VITAMIN B (Bi) 49 



pounds for antineuritic properties (Williams, 1916, 1916a, 1917, 1921 ; 

 Harden and Zilva, 1917). 



Sahashi (1925, 1926, 1926a) prepared 300 grams of the /3-acid 

 of oryzanine (Suzuki) and attempted to determine its chemical nature. 

 On heating crystals of the acid to 150° C. water of crystallization was 

 driven off, leaving an anhydrous salt conforming to the empirical 

 formula C10H7NO4 instead of CioHgNOi ascribed to it by Suzuki. 

 From various tests the constitution of the compound was considered 

 to be 2,6-dioxy-quinoline-4-carboxylic acid. 



The a-acid was thought to be the |5-acid containing one molecule 

 of water of crystallization. The resemblance of this compound to 

 a-hydroxypyridine and its derivatives found by Williams (1916a) to 

 have temporary antineuritic properties, led Sahashi (1927) to deter- 

 mine whether the 2,6-dioxyquinoline group in the ^-acid has any rela- 

 tion to the curative property of the compound for polyneuritic pigeons. 

 Pure crystalline 2,6-dioxyquinoline hydrochloride obtained from the 

 |5-acid of oryzanine showed marked curative properties, suggesting that 

 the nucleus of the antineuritic factor may consist either of the 2,6-dioxy- 

 quinoline group or its tautomeric form. 



Ho/NrS ^ Ho/\r^ 



N NH 



Later, in an attempt to synthesize a glyoxaline derivative of the 

 same empirical formula as that of Jansen and Donath's (1927) crystal- 

 line compound, Sahashi (1928) prepared 4- (or 5 ) -glyoxahnemethyl- 

 ethyl carbinol. Injections of this he found to be temporarily curative 

 for polyneuritic pigeons. 



Resuming earlier attempts by Williams to determine the nature of 

 the antineuritic vitamin by testing various synthetic substances for 

 antineuritic activity, Williams and Eddy (1927-28) tested various 

 di-keto piperazines as sources of the antineuritic vitamin for rats. The 

 enol forms of certain of these compounds were found to have a definite 

 effect on growth and the keto forms to be entirely without effect, thus 

 conforming to the earlier theory of isomeric changes as a factor in 

 the lability of antineuritic substances. "The chemical features asso- 

 ciated with the apparent curative activity of the earlier synthetic prepa- 

 rations are also present in the di-keto piperazines. Feeding results to 

 date with pigeons are not convincingly positive, and those with rats, 

 while apparently beyond the limits of experimental error, still lack 

 the excellence obtainable with natural antineuritics. We seem to be 



