August 3G, 1898.] 



SCIENCE. 



241 



In the present state of our knowledge it 

 is not possible for us to assign to more than 

 a few acids the classes to which they be- 

 long, and a vast field of chemico-physical 

 work is open to the investigation of this 

 subject alone. 



We have also to determine by methods 

 now well known to chemists the molecular 

 masses of metatungstates and metamolyb- 

 dates and of the complex inorganic acids, 

 which may be considered as derived from, 

 or at least as definitely related to them. 

 Certain double salts appear to be composed 

 of metatungstates or metamolybdates united 

 to salts containing subdeterminants, as, for 

 instance, and more especially, platinic 

 oxide. I may, for the sake of clearness, re- 

 fer to such formulas as 

 10 WO,.4Na,O + 10 W03.PtO,.5Nap + 58aq. 

 and 



4f 12M:oOs.2K,0.3H,OJ + 512Mo03.PtO,. 



2Kp.4H,0|+24aq. 



These salts appear not to be decomposed by 



water and to belong therefore to class first. 



The compounds of arsenic oxide with 

 tungstic and molybdic oxides closely re- 

 semble those of phosphoric oxide and do not 

 now require special notice as regards their 

 structure and properties, nor as to the ques- 

 tion of their position as unitary or double 

 salts. 



On the other hand, the salts which contain 

 arsenious oxide present some special points 

 of interest, since this oxide exhibits distinct 

 basic and, in its salts, distinct acid proper- 

 ties. The general formulas for this class of 

 compounds, so far as they have been studied, 

 are 



mWOj.nAs.Oj.pROandm'MoOjn'As.Oj.p'EO 

 As characteristic salts we may take the 

 arsenoso- moly bdates 



6Mo03.3As,03.2EO 



of which four have been obtained in fine 

 crystals. 



The copper and nickel salts of this type 

 have a fine green color and, therefore, con- 

 tain the two metals as ions, at least in part. 

 It is, therefore, most probable that the salts 

 are unitary. This view is supported by the 

 fact that the alkaline salts from which the 

 salts of manganese, copper, zinc and nickel 

 are obtained by double decomposition give 

 reactions which are different from those of 

 the constituent molybdic and arsenious 

 oxides. The same statement applies to 

 other known arsenoso-molybdates and to 

 arsenoso-tungstates. 



I pass over without notice a number of 

 interesting salts, some of which contain for 

 a single determinant at least three different 

 sab-determinants, as, for example, phos- 

 phoric and vanadic pentoxides and vanadic 

 dioxide. We are only at the beginning of 

 our knowledge of these compounds. From 

 their extreme complexity it would seem 

 most probable that they belong to the 

 second class and are in some cases double, 

 triple or quadruple salts, of which possibly 

 the components may belong to the first 

 class. Long and laborious investigations are 

 required, sure to be rewarded by an abun- 

 dant harvest of new facts, of little or no 

 practical value perhaps, but to be looked 

 at from the higher point of view of pure 

 abstract knowledge. 



The term ' complex inorganic ' was at first 

 intended to embrace all the compounds con- 

 taining a relatively large number of mole- 

 cules of tungstic and molybdic oxides as 

 determinants, beginning with the silico- 

 tungstates of Marignac. Hittorf, in 1859, 

 appears to have first drawn the distinction 

 between double salts which are decomposed 

 by solution and those which are stable un- 

 der the same circumstances. Finally, Ost- 

 wald proposed to restrict the term ' com- 

 plex ' to the salts which are not decomposed 

 by solution and which gave none of the 

 chemical reactions of the constituents. 



Of these two classes it is difficult to say 



