EDWTN JOSEPH COHN 699 



Experience has shown that neither of these methods of determining 

 the isoelectric point of a globulin is altogether satisfactory. It is true 

 that the precipitation of a globulin passes through a sharper maxi- 

 mum the freer it is from salt. Freedom from salt, at least as 

 appHed to the proteins is, however, relative. Probably no protein 

 has even been prepared that was completely "ash-free." Moreover, 

 three distinct sources of error may make, and in the experience of the 

 writer have made the point of maximum precipitation of a protein 

 appear at a hydrogen ion concentration other than its isoelectric point. 



The first of these is the presence of another protein with a slightly 

 different isoelectric point. The observed precipitation is in this case 

 the sum of the precipitations of the two proteins. As a result the zone 

 is usually widened, and the point of maximum precipitation shifted 

 in the direction of the isoelectric point of the second protein. The 

 magnitude of the shift, and therefore of the error, depends upon the 

 difference in the isoelectric points of the two proteins, upon their 

 relative concentrations, and upon their relative solubilities. 



The presence of salts containing either bivalent or trivalent cations 

 or anions may also lead to error. For multivalent ions, even in low 

 concentration, shift the point of maximum precipitation to a hydrogen 

 ion concentration other than the isoelectric point of the protein. 

 Hardy (8), Mellanby (9), Osborne and Harris (10), and later Hopkins 

 and Savory (11), have shown that the solvent action of salts upon 

 globulins increases greatly with the valence of their ions. Moreover, 

 cations of high valence are more effective in dissolving globulins on one 

 side of the isoelectric point and anions on the other. For this reason 

 the salts of monovalent acids and bivalent bases, (or the salts of 

 monovalent bases and bivalent acids) exert a greater solvent action 

 upon globulins at certain hydrogen ion concentrations than at others. 

 This also results in a shift in the precipitation zone. 



In greater concentration salts precipitate proteins of all classes. 

 The precipitating action of electrolytes, hke the solvent action, in- 

 creases greatly with the valence (1). Michaelis has recently shown 

 theoretically and experimentally how the presence of salts of high 

 valence shifts the point of maximum precipitation of proteins (12). 



Finally, I have observed that even uni-univalcnt salts of the type of 

 NaCl in relatively low concentrations, may shift the precipitation 



