PHOSPHORIC ACID.
23
T h e table on the preceding page shows the effect of the peaty matter and also a comparison of the methods, together with that of the m-phenylene diamine. I n some cases a pink color was obtained and no blue, due probably to the greater delicacy of the napthylamine test, it being competent to detect 0.0001part of nitrogen as N,O, in 100,000. This we think is the extreme limit of the test, as different shades, not depths of color are obtained upon adding different quantities of the reagents, as Dr. J. T. Tanner' found. T h e iodo-zinc starch method is incapable of detecting less than 0.0002 part of nitrogen as N,O, per 100,ooo. I n a few cases a blue color appeared, but no pink, but upon passing carbon dioxide through the water no blue was obtained. This may possibly have been due to hydrogen peroxide. Where large quantities of nitrites are present, a purple color instead of a blue is obtained, which is difficult to estimate ; in such cases the water should be diluted before applying the test. I n conducting the Griess test, the directions given by Dr. Tanner' were followed with the additional precaution of using water free from nitrites in the preparation of the reagents. T h i s was prepared by distilling the middle portion of ordinary distilled water with an excess of alkaline permanganate, collecting the middle portion of the distillate thus obtained. Water prepared in this way gives no test upon eighteen hours' standing, even when tightly stoppered.
-
A GRAVIMETRIC flETHOD OF ESTIMATING PHOSPHORIC ACID A S AflflONIUI1 PHOSPHOMOLYBDATE.3 BY THOMAS S. GLADDING. Received November
T
xi,
1895.
HE estimation of phosphoric acid by weighing the yellow
precipitate of amnionium phosphomolybdate has often been attempted, but, except in iron analysis, where the amount of phosphorus is very small, such a method has never yet been successful. T h e reason of such failure is evident when we consider the analyses that have been made of the yellowprecipitate. A fern7 only need be presented. 1 Report National Board of Health, 2882, 280. 2 Lac. cif 8 Read before the New York Section of the American
.
Chemical Society, Nov. 8, ~895.
T H O M A S S. GLADDING.
24
Ranimelsberg.
H,O ............ 5.77 ........ 3.25 ",OH P,O,. ........... 3.90 MOO, .......... 86.45
Struve and Svanberg.
Sonnenschein.
Gibbs.
9.49 9.49 3.63 86.88
11.23 11.23 3.03 86.87
3.94 3.35 3.66 89.05
Gibbs gives the following formula for the salt :
24MoO,, €',Oh,3(NH,),O +24XloO,, P,O,2 ( N H 4 ) , 0 . H , 0 + ~ 6 A q . H e prepared the salt by mixing solutions of ammonium molybdate (seven parts of water to three of the salt) and phosphate, adding nitric acid in excess to the solution and boiling. Such a method would give more or less occluded molybdic oxide. His analyses gave 3.70 per cent. and 3.83 per cent. of phosphorus pentoxide against the theoretical percentage of 3.66. Such results were sufficiently accurate for his purposes, but would discourage any suggestion of using the yellow salt as the basis of a gravimetric method. His formula is, however, without any doubt, the correct one, with the exception of the water of crystallization. M y own analysis of the salt precipitated in the manner described later on, and dried at a temperature of 10.5' C. to a constant weight, gives me the following composition :
.......................
48M00,. 2P,0j .......................... IO",.
IIH,O
.......................
.........................
Theoretical
91.38 3.76 2.25 2.61
By analysis
91.36 3.76 2.31 2.57
For ammonia 1.015 grams gave 0.0234 gram N H , = 2.30 per cent. For ammonia 20.193 grams gave o.qGgo gram N H , = 2.32 per cent. For moisture 7 . 2j grams gave -= 2.53 per cent. '( = 2.61 " 10.30 ' I T h e water was determined by estimating total hydrogen by combustion with copper oxide. For phosphoric acid, repeated analyses and syntheses gave almost exactly 3.76 per cent. T h e molybdic acid was found by difference. T h e fact that drying at Ioj" C. expels all the water except (
