268 M. Beudant on the Connexion between [A tint, 



classing this compound under each of its component species, 

 and world thus render his classification much confused. 



If, on the other hand, we class them erystallographically, they 

 will all be arranged with the sulphates of iron ; but hence would 

 result the inconvenience of classing with one substance a crystal 

 containing an accidental mixture of as far as 97 per cent, of other 

 substances, although distinguished from pure sulphate of iron, by 

 addmg the terms mixed with such or such other substances. 



On comparing the modes of reasoning of the chemist and the 

 crystallographer, a decided advantage appears in favour of the 

 latter. The chemist would be undecided whether to consider 

 the compound as a mixture or as a triple salt, were it not for 

 the conformity observed between the proportion of acid contained 

 in it and that required to saturate the bases separately, which 

 proportion may be merely conjectural. But the crystallographer 

 would decide on its being a sulphate of iron, containing a mix- 

 ture of other elements . Knowing that similar compounds crys- 

 tallize in similar forms, and dissimilar compounds in dissimilar 

 forms, he would be led to conclude from identity of form, in the 

 instance under consideration, that the sulphates of copper and 

 zinc were simply mixed, and not chemically combined with the 

 sulphate of iron. 



But both these methods of classification possess their incon- 

 veniences. In one are ranged, under the sulphates of zinc and 

 copper, salts presenting the form of sulphate of iron ; and in the 

 other, the sulphate of iron is made to embrace salts almost 

 wholly composed of sulphates of copper and zinc. It appears, 

 however, that these anomalies may be, in some degree, corrected, 

 by classing with the sulphates of iron all the crystals presenting 

 the form of that salt ; and if the proportions of that and the 

 other salts be equal, giving it that place alone ; but where the 

 sulphates of either zinc or copper predominate, then placing it 

 also under the class of the predominating salt. 



If this principle of classification be admitted in the instances 

 alluded to, there appears no reason why it should not be extended 

 to mineral substances presenting similar discrepancies between 

 •4heir form and composition. 



But there is a wide difference between our mixed salts and 

 mineral mixtures. The elements which the chemist finds in the 

 salts, are known to combine in definite proportions, to form 

 compounds which are well understood from repeated analysis 

 and synthesis, and of which the crystalline form may be deter- 

 mined with precision. But mineral bodies do not possess this 

 advantage ; for no mineral compound, except some few natural 

 salts, has been produced in our laboratories. We know the 

 composition of minerals only from analysis, which frequently 

 exhibits difference of composition in the same species, and simi- 

 larity of composition in different species. We do. not know the 



