278 SCIENCE PROGRESS. 



article is more especially concerned, namely, the inferences 

 drawn from pseudomorphs as to the chemical constitution 

 of minerals. 



It is a curious fact that, disregarding those which are 

 obviously mere casts or moulds, and therefore represent no 

 necessary chemical relations, the true alteration pseudo- 

 morphs (metasomatic products) in which one mineral has been 

 gradually transformed into another are few and simple. 



If well proved and indubitable cases alone are examined 

 it will be found that any given mineral usually alters into 

 only a few definite compounds, and that any given com- 

 pound is found pseudomorphous only after a few definite 

 minerals. It becomes, therefore, important to consider 

 whether there be not some relationship between the initial 

 and final product independently of the mode in which the 

 pseudomorph has been produced, and whether there may not 

 be some community of composition between those minerals 

 which result in the same product. 



Chemical analysis of the natural silicates explains, as a 

 rule, nothing but their percentage composition, and any 

 attempt to understand their constitution more fully is 

 necessarily based upon highly speculative reasoning. For 

 a long time the only clue available was that afforded by the 

 mutual replacement of one radicle by another in different 

 samples of the same mineral, or in closely allied minerals 

 belonging to the same isomorphous group. 



In a recent memoir upon the constitution of the Silicates 

 {Bulletin of the United Stales Geological Survey, No. 125, 

 1895), F- W. Clarke points out that the mineralogist does 

 not, like the organic chemist, deal with bodies ol known 

 molecular weight which can be measured by the density of 

 a vapour or by cryoscopic methods ; consequently the only 

 means available are: (i) isomorphism, which indicates 

 similarity of chemical structure; (2) dissimilarity of form 

 with identity of composition, e.g., in Andalusite and Kyanite, 

 which must be taken to indicate isomerism ; (3) pseudomorphs; 



(4) the artificial production of alterations in the laboratory ; 



(5) the thermal decomposition of silicates ; (6) their artificial 

 synthesis. 



