other, as well as to that structure which forms the 

 leading object of geology. On this subject a 

 mineralogical arrangement can throw no light ; 

 as the same limited number of minerals is re- 

 peated, in different proportions or combinations, 

 throughout every part of the whole series of 

 rocks. 



If a mineralogical arrangement of rocks is to 

 be adopted, it should be combined with a classifi- 

 cation of minerals ; of which the object is, to des- 

 cribe accurately, and to class in a manner the 

 most convenient for investigation, all the subjects 

 concerning which it treats. And, as in systems 

 of mineralogy, the simple rocks are described 



among the simple minerals, so the compound 







ought to form a part of the catalogue of accidents 

 in the history of each mineral, and should be enu- 

 merated among the circumstances of position un- 

 der which that occurs. 



If, in such an arrangement, some plan of this 

 nature were not adopted, the rocks would be se- 

 parated under two distinct heads : a plan prod uc- 



