892.] On the Estimation of Uric Acid in Urine. 97 



tate always contains a trace of pigment, the results are somewhat 

 high, but the method will be found sufficiently accurate for clinical 

 purposes. 



Evidence as to the Accuracy of the Ammonium Chloride Separation when 



applied to Urine. 



(1.) After urine has been saturated with the chloride, allowed to 

 stand, as described, and filtered, uric acid cannot be detected by any 

 process in the filtrate. 



(2.) The uric acid as weighed gave the following figures on com- 

 bustion. Mixed residues from actual determinations were used : 



Found. 



(a.) (J.) Theory. 



C 35-460 35-990 35-71 



H 2-561 2-590 2-381 



N 32-400 31-980 33'3 



The above nitrogen figures show some variation from theory, but it 

 should be remembered that elementary analysis is a very rigorous 

 test when applied to such unpurified products. I have never obtained 

 figures so near to the truth from the product weighed in the Ludwig 

 process. 



(3.) Comparison with the standard Ludwig-Salkowski process : 



Series 1. 



The experiments in Series 2 will be seen to afford nearer agree- 

 ment than the earlier series, carried out before the same amount of 

 experience had been obtained. The better agreement depends partly 



VOL. LII. H 



