﻿F. W. Clarke — Lithia Micas. 359 



It will at once be seen that the three samples examined by 

 Mr. Riggs were sensibly identical, and not far in composition 

 from the material analyzed by Cooke. In the latter the fluor- 

 ine amounted to 249 per cent, which is only about one-third 

 of the figure found in this laboratory, while the silica, plus that 

 represented in the SiF 4 , is sensibly higher. The integrity of 

 the species, however, seems to be clear; and it cannot be iden- 

 tified with zinnwaldite. Empirically, as deduced from the new 

 analyses, the formula of cryophyllite may be written 



B , eH,K,Li t Al.F 4 (SiO 1 ) 10 ; or, R',Al 8 F 3 (SiO s ) 5 . 



With the "lepidomelane " or annite, quite unexpected re- 

 sults were obtained. The material analyzed by Mr. Riggs was 

 black, brilliant, broadly foliated, and apparently very pure. 

 His figures are given in contrast with those published by Cooke. 



Eiggs. Cooke. 



Si0 2 



31-96 



39-55 



TiQ 2 



3-42 







A1 2 3 



11-93 



16-73 



Fe 2 3 



8-06 



12-07 



FeO 



30-35 



17-48 



MnO 



•21 



Mn 2 3 -60 



CaO 



•23 







MgO 



•05 



•62 



Li 2 G 



trace 



•59 



Na 2 



1-54 



trace 



K 2 



8-46 



10-66 



H 2 



4 25 



1-50 



F 



trace 



SiF 4 -62 



4 



100-46 100-42 



It is at once evident that two entirely distinct micas are here 

 represented, and the question is raised whether the Rockport 

 granites may not contain a series of complex isomorphous mix- 

 tures. Cooke, indeed, pointed out the isomorphism of cry- 

 ophyllite with his samples of annite, and showed that the lithia 

 and fluorine in the latter were probably due to admixtures of 

 the former. We now see that at least three micas are involved 

 in the problem to be solved, and the difficulty of establishing 

 definite formulas is enormously increased. For the present, 

 approximate formulae only can be assigned to these micas, in- 

 volving several assumptions, and representing probabilities 

 rather than complete interpretations of the facts. If we unite 

 the groups Si0 2 and TiO a in the new annite, and regard the 

 ferric iron as belonging partly with the alumina, and partly as 

 having been derived by oxidation from the ferrous state, we 

 have the following general formulas for the two analyses. 



