312 CARL HALLQVIST 

27 blue: 16 violet :8 white (table 16) 



expected: 28,09 9,56 12,75 ; T4 
= 
1,69 6,44 4,75 À 
D/M ——=0,48 ~—=2,31 —— = 1,53 
3,54 2,79 3,09 
The number of individuals in the whites and in the violets seems 
to indicate quite other ratios than 9:3:4. The results obtained in 
F; show, however, that this ratio must be correct, and that the ‘4 
abnormal number in F3 is a case of an unusual great deviation, although | 
within the range allowed by the theory. From blue parent plants re 
partly di-hybrids have been obtained showing the following numerical 
relation: 

216 blue : 76 violet : 92 white (table 17) 
expected: 216 72 96 
D/M 0 : = 0,52 Si — (0,47 : 
1,65 8,49 
partly mono-hybrids with the following numerical relations: 
135 blue : 50 violet 
expected: 138,55 46,25 D/M Oe 
90 
169 blue : 43 white 

expected: 159 53 D/M —— = 1,59. 
6,30 
No constant blue families have been found, however. This is probably 
due to the small number of families. 13 families from blue mother 
plants, distributed in the following classes have been followed: 
0, constant blue : 5 9:3:4-segregates : 4 blue and violet : 4 blue 
and white 
expected: 1,44 5,74 2,89 2,89 
3 constant violets and 5 3:1-segregating families have been obtained 
from violet plants. The proportion expected is 2,67: 5,33 and the aa 
total number of individuals in the segregating families was; | 
242 violet : 93 white 
9,95 
expected: 251,25 : 83,75 DM 
= 1 ir 
7,98 
One white plant gave a constant white progeny. 
» u 
The seed colour in the F.-plants has been examined in 26 blue 
plants, 14 violets and 6 whites. The colour of the seeds in the blue 
