RIBOFLAVINE 



compound, not identified. On acetylation it yielded a tetra- acetate, 

 indicating the presence of four hydro xyl groups ; 2' ^ since oxidation 

 with lead tetra-acetate yielded formaldehyde,^ it followed that a 

 primary hydroxyl group was present in the a-position to a secondary 

 hydroxyl group. Confirmation of this was subsequently obtained by 

 the formation of a diacetone compound.* Riboflavine gave a positive 

 murexide test,^ indicating the presence of a purine group, whilst 

 alkaline hydrolysis gave urea.^ 



Lumiflavine 



On irradiation in alkaline solution, riboflavine yielded lumiflavine,^'^ 

 with the empirical formula, C13H12N4O2. Thus, photolysis removed 

 from riboflavine the elements C4H8O4 and, since lumiflavine was in- 

 capable of acetylation and failed to give formaldehyde on treatment 

 with lead tetra-acetate, it was concluded that a hydroxylated side- 

 chain, -CHOH— CHOH— CHOH— CH2OH, had been removed. 



Lumiflavine was sparingly soluble in water, but soluble in chloro- 

 form and, like the parent substance, it yielded urea on alkaline hydro- 

 lysis, together with an acidic substance, C^gHiaNgOg, which lost CO2 

 on heating to give a substance, CnHigNaO.^' ^ Since two moles of 

 water were taken up during the alkaline hydrolysis, R. Kuhn and 

 H. Rudy ^ concluded that the latter must arise, not from a ureido or 

 guanidino group, which would require only one mole of water, but 

 from a ring. This must, therefore, contain two carbonyl groups, one 

 forming urea and the other being hydrolysed to a carboxyl group. 

 These changes can be represented as follows :— 



C 

 I 



\c CO ,2„o \ CO CO 



I I ^^^ CO 1 _^ I 



C NH + C ^ CH 



The decarboxylated compound, CnHigNgO, which had the pro- 

 perties of a lactam, yielded 4-amino-i : 2-dimethyl-5-methylamino- 

 benzene (I) ^ when heated with sodium hydroxide, so that lumiflavine 

 must be an alloxazine derivative (II). 



