9 



a system for the classification of rocks. But even 

 in this department, the involved, obscure, and 

 often unintelligible nature of the affinities and 

 qualities of the substances appertaining to it, have 

 hitherto prevented, or rather defeated, all attempts 

 at a natural classification ; and will probably long 

 deprive us of an arrangement which shall be na- 

 tural, and, at the same time, useful. 



These apparently insuperable obstacles, which 

 as yet impede a classification of minerals, equally 

 interfere with a classification of rocks. But even 

 those are not the whole ; since we must add to 

 them the additional difficulties which arise from 

 the innumerable and apparently capricious modes 

 in which these minerals are intermixed and varied, 

 and from the consequences in the composition 

 and structure of rocks thence resulting. 



If we first examine the effects which would arise 

 from assuming the composition of rocks as the basis 

 of an arrangement, we shall find, that, although 

 nature presents us with many mineral species, 

 a very limited number only of these enters into 



