1907.} Flower Colour in Antirrhinum majus. 301 



The reds in F 2 are of four kinds and would on self -fertilisation give in F3 

 the following results : — 



1 FD — * FD red 

 2 FDw' — > FD (red)-f-Fw' (flesh-colour) 

 2 FD w —* FD (red) + Dw (delila) 

 4 FDww' — ► the whole series again 



The delilas are of two kinds : — 



lDw-^Dw (delila) 

 2 Dww' — > Dw (delila) + ww' (white) 



The flesh-colours also : — 



1 Fw' — * Fw' (flesh-colour) 

 2 Fww' — ► Fw (flesh-colour) + ww' (white) 



Professor De Vries showed that a delila and a flesh-colour in F2 gave, on 

 selfing, an F 3 consisting of delilas only and flesh-colours only respectively. 

 Other F 2 delilas gave delilas and whites and other F 2 flesh-colours gave flesh- 

 colours and whites. Further, among the F 2 reds, some gave reds and flesh- 

 colours in the proportion 3:1, while others gave reds and delilas in the same 

 proportion. 



Professor De Yries has kindly informed me that his types correspond to the 

 following types mentioned in this paper, namely : — 



Eoth to Magenta 



Delila „ „ delila 



Fleischfarbig „ Eose 



Weiss „ Ivory 



„ (tinged) „ Eose delila 



Eose x a delila form has been found to give a deep magenta whole form 

 (p. 299) and the result is consistent with the supposition that there is an inter- 

 dependency between the shade of colour in lips and tube ; the magenta tube- 

 colour, introduced by rose, takes, in the cross-bred, the same shade as the 

 lip-colour. This result corroborates Professor De Vries' suggestion that roth 

 consists of fleischfarbig and delila* 



Translated into the types used in this paper, the simplest expression for 



* [In 'Camb. Phil. Soc. Proc.,' vol. 12, 1902, p. 50, I suggested an interpretation of 

 De Vries' results with Antirrhinum, which then seemed to me preferable to that put 

 forward by him. Further evidence as to this and other cases of the resolution of 

 compound characters has shown that my suggestion was incorrect (see also ' Eep. Evol. 

 Com.,' Ill, p. 12).— W. Bateson.] 



Z 2 



