810 THEORY OF THE SYSTEM. . 216. 



consideration of mineral productions, and what would be 

 the consequence if, on the contrary, we should meet every- 

 where with exactly the same degrees of difference. 



These degrees of difference must not be valued accord- 

 ing to the number or the kind of properties in which those 

 individuals differ which are not identical. They depend 

 rather upon certain relations of these properties with each 

 other, which will be explained afterwards. A crystal of 

 hexahedral Iron-pyrites is much more different from a 

 crystal of prismatic Iron-pyrites, though they should ex- 

 actly agree in every property except in the form, and what 

 depends upon it, than one crystal of rhombohedral Corun- 

 dum, of the variety called Sapphire, from another of the 

 same species called Adamantine spar. And yet the form 

 of the first is an isosceles six-sided pyramid, that of the 

 other a regular six-sided prism. The first presents almost 

 no trace of cleavage, while the other cleaves very easily 

 parallel to the faces of a rhombohedron : they differ more- 

 over in colour, in transparency, and in many other charac- 

 ters. Such examples are common ; and whoever therefore 

 would determine the degrees of the natural-historical affi- 

 nity according to the number, or even to the kind of pro- 

 perties not agreeing, considering the one as more essential 

 than the other, would act contrary to the principles of 

 Natural History. 



If it were possible to invent a scale for measuring with 

 accuracy the degrees of difference among the non-identical 

 individuals, this would afford most useful assistance in clas- 

 sifying the productions of inorganic nature. But there 

 exists no such scale. We must contrive, therefore, to col- 

 lect several of the non-identical individuals, adapted to this 

 purpose in respect to their properties, and to bring iliem 

 under tlic idea of identity (. 214.). This will enable us to 

 extend the inferences which may be drawn from identi- 

 cal individuals, to such as by themselves do not exactly 

 agree in all their properties. Without this process of ex- 

 tending the idea of identity, it would be impossible to 

 derive from it a sufficiently useful employment in Natural 



