168 THE VITAMINES 



resulting filtrate, after removal of the excess baryta, was tested for 

 its curative power on beriberi pigeons. The solution was very active 

 and was free of phosphorus, protein, and carbohydrates (I.e. 463). 

 The first difficulty encountered was that of determining the exact 

 dose for a sick pigeon. As will be shown later, the solution con- 

 tained much free choline which is very toxic for pigeons. This 

 difficulty could be overcome by calculating the dosage on the basis of 

 the quantity of rice polishings used. With every additional fraction, 

 the naturally occurring loss must be made up by an increased dosage. 



The fact that the solution was phosphorus-free served as a final 

 blow to Schaumann's hypothesis regarding the lack of phosphorus, 

 since it was demonstrated for the first time that a solution completely 

 free from phosphorus, was active. Since the phosphotungstic acid 

 filtrate was absolutely inactive, all of the active substance must have 

 been contained in the precipitate. 



The usual methods were used in the working up of such precipitates. 

 The resultant solution, which contained a large amount of potassium 

 and hence reacted strongly alkaline, was neutralized with hydro- 

 chloric acid and then concentrated to a syrup in vacuum. The 

 residue was taken up in alcohol, leaving the greatest part of the 

 potassium chloride behind. The alcoholic solution was precipitated 

 by sublimate, the precipitate being recrystallized from water, with 

 the addition of some sublimate. The crystals were then suspended 

 in water and decomposed with hydrogen sulfide. This solution, as 

 well as that obtained after the elimination of excessive sublimate, was 

 active (the filtrate, however, being more so). Hence it was not 

 possible to effect a separation with sublimate. The solution obtained 

 from the precipitate was concentrated in vacuum, the residue dis- 

 solved in alcohol and precipitated with an alcoholic solution of 

 platinic chloride. A series of fractions were obtained which, on 

 analysis, proved to be pure choline. The alcoholic solution was still 

 active after complete elimination of choline. It could not be doubted 

 therefore that the curative substance did not belong to the choline 

 group. In the mother liquor remaining after precipitation with 

 sublimate, various precipitants were tried in order to precipitate the 

 active substance. 



For this purpose, silver nitrate in alkaline solution was used by the 

 author (corresponding to the precipitant for the histidine group). 

 The solution was freed from chlorine by silver sulfate. After the 



