SSB Reflections oil some Mincratogicat Systems* 



character, they are varieties ; if in two or three, they form 

 subspecies. If the number of dissimilar specific characters 

 exceed three, the minerals thus characterized must belong 

 to different species. Here, then, is a second idea of the 

 species which has just presented itself; and to reconcile it 

 with the first, we have only to believe that the variation of 

 more than three specific characters is an inseparable conse- 

 quence of an essential difference in the chemical composition 

 of minerals. 



We cannot be perfectly sensible why the number three 

 should be that which characterizes a change in the che- 

 mical composition, and we are tempted to believe that it 

 has been taken at random. It is nevertheless founded on 

 a motive. The species is subdivided into subspecies and 

 variety ; the variety, therefore, is the last subdivision of 

 the species ; and to determine the last subdivision but one, 

 the variation of two or of three has been chosen. It is, 

 therefore, necessary to carry beyond three the difference of 

 species ; that is to sav, (according to the passage already 

 quoted,) the essential difference in the chemical composition. 

 1 have said that there was a very good motive for choosing 

 the number three; but I have not said that it was suscepti- 

 ble of being acknowledged by nature. 



In establishing the difference of species solely on the dif- 

 ference of any determinate number of specific characters, 

 we render it independent of every consideration either of 

 the value of the characters, or of the extent in which the 

 shades of these characters differ. In this hypothesis if. 

 matters not, that two minerals have two or even three cha- 

 racters so different that they form, if we may so speak, the 

 extremes of the series of characteristic analogies: if we can- 

 not discover a fourth which is also different, these two mi- 

 nerals must belong to the same species : or let any number 

 whatever of minerals that I can divide into two series form 

 a difference in three specific characters; let all those of the 

 first portion be opake, ductile, and so soft as to be cut with 

 the nail ; let those of the second be transparent, not ductile, 

 and so hard as to resist the file; let all the other characters 

 be entirely similar in every respect, still the individuals 

 which compose these two portions are of the same species, 

 since the number of different characters does not exceed 

 three. Moreover, let us suppose all these minerals of a 

 brownish' red colour, except one only, which maybe a red- 

 dish brown. Here we must now renounce these principles 

 in which the single shade between brownish red and reddish 

 brown, (for M. Werner places them in two different species 



among: 



