136 THE VITAMINS 



active as the untreated yeast. Extraction with 60 per cent alcohol re- 

 moved between one-half and three-fourths of the vitamin G (assuming 

 no destruction of vitamin in the residue). The extract in the form in 

 which it was fed was much less potent than the residue ; it appeared 

 that over half of the vitamin thus extracted and dried was inactivated. 

 Hence in interpreting their results, the degree of solubility was esti- 

 mated largely from the vitamin potency of the residual extracted 

 material. From this viewpoint it could be shown that there is pro- 

 gressively greater solubility of vitamin G in the more aqueous solutions 

 of alcohol. From dried bakers' yeast, alcohol 95 per cent by weight 

 extracted no measurable amount of vitamin G; alcohol 80 per cent by 

 weight extracted about one-quarter ; and alcohol 60 per cent by weight, 

 about one-half of this vitamin. 



As noted above, considerable of the extracted vitamin G was in- 

 activated under these experimental conditions. Bisbey (1930) found 

 that such loss could be largely obviated if the exposure to air was 

 reduced to the minimum during the extraction and concentration proc- 

 esses. In her work it was found that absolute alcohol at room tempera- 

 ture (21° to 25° C.) extracted only a small proportion of the vitamin 

 G (B2) from skimmed milk powder, whereas in parallel experiments 80 

 per cent alcohol extracted about half. 



Alleman (unpublished data) interpreted the finding of McCollum 

 and Kruse (1926) that a more potent extract of water-soluble B could 

 be obtained, if the alcohol were acidified with gallic acid, not as 

 due to superior solvent property associated either with the acidity or 

 molecular structure of the acid used as suggested by them, but as 

 due to a conservation of the extracted vitamin, perhaps protected 

 by a higher reduction potential of the gallic acid present. Both Bisbey 

 and Alleman observed that acid in an alcoholic medium appears to 

 catalyze the inactivation of the vitamin, thus confirming similar obser- 

 vations of Chick and Roscoe (1929) and Narayanan and Drummond 

 (1930). 



Halliday (1929) found that vitamin G of "protein-free milk" was 

 adsorbed to about the same extent whether the adsorption took place 

 at pH 3 or pH 4, and with widely varying proportions of Lloyd's 

 reagent. Under the conditions employed, about one-half the vitamin 

 was inactivated, probably due in part at least to oxidation, and of the 

 remainder about two-thirds was found on the adsorbent and the rest in 

 the filtrate. This relationship appeared to hold fairly constant regardless 

 of the amounts of adsorbent used within the limits investigated (5 to 

 80 grams per liter). 



