EXPERIMENT IN MINERALOGY. 251 



still because the relations between isomorphous substances 

 are but imperfectly understood. The present state of our 

 knowledge on this subject is summarised in a valuable 

 treatise recently published by Arzruni (2) as a section of 

 Graham-Otto's Lehrbuch der Anorganischen Chemie ; its 

 application to mineralogy is to be found in the great treatises 

 of Dana (3) and Hintze (4). 



But reference to such books will only convince a critical 

 reader of the necessity of further experimental research. 

 Instance upon instance may be found in the very best ex- 

 plored regions of the science. The felspars by all their 

 characters declare themselves to be an isomorphous group, 

 and we have therefore to harmonise the empirical formulae 

 KAlSi 3 O s of orthoclase and CaAl 2 Si 2 O s of anorthite ; but 

 whether the formula of the latter is to be doubled and CaAU 

 is to be supposed equivalent to Si. 2 in the former, or whether 

 we are to suppose that Si 3 8 is a radicle which can replace 

 Si0 4 , is, without further evidence, a matter of the purest 

 speculation ; it is true that an appeal can here be made to 

 the products resulting from the decomposition of felspars, 

 and the constitution formulae generally adopted are based 

 upon the kaolinisation of these minerals ; in this connection 

 reference may be made to a suggestive paper by Scharizer 

 (5) on the metasomatic products of the silicates. 



EXPERIMENTS ON CHEMICAL COMPOSITION. 



A preferable method would be to investigate the 

 action of various physical and chemical processes upon the 

 mineral, in order to determine what proportions of its con- 

 stituents are combined in different manners within the 

 molecule. Very important in this respect are the numerous 

 observations of Doelter upon the solubility of minerals and 

 their behaviour in the presence of various solvents (6). 



Particularly striking and promising among recent re- 

 searches are those of Clarke upon the silicates, especially 

 the mica group (7) ; the experiments were made to ascer- 

 tain how far they conform to his view that such stable com- 

 pounds as the natural silicates must be far less complex 

 in constitution than is commonly supposed, and that they 



