﻿F. W. Clarke — Lithia Micas. 



361 



tion is made that double salts are to be represented as substitu- 

 tion derivatives of normal salts of similar type, and the results 

 appear as follows : 



Muscovite. Annite (Cooke). 



/SiQ 4 =R 3 ^ /Si(X=R, 



Al^-SiO = 

 X SiO = 



Al 

 Al 



Alf-SiO. 



X 



3 



Al 



Annite (Riggs). 



/ 



sio; 



R 2 



AK-Si0 4 =Al 

 \SiO =R. 



/SiO 



Al^-SiO 



;sio 



AlfSiO 



= R. 



Lepidolite. 



SiF0 3 = 

 Al— Si0 3 . 



R 2 



Al 



Cryophyllite. 



/SiF0 3 = R 2 

 Al^-SiO.-R 



>Si0 3 

 AlfSi0 3 -R 



\SiFO -R n 



To zinnwaldite, as represented by the published analyses, a 

 probable structure is not easily assignable ; and the variations 

 in its composition suggest that the specimens examined have 

 been mixtures of isomorphous compounds. It needs to be 

 thoroughly reinvestigated, not only for itself, but also in its re- 

 lations to whatever other micas may be associated with it. The 

 matter of association can never, in the study of the micas, be 

 safely neglected. 



Although the formulas herein assigned to annite and cry- 

 ophyllite are to be regarded merely as trial hypotheses, there 

 can be little doubt that the ratios between the sesquioxides and 

 the silica are correctly given. These ratios are the ratios indi- 

 cated by the published analyses of phlogopite; with which, 

 therefore, rather than with ordinary lepidomelane, the Rockport 

 micas are chemically to be classed. Zinnwaldite is already so 

 classed by Tschermak and others, and cryophyllite falls easily 

 into the same category. The subjoined formula fairly repre- 

 sents a typical phlogopite. 



.SiFO s =:Mg 

 Al— Si0 4 = MgH 



\ 



/o-^ 4 - 



Si0 4 = Mg 



Al^Si0 4 = MgK 



\SiFO 



Mg 



