far as 

 the boundary of the secondary strata. 



It is obvious that the consequences of this 

 proceeding, as far as it relates to the determina- 

 tion of the class of any series of rocks from this 

 principle, is to reason in a circle ; an assumed 

 class first giving the name to a certain number of 

 species, and those again being used to prove the 

 existence of that very class. Whatever resem- 

 blance to this process may ocaasionally occur in 

 describing the primary and secondary classes, it 

 is evident that, in these, we can always have re- 

 course to the infallible test of geological position ; 

 if not actually in the spot under consideration, 

 yet deduced from ample experience in other 

 situations. To this test the transition class is 

 seldom or never properly amenable. 



The arrangement here to be adopted is 

 therefore the first, or the division into two classes, 

 with certain modifications. For the form pri- 

 mitive, there is also substituted the word pri- 

 mary, which is merely of a relative nature, 



