490 PROTOPLASM 



when conditions are more favorable to it. Gelatin in very 

 dilute concentration and in hot solution is very probably molecu- 

 larly dispersed. In more concentrated and cold solutions, 

 molecular aggregates or colloidal particles certainly exist. The 

 former solution would show molecular (valence) reactions; the 

 latter, colloidal (adsorption) reactions. 



Jacques Loeb held the classical (molecular) viewpoint in 

 regard to all protein reactions and attributed the apparent 

 colloidal behavior of proteins to a 



. . . failure to measure the hydrogen-ion concentration of the protein 

 solutions, which happens to be one of the main variables. When the 

 hydrogen-ion concentrations are duly measured and considered, it is 

 found that proteins combine with acids and alkalies according to the 

 stoichiometrical laws of classical chemistry and that the chemistry of 

 proteins does not differ from the chemistry of crystalloids. 



In addition to maintaining that the physical properties of 

 proteins, their viscosity, their isoelectric point, etc., are deter- 

 mined simply and solely by the hydrogen-ion concentration, 

 Loeb stated further that proteins form true solutions, i.e., 

 molecular and not colloidal dispersions; that they exhibit no 

 selective adsorption ; and that, therefore, there are no Hof meister 

 (lyotropic) series of ions. Loeb was partially right and partially 

 wrong. Gelatin may be molecularly dispersed when in hot and 

 dilute solutions. On the other hand, the colloidal, micellar 

 structure of cellulose, if not of gelatin, has been established 

 beyond all reasonable doubt. 



As for the disappearance of hydrogen from hydrochloric acid 

 when gelatin is immersed in it, the question cannot be definitely 

 answered, i.e., whether the hydrogen is removed by adsorption 

 (colloidally) or by chemical union with the protein to form a 

 salt (by a primary valence bond). It appears that some proteins 

 do form salts with acids and bases, while others do not. Whether 

 they do or do not may rest on whether or not there are exposed 

 any free amino or carboxyl groups. The water solubility of 

 albumin suggests an abundance of such free groups. The poor 

 solubility of globulin suggests an absence of such free groups. 

 In the first case, one would expect stoichiometric behavior; and 

 in the second, colloidal behavior. 



Evidence in support of the Hofmeister series is now so abundant 

 (with pH control) that their existence can no longer be ques- 



