S. C. Harland 225 



5. Purples from F^ families which exhibited the 3 : 1 type of ratio 

 behaved according to expectation. Of 18 families grown, 12 segregated 

 into purple and white in the 3-1 ratio, while 6 bred true. 



6. Purples from F^ families which produced the 9 : 7 ratio were not 

 grown. 



It is clear that the hypothesis advanced to explain the experimental 

 results is confirmed at practically all points. 



Correlation of Flower Colour with Anthocyanin Colouration 

 OF Seed and Vegetative Parts. 



The purple flowered F^ of the cross between the two white flowered 

 forms St Vincent Bush and Vilmorin's Stringless exhibited (a) purple 

 colouration of tJie nodes of the main axis, and to a slight extent of the 

 veins of the leaf, (6) black seeds. Jn subsequent generations all plants 

 with purple flowers possessed purple nodes, and vice-versa. One of the 

 parents, St Vincent Bush, is devoid of anthocyanin colouration in all 

 parts of the plant, including the seeds which are cream. The other 

 parent, Vilmorin's Stringless, possesses an extremely minute degree of 

 pigmentation of the stipular hairs. The presence of colour can only be 

 detected microscopically. The seeds, however, are of a purplish brown 

 colour, and sections of the seed coat showed that the pigmentation was 

 partially anthocyanic. 



It has been pointed out by Morgan et al. (1915), that a single factor 

 difference may have not one, but many effects. On flower colour alone 

 it is not possible to distinguish between the phenotypes Or, cR, and cp, 

 for they are all white flowered. The factor C, however, produces visible 

 effects in other parts of the plant body, and the combination Cr may 

 thus be easily distinguished from cR and cr. So far as we know there 

 are no means of distinguishing cR and cr. 



The factor C produces a purplish brown anthocyanin colouration of 

 the seed coat, and a minute amount of pigmentation of the stipular 

 hairs. 



The Fi of the cross Cr by cR is coloured purple in the region of the 

 nodes along the main axis, while the seeds are black. In subsequent 

 generations all purple flowered, i.e. all plants with both C and R, have 

 black seeds and purple nodes. It thus appears that the combination of 

 C and R determines a particular mode of distribution of anthocyanin 

 pigmentation in the plant body. The factor C determines the production 

 of anthocyanin in seed and in stipular hairs. The factor R has no effect 

 except in presence of C, when it converts white flower into purple, brown 

 seed into black, at the same time extending the pigmentation of the 



