1813.] upon JVeiiier's Mineralogica! Method, 240 



siveness of the suites which constitute particular characters (a* 

 colour) will necessarily vary in different species ; hecause it will 

 depend in some measure on the permanency of the other cha- 

 racters. In some species the colour suite is very extensive; 

 because the other characters may nearly coincide in a great 

 variety of specimens, which, notwithstanding, may ditFer in 

 their colourso The same observation may be applied to the other 

 characters. But after a species is once defined or described (for 

 the description is the only possible definition of it), then every 

 suite of characters becomes just as charjacteristic of the species 

 as if it were not susceptible of the least variation. 



Such is the method which Werner has prescribed to himself 

 in forming species. -And I appeal to Mr, Chenevix himself 

 whether it is not the best, or rather whether it is not the only 

 method possible ; provided it be true that most stony bodies are 

 not in fact divided into species by nature, but that the division of 

 them into species is entirely conventional and artificial. This, 

 however, is a conclusion which I am sensible Mr. Chenevix will 

 not readily admit ; for he affirms that all minerals which deserve 

 a place in a mineral system are divided into species by nature. 

 This also must be the opinion of Haliy. It is an opinion which 

 might be plausibly supported if we were to adopt their method 

 of proceeding, that is to say, reject -|-|ths of the whole mineral 

 kingdom as unworthy of notice, and make our selection ad 

 libitum out of the remaining 20th. But it is an opinion which 

 every person, who conceives mineralogy intended to convey 

 information respecting minerals in general, and not confined to 

 small spots and corners of the system, will find it impossible to 

 establish. 



That we may be enabled to form some kind of judgment, let 

 us compare the definition of Haiiy with the minerals as he has 

 arranged them into species. That definition consists of two 

 parts : let us consider them in succession. 1. As to the first 

 part, it is obvious that, as far as chemical analysis goes, there 

 does not exist any identity in the composition of those minerals 

 which he places under the same species. Thus in three speci- 

 jriens of mica analysed by Klaproth the proportions of alumina 

 were 20, 34, 12, respectively; the proportions of iron, 15, 5, 22; 

 and those of potash, 15, 9, 10. On^e of the species contained 9 

 per cent, of magnesia, while the other two exhibited no trace of 

 that earth. But it is needless to multiply examples : the fact is 

 universally known and acknowledged. I refer Mr. Chenevix to 

 the various analyses of felspar by Vauquelin, Klaproth, and 

 others ; to his own analyses of sapphire compared with those of 

 Klaproth ; and, indeed, to almost all the analyses of stonpr 

 bodies hitherto published. Mr. Chenevix, indeed, admits this 

 perpetual diversity ; but he has fallen upon a method of account^ 



