236 



SCIENCE. 



[N. S. Vol. VIII. No. 191. 



In considering tlie theory of the complex 

 inorganic acids it will be most advantageous 

 to begin with the group of silieo-tungstates 

 discovered by Marignac. Of these there 

 are four classes, two of which are isomeric. 

 "We may write the general formulas of the 

 neutral salts as follows, omitting water of 

 crystallization : 



10 W03.SiO,.4EO 10 W03.SiO,.2E0.2H,0 



11 WO,.Si03.4 EO 



f 12 W03.SiO,.4EO 12W03.SiO,2R0.2H,0 

 1 12 W03.SiO,.4 EO 



It is, of course, possible that the second 

 group, containing eleven molecules of 

 tungstic oxide, may have twice the molec- 

 ular mass given, so that the general for- 

 mula is : 



12 WO3.SiO,.4RO + 10 W03.SiO,4EO 



The potassic salt, which is the only one de- 

 scribed by Marignac, will then be the ana- 

 logue of the double sodium and zinc meta- 

 tungstate, which I have described, and 

 which has, at least probably, the formula: 



12 WO3.5Na,O+10WO34ZnO 



Taking now the first series or group, the 

 general term may be written in one of sev- 

 eral ways, as, for example : 



(1) lOWOj.SiO.CEO), 



(2) 10WO3.EO+SiO,(RO)3 



(3) 10WO32EO + SiO,(EO), 



(4) lOWOjSEO+SiO.RO 



(5) 10WO34EO + SiO, 



Formulas (1 ) and (2) appear to be ex- 

 cluded by the fact that silicates of the 

 forms SiO,(EO), and SiO.XEO), are not 

 known to exist, the only forms of silicic acid 

 occurring in silicates containing a single 

 base, in other words, in single as distin- 

 guished from double salts, being metasilicic 

 acid SiOjHj and ortho-silicic acid SiO^H^ or 

 SiOjHjO and Si022H,0 or corresponding 

 salts. On the other hand, formulas (3) 



and (4) are admissible so far as the silica 

 is concerned, but considered as double 

 salts they require us to assume the exist- 

 ence of 10 WO32EO and 10 WO33EO, 

 which are not known to exist. Formula 

 (5) is also admissible, if we suppose that a 

 molecule of water is present, and thus we 

 have 10 W03.4EO + SiO,OH3. 



Marignac considered all the acids of this 

 series as quadribasic and consequently that 

 two molecules of basic water were present 

 in each of a number of well defined salts. 



From the above it appears diflicult to ex- 

 plain the constitution of the various silieo- 

 tungstates and tungsto-silicates if we con- 

 sider them as simply double salts. We 

 must assume either that tungstates not 

 known to exist or that silicates which have 

 no analogues or that both these classes enter 

 together into the structure of the complex 

 molecules. 



We get no assistance from a considera- 

 tion of the mode of preparation of the com- 

 plex salts in question. Marignac obtained 

 them by boiling gelatinous silica with meta- 

 tungstates. Gelatinous silica has no defi- 

 nite constitution, but appears to be simply a 

 mixture of two or more forms of silicic acid. 

 It has not been shown, I believe, that solu- 

 ble alkaline silicates dissolve tungstic oxide 

 ■ on boiling, but Parmentier found that silico- 

 molybdates are formed by the action of al- 

 kaline silicates upon alkaline molybdates 

 in presence of silicic acid, and Pechard has 

 shown that a silico-molybdate of the twelve 

 molecule series is formed by the action of 

 fluosilicic acid upon the 14:6 metamolybdate 

 of ammonium. 



The same reasoning applies to the silieo- 

 tungstates and tungsto-silicates, which con- 

 tain twelve molecules of tungstic oxide, 

 since the last named classes of acids also 

 unite with four molecules of base. 



In this connection three remarkable 

 double salts deserve notice. These have 

 respectively the formulas: 



