The Fractional Precipitation of Serum Globulin at Different Hydrogen Ion Activities. Experi- Ments with Globulin Obtained from Normal and Immune (antipneumococcus) Horse Serum

The problem as to whether or not the globulin fractions obtained from blood serum are chemical entities has been the subject of much discussion recently. Svedberg (l), basing his views on particle (molecular) weight determination, came to the conclusion that there is only one serum globulin, the molecular weight of which he determined to be 103,800. Sorensen (2) concluded, however, from solubility measurements that the globulin fractions, like many other proteins, consisted of two components, each of which was of protein nature. He assumed that one of these components was the water-soluble globulin (pseudoglobulin) and that the other was insoluble in distilled water (euglobulin). The peculiar fact t,hat comparatively less globulin was precipitated from concentrated than from dilute solutions under otherwise identical conditions was attributed by Ostwald (3) to the colloidal character of the globulin. He pointed out that the dispersion of colloids in general depends on the amount of solid phase present, for the amount of peptizing agent as a rule is increased in proportion to the total amount of substance used in a solubility determinat,ion. This explanation also involved the assumption that the globulin consisted of at least two components. In the course of an attempt to purify serum globulin by means of electrodialysis (4) it was found that this procedure, even when repeatedly applied to the same substance, always yielded two fractions: one which was insoluble at its isoelectric zone and