Dr. J. E. Reynolds on certain Mineral Silicates. 281 



(b) The next hydrates are those of 2Si0 6 ; and Salvetat's 

 \Siy 

 mineral; Randanite*, H-Yo, exactly corresponds to one hydrate 



of this body. Fuchs's compound f is probably another hydrate 



\SiV 

 of the same acid, viz. H 6 \^0 3 ; and we may have Ebelmen's 



\sry 



compound J as a third of the series, H^^Yo 3 . 



(c) In this subdivision we have the hydrates of 3Si0 6 , 

 which are few in number. Some varieties of semiopal appear to 



\siv 

 consist of the first hydrate, Hy^O. It will be seen further on 



/O ls \ 

 that orthoclase can be best represented as constructed on the 



\si 3 / 

 type of the hydrate H 12 Yo 6 . 



It is a singular fact, and one well worthy of remark in this 

 connexion, that the proportion of water found in all the native 

 hydrates of silica hitherto analyzed is such that the specimens 

 containing most water would be best represented by the formula 



\Si/ 



H^Yo, and those containing least water fall into the c series 



/O b \ 



\SiV \SiV _ 



above, as H 2 Yo, while Randanite, IT^Yo, is the intermediate 



/0 1S \ /0 12 \ 



term required. It is quite true that we know little of the func- 

 tion of the water in these compounds; but the above proportion 

 would appear to be the result of more than mere accident, espe- 

 cially when viewed in connexion with the relations exhibited be- 

 low between some of the silicic hydrates and their corresponding 

 metallic salts. 



* Quoted in Dana's ' Descriptive Mineralog)^,' new edition, 1868, p. 200. 

 The above is the formula of the mineral dried at 100° C. 

 f Ann. der Chem. und Pharm. vol. xxxii. p. 119. 

 X Ann. de Chim. et de Phys. (3) vol. xvi. p. 129. 



