Dr. J. E. Reynolds on certain Mineral Silicates. 283 



I am well aware that modern views on the molecular construc- 

 tion of condensed compounds would oblige us to deny the possi- 

 bility of obtaining some of the above hydrates in the free state. 

 It need scarcely be remarked that though this is admittedly the 

 case with such unstable bodies as the poly silicic hydrates, yet it 

 is well known that the metallic salts of condensed acids may exist 

 though the isolation of their corresponding hydrates would be a 

 matter of doubt, or (according to present ideas) an impossi- 

 bility. Whatever views on the subject may ultimately obtain 

 acceptance, they are not likely to alter the above general modes 

 of expression for the atomic ratios of the best-known native sili- 

 cates. In the cases of the basic or subsilicates, it is shown above 

 that they correspond to one or other of three hydrates of the 

 first stage of condensation of SiO 2 dealt with in this paper, viz. 

 SiO 6 '. In this way their relations can be clearly brought out, 

 though the cause of the peculiar ratio of R( = H) to Si is difficult 

 to understand. The normal or unisilicates ought to be the most 

 highly basic of these salts ; and yet we find the subsilicate of 

 Class 3 containing twice as much basic element as an unisilicate. 

 According to Dr. Odling this anomaly may be explained thus : — 

 " If we consider alumina as always basic, then there are sili- 

 cates " (such as those above typified) "in which the basicity ex- 

 ceeds that of the tetrabasic group; but the basic or acidulous 

 function of alumina is yet sub judice ; and if we exclude alumi- 

 nous minerals, we shall find that the tetrabasic are the most 

 highly basic of all known silicates, natural or artificial.'" These 

 remarks are specially directed to the proof of the conformity of 

 silicic to other acids ; but they obviously apply forcibly to the 

 question of the constitution of the subsilicates on the types I 

 have given above. 



Bisilicates. 



Class 1. 



\Si/ 



Enstatite . . . Mg"Vo 

 /0 2 \ 



= SiO||0 2 ||Mg" (Dana*). 



\Si/ 



Hypersthene (Mg, Fe)"X° 



Wollastonite 



* With each class I give Dana's new formula for the first member of the 

 group. 



