193 On Cnjstallograpliy, 



conclude that they indicate perceptihlc dlfTerences of com- 

 position between the specimens analysed. I could produce 

 other examples of a mineral, the different analyses of which 

 made by skilful operators have given difierent products ; and 

 we shall presently sec that this must necessarily take place 

 in a variety of circumstances. 



Now I return, and I ask, on what foundation Mr.Kirwan 

 gives the name oF feldspar to the various specimens which 

 have been the subjects of the analyses above mentioned. 

 This intimacy surely \% not founded on the results of these 

 analyses; since we should rather be inclined to infer, from 

 the differences which they have presented, that a \'q\y at 

 least of the substances to which they are referred, con- 

 stitute distinct species, Iti a word, it is visible that Mr. 

 Kirwan has tacitly supposed that, abstracting all considera- 

 tion of the analyses, the substances in question had been 

 regarded as feldsj->ars. 



On perusing what has been written by naturalists on the 

 subject of this mineral, we see that its place was assigned 

 according to a certain assemblage of characters, such as a 

 hardness capable of producing sparks with steel ; a leafy 

 texture added to its breaking into rhomboidal fragments : 

 a specific gravity of about 2*5 ; a fusibility into a white 

 enamel ; &c. 



But these characters arc for the most part variable to a 

 certain point; and this variation may even be extended far 

 enough, in certain cases, by a consequence of that which 

 the component substances undergo. These are so many 

 useful ways of assisting us to recognize bodies which be- 

 long to one and the same species ; but in addition to their 

 not presenting to the mind a sufficiently simple and precise 

 idea of what constitutes this species, their results are not 

 always proper for tracing the limit which separates one 

 species from another: and it is on account of our being re- 

 stricted to consulting them only, that we have confounded 

 the pyroxene with the amphibolus, the chabasie with 

 the mesotvpe, and so on with several other relative con- 

 nexions, the fault of which will become evident from what 

 wc shall ^ay under the respective heads of the substances 

 to which tht-y are referred. 



There exists a character much more solid and much 

 more proper by its invariability, to serve as a rallying point 

 to the different bodies, which belong to one and the same 

 species. This is derived from the exact form of the inte- 

 grant molecule, because this form subsists, without any 

 sensible alteration, independently of all the causes which 



may 



