126 Prof. Keeble, Dr. Armstrong, and Mr. Jones. 



flowered races (dominant and recessive whites), the horticultural varieties 

 of this species comprise races with blue, red, and magenta flowers, and our 

 purpose is to put forward a biochemical hypothesis to account for the 

 production of these distinct colours and for the genetical relations which 

 obtain between them. 



It has been suggested by Miss Wheldale* that each of the chief colours of 

 such a series is determined by a special oxydase, but neither general 

 considerations nor such observations as we have been able to make lend 

 support to this view. 



It is true that the flowers of different varieties of P. sinensis contain 

 different amounts of oxydase, but we find no constant relation between 

 amount of oxydase and type of coloration. Moreover, the recent researches 

 of Bachf point definitely away from the hypothesis that oxydases are specific. 



If hypothesis of specific oxydases be rejected, we nrust ascribe specific 

 coloration either to the action of an oxydase on different chromogens or to 

 the interaction, with a chromogen or an oxydase, of specific substances which 

 modify decisively the colour produced in the course of the reaction. 



Any discussion of these alternatives must take into account the observations 

 of A. G. Perkin, that the hydroxy flavone glucosides of plants occur, as a rule, 

 not singly but in groups. There is some ground for regarding these glucosides 

 as constituting the prochromogens from which the anthocyan chromogens are 

 derived, and it is therefore a matter of great significance to the student of 

 genetics that the plant is, as it were, offered a choice of several pigment- 

 forming materials on which its hydrolysing and oxidising enzymes may act. 



Pending fuller investigation of the possibility that the colour of a variety 

 may be determined by a selective action on one of a group of allied glucosides, 

 we are inclined to adopt the latter of the two alternatives, and to suggest that 

 the serial colours of flowers are due each to the intervention of specific 

 substances in the reaction of oxydase on chromogen. 



This hypothesis is rendered plausible by the following observations, first 

 on the colours produced when a mixture of phenols is treated with oxydase 

 and second, on the behaviour of our artificial chromogen benzidine when 

 acted on by oxydase in the presence of various phenols. 



When a mixture of phenols is treated with a plant oxydase a competition 

 for oxygen ensues. For example, if oxydase be caused to act on guaiacol 

 until the red colour is produced, the addition of other phenols brings about 

 a more or less quick change of colour. Thus a-naphthol converts the red 

 into mauve, and the ultimate colour which is produced is of a far deeper tint 



* ' Prog. Eei Bot.,' 1910, p. 469. 



t 'Arch. Sci. Phys. Nat.,' June, 1912, vol. 33. 



