284 Professor Necker on Mineralogy considered 



two principal systems of characters; for as we have just seen^ 

 we are not allowed to assemble in the same division, minerals 

 differing in their chief characters. In this way separate groups 

 will be formed of such anomalous individuals, species, genera^ 

 orders, &c. 



§ 41. But now comes the question, — When we are called 

 upon to dispose of these anomalous groups in the classificatioUj 

 and to join them in superior divisions with some other groups^ 

 shall we associate them with those resembling them by their 

 physical or by their chemical properties ? Now such a question 

 cannot be answered a priori^ for, as we have seen before, the 

 chemical properties and composition, as well as the physical 

 characters, are both essential to the existence of a mineral indi- 

 vidual, and in this respect none has a right of pre-eminence 

 over the other. 



It is then necessary to examine these two distinct systems 

 of properties, and to see whether they possess the same degree 

 of accuracy, of certainty, and of independence, of every consi- 

 deration foreign to themselves, and to give in doubtful cases the 

 pre-eminence to the most certain, and the most independent con- 

 siderations. Now in this respect, we do not hesitate to say that 

 the physical characters must be preferred. 



In these characters there exists an evidence so simple, so im- 

 mediate, so palpable, that their manifestation becomes obvious 

 to all those who consider them in the manner pointed out by 

 the first observers; and hence it happens, that although now 

 an unprecedented degree of accuracy has been given to the 

 physical and external characters, it is nevertheless to be observ- 

 ed that the descriptions and observations of more ancient mine- 

 ralogists as Werner and Haiiy, deficient as they are in that mi- 

 nute accuracy now required, continue still to be true as far as 

 they go, so that Werner's descriptions, as given by his disciples, 

 Jameson, Brochant, &c. may still be applied to the minerals for 

 which they were intended. They have been extended, impro- 

 ved, and rendered more accurate, but they have never been 

 proved to be false and consequently useless. 



The case is different with the notions entertained in regard to 

 the chemical composition of minerals, for the progress of know- 

 ledge on this subject has shown not only that the ideas of more 



