THE INHERITANCE OF FLOWEK COLOUR IN PISLM / / 



results in Fi when crossing them with 01001 are to be seen in the 

 table on the preceding page. 



Thus, nine whites differently constituted and nine different possibi- 

 lities in Fl. I have carried out 46 such crosses in all, and I have found 

 all the nine possibilities realized. However, in some cases the number 

 of individuals in Fj has been rather small. It is my intention to 

 carry out such crosses in a great extent for control. 



Two of the above mentioned 46 crosses have been followed up in 

 Fa. One of them is the one mentioned in table 17, whose white parent 

 were constituted aaBBcc. Another white (from Fa of cross I, table 1) 

 yielded light purple Fi, and light purple and white in the ratio 3 : 1 

 in Fa, and were, therefore, constituted aabbcc (table 18). 



A third white (from cross I, table 1) yielded in both cases, when 

 crossed with two homozygous purple (0401 and 0351), purple flowered 

 Fl, and in Fo purple, rose, and white in the ratio 9:3:4 (table 19). 

 Thus, Fl was AaBBCc and the white parent aaBBcc, just as in the 

 cross in table 17. • 



Finally, a fourth white (also from cross I, table 1) yielded when 

 crossed with 0142 (homozygous rose AABBcc) purple Fi, and in Fa the 

 same segregation took place as in the case of the last mentioned cross 

 (purple, rose, and white in the ratio 9:3:4). Consequently, this white 

 had the usual constitution of ordinary white or aaBBCG (table 20). 



Lastly, I have also crossed 01001 with the two above mentioned 

 homozygous purple (0401 and 0351). In both cases Fa segregated in 

 purple, violet, rose, and light purple, as was to be expected. Although 

 the numerical relations could not be ascertained with sufficient exact- 

 ness on account of the attacks of thrips, the segregation of these two 

 crosses in Fa speak also strongly in favour of the correctness of the 

 results otherwise arrived at in my study of the inheritance of the flower 

 colours in Pisum. 



