Casimir Funk 5 



that could be regarded as vitamine itself or a decomposition pro- 

 duct of it. The results of the extractioii of such enormous qiian- 

 tities have shown beyond doubt that, because of the instabihty of 

 vitamine, ordinary methods are inadequate. 



Yeast undoubtedly contains more vitamine than do any of the 

 known food-stuffs. Extraction with alcohol, however, removes 

 only a very small part of the total vitamine content. More promis- 

 ing results with yeast than with rice were described in the earlier 

 paper (6). The alc. extract, after hydrolysis with acid, was frac- 

 tioned according to the methods described in my earlier papers. 

 From the silver-and-baryta fraction a crystallin product was ob- 

 tained which melted at a temp. above 200° C. and which, when ad- 

 ministered to beriberi pigeons, induced very quick recovery. When, 

 however, maintenance experiments were performed, i. e., with 

 pigeons on a diet of polished-rice with daily injections of the pro- 

 duct, the birds could not be kept alive for more than a few days(8). 

 This shows that the vitamine originally in this product had already 

 undergone chemical changes that were so profound as to cause it to 

 lose some of the pharmacological properties of the original yeast. 

 The product of this first crystallization was divided into three sub- 

 stances, one of which was nicotinic acid. The administration of 

 these substances, individually or collectively, to beriberi pigeons did 

 not show appreciable curative action. 



Summarizing the results of previous chemical investigation of 

 vitamine, we find that this substance is precipitated by a series of 

 reagents; the curative fraction always contains, especially in the 

 case of yeast, the three substances mentioned above (including nico- 

 tinic acid), which, however, when purified, do not exert curative 

 action. Vitamine is extremely unstable and the most delicate 

 methods destroy it completely. The destructive factor remains to 

 be elucidated. Heat and acid (4) are not very important destruc- 

 tive agents. There is the further difficulty that vitamine is very 

 readily adsorbed by colloidal precipitates. This tendency cannot, 

 however, account for the total loss of vitamine, since the fractions 

 thus obtained are markedly curative. Other probabilities are de- 

 struction by alkali, loss by oxidation, and the presence of one or 

 several labile chemical groupings in the vitamine molecule. 



