THE ANTHOCYAN PIGMENTS 609 



pigment can only be taken as evidence of the presence of a small 

 quantity of impurity. 



That luteolin and morin give red pigments on reduction in 

 acid alcoholic solution by means of sodium amalgam has been 

 known for many years (cf. Rupe, Die Chemie der natilrlichen 

 Farbstoffe, vol. i. pp. 77 and 85). Quite recently, Watson and Sen 

 (Journ. Chem. Soc. 1914, 389) obtained a red pigment, from quer- 

 cetin in like manner, and this was followed by production by R. 

 Combes {Compt. rend. 1913, 1002) of a red pigment, identical 

 with that which he had obtained from the red leaves of ampe- 

 lopsis hederacea, by reducing in the same way, the yellow 

 pigment he obtained from the green leaves of the same plant. 

 A further paper of Combes (C. rend. 191 3, 1454) described the 

 reverse change, viz. from red to yellow, by means of oxidation 

 with hydrogen peroxide. In each case he obtained crystalline 

 compounds and compared their melting-points ; he did not, 

 however, give analyses, nor state whether he worked with 

 glucosides or no. 



That a series of red pigments, whose properties coincide in 

 every way with those of the anthocyanidins, may be produced 

 by reduction of the flavonol derivatives by various methods, the 

 best of which appears to be treatment of the pigment dissolved 

 in a mixture of five volumes absolute alcohol and one volume 

 concentrated hydrochloric acid, with magnesium, has been con- 

 firmed by Everest (Proc. Roy. Soc. 1914, B, 87, 444), and, more- 

 over, in the same paper he described the production by the 

 same means of anthocyanins from the glucoside flavonol deriva- 

 tives present in various flowers, showing the direct formation of 

 red glucoside pigments from the yellow flavonol glucosides with- 

 out intermediate hydrolysis. This important observation makes 

 the hypothesis of Miss Wheldale and others, in which hydrolysis 

 of the flavonol glucoside is an essential factor, unnecessary, and, 

 moreover, the above-mentioned observations show that reduc- 

 tion, not oxidation, appears to be necessary for the passage from 

 flavonol to anthocyan. 



The results obtained, coupled with those of Willstatter and 

 Everest, in their investigation of the cornflower pigment, led 

 Everest to suggest the following scheme as a typical repre- 

 sentation of the passage from flavonol to anthocyan, and to 

 draw attention to the fact that all available evidence pointed to 

 the formula III being that of a typical anthocyan pigment. 



