372 Chemistrji of MendeUau Factors for Flower-Colonr 



In water and alcohol solution both the anthocyanins and their salts 

 tend to become colourless with the formation of a colourless iso-fonn. 

 Willstatter states that the analysis of the colourless iso-form in the 

 case of delphinidin has shown that the change from the quinonoid to 

 the non-quinonoid combination takes place with the addition of a 

 molecule of water, though no more light is thrown on the reaction than 

 is given in the following equation : 



Ci^H.oCHCl + H,0 = HCl + C,.,H,.A. 



In acid solution, on the other hand, and in the solutions of some 

 neutral salts', the anthocyanins are stable owing to the formation of 

 the oxonium salts, and the stability is maintained in the absence of 

 excess of water. 



Formulae of the type II, as shown above, are discarded by Will- 

 stiitter since, for example, if such a formula were correct for cyanidin, 

 it should be possible to obtain maclurin as an intermediate product of 

 decomposition and this he did not achieve. 



Adopting therefore formulae of type I, Willstatter then points out 

 that the relationship between anthocyanins and Havones is such that 

 it should be possible to obtain the red pigments from the yellow by 

 reduction and, converselj-, the yellow from the red by oxidation. 



The question as to the formation of anthocyanins from flavones on 

 reduction by means of nascent hydrogen has been considered in the 

 author's previous paper. The phenomenon was first investigated by 

 Combes- who maintains that he obtained a cr3'stalline red pigment 

 identical with natural anthocyanin by reduction of a naturally occurring 

 flavone : these statements, however, are not supported bj' analyses. 

 The same view of the identity of products of flavones with natural 

 anthocyanins has been held by Everest''. 



Willstatter, also, has investigated the reduction products of flavones 

 in view of his hypothesis as to the origin of anthocyanins and he claims 

 to have shown that in the process two products are formed, of which 

 one is a true anthocyanin, the other not. In his experiments, an 

 alcoholic solution of quercetin acidified with hydrochloric acid is reduced 

 with sodium amalgam or magnesium. The bulk of the purplish-red 



1 See Willstatter's first paper. 



- Combes, E., C. It. Acad. Sci. , Pans, 1913, T. cLvn. p. 1002 and p. 1454, and T. cLviti. 

 p. 272. 



3 Everest, A. E., "The Production of Anthocyanins and Anthocyanidins." Proc. 

 Roy. Soc, London, 1914, B, Vol. Lxxxvii. p. 444. Ibid. Pt II. 1915, B, Vol. r.xxxvm. 

 p. 326. 



