DliTERMINATION OF PHOSPHORIC OXIDE. 359 



phosphoric oxide ( P2O5). This factor is that used by Pemberton* 

 who used a sodium hydroxide solution made by dikiting 323.8 c.c. 

 of normal solution to i litre. 



Prescottf has come to the conclusion that this factor is too 

 high. 



When Vipond introduced the above method into the Pretoria 

 Laboratory, he empirically determined the factor in two series 

 of experiments. The first series led him to adopt 0.0002775 as 

 the factor for N/io alkali. Later, as the result of his second 

 series of experiments, he informed the writer that he was of 

 opinion that 0.00028 was more nearly correct. This being rather 

 unsatisfactory, the writer was led to making the following deter- 

 minations with a view to ascertaining the exact relation between 

 N/io alkali and the ammonium jihosphomolybdate precipitate as 

 obtained by von Lorenz's method. It is evident from the results 

 obtained by Vipond and by the writer, that the above equation 

 does not represent the reaction when the precipitate is obtained 

 as above. According to Hundeshagen, the precipitate obtained in 

 the presence of nitric acid and ammonium nitrate has the com- 

 position, (NH4)3P04.i2Mo03.2HN03.H20. A precipitate of 

 this composition would recjuire 25 NaOH for solution, or, 

 doubling we have the equation : 



(NHJ«P20g.24Mo03.4HN03.2H,0 -f 50 NaOH 

 = 2(NH,),HPO, -f (NHJ2M0O4 -f 23Na2MoO, + 4NaNO, 



+ 26H2O. 



It will be shown that the relation between sodium hydroxide 

 and phosphoric oxide in the above method of titration of von 

 Lorenz's precipitate is in practice that indicated by the above 

 equation. 



A difficulty, however, arises in that ammonium phospho- 

 molybdate of the formula given above contains 3.470 per cent, 

 of phosphoric oxide, while according to von Lorenz his i)recipitate 

 contains only 3.295 per cent. In a number of determinations of 

 phosphoric oxide by the orginal von Lorenz method, we have 

 found that the factor 0.03295 is the correct one. The discrepancy 

 can only be accounted for by analysis of the precipitate. Investi- 

 gations on this point are being carried out. 



In the experimental determination of the relation between 

 sodium hydroxide and phosphoric oxide, the phosphates used 

 were, (i) specially prepared di-calcium hydrogen phosphate, (2) 

 Kahlbaum's potassium di-hydrogcn ])hosphate and (3) samples 

 of fertilisers as received for analysis. 



The sodium hydroxide solution used was standardised 

 against recrystallised and sublimed benzoic acid, and also against 

 recrystallised salicylic acid, succinic acid being unfortunately 



* Journ. .'liner. Chem. Soc. (1894), 16, 278. 

 i Jouru. Agr. Sci. (1914), 6, in. 



