GIBBS. — COMPLEX INORGANIC ACIDS. 261 



of acid tuugstates or molybdates gives such well defined and beautifully 

 crystalline salts that the subject will doubtless attract the attention of 

 other chemists. I will make the suggestion that possibly the salts 

 which I described in my preliminary notice may be obtained by add- 

 ing ammonic chloride to solutions of the double salts described in this 

 paper, so as, in the case of the tungstic compounds at least, to pre- 

 cipitate the tungstic oxide not combined with platinic oxide in the 

 form of 



10 WO3 . 4 Na^O + 4 {10 WO3 . 4 (NHJ.O} + ,50 aq., 



or an analogous salt. The solution should then contain only a platino- 

 tungstate. The application to the platiuo-molybdates described is less 

 probable. 



Second Series of Pyrophospho-Tungstates and Pyrophos- 



pho-molybdates. 



When sodic pyrophosphate is added in excess to a solution contain- 

 ing a metallic salt, the precipitate which is at first found is in many 

 caseg redissolved with formation of a double salt of sodium and the 

 metal in question. In a certain number of cases, the heavier metal 

 in the new compound is not replaceable under ordinary conditions, and 

 does not exhibit its characteristic reactions with tests. These facts 

 are of course familiar to all chemists. Persoz supposed that these 

 salts might be represented, in the case of divalent metals, by the 

 general formula, as we should now write it, 



PAR" + P20,Na„ 



and that the group P2O7R" was to be regarded as electro-negative to 

 the group PoOjrNa^, so that the compound would be simply analogous 

 to ClNa. If we write the double salt PsO-R'^Na,, we may regard 

 the complex PoO-R" as playing the part of a relatively electro- 

 negative group, as in the case of double cyanides, so called. Whether 

 this view is to be considered as identical with that of Persoz, is a 

 question about which opinions may differ, and wliich is not important 

 for my present purpose. Admitting that the groups PgO-R" or 

 P2O-R'" are transferable as such in their relatively simple alkaline 

 salts, we may inquire whether they enter into the composition of com- 

 plex acids, and if so, whether the compounds so formed differ from 

 ordinary pyrophospho-tungstates and pyrophospho-molybdates. As 

 the initial point in this investigation I have selected mangano-disodic 

 pyrophosphate, PgO^MngNaa . 



