46 Of ChJJ'if.cailon. why a body fhould occupy one particular place rather than another place, there being no relation to connedt it with the preceding. Hence ic follows, that chofe bodies which re-* femble each otber moft:, fhould be grouped toger- ther; and, Confequently, that there fhould be as many heads of general divifion as there are ge neral grounds of refemblance, If it be afked whence this refemblance fhould betaken; whether from the external or internal charadters ? I attfwer, it fhould be taken from both, it being the fefult of the joint confideration öf both. Now npon exatnining the iotality of homoge- rreous earths and ftones, it will be found that thofe refemble each otber moft that contain the •Iargeft proportion of the fame Jimple earth, or moft of the charadlerizing properties of the fame fimple earth ; and, as there are nine fimple earths, it follows that there mnft be nine genera, or pri- mary divifions of bomogeneous earths and ftones. In the fame manner, under each head or ge/ius, we may conftder thofe fubftanccs as fpectficaliy dif ferent that refemble each other lealt, or that dif- fer in fome important property relatively to hu man ufes; for lyftems being fabricated to help the memory, by pointing out the moft important diftindlions, and none being more fo than thofe that are applicable to human ufes, fhould never lofe fight ot their primary end and defignation. I am well aware thatfome very eminent mineralo- gifts would wifh to referve diftindlions relative to pradfica 1 utilitv, for a brauch of mineralogical knowledge, which they intitle Gh.conomical Mjneralogy; and when thefe diftindlions are not ot confidcrable importance, and would intro- duce