6o 
Journal of Agricultural Research 
Vol. XXVII, No i 
All the F 2 heads could be placed in one of the above color classes with 
a fair degree of accuracy. The results of this classification of the F 2 
progenies of the feterita X Red kafir cross are shown in Table V. 
Table V.— Distribution of seed color in the F 2 generation of the cross feteritaXRed kafir , 
grown in 1922 
Plants. 
■; 
Having seeds with— 
Row number. 
Total 
number. 
j Brown nucellar layer 
! present. 
i 
Brown nucellar layer 
absent. 
Seeds, 
brown. 
Seeds, 
white. 
Seeds, red 
or pink. 
Seeds, 
white. 
165 
114 
7 
33 
II 
138 
103 
7 
20 
8 
3 . 
174 
126 
9 
26 
13 
4 . 
121 
87 
6 
22 
6 
Total. 
S 9 » 
430 
29 
IOI 
38 
Calculated (45:3:12:4 ratio). 
421 
28 
112 
37 
X 2 =i,335. P =0.72468. 
A study of the color distribution in the F 2 generation of the cross 
between feterita and Red kafir indicates that the two color factors 
involved in the cross between feterita and White kafir are concerned, 
and that in addition a third color factor for red, introduced by the Red 
kafir parent, must also be considered. The color factors in feterita 
which are concerned in this cross may be designated as BBssrr f while 
those of Red kafir may be designated as bbSSRRj R being the factor 
for red color, as present in Red kafir, and r its recessive allelomorph for 
absence of the red color. The F x cross then would be BbSsRr. The 
probable effect on seed color of the B , 6, and S } s factors was stated in 
the discussion of the previous crosses and the R, r, factors will be con¬ 
sidered more fully later. A study of the feteritaXRed kafir cross 
indicates that when the factor for brown nucellar layer is present with 
the factor for red, R> the color of the kernel is brown rather than red, 
and the phenotype then is brown. 
On the basis of three color factors in 64 individuals of the F 2 generation 
there should be 45 plants with brown seeds with brown nucellar layer, 
3 plants with white seeds with brown nucellar layer, 12 plants with red 
or light red seeds and no brown nucellar layer, and 4 plants with white 
seeds without brown nucellar layer. It must be assumed that the 
presence of either the 5 or the R factor will cause the brown color to 
appear in the epidermis of the seed. The actual color distribution 
obtained in the four F 2 rows, as compared with the theoretical results, 
is shown in Table V. A deviation of this magnitude might be expected 
to occur once in 1.38 cases as the result of random sampling. 
The fit of the observed to calculated distribution of the color pheno¬ 
types is very close, in fact much better than usually would be expected, 
considering the relatively small number of individuals, 598. The con- 
