Constitution of the Natural Silicates. 409 



Taking ripidolite as the mineral of the three which has been 

 most completely examined, we may recall that two concordant 

 experiments with gaseous hydrochloric acid gave 13*36 and 

 13 - 46 per cent of removable magnesia, presumably representing 

 the group MgOH. In mean, these percentages correspond to 

 34 atoms of magnesia. Regarding this as an index of the 

 MgOH present, we may combine the remainder of the 

 hydroxyl with the sesquioxides to form the univalent group 

 A1H 2 2 , and the ripidolite formula now becomes (A1H 2 2 ) 38 

 (MgOH) 34 B/ / M H !n (SiO 4 ) 60 ; with three oxygen atoms unac- 

 counted for and negligible. Generalizing this expression we 

 have 



R"J-*'..(SiOJ 10 . 

 or almost exactly, R '„(Si0 4 ) 2 R' 4 . 



This is an olivine formula, with half of the R" replaced by 

 R' 2 , and is strictly comparable with the formula of serpentine. 

 It will be remembered that von Wartha* some time ago ad- 

 vanced the opinion that the chlorites and serpentines form one 

 continuous series of minerals, and his view is by this discussion 

 curiously supported. Furthermore, the probable juxtaposition 

 of the groups A1H 2 2 and MgOH in ripidolite accounts in 

 great measure for the apparent formation of spinel when the 

 mineral is decomposed by heat. 



The ratios found by analysis between H, MgOH, and A10 2 H 2 , 

 indicate that ripidolite is probably a mixture of two isomor- 

 phous molecules ; and the observed data are best satisfied by 

 assuming the compounds Mg 2 (Si0 4 ) 2 (MgOH) 3 H and Mg 2 (Si0 4 ) 2 

 (A10 2 H 2 ) 3 H in equal proportions. For a mixture of these 

 molecules in the ratio of 1 : 1, the composition is easily calcu- 

 lated ; and the results agree well with the analysis. If, in the 

 latter, we recalculate the ferric and chromic oxides to their 

 equivalent in alumina, and compute the ferrous oxide as mag- 

 nesia, reducing the summation afterwards to 100 per cent, we 

 get the following direct comparison between analysis and 

 theory : 



Found. Theory. 



Si0 2 31-18 31-09 



A1 2 3 19-87 19-82 



MgO 35-74 36-27 



HO 13-21 12-82 



100-00 100-00 



A closer concordance could hardly be expected. 



For prochlorite, notwithstanding the uncertainty as to the 

 behavior of the ferrous iron, similar ratios appear. The ex- 



*Groth's Zeitschrift, xiii, p. 71, 1887. 



