386 F. W. Clarke— Theory of the Mica Group. 



/SiO=MgR /SiO=MgR /Si0 4 =MgR 



Al-SiO=Al Al-Si0 4 =MgR Al-SiO=MgR 



\SiO=Al \SiO=Al \SiO=MgR 



To these we may add, as No. 7, the compound A] 2 (Si0 4 ) 6 Mg 9 , 

 the bivalent analogue of No. 3, and identical in type with it. 

 Now, so long as we have only orthosilicate micas to consider, 

 these seven formulae cover all their variations in composition ; 

 provided that fluorine, when present, is represented either by 

 — Mg— F or — A1=F 2 , univalent groups which are included 

 under the general symbol W. Most of the micas appear as 

 intermediate mixtures of these presumably isomorphous types. 

 No. 1 represents muscovite and paragonite, and No. 6 agrees 

 tolerably with some phlogopites. No. 2 may be resolved into 

 a mixture, in equal molecules, of No. 1 and 3 ; and similarly 

 No. 5 may be regarded as composed of Nos. 4 and 6. Nos. 

 5 and 6, moreover, may be simplified into mixtures between 3 

 and 7, so that numbers 1, 3, 4 and 7, represent all the necessary 

 relations. Even No. 4 is possibly superfluous. 



So much for the normal orthosilicate micas. But in the 

 lepidolites, phlogopites, and some muscovites, the ox} 7 gen- 

 silicon ratio is low ; and in the lepidolites especially it approx- 

 imates more or less closely to the metasilicate type. This 

 order of variation is clearly established, while variations in the 

 opposite direction, that is toward excess of oxygen may be 

 questionable. If, however, in any mica the oxygen, can be 

 properly in excess of Si0 4 , that excess may be fairly regarded 

 as present in the group — A1=0, which is obviously equivalent 

 to — A1=F 2 , and takes place with the latter as a part of R'. 

 Examples of this kind are given in one of my former papers.* 

 In all such cases the system of formulae proposed above applies 

 perfectly, and needs no qualification. The variations in R ; 

 always fall within its limits. 



The lower values for the silicon-oxygen ratio are explicable 

 as follows : The polysilicic acid H 4 Si 3 8 , which, like H 4 Si0 4 is 

 tetrabasic, is represented in nature by orthoclase and albite. 

 In anorthite we have an orthosilicate, and its mixture with 

 albite gives, as is well known to all mineralogists, the inter- 

 mediate triclinic feldspars in which pseudo-metasilicate ratios 

 often appear H 4 Si 3 8 +H 4 Si0 4 =H 8 Si 4 12 =4H„Si0 3 . If we 

 assume a similar state of affairs among the micas, and regard 

 orthosilicates and polysilicates as isomorphously miscible, the 

 lepidolites and other low-oxygen micas are completely 

 accounted for. We have then the same system of general 

 formulee for all micas, the normal salts Al 4 (Si0 4 ) 3 and Al 4 (Si 3 8 ) 3 

 being the theoretical starting points for derivation. In every 



* This Journal, Aug. 1887, p. 131. 



