THE ANTIBERIBERI VITAMINE 191 



the protein and obtaining an extract free from the products of auto- 

 lysis. Concretely, 4.5 kilos of fresh yeast, corresponding to 830 

 grams of dry yeast, was crumbled into boiling water, the mass centri- 

 fuged, the residue washed with water and again centrifuged. Alto- 

 gether, 15 liters of water were used, the combined extracts were 

 evaporated to two liters and poured into three liters of 93 per cent 

 alcohol, giving a 53 per cent alcoholic content by weight. The pre- 

 cipitate, which was inactive, was filtered off and the filtrate, together 

 with the washings, evaporated to 300 cc. This liquid was poured 

 into 1960 cc. of 93 per cent alcohol, giving a 79 per cent alcoholic 

 content. The second precipitate was dissolved twice in water and 

 precipitated with alcohol so that the alcoholic content should be 

 90 per cent. This fraction was supposed to contain almost the 

 entire quantity of vitamine B present in the yeast, and amounted to 

 6.2 per cent of the dry yeast. The preparation had an acid reaction 

 and gave a heavy precipitate with lead acetate and sublimate; 25 

 per cent of this fraction could be precipitated with silver nitrate- 

 baryta. This work shows that the vitamine of non-hydrolyzed yeast 

 is insoluble in strong alcohol. In our experiments this method did 

 riot prove successful, as in each fresh sample of yeast the limits of 

 precipitation have to be established anew. 



There is also the report on the chemistry of yeast by Myers and 

 Voegtlin (503). They showed that vitamine B could be extracted 

 from yeast by shaking the latter with olive oil; the procedure obvi- 

 ously possesses no particular advantage over the usual extraction, so 

 that later they used methyl alcohol acidified with hydrochloric acid 

 for extraction. The resulting extract was concentrated in vacuum 

 and the residue extracted several tunes with dilute hydrochloric acid. 

 In this extract, the purine bases were precipitated with silver acetate. 

 The precipitate was filtered off and more silver acetate added to the 

 filtrate, followed by a saturated solution of baryta. This precipitate 

 was decomposed in the usual manner, and the histidine removed from 

 the filtrate with mercuric sulphate. The vitamine was then pre- 

 cipitated from the filtrate with absolute alcohol. The precipitate 

 was dissolved in water and freed from mercury, whereupon a very 

 active solution was obtained. This solution gave a brown diazo 

 reaction and contained a product similar to histamine. When the 

 solution was concentrated, crystals were obtained which were active 

 only so long as they contained some mother liquor. On washing the 



