ACCURACY OF NEUMANN’S METHOD. 89 
Table VI. 
Influence of amount of phosphate estimated on error of Neumann’s 
method. 
Estimation. | P,O, present.) P,O, found. gatas oe Pee clas 
31 51:0 mg. + 14:3 +0°6 mg. 
32 Osos: ET = 0:1. ,, 
33 44°5 mg. 50-08 12°1 —0O-4 ,, 
34 AQ OMe. 11:9 —0°5 ,, 
Mean 50°4 ,, 1322 
12 oO + 3:4 OO 
13 U3d92" 5, 3°8 +0°2 ,, 
14 22°3 mg. DS 2°2 —0:2 ,, 
15 ZOO a 3°6 Ooi. 
Mean ZOO, a. o°4 
35 ily i ae + 0:9 Onl; 
36 oy aaa + 3°5 +0°2_,, 
37 bh 25emes | yl lie Fs + 3:5 aOe Digs, 
38 PD — 0:9 —0°3,, 
Mean 12a +1°8 
These results show that the error increased rapidly with 
the amount of phosphate estimated. For 44°5 mg. of P.O; 
the average error is 13°27%, for 22°3 mg. it is 3°4%, for 
11°25 mg. itis 1°8/%. The results for 11°25 mg. of P.O; are 
peculiar as among them there is one with a negative error. 
The percentage error of these results shows a considerable 
variation, but the actual differences of the results from the 
mean value show that the ‘absolute error is no greater than 
those shown by the other series of estimations. 
2. Rate on addition of Precipitant.—The rate of addition 
of the precipitant is stated by Baxter and Griffin and by 
Artman to exert a marked influence on the excess of 
molybdenum carried down by the precipitate of ammonium 
phosphomolybdate ; rapid addition of the ammonium molyb- 
date leads to the appearance of a precipitate containing 
more molybdenum than is present in the precipitate formed 
