2 PROCEEDINGS OP THE AMERICAN ACADEMY 



employed a solution of amnionic molybdate as a test for the presence 

 of phosphoric acid. Sonnenschein* appears to have first shown that 

 phosphoric oxide was an essential constituent of the yellow compound 

 formed. Finally Debray assigned to the ammonium salt the formula 



20 Mo0 3 • p 2 5 • 3 (MH 4 ),0 -f 3 H 2 0, 



and separated the corresponding acid. 



In a paper presented to the Association of German Naturalists and 

 Physicians in August, 1872, f Scheibler described salts of two different 

 phosphotungstic acids, and gave formulas for the acids themselves, as 

 well as for a sodium salt belonging to a third series, all of which, how- 

 ever, he regarded as provisional. Since then nothing further has 

 appeared upon the subject from Scheibler's pen, and I have conse- 

 quently felt at liberty to include the phosphotungstates in my own 

 work. 



My investigation of the complex inorganic acids had advanced but 

 little before I found it necessary to study the alkaline salts of tungstic 

 acid with special care. This study has alone occupied a great deal of 

 time, and has proved one of extraordinary difficulty, in spite of the 

 previous labors of Laurent, Lotz, Scheibler, Zettnow, Marignac, and 

 others. The difficulties in question are mainly these : — 



1. The alkaline tungstates are numerous and unusually complex. 

 Salts of essentially different formulas approach so closely in percentage 

 composition, that the differences lie very near the unavoidable errors 

 of analysis. Thus Scheibler maintains that the formula of a particular 

 sodium salt is 



7W0 3 . 3Na 2 0-f 16aq, 



while, according to Marignac, the same salt must be represented by 

 12W0 3 . 5Na 2 + 28aq. 



The analyses are hardly sufficiently close to decide the question upon 

 purely analytical grounds. 



2. Almost all the alkaline tungstates are efflorescent in a very 

 marked degree. 



3. The salts of one series agree so closely in "chemical properties 

 with those of the next, that distinctive tests are wanting, and analysis 

 does not always suffice to distinguish two salts even when unmixed. 

 Mixtures are naturally very hard to deal with. 



* Journal fur Prakt. Chemie, liii. 342. 



t Berichte der Deutschen Chem. Gesellschaft, v. 801. 



