226 APPENDIX 



in the blue compound, it might be possible to explain the non-appearance 

 of a blue derivative of some anthocyanins on the assumption that all 

 available hydroxyls are replaced by sugar or methoxyl groups. But 

 to the case of Rosa, apparently, this suggestion cannot be applied. 

 Equally difficult is an explanation for the occurrence of the soluble 

 yellow variety in Dahlia variabilis, when from Centaurea Cyanus it is 

 absent ; for both these species are capable of forming the same antho- 

 cyanins, cyanin and pelargonin. At one period of investigation it 

 seemed as if the whole range of colour-variation in a species might 

 depend upon the molecular structure and chemical potentialities of 

 the initial chromogen or flavone. But it now seems possible that the 

 explanation of difference in colour-range may be found to lie in the 

 difference in type of the oxybenzoic acids formed in variation by the 

 genus or species in question. From some oxybenzoic acids certain 

 anthocyanins and flavones will be synthesised, from others, a totally, 

 or partially, different set of anthocyanins and flavones. So that the 

 possible extension of the colour-series may depend rather upon these 

 types of oxybenzoic acids than upon the subsequent inter-relationships 

 between the products synthesized flavones and anthocyanins. 



