1826.] affecting Chemico-Mineralogical Systems. 385: 



must be less, therefore, in arranging bodies according to their 

 electro- negative principles. However, when we attempt to place 

 compounds that are liable to vary from the isomorphism of their 

 bases, we meet the same difficulties in arranging them together 

 as in the mode of classing them according to the most electro- 



fositive elements ; but the difficulties are of less importance, 

 have shown in my former essay, moreover, not only that the 

 electro-negative disposition has many advantages, but also that 

 all the oxygenated compounds being in that method arranged 

 under oxygen, the first class in mineralogy, that which is purely 

 inorganic, is thereby subdivided into two parts, one containing the 

 oxidated minerals, and the other those which are not oxidated. 



In all former systems, including those of Werner and Haiiy, 

 attempts have been made to preserve the advantage resulting 

 from a classification according to the positive element. Each 

 metal, properly so called, constitutes in those systems a family 

 which embraces all its combinations. We must renounce this 

 advantage, in a classification according to the negative prin- 

 ciple. More than one mineralogist, perhaps, will not like to 

 seek for iron, copper, or silver, in several families in which they 

 are found dispersed. I must, therefore, show how the classifi- 

 cation according to the positive element may be adapted to the 

 isomorphous changes. The compounds of garnet, tourmaline, 

 pyroxene, &c. may be placed under several bases just as any 

 metal may be placed under its sulphate, 8cc. ; and thus the names 

 of alum, garnet, or tourmaline, no longer denote mineralogical 

 species, but merely indicate modes of combination ; still, how- 

 ever, the exchange of bases, amongst compounds in indefinite 

 proportions, throws some doubt on the place that should be 

 assigned to such or such a garnet; and sometimes the same 

 individual may, with equal propriety, be arranged in more 

 places than one, which always shows a vicious principle in clas- 

 sification. Thus, adopt what method we will, we cannot, if we 

 follow the principle strictly, avoid meeting with something which 

 shocks us by its novelty ; but 1 must add that we are not to 

 reject a thing as false merely because it is new. 



It seems then to be demonstrated, first, that in the present 

 state of our knowledge, it is impossible to determine satisfacto- 

 rily, with regard to minerals in which isomorphous substitution 

 prevails, those that compose mineralogical species ; and 

 secondly, that as the interchanges occur chiefly amongst the 

 electro-positive principles of minerals, their classification, 

 according to the most electro-positive principle, cannot be 

 employed without great difficulty. 



In a system which classes minerals according to their most 

 electro-negative element, compounds in which isomorphous 

 bases replace one another may be naturally arranged near each 

 other, and it is of less consequence whether we separate the 



New Series, vol. xi. 2 c 



