Claſſes, Genera and Species are therefore to be formed upon the internal nature and compoſition; the varieties upon the external appearances. In ſuch a ſyſtem both methods conveniently agree.
Cronstedt firſt attempted this method, and with great ſucceſs; but afterwards the liquid analyſis, in which the illuſtrious Margraaf took the lead, better opened the internal ſecrets of nature; ſo that the excellent work of Cronſtedt now appears to contain many errors; theſe however are not to be attributed to any fault in the author, but to the inſufficiency of his experiments. The attempts of Mr. Pott by fuſion have long been known; but theſe however uſeful in other reſpects, rather tend to confound than to lay open the component parts of bodies.
In methodizing foſſils, compounds ſhould rank under the moſt abundant ingredient. Thus let a and b repreſent the component parts; if the former be the heavier, the compound muſt be placed under the genus of that: but this rule admits of ſeveral exceptions.
Thus, the properties of all ingredients are not of the same intenſity, if I may be allowed the expreſſion; ſome are more powerful or efficacious, ſo as to impreſs the maſs with their own genus and character, though forming leſs than half the weight. In ſuch a caſe the qualities are rather to be conſidered than the quantity, eſpecially if b ſo far from preponderating hardly ever amounts to half the weight.
Argillaceous earth (earth of allum) and magneſia are never found ſeparate, but almoſt always mixed with other things ſo that their weight conſtitutes the ſmaller part of the maſs: therefore if the above rule (§ 14.) was rigourouſly adhered to, theſe primitive earths would not be found amongſt the Genera, which would doubtleſs be an abſurdity.