ANTHOCYANINS AND GENETICS 215 



The suggestions which the author has made from time to time are 

 first, that the blues and purples are more oxidised conditions of the 

 reds, and secondly, that the blues and purples may be formed by the 

 condensation of the red with some other aromatic substances in the 

 cell. The hypothesis of further oxidation, though possible, has as yet 

 no definite evidence in its favour, and would involve the postulation 

 of a second oxidase in addition to the R factor. The condensation 

 hypothesis is perhaps more probable; in this case the blueing factors 

 Bj and B 2 would be the power to form the substances with which 

 condensation takes place. It will be necessary also that any hypothesis 

 for the blueing factors should explain the difference between recessive 

 blues, like those of Primula, Cineraria, and Anagallis, which are recessive 

 to red or purple, and dominant blues, like those of Centaurea and Cam- 

 panula, which are dominant to red, pink and purple. 



The identification of luteolin in the yellow variety of Antirrhinum 

 has made clear the difference between crimson and magenta in that 

 species. Crimson is, in fact, magenta pigment mixed with the yellow 

 luteolin, whereas in the magenta type the mixture is with apigenin 

 only which does not affect the colour. The same explanation can 

 probably be applied to the crimson and magenta series in Althaea rosea, 

 Dianthus Caryophyllus and Dahlia variabilis in which both ivory, and 

 a yellow variety with soluble pigment, appear. The same range of 

 colour, it is true, is found in Primula sinensis and Dianthus barbatus, 

 and in these there is no soluble yellow pigment; hence the difference 

 between magenta and crimson in these latter species cannot be explained 

 on the ground of mixture of pigments. 



The most reasonable interpretation of the blueing factors at the 

 moment seems to be to regard local differences of colour in the flower 

 as due to local presence of various organic substances ; this may explain 

 also the change of colour with age in the same flower (Myosolis). But 

 the change of the pigment of the whole plant from magenta, for instance, 

 to red in Antirrhinum, and from blue to pink, as in Centaurea, is likely 

 to be something more fundamental than change in alkalinity and 

 acidity of the cell-sap 1 . 



Another line of speculation is opened out by some further results 

 obtained by Willstatter. He notes a fact, which indeed had been 

 noted before, that the blue pigment of the Cornflower is unstable in 

 water solution. The colour disappears fairly rapidly, and a colourless 

 solution is left. This he explains on the hypothesis of the formation 



1 See Appendix. 



