OF THE CLASSIFICATORY SCIENCES. 213 



obtain FIXED CHARACTERS of the kinds of things; 

 and the criterion of the fitness of names is, that 



THEY MAKE GENERAL PROPOSITIONS POSSIBLE. 



I proceed to review the progress of certain sci- 

 ences on these principles, and first, though briefly, 

 the science of Mineralogy. 



Sect. 2. Of Mineralogy as the Analytico- 

 classificatory Science. 



MINERALOGY, as it has hitherto been cultivated, is, 

 as I have already said, an imperfect representative 

 of the department of human knowledge to which it 

 belongs. The attempts at the science have gene- 

 rally been made by collecting various kinds of 

 information respecting mineral bodies ; but the sci- 

 ence which we require is a complete and consistent 

 classified system of all inorganic bodies. For che- 

 mistry proceeds upon the principle that the consti- 

 tution of a body invariably determines its proper- 

 ties ; and, consequently, its kind ; but we cannot 

 apply this principle, except we can speak with pre- 

 cision of the kind of a body, as well as of its 

 composition. We cannot attach any sense to the 

 assertion, that " soda or baryta has a metal for its 

 base," except we know what a metal is, or at least 

 what properties it implies. It may not be, indeed 

 it is not, possible, to define the kinds of bodies by 

 words only ; but the classification must proceed by 

 some constant and generally applicable process; 



