DISTRIBUTION OF NITROGEN 



107 



This long method was rigidly tested by Van Slyke upon pure 

 amino acids singly, and when mixed together, and upon typical 

 proteins. The accuracy of the method was borne out by the figures 

 under the column of total nitrogen (p. 132); usually, the nitrogen re- 

 covered was within I per cent. In the duplicate analyses of proteins 

 the maximum and average differences were : 



The differences of 2*14 per cent, for the histidine (edestin) and i'6o 

 per cent, for the mono-amino nitrogen (hair) were more than twice 

 any other deviations from the figures in the series. The highest 

 average difference was 079 per cent, for histidine and the lowest 0*05 

 for cystine. 



The correspondence between the figures and the actual amounts 

 of amino acid isolated from the protein is fairly good. The high pro- 

 line content of gelatin was well known, but the non-amino nitrogen 

 content of 1 5 per cent, of the total nitrogen of this protein is very 

 striking. Further, the large amount of lysine in haemoglobin was un- 

 expected ; haemoglobin has always been supposed to be composed 

 chiefly of histidine. 



Van Slyke [1913, 2] published revised data for- caseinogen. They 

 differ considerably from his earlier data and may be partly accounted for 

 by the use of the new apparatus for determining amino nitrogen. The 

 data for gliadin by Osborne, Van Slyke, Leavenworth and Vinograd 

 [191 5] are in closer agreement with the older data of Van Slyke's ; the 



