August 26, 1898.] 



SCIENGE. 



239 



(14Mo03.6P03Ba3)BaO + 55 H,0. 



and we may safely infer that the type is 



(mMo03.6P03E3)pEO, 



there being an internal and external basi- 

 city, as in the cases above noted. 



It is worthy of note that in all the known 

 acids containing phosphorus in some oxi- 

 dized form other than orthophosphoric 

 acid the number of molecules of the acid 

 united to tungstic or molybdic oxide is 

 greater than in the ortho-phosphotungstates 

 or ortho-phosphomolj'bdates containing an 

 equal number of molecules of tungstic or 

 molybdic oxides. 



Two classes of phosphorus compounds re- 

 main to be considered — those which contain 

 phosphorous and those which contain hypo- 

 phosphorous acid or corresponding salts. 

 Chemists are not wholly agreed as to the 

 rational constitution of either of these acids, 

 but it is perhaps more probable that the 

 phosphorous molecule is represented by the 

 formula 



5H.P0(0H)J 



and the hypophosphorous molecule by the 

 formula 



{H^PO.OHJ. 



The typical formula for salts containing 

 the former is 



mW03.u [H.PO(OR),pE'0 

 or m'Mo03.n'JH.PO(OR),p'E'0. 



and for those which contain the hypo-phos- 

 phorous molecule 



mW03n {H.PO.ORJ pRO 

 m'MoOjU' \ H,PO. OR \ p'R'O 



Both groups of salts appear to belong to the 

 first class of complex acids or salts. It is 

 worthy of notice that the phosphorous and 

 hypo-phosphorous salts do not reduce cop- 

 per from its solutions even on boiling, and 

 they can, therefore, hardly be supposed to 

 be simply double salts in the ordinary ac- 

 ceptation of the term. 



We owe to Blomstrand the discovery of 

 a specially remarkable and interesting class 

 of complex acids, the molybdo-periodates 

 or, in the notation which is more commonly 

 used,periodo-molybdates. The acid of this 

 series has the formula : 



12Mo03.ip,.5H,0. 



and is the analogue of phosphomolybdic 

 acid. 



24Mo03.P,05.3H,0. 



The salts of this series are 5-basic, but, as 

 in the case of metatungstates and meta- 

 molybdates, these are composed of two or 

 more normal salts. The best defined of 

 these has the formula : 



(12Mo03.ip,.5K,0) 

 -f (12Mo03l,0,-2K,0.3Hp) + 30 H,0. 



which we may compare with 



(24Mo03.PP53(]SrHJP) 

 -f (24Mo03.PA-2(NH^),O.H,0)16 H.O. 



In the complex acids, of which the present 

 paper treats, we may consider the teroxides 

 of tungsten and molybdenum as the de- 

 terminants. Oxides, of which SiO^, P^O, 

 EjOj and 1^0, may serve as types, may be 

 regarded as sub- determinants. The chem- 

 ical potential of the compound will, of 

 course, be a function of the chemical poten- 

 tials of both determinants and sub-deter- 

 minants. Protoxides appear not to form sub- 

 determinants, but in all cases to act simply 

 as bases. One of the difliculties of our sub- 

 ject consists in determining how far the 

 sub- determinants proper may act as bases, 

 and especially whether they are not some- 

 times partly bases and partly sub-determi- 

 nants in the same salt. 



In my published paper I have made two 

 assumptions as regards the role of vanadic 

 pentoxide in the vanadio-tungstates and 

 vanadio- molybdates. 



The first is that VJO2.O3 may replace "WO3 

 so that chemically 



mWO, + n(VA)03 



