OF ARTS AND SCIENCES. 71 



them or to assign to them their true position as double chlorides, etc., 

 corresponding to double oxides capable of forming complex acids. 

 Examples of this class are the following : — 



SnCl4 + PCls Casselmann. 



SnCl, + POCI3 



TiCl, + FOCI3 AYeber. 



TiCl^ + ^^h Tuttschevv. 



SeCl^ + SbClg Weber. 



VOCL -j- ^^^K Brauuer. 



AW]^ -\- 2 POCI3 Casselmann. 



Yelci + 2POCI3 



SbCL + PCI3 Weber. 



SbCh + POCI3 



It is at least possible that in these and many similar cases complex 

 or double acids correspond to the double chlorides, and it is probable 

 that the action of such double chlorides upon solutions of acid tung- 

 states and molybdates would produce complex acids containing tung- 

 sten or molybdenum and the two acid-forming oxides corresponding to 

 the two united chlorides. I shall at another time specially consider 

 the compounds of platinous and phosphorous chlorides discovered by 

 Schlitzenberger. 



In my paper upon the vanadium compounds of tungsten and 

 molybdenum I have endeavored to show that many formulas may be 

 reduced to the same types as those of the phospho-tungstates and 

 phospho-molybdates, provided that we assume that a part at least of the 

 vanadic pentoxide is to be regarded as replacing tungstic or molybdic 

 oxide, or that we have the chemical equivalence expressed by the 

 equation 



m Vp.^ . O., =: m WO, = 7?^ M0O3. 



This hypothesis had the advantage of simplifying many formulas 

 materially, but was wholly unsupported by any positive evidence. 

 I have now siiown that there exist classes of phospho-vanadates and 

 arsenio-vanadates which do not contain tungstic or molybdic oxide, 

 but which are exactly parallel to phospho-tungstates, vauadio-tung- 

 states, etc. in composition. Thus we have the salt, 



12 V2O5 . P2O, . 3 BaO + 45 aq ; 

 and the acid, 



20 V2O, . P2O, . G ILO + 53 aq ; 



