1913] Stanley R. Benedict and Emil Osterberg 43 



through the mixture is immaterial. As aeration proceeds, however, 

 the phosphate is gradually decomposed, even by the carbonate, so 

 that at the end of an hour the phosphate crystals may no longer be 

 readily found. But once having had these crystals formed in con- 

 siderable quantity we must conclude that the ammonia precipitated 

 in this form is not quantitatively liberated again, because, as Steel 

 has shown, starting with larger amounts o£ magnesium ammonium 

 phosphate, one can scarcely recover fifty percent of the ammonia 

 where carbonate is used. The abundant formation of the triple 

 phosphate in the early stage of Folin's original method accounts 

 fully for the slightly higher figures we have obtained where hydrox- 

 ide was the alkali employed. It also accounts for the long time fre- 

 quently necessary to obtain a maximal yield of ammonia in the Polin 

 process. Using the same air-current, we have observed that either 

 carbonate or hydroxide would give theoretical results at the end 

 of an hour, using pure ammonium chloride Solution, whereas in the 

 case of urine, sodium hydroxide would still give a maximal yield in 

 one hour, while sodium carbonate required nearly two hours to give 

 its maximal yield. 



In the light of the above-mentioned facts it is obvious that if 

 Steel's Observation that magnesium ammonium phosphate is not 

 quantitatively decomposed by carbonate be correct (and we have 

 ourselves repeatedly verified this conclusion) , it f ollows as a matter 

 of course that Steel's modification must give higher figures with 

 all urines, and this is exactly what we have found to be the case. 

 The difference appears to be in favor of Steel's modification. We 

 cannot account for Steel's results in this connection. 



As a conclusion we wish to point out that the question of triple 

 phosphate formation, and its incomplete decomposition by carbonate, 

 is one which applies to all urines in varying degrees, and we are of 

 the opinion that whatever procedure is used, one should be adopted 

 which will ensure complete decomposition of the phosphate whether 

 originally present in the urine or not. In discussing Steel's findings 

 Folin^ recommended that urines which contained crystalline mag- 

 nesium ammonium phosphate should first be treated with sufficient 

 acid to dissolve the crystals, after which seven to ten grams of potas- 



3Folin: Jour. Biol. Chem., 8, p. 497 (1910). . 



