H. Martin Leake and B. Ram Pershad 17 



of Groups III and IV build a series which cannot be separated into its 

 components (PL I, fig. 12 c). The true ultimate colour constitution can, 

 thus, only be determined by breeding. 



The Seed ColouRj 



The colour of the seed is derived from a pigment of the testa and 

 is, therefore, a parental character. The following colours occur : 



1. White. 



2. Straw. 



The above two colours have not been distinguished in any breeding 

 experiments, and may be considered for the present to constitute a 

 single group — the white seeded group. 



3. Pink. 



This colour covers two forms, of which the seed coat appears respec- 

 tively opaque and transparent. Both forms occur in the same capsule, 

 and the cause of the difference, whether dependent on maturity or a 

 racial character, has not been worked out. In the few observations on 

 seed colour the pink is considered a simple character. 



4. Blue. 



Here again there are two forms, the one light, the other deep, blue 

 and both may occur in the same capsule. Races with pure light, and 

 with pure dark, blue seed have been isolated. As in the last case, how- 

 ever, the blue character is treated as a simple one. 



5. Brown. 



This colour has only been found pure in a single culture from a 

 single plant, and in a few stray plants. For the present, beyond noting 

 the existence of the form, nothing can be recorded about it, and the 

 character forms the basis of a series of experiments as yet incomplete. 



The first striking feature connected with the seed colour is that the 

 division between the group with colourless, and that with coloured seed 

 coincides with the division that can be drawn by the eye of the petal ; 

 Groups I and V, with the eye colourless, being white seeded, while 

 Groups II, III and IV, in which the eye is coloured, invariably have 

 coloured seed. Among the hundreds of thousands of plants which have 

 passed under observation no single exception has been observed. It 

 follows from this that the ratio of coloured to colourless seeded plants 

 is in all cases 3:1. 



Joarn. of Gen. x 2 



