88 H. S. H. WARDLAW. 
date, each formula weight of the precipitate containing | 
the proportion of molybdenum found in the present case 
should require for its neutralisation 29°5 mols of NaOH 
instead of 28 as assumed by Neumann, so that the results 
calculated by means of Neumann’s factor would be (29°5/28 
—1) 100 = 5°2% too high. This excess of molybdenum is 
therefore sufficient to account for the high results shown 
in Table lV. The error of these results is + 4% + 1°3%, 
Influence of some conditions on the Error. 
The influence of the following conditions upon the error 
of the results obtained in the estimation of phosphorus by 
Neumann’s method have been observed :— 
. Amount of phosphate estimated. 
. Rate of addition of precipitant. 
. Length of time between precipitation and filtration. 
Pe Wh 
. Temperature of precipitation. 
1. Amount of Phosphate estimated.—Forty cc. of 107% 
ammonium molybdate solution are stated by Neumann to 
be a suitable amount for the precipitation of any quantity 
of P.O; between two or three milligrammes and sixty 
milligrammes. As, however, errors had already been 
encountered in the use of Neumann’s method, three series 
of estimations were performed on different amounts of 
standard phosphate solution to ascertain whether these 
errors remained about the same size or depended upon the 
amount of phosphate precipitated. The proportions of the 
reagents used for the precipitation were those given pre- 
viously; the amounts of sodium hydroxide used to dissolve 
the precipitates were, of course, increased in proportion to 
the amount of phosphate present, but the procedure was 
otherwise as described above. The following are the results 
obtained :— 
