622 STATE BOARD OF AGRICULTURE. 



Patten and the author (42) reported the results of the analysis of 

 two samples of superphosphate, using four citrate solutions of different 

 compositions, in which there was a marked increase in the insoluble 

 P2O5 with an increase in alkalinity. 



Eudnick and Latshaw (45) prepared two solutions having ratios of 

 ammonia to citric acid of 1 : 3.786 and 1 : 3.748 respectively. With 

 these solutions, which according to our measurements would have re- 

 actions of approximately 7.2 and 7.8, they found no significant differ- 

 ences in the results of the analysis of one sample of phosphate. 



Shepard (4G) found some tricalcium salt dissolved by ammonium 

 citrate. 



Zulkowski and Cedivoda (54) claimed that dicalcium phosphate is 

 soluble in an excess of triammonium citrate and in the theoretical 

 amounts of mono- and diammonium citrate while tricalcium phosphate 

 is soluble only with difiQculty in a large excess of any of these solvents. 



The originators of the aininonium citrate method designated that the 

 reaction of the reagent should be neutral but the specification was not 

 emphasized further than to emphasize the one word neutral. There is 

 no discussion of the evils resulting from a deviation from this recom- 

 mendation. 



As a matter of fact, as pointed out above, within a short time after 

 the publication of the original method, the use of an alkaline solution 

 was advocated and rather Avidely adopted. Joulie (31) in France and 

 Petermau (43) in Germany suggested such solutions and these acquired 

 considerable popularity in Europe. Similar solutions were also advo- 

 cated by Landrin (34) and Millot (39).^ Very recently Eastman and 

 Hildebrand (15) have recommended the use of a solution of the tri- 

 ammonium citrate as being more accurately prepared than a truly 

 neutral solution though the former solution would be slightly alkaline. 

 In this connection it is of some significance that Luck (36) in his 

 acrimonious reply to Joulie (31) states that, in attempting to avoid the 

 povssibilit>' of the solution becoming acid in tlie course of tlie digestion 

 through the escape of ammonia and to insure the presence of the ''neutral 

 salt" the solution was, after being adjusted to a density of 1.09, brought 

 ''to a iveakly alkaline reaction with a few drops of ammonia." It fol- 

 lows from this admission that the original Fresenius, Neutauer and 

 Luck method involved the use, not of a neutral solution of ammonium 

 citrate, hut of a solution approximating that of the so-called "neutral" 

 salt, i. e., the triatnmonium citrate, which solution icould he slightly 

 alkaline. 



To insure the solution not becoming acid during the digestion, Fresen- 

 ius, Neubauer and Luck (16) further conducted the digestion at the 

 low temperature of 35° in order to minimize the escape of ammonia. 

 This point was later discussed by Gladding (18) and by Williams (53). 



This effect of temperature is two-fold, an increase tending to increase 

 the rate of solution and thus to permit of the more accurate estimation 

 of the difficultly soluble ph'osi)li;)tes of iron and alumiiiiim l>ut also tend- 

 ing to drive off ammonia and thus to give rise to an acid solution which 

 will dissolve an undue amount of tricalcium phosphate. The method 

 originally adopted by the A. O. A. C. called for a bath temperature of 

 40° attained as rapidly as possible after introducing the flask contain- 



1 Who was, however, not its originator. 



