OF ARTS AND SCIENCES. 119 



•24': 1 sodic phospho-tungsfate (2-i WO3 . P2O5 . 2 Na^O + 27 aq), a 

 beautiful white crystalline precipitate is formed at once. This is but 

 slightly soluble in boiling water. By long boiling with baric chloride, 

 argentic nitrate, or mercurous nitrate, the salt gives crystalline com- 

 pounds, which in the cases of the barium and silver salts are colorless ; 

 in that of the mercurous salt, yellow. So far as can be determined 

 without quantitative analyses, there appears to be in these cases only 

 a simple double decomposition. For analysis the ammonium salt was 

 thoroughly washed with cold water on the filter-pump, drained, and 

 dried on woollen paper. Of this salt, 



1.3577 gr. lost on ignition with W04Na2 0.0833 gr. = 6.14 % 

 1.2921 gr. " " without " 0.0812 gr. = 6.29 % 



1.6822 gr. gave 0.0G46 gr. Sn = 4.88 % SnOg 



1.1596 gr. " 1.0850 gr. oxides =93.57% 



1.1826 gr. " 0.0443 gr. PoO-Mga = 2.39 % 



0.9575 gr. « 0.0182 gr (NHi)20 = 1.89 % 



The analyses correspond well with the formula 



22 WO3 . 2 SnO^ . PA • 2 (NH,)20 + 15 aq, 



5920 100.00 99.70 



STANNO-PHOSPHO-MOLYBDATES. 



The salts of this series are formed under precisely the same con- 

 ditions as the corresponding stanno-phospho-tungstates. It is at least 

 probable that it will be found that the two classes of salts may be 

 formed under other conditions, as, for example, by boiling stannic 

 oxide with phospho-tungstates and phospho-molybdates, or by digesting 

 stanno-tungstates and stanno-molybdates with alkaline phosphates or 

 with phosphoric acid ; but I have made no experimtMits to determine 

 these points. 



Analytical Methods. — Tin may be determined in these salts by 

 fusion with potassic cyanide, provided that, as in the case of the sepa- 



