Mendelian Factors for Flower- Colour. 



309 



Since anthocyanin can thus be decolorised by oxidation as well as by 

 reduction, in each case giving a product in which the colour is easily restored, 

 there is as much reason, on the evidence of these experiments, to postulate 

 one process as the other for the cause of decolorisation by treatment with 

 alcohol. As a matter of fact, the conditions in both experiments are so 

 different from those obtaining when petals are treated with alcohol, that 

 probably neither experiment has any real bearing on the question at all. 



That an alternative to the reduction and oxidation hypothesis can be 

 offered, is shown by the parallel series of changes produced by using phenol- 

 phthalein solution, made red by ammonia, as a pigment. This red solution is 

 decolorised by alcohol and the colour restored by diluting largely with water 

 or by addition of a drop of alkali.. On evaporating the decolorised alcoholic 

 solution to dryness, a red residue is obtained. As it happens, phenol- 

 phthalein is colourless with acids, while anthocyanin gives colour reactions 

 with both acid and alkali. Apart from this accidental difference, the two 

 oases are strikingly similar. 



Without wishing to insist on the parallel too rigidly, it would seem that 

 the two series of phenomena might well have similar explanations. The 

 present authors tentatively offer two alternative suggestions without 

 attempting to decide between them. 



It may be that strong alcohol dehydrates the anthocyanin, giving a colour- 

 less compound, and that colour is restored by subsequent addition of two 

 radicals, either H and OH, or some other pair, such as H and I. Such an 

 effect might perhaps be accounted for by the production in anthocyanin of a 

 lactone grouping. A somewhat similar explanation has been advanced to 

 account for the phenolphthalein changes.* 



Or, the loss of colour when the petals are treated with alcohol may be due 

 to combination of the anthocyanin with alcohol to an unstable colourless 

 compound, which is easily decomposed by various reagents. A similar 

 explanation has been advanced by Hantzsch to account for the differently 

 coloured solutions given by certain substances in different solvents. 



A few minor points in connection with the above work may be considered. 



First, Keeble, Armstrong, and J ones state that the restoration of colour to 

 petals is accelerated by a drop or two of hydrogen peroxide either in acid or 

 alkaline medium, and, further, that the reappearance of colour is not due to 

 the acidity or alkalinity of the medium, because the original colour, purple, 

 red or pink in differently colored petals, is first restored, and the acid or 

 alkaline anthocyanin colour only appears later. 



* Meyer, E., u. Spengler, 0., " Zur Constitution cler Phtale'insalze," ' Ber. D. Chem. 

 Ges.,' 1905, vol. 38, p. 1318. 



