14 



rest for some time the gas was measured, due allowance being made for 

 the nitrogen present in the flask at first and for the small additional 

 amount afterwards introduced. Of course, the proper corrections for 

 pressure, temperature, and tension of aqueous vapor were made in con- 

 nection with all the measurements of gas, and the quantity of nitrogen 

 obtained was divided by 2, one half being yielded by the organic sub- 

 stance and the other by the nitrous acid. 



With a view to guard against the retention of any nitrogen in the 

 form of di-azo derivatives from amids, a moderate excess of hydro- 

 chloric acid was used. In the case of kreatinin the results are com- 

 plicated by the formation of the supposed nitroso compounds observed 

 by Dessaignes. With several of the proteids the bright yellow color 

 of the so-called xantho-proteic acid was well marked, as the conse- 

 quence, doubtless, of the action in the flask of nitric acid produced by 

 the breaking up of nitrous acid into nitric acid, nitrogen dioxid and 

 water. Too little is known of the yellow substance formed, Mulders 

 xantho-proteic acid, to indicate how far its nitrogen comes from the 

 proteid and how far from the nitric acid, or in what direction or to 

 what extent its formation may influence the amount of elementary 

 nitrogen given off. In the case of proteid solutions coagulable by heat 

 it was manifestly important to allow the action of the hydrochloric acid 

 on the silver nitrite to become complete below the temperature of 

 coagulation, as otherwise silver nitrite was entangled in clots of the 

 coagulating proteid. In some, at least, of these experiments, it seemed 

 probable that a part of the nitrogen dioxid given off, and perhaps of 

 the free nitrogen, was due to the reducing action of the carbon or 

 hydrogen of the organic substance upon nitrous or nitric acid, and not 

 solely to the normal interaction of nitrous acid and the amidogen radi- 

 cle present. 



The process was varied, not only in respect to the temperature 

 applied, but also, within moderate limits, in respect to the pressure on 

 raising or lowering the gas-measuring tube in the water. It was further 

 varied by increasing the proportion of nitrous acid to different extents 

 in excess of the theoretically necessary amount. 



In all some fifty or sixty experiments were made. In a number of 

 cases the simpler amidic compounds gave a fair approximation to the 

 quantity of nitrogen theoretically to be expected from them, though 

 even with such substances as leucin, asparagin, and aspartic acid the 

 results were not as sharp as the claims of Sachsse and Kormaun would 

 lead one to suspect. One of the best experiments yielded for aspartic 

 acid 9.57 per cent of nitrogen, instead of the 10.53 per cent actually 

 present. Kreatin did not give one-half of its nitrogen, as has been 

 heretofore recorded as the fact, but a somewhat nearer approach to 

 one-third 10.19 per cent as against 32.00 per cent the probability of 

 which result is indicated by the accepted structure of the molecule. 

 In all cases- the proteids gave some nitrogen, but the proportion was 



