414 Clarke and Schneider — Experiments, etc. 



all these salts should be substitution derivatives of normal 

 aluminum orthosilicate, from which the more definite micas 

 develop as follows : 



Normal orthosilicate Al 4 (Si0 4 ) 3 . 



Muscovite Al„(Si0 4 ) 3 KH 2 . 



Normal biotite Al 2 (Si0 4 ) s Mg s R' 3 . 



Normal phlogopite Al (Si0 4 ) 3 Mg 3 R' 3 . 



Applying these formulae to the expression given above for 

 the Burgess phlogopite, and regarding the fluorine as present 

 in a group — Mg — F, we have for the composition of that 

 mineral: 



Al(SiOj 3 Mg 3 KH 2 + Al i (Si0 4 ) 3 Mg 2 K(MgF), 



the two molecules being mixed in the ratio 1:1. Recalculat- 

 ing the original analysis to 100 per cent, uniting Ti0 2 with 

 Si0 2 , Fe 2 3 with Al 3 6 3 , FeO and BaO with MgO, and Na s O 

 with K a O, we have this comparison : 



Found. Calculated. 



Si0 2 . 41-04 41-09 



A1 2 3 17-59 17-46 



MgO 27-39 27-39 



KjO 10-62 10-73 



H 2 0. 2-03 2-07 



F._ 2-29 2-17 



10096 100-91 



Less oxygen -96 -91 



100-00 100-00 



The results for the non-fluoriferous phlogopite from Edwards 

 are less satisfactory. Its formula, condensed a little from that 

 given above, is 



in which the ratio between R ;// and Si is 1 : 3*5 nearly. But 

 in this mica, three atoms of magnesia are removable by gaseous 

 HC1, corresponding to 3MgO£L If we assume that this rep- 

 resents a small admixture of a foliated serpentine, and deduct 

 proportionally, there remains 



Al 2 ,Mg 65 H 40 K 20 Si 69 O 



2 66' 



which is very nearly Al(Si0 4 ) 3 Mg2KH 2 , or normal phlogopite. 

 At first, as the mineral occurs in a talc mine, we suspected 

 that its anomalies might be due to intermingled talc; but its 

 complete decomposability by hydrochloric acid showed that 

 supposition to be incorrect. If the mica theory is correct, this 



