EXPERIMENT STATION BUiLLETINS. 625 



It will be seen that solutions "A" and "B" remain practically con- 

 stant, but that the alkaline ones "C" and '"D" lost enough ammonia 

 during the digestion and subsequent filtration to reduce their Ph values 

 approximately 0.2. It would therefore appear to be a futile procedure 

 to adjust with great care an alkaline solution to be used in this work 

 since its reaction changes so much during tlie analysis. This perhaps 

 constitutes a rather strong argument in favor of the truly neutral solu- 

 tion, since this maintains its original character throughout the determi- 

 nation except in certain cases, one of Avhich was met with in Sample No. 

 3216. This sample, it will be observed, consistently caused an increase 

 in the acidity of the citrate solution. 



These results are notably at variance with those of Bosworth (7) 

 whose method of attacking the problem seems however to be open to 

 criticism in that he passed a current of air over the digestion liquid. 

 Obviously the continued removal of the air from the flask with the con- 

 sequent removal of ammonia would tend to increase the extent of the 

 dissociation of the ammonium salt and hence magnify what under the 

 prescribed conditions of the determination evidently is a matter of small 

 account. Even on a priori grounds, without the support of experiment 

 it is difficult to conceive of a great change in the reaction of the liquid 

 when the escape of the ammonia is reduced to the extent that it is in 

 the procedure under discussion. In solutions C and D one would 

 naturally expect the magnitude of this tendency to be increased but it 

 will be observed that even in these solutions only one samjjle yielded an 

 acid filtrate and that was unquestionably due to some acid in the sample. 

 Hence we must conclude that the old bogie of an acid solution resulting 

 from this procedure must be abandoned. 



On pure calcium pliospUates: Not only have experiments been con- 

 ducted upon the action of ammonium citrate solutions on commercial 

 fertilizers but supposedly pure di- and tricalcium phosphates have also 

 served as materials for investigation. Considerable mention has already 

 been made of this, and several articles from the literature referred to. 

 Briefly, the results seem to show that an acid citrate solution attacks 

 the tribasic salt to a much greater extent than does an alkaline one but 

 as the alkalinity is increased the solubility of the dibasic phosphate 

 also diminishes. Both the neutral and alkaline solutions appreciably 

 dissolve the tricalcic as well as the dicalcic salt and hence do not give 

 a clear separation of the two. 



Gladding (19) as a result of his work, makes the following interesting 

 comment : 



''That artificially precipitated phosphate of lime, whether the dicalcic 

 or tricalcic form, when mixed with sulphate of lime (as is the case in 

 a superphosphate) and exposed to atmospheric conditions in shade in 

 open vessels for several days, and dried to a pulverulent powder, is 

 readily and conveniently soluble in a neutral solution of citrate of 

 ammonia, at the temperature of 40° C." 



Could the truth of this contention be conclusively demonstrated, it 

 would give an added importance to the results obtained by the citrate 

 method since it could be argued that the distinctive difference in the 

 conduct of tricalcium phosphate towards ammonium citrate before its 

 conversion into acid phosphate and after it has gone through this 

 79 



