THE ANTIBERIBERI VITAMINE 197 



Dumas and to Kjeldahl, as an indication of the vitamine fraction. 

 However, since this fraction contained a relatively large amount of 

 pyridine derivatives, the method is valueless so long as we do not 

 know the relationship between these substances and the vitamines. 

 Brill and Alincastre (520) tried, in this way, to average the maximum 

 vitamine content from the phosphotungstic acid precipitates of 

 various vegetables. Seidell (521) sought to determine the vitamine 

 content of yeast by estimating the nitrogen in activated fuller's 

 earth. Eddy (522) tried to do the same with an activated fuller's 

 earth out of pancreas. Since fuller's earth adsorbs also a large 

 amount of inactive material, this method is of no special consequence. 

 In discussing the chemistry of vitamine B, it will be recalled that 

 we have often made use of a reaction introduced by Folin and 

 Macallum (I.e. 486) for the determination of uric acid and phenol. 

 It was shown by us, and later by others, that all vitamine-containing 

 extracts give this reaction. It depends upon the development of a 

 blue color with phosphotungstic and phosphomolybdic acids on the 

 addition of sodium carbonate. The chemical nature of the sub- 

 stances giving this reaction was investigated by Funk and Macallum 

 (523), as well as by Lewis and Nicolet (524). With Macallum, we 

 also showed that if an incision is made in the maize kernel and the 

 test applied in situ the blue color develops in the part known to be 

 rich in vitamine. Naturally, these parts are even richer in other 

 extractives. As this method came into use for the chemical analysis 

 of various corn meal products, we (525) showed that in extracts, 

 prepared either cold or warm, the developed color increased with the 

 vitamine content. In the warm extracts, the color was weaker, 

 showing that the substances responsible for the color reaction are 

 destroyed by heat. If it should appear later, with greater certainty, 

 that this reaction has nothing in common with the vitamine, it might 

 still be useful in vitamine fractionation as an index of the purity of 

 the isolated substances. These reactions could conceivably also 

 serve as an index of the purity of the isolated lipoids. 



Green (526) tried to estimate the vitamine content of various food- 

 stuffs by means of a biological method (animal experiments) . The 

 method depends on the length of time pigeons survive on various 

 diets, rich or poor in vitamines. As a result of these studies, Green 

 proposed the following formula : 



C 1 



S = 



V - X K 



