52 THE TRUE PRINCIPLES OF CLASSIFICATION. 



4. The classification should be a natural one, therefore empirical, euiljod}-- 

 ino" all known characters of the rocks. A natural niineralocrical classification 

 of rocks is an impossibility, as it is based on part of the characters only — 

 characters which are unstable. Minerals may serve for the establishment of 

 varieties, but not of species. 



5. Geological age has no value in the classification of rocks, and is not to 

 be employed except incidentally in varietal forms. 



6. The addition of new names in any science, unless they are absolutely 

 necessary for its advancement, is a detriment : the needed names should be 

 taken from those now in use, so flir as possible, and they should be emploj^ed, 

 as nearly as may be, in their most approved sense ; when they belong to 

 varieties they should be defined as such, and placed in their natural relations 

 bo the species of which they form a part. 



7. In the classification of rocks, the original characters ought to hold 

 priority over any of the secondary ones ; and they should give name to the 

 rock, and decide its relations, so long as they exist in a determinable state. 

 If necessary, or convenient, variety names or adjective terms can be added, 

 to mark the special peculiarities of secondary or other origin — but only in 

 especially important cases. 



8. If a rock is found to have the characteristics of any species as its 

 prevailing characters, it should be referred to that species. 



9. Chemical analysis alone, as a general rule, is insufficient to furnish 

 data for naming a rock, since it is to be expected that rocks originating in 

 different ways should have the same composition. 



10. The relation of a rock to its associated rocks in the field is the 

 principal criterion for determining its origin, especially in the altered rocks. 



11. Association alone is an insufficient guide in determining the origin 

 of a rock. 



12. The origin of a rock should have an important bearing upon its 

 classification. 



13. The classification should be the exponent of some general law, which 

 should embody all that is known at present of the rocks, and give promise 

 for the future. 



