or 



Dr. J. E. Reynolds on certain Mineral Silicates. 279 



and Si ,v i OR ^ with Naquet, 



LORj 



Si II 4 II R 4 and SiO || O 2 1| R 2 with Dana, 



I employ the following form, constructed, as will be readily seen, 

 on a modified water- type, 



\Si/ \Si/ \HV \H*/ 



H*Vo 2 and IpVV = hVo 2 and hVo, 



1 : 1 and 1 : 2 being the oxygen ratios. 



The use of this construction enables me to avoid the anomaly 

 already referred to. This plan is also adopted in order to show 

 the mutual dependence of the elements constituting the mole- 

 cule of a silicate, the mode in which it is capable of breaking np 

 under the influence of certain reagents, and the oxygen ratio of 

 the compound, and, further, for the purpose of clearly exhibiting 

 the relations of the salt of a condensed acid to its analogues 

 amongst the lower terms of the series. It appears to me that 

 these advantages are gained by the above construction, without 

 involving our acceptance of any particular theory as to the abso- 

 lute structure of the molecule. 



In adopting this plan in the cases of the silicates the formulae 

 of which are given below, 1 seek to show much more than the 

 mere composition of some of the minerals belonging to the 

 several groups. If we disregard the minor subdivisions, it is 

 shown that all the anhydrous mineral silicates may be divided 

 into two great classes — (1) the silicates which contain in each 

 molecule the elements of SiO 2 only, and (2) those which contain 

 but 3 (SiO 2 ) =Si 3 t O 6 , or a multiple of this quantity. 



When the force of this division is realized, I believe that much 

 of the difficulty at present attending the study of the mineral 

 silicates will be found to have disappeared ; moreover the rela» 

 tions of groups of minerals become quickly evident when their 

 formulas are constructed in accordance with the principles above 

 stated. 



* It is obvious that this is equivalent to writing the first hydrate H 4 / 

 Si. /O 4 



and the second H 2 / , but by the construction adopted in the text it 



/O 3 

 becomes possible to express the " harmonies of variations " of the molecule, 

 t Hereafter in the text Si ( = 28) is used to represent the silicon as it oc- 

 curs in SiO 2 , and Si ( = 84) as in SiO 6 , the condensed anhydride. The 

 convenience of this distinction is very considerable. 



