112 PROCEEDINGS OP THE AMERICAN ACADEMY 



completely. The yellow precipitate formed becomes more dense and 

 compact by boiling with the supernatant liquid. It is nearly insoluble 

 in pure water, but dissolves to some extent even in very dilute nitric 

 acid. Dilute chlorhydric acid decomposes the salt, phospho-tungstic 

 acid being set free. The mercurous salts of the series are the only 

 ones which are sufficiently insoluble to be available in analysis. 



Soluble salts 'of the different series of phospho-tungstates usually 

 possess a strongly marked bitter taste. In a single instance the taste 

 is at once sweet and astringent. The salts which are acid in constitu- 

 tion exhibit a strongly marked acid reaction. 



Analytical Methods. — To determine the sum of the percentages of 

 tungstic and phosphoric oxides the salt was dissolved in water or 

 dilute nitric acid, and the two oxides precipitated together by means 

 of mercurous nitrate with the addition of mercuric oxide, as in the 

 estimation of tungstic oxide in the alkaline tungstates already de- 

 scribed. This method gives good results only when used with great 

 care, as my assistant, Dr. Gooch, observed that the high temperature 

 and long-continued heat required to expel the whole of the mercury 

 also drove off phosphoric oxide, so that the results frequently varied 

 very materially, and were almost always too low. I have sometimes 

 preferred to precipitate the two oxides with mercurous nitrate in 

 small excess from the boiling solution, — as in the case of the tung- 

 states, — and then to evaporate to perfect dryness on a water-bath, 

 continuing the heat until the whole of the free nitric acid was ex- 

 pelled. The dry mass of mercurous phospho-tungstate and basic nitrate 

 adheres somewhat to the evaporating dish, but not so as to render its 

 removal very difficult. Dr. Gooch found that the adhesion was almost 

 entirely prevented by first evaporating the solution and precipitate to 

 a small volume, and then adding water in considerable quantity and 

 evaporating again, this time to perfect dryness. No phosphoric oxide 

 is vaporized on ignition for an hour at a cherry-red heat, and it is not 

 usually necessary to ignite the mixed oxides a second time so as to 

 insure a constant weight, though it is always better to do so. The yellow 

 powder remaining after ignition is a mixture of tungstic oxide, W0 8 , 

 and phosphoric oxide, P 2 5 , from which water removes a portion, but 

 not the whole, of the latter. The quantitative determination of the 

 sum of the two oxides by this method is very nearly, but not abso- 

 lutely, accurate. A trace of phosphoric acid almost always remains 

 in the filtrate from the mercurous salt, and may be detected by evap- 

 orating this to dryness, expelling the mercurous oxide by heat, and 

 then igniting a portion of the residue with magnesium wire. The 



