166 THE VITAMINES 



that these pyridine derivatives are indicative of decomposition 

 products of vitamines, then it would be possible to demonstrate their 

 formula and chemical constitution with comparative ease; subse- 

 quently the synthesis could be made. Although the amount of these 

 pyridine derivatives is small enough to permit of a possible relation- 

 ship with the vitamines, our facts are not sufficiently certain to 

 proceed to such work, since in the end we may be dealing with 

 incidental impurities. 



The second method of isolation, by extraction with various solvents 

 or treatment with adsorptive agents, has given some interesting 

 results, though it could finally be shown that such fractions still 

 contained complexes. Hence a further fractionation would again be 

 necessary, whereby the above mentioned difficulties would again have 

 to be dealt with. . 



The third way, naturally only adapted to exclude from our con- 

 sideration substances of known composition, did not prove of value 

 in the studies of the nature of vitamines. Obviously, a positive 

 result might have been due to a contamination with vitamine, or to 

 a secondary effect that might be able to simulate a vitamine action. 



The fourth, or synthetic, procedure has yielded no results till 

 now, since the investigations have been based on weak foundations. 

 What then is the best method for the chemical study of the vita- 

 mines? First of all, we must stabilize the vitamines, steps towards 

 which have already been taken. If this should prove to be impracti- 

 cable, we must look for another vitamine in Nature which is more 

 stable. We shall see directly that in general, vitamine B is more 

 stable than the two other vitamines. It would suffice if, after the 

 chemical purification of this substance, there remained a single 

 function which would characterize the substance as a vitamine. We 

 have already learned that all three vitamine types are present in 

 seeds, or that they may eventually arise through chemical re-arrange- 

 ment. Although we have no positive proof that the three vitamine 

 types are related chemically, it is not impossible that the identifica- 

 tion of a single vitamine may be of immeasurable help in the study 

 of the other vitamine types. 



The vitamines have this in common, that they are all necessary 

 to complete a synthetic diet. Further, none of the vitamines can be 

 identified with any of the dietary constituents already known; they 

 are active in very small quantities and none of them possess any too 



