﻿360 F. W. Clarke— Lithia Micas. 



Annite, Cooke, R' 8 Al 4 (Si0 4 ) B . 



Annite, Riggs, R' 14 Al 2 (Si0 4 ) 5 . 



For the former, the equivalent of R' 8 is approximately K 2 H 2 Fe" 2 , 

 and for the latter it is K 2 H 4 Fe 4 . These values correspond to 

 the following percentage compositions. 





Cooke. 



Riggs. 



Si0 2 



39'5 



36.6 



A1 2 3 



26-8 



12-4 



FeO 



18*9 



35-1 



K 2 



12-4 



11-5 



H 2 



2-4 



4-4 



100-00 100-00 



Attention has already been called to the fact that some speci- 

 mens of cryophyllite form borders upon plates of annite, pre- 

 cisely as the lepidolite of Auburn is arranged about nuclei of 

 muscovite. It accordingly becomes quite probable that similar 

 relations connect the two pairs of minerals ; and upon these 

 relations the formulae so far deduced shed some light. In each 

 case we have a mineral with metasilicate ratios implanted upon 

 an orthosilicate ; and a derivability of the one from the other 

 is forcibly suggested. Structural analogies also appear ; for in 

 each couplet we have evidence of a common type of nucleus, 

 which may be represented thus: 



Lepidolite nucleus, Al(SiO s ) 8 . Cryophyllite nucleus, Al 2 (Si0 3 ) 5 . 

 Muscovite nucleus, Al(Si0 4 ) 3 . Annite nucleus, Al 2 (SiOJ 5 . 



The development of complete structural formulas from these 

 nuclei is rendered difficult by our ignorance of the part which 

 fluorine plays in such compounds. In the ordinary lepidolites 

 it is easy to conceive of it as united with the aluminum in the 

 univalent group A1F 2 ; but in the iron micas and phlogopite a 

 similar representation is unsatisfactory. A different solution 

 of the problem is therefore to be sought, and it may be found 

 by an application of the well-recognized principle that fluorine 

 and hydroxyl can replace each other isomorphously. 



If now we start from orthosilicic acid, Si(OH) 4 , and regard 

 the hydroxyl groups as successively replaceable by atoms of 

 fluorine, we can conceive of a series of acids ranging from 

 Si(OH) 4 to SiF 4 ; and with the aid of such a supposition many 

 of the fluoriferous silicates may be rationally explained. For 

 example, the acid SiF(OH), may be considered, and its nucleus, 

 SiF0 3 , a trivalent residue, can be applied to the discussion of 

 the lithia micas. Upon this basis the formulas for lepidolite 

 and cryophyllite become curiously significant, especially when 

 written in direct comparison with the structures assignable to 

 muscovite and annite respectively. In every case the assump- 



