Chemical and Mineraloglcal Nomenclature. 207 



difdairi, Guytoa tells us, " it is fo much more difficult to 

 conceive, as it is an evident facrifice of principles to habits *;" 

 as if the ground of their fyltem of nomenclature were univer- 

 fally allowed, and afforded rules fo Uriel and general as to 

 enjoy the Angular privilege of admitting no exception ! as it 

 there was no fuch thing as principles of convenience, or, if 

 there were inch principles, that they were to be lacrilieed to 

 mere fpeculative truths (if truths) of much lei's importance: 

 a truly har(h, intolerant, and defpotic maxim ! as ill calcu- 

 lated to point out the moll advantageous road to fcience, as 

 the maxim that a ilraight line ihould always be followed, 

 would be to infure us the beft road to the fummit of a 

 mountain, though prefenting in that direction a feries of 

 feabrous and abrupt precipices ; and hence departed from by 

 Lavoifier himfelf, as we have feen in the inftanccs otfallpetre 

 and borax, and indeed by the whole French l'chool in the 

 inftances alt water and diamond, as already mentioned : for 

 common fenfe, in fome initance or other, ieldom tails of af- 

 ferting its rights : yet Lavoifier tells us he was cenfured (by 

 fome chemical bigots) for this condefcenfipn. The principles 

 of religion and jultice are the only that can in no poili'blecafe 

 yield to conveniency. 



Among many juil reflections that occur in the preface 

 to Lavoilier's celebrated elementary treatile of chemiftry, 

 there are fome connected with this fubject that appear to 

 me not quite correct. Thus, p. x. (of the original) he tells 

 us, " that the only way of avoiding thefe errors (unfounded 

 hypothefes) is to fupprefs reaioning, or at leaft to Amplify 

 it as much as pollible, as it proceeds from us, and can alone 

 lead us altrav ; to try it always by the telt of experiment, to 

 preferve only the facts, which are the data given by nature, 

 and which cannot deceive us; to feek for truth only in the 

 natural concatenation of experiments and observations, as 

 mathematicians arrive at the folution of a problem by the 

 fimplc arrangement of the data, &c." From this paragraph 

 wc might be led to conclude that all reafoning fhould be ba- 

 nifhed from chemical inveltigation, or, at lcalt, that only the 

 fimpleft fhould be admitted ; yet it may eafily be fhown that 



* Ann. Chem. XXV. p. 207. 



the 



