VITAMIN G (B2) 113 



same solubility relations toward several solvents was criticized. "In 

 fact it is quite conceivable that the treatment of natural foods v^ith 

 boiling alcohol breaks up combinations of vitamins with other sub- 

 stances so that the removal of the water-soluble B by this solvent may 

 have been a combination of chemical and purely solvent action." 



In reviewing the literature on the solubilities of the water-soluble 

 growth-promoting and the antineuritic vitamin, both were shown to be 

 soluble in water and dilute alcohol and, generally speaking, insoluble 

 in fat solvents. Water-soluble B was said by Osborne and Mendel 

 (1917a) and by Drummond (1917) to be insoluble in absolute alcohol, 

 and by McCoUum and Simmonds (1918), only incompletely soluble in 

 95 per cent alcohol. On the other hand, the antineuritic vitamin had 

 been said to be soluble in absolute alcohol by Eijkman (1911), Funk 

 (1911), Fraser and Stanton (1910) and others who had worked on its 

 extraction and concentration. The conclusion that the growth-promot- 

 ing factor is of somewhat lower solubiHty than the antineuritic was 

 thought to be even more clearly shown in the case of acetone and ben- 

 zene. From McCoUum's laboratory, it had been reported that acetone 

 and benzene extract the antineuritic vitamin from fat-free wheat em- 

 bryo (McCollum and Kennedy, 1916) but that the same solvents do 

 not extract the water-soluble growth-promoting substance to any appre- 

 ciable extent (McCollum and Davis, 1915b; McCollum and Simmonds, 

 1918). After extraction with alcohol, however, the latter appears 

 to be only slightly soluble in acetone but more readily soluble in ben- 

 zene. 



(Recently, however, Williams and Waterman (1926) repeated the 

 experiments of AlcCollum and Simmonds, using rice polishings as the 

 raw material and pigeons as the experimental animals in preventive 

 tests, and found the active material to be soluble in benzene only in 

 the presence of small amounts of water such as might readily be present 

 in an alcohol extract. "It seems safe to assume," they concluded, "that 

 the effective solvent is the low boiling ternary mixture of alcohol 18.5 

 per cent, water 7.4 per cent, and benzene 74.1 per cent, which distils 

 off at 64.85° C. from a mixture of any proportions.") 



In discussing the attempts at isolating the two vitamins, Mitchell 

 criticized the work of Funk and Macallum (1916a) and of Eddy 



(1916) on the ground that the basal rations used in making the feeding 

 tests, besides their deficiency of water-soluble B, "could not have con- 

 tained more than a trace of fat-soluble A." The results of Drummond 



(1917) while supporting "the view that water-soluble B and the anti- 

 neuritic vitamin are identical do not constitute a demonstration. The 



