74 H. S. H. WARDLAW. 
acceptance for any but routine work and where great 
accuracy is not required. 
Of the volumetric methods for the estimation of phos-— 
phate from the ammonium phosphomolybdate precipitate, 
those of Hundeshagen and of Grete depend on the direct 
titration of the phosphate against standard molybdic acid, 
but the majority are simple acid-alkali titrations, the 
ammonium phosphomolybdate being dissolved in excess of 
standard alkali, and the excess determined by titration 
with standard acid. The amount of phosphate present is 
obtained from the amount of standard alkali used up in the 
interaction with the precipitate. In addition to all the 
disadvantages due to uncertainties as to the composition 
of the precipitate, these methods are also generally faced 
with the difficulties attendant on an indistinct end-point. 
Taking the formula given to the precipitate by Hundes- 
hagen, who was the first to investigate the conditions of 
formation of ammonium phosphomolybdate, the reaction 
which occurs with caustic soda is given by the equation: 
(NH,),-PO,.12 MoO,.2 HNO, + 28 NaOH = 
3NH,0OH+Na.HPO,+12Na,Mo0,+2NaNO,+13H,.0. 
On titrating back the excess of acid, therefore, an indicator 
to which NaH,PO, reacts acid, i.e.,a feebly acid indicator 
such as litmus or phenolphthalein, must be used. The 
“ammonia formed in the reaction, however, will exert its 
usual disturbing influence on the sharpness of the end-point 
shown by this class of indicators, and will materially 
influence the accuracy of the determination. Karly methods 
of this type are those of Thilo, Handy, and Pemberton jun. 
In 1902, however, a method which showed a distinct. 
advance on the previous alkalimetric methods was published 
by A. Neumann. In this method the ammonia formed in 
the reaction with caustic soda is eliminated by boiling 
before titration. Neumann therefore avoids all errors 
