46 THE VITAMINS 



with barium chloride, sodium hydroxide, and hydrogen sulfide. The final 

 alcohol fractions were about as active as the preparation which Jansen 

 and Donath obtained at the phosphotungstic acid stage. Although Kin- 

 nersley and Peters obtained very active fractions by the use of the 

 silver nitrate method (Seidell, 1922) they found that this method does 

 not always prove reliable. The use of phosphotungstic acid with a pre- 

 liminary charcoal adsorption step as suggested by Jansen and Donath 

 gave fractions of 0.07 to 0.08 milligram per day activity, but with 

 marked losses, particularly when acetone was used as a solvent. The 

 authors think that precipitation with phosphotungstic acid is a "genu- 

 ine property of the active principle." With chloroplatinic acid they have 

 obtained very active preparations but with large losses and variable 

 yields. 



The active preparations give a slight Pauly test, contain 13 to 20 

 per cent nitrogen, give qualitative tests for sulfur, negative Benedict 

 test, and positive MoUsch reaction. 



In discussing the properties of this yeast vitamin Bi Kinnersley and 

 Peters emphasized that the properties of preparations change during 

 the process of concentration. Thus while in the early stages of frac- 

 tionation from yeast, vitamin Bi is thrown out upon alcohol precipi- 

 tation, at a later stage after the removal of certain substances rela- 

 tively insoluble in alcohol, vitamin Bi is soluble in alcohol of high 

 concentration; in their experience vitamin Bi (activity 1.0-0.03 milli- 

 gram per day) is soluble in alcohol or water, but insoluble in chloro- 

 form, carbon tetrachloride, ether, acetone, or ethyl acetate. When frac- 

 tionated from mixtures of benzene and alcohol or ether and alcohol in 

 which water is reduced to a minimum, active preparations are thrown 

 out of solution, but much seems to depend on the acidity. 



The stability to alkali appeared to depend upon various factors. 

 Alcohol and hydrogen peroxide appeared to accelerate the destruction 

 by alkali. It is considered advisable to carry out any alkaline stage in 

 the process of concentration and purification at a low temperature 

 and in the absence of alcohol. In general the active material proved 

 stable to oxidizing agents and fairly resistant to reducing agents. 



On account of criticisms from various sources on the reliability 

 of short curative tests upon pigeons in vitamin B studies, Kinnersley, 

 Peters, and Reader (1928) subjected the test as originally described by 

 Peters (1924) to rigid examination as to its quantitative value with 

 the conclusion that if certain factors are taken into account the test 

 can be used with confidence. Two possible cures other than that due 

 to the curative substance were recognized. One is the so-called "heat" 



