104 THK HOLE OF NITKOGEN IN PLANT-METABOLISM iii., 



ing for a considerable time. The acid added no doubt plays 

 a considerable part in the denaturing, for if an extract be 

 previously neutralised to phenolphthalein with alkali, and 

 then boiled, scarcely any coagulation takes place. When a 

 1% sodium chloride extract is heated very slowly, coagulation 

 begins about 65'' C, and the solution is distinctly acid. The 

 protein has formed a compound with the acids during the 

 exti'action, and the conditions are then assured for heat- 

 coagulation. The precipitate obtained by carbon dioxide in a 

 dilute saline solution is now considered to be composed 

 entirely of globulins. The extent of dilution necessary is of 

 great importance. In this case, 10 vols, of water were re- 

 quired for complete separation. The nitrogen is equivalent tu 

 63% of the tannic acid nitrogen. 



Trichloracetic acid is usually stated as a precipitant for 

 globulins, albumins, and proteoses. The latter, however, 

 may be kept dissolved by boiling and filtering the liquid hot. 

 The filtrate, after cooling, deposits these proteoses, which 

 may then be separated from the perfectly cold liquid by 

 filtration. This small deposit easily redissolved on boiling 

 with water, and was as easily recovered by cooling. Jt 

 amounted to 1-6% of the total nitrogen. By saturation of the 

 extract with sodium chloride, a very small amount of protein 

 is salted out. Finally, phosphotungstic acid was used to pre- 

 cipitate the basic constituents of the non-protein part. Jt 

 was not used as a protein-precipitant. 



A number of attempts were made to dialyse a solution of 

 proteins obtained by salting out with ammonium sulphate, 

 but it was not found an easy task to completely prevent the 

 changes due to fungoid and bacterial contamination from tak- 

 ing place, and they had to be abandoned. The proteins on being 

 salted out by complete saturation with ammonium sulphate, 

 only partially went into solution again, when the precipitate 

 was diluted with water. The figures obtained for albumins 

 and globulins were, therefore, not reliable, and are omitted. 



