Dr. J. E. Reynolds on certain Mineral Silicates. 277 



presently seen that no such sacrifice is really needed. I have 

 little doubt that if a reasonable compromise could be effected, the 

 transition from the notation of the old to that of the new system 

 would become so easy that many mineralogists might be induced 

 to adopt the latter who are now opposed to the change. I hope 

 to show presently that this may be done in a satisfactory manner. 



It might be objected that the exhibition of the oxygen ratio of 

 a silicate by a formula is not an accurate statement of fact — that 

 the oxygen acts merely in binding together the acidific and basic 

 elements of the compound, and that the amount of oxygen em- 

 ployed in this way can be only measured by the atomicities of 

 the principal constituents of the salt. To this it might be fairly 

 replied that but little is really known of the function of the oxygen 

 contained in a silicate ; and though the above statements may 

 be true, all the decompositions with which the mineralogist is 

 acquainted lead him to the general conclusion that each atom of 

 silicon ( = 28) in a silicate has two of the atoms of oxygen more 

 intimately associated with it than the next one. As a striking 

 example of such a decomposition, we might instance the direct 

 liberation of SiO 2 ou fusion of an anhydrous silicate with the 

 residue of the ignition of microcosmic salt. It appears to me to 

 be quite as reasonable to separate the oxygen of a silicate into 

 two sections in accordance with such a decomposition, as to di- 

 vide the oxygen into portions according as part may be supposed 

 to bind together the molecular edifice, and the residue not to be 

 so engaged*. In the absence of distinct proof to the contrary, 

 the utility of the former course would appear to render its adop- 

 tion preferable. 



Accepting SiO 2 (Si =28) instead of SiO 3 (Si =21) for silica 

 as the constitution of that body, which has been established 

 beyond a doubt by the recent admirable researches of MM. 

 Friedel and Crafts f, I take, in a restricted sense, the view now 

 generally admitted by chemists, viz. that silicic anhydride is 

 capable of almost indefinite polymeric modification ; that is to 

 say, we may have not only SiO 2 , but Si 2 4 , Si 3 6 , Si 4 8 , 

 &,c, all being simple multiples of the primary SiO 2 . The 

 chief evidence in favour of this theory is afforded by the com- 

 position of some of the silicic hydrates, the constitution of 

 which can only be satisfactorily explained on the above view. 



* It is necessary to state distinctly here, that in taking this view of the 

 matter I in no way seek to represent that the molecule, in its normal state, 

 is made up of a metallic oxide and silica according to the plan of Berze- 

 lius ; and it will be seen that I guard against such an idea in the construc- 

 tion of the formulae adopted, which shows the mutual dependence of the 

 elements constituting a compound. 



t Ann. de Chim. et de Phys. (4) vol. ix. p. 5. 



