16 
Journal o f Agricultural Research 
Vol. XXIX, No. 1 
The F 3 results show that some of the 
plants classed as white segregated for 
white and brown. The percentages of 
these were larger at Mandan than at 
Davis. Correcting the F 2 Davis re¬ 
sults on the basis of the F 3 plants 
grown at that station, the following 
results are obtained: 
Brown 
White 
Obtained (uncorrected)- 
735 
295 
Corrected on basis of F 3 _ 
767 
263 
Expected on 3:1 ratio_ 
__! 772 
258 
Probable error.._! 9.38! 
I 
The deviation of 5 ±9.38 shows a 
very close fit. 
A similar correction of the Mandan 
F 2 results on the basis of the F 3 plants 
grown there gives the following: 
! 
! 
1 
Brown ; 
1 
White 
. 
Obtained (uncorrected)- i 
1, 291 
510 
Corrected on basis of F 3 . 
1,384 ; 
417 
Expected on 3:1 ratio- 
1,351 j 
450 
Deviation- 
33 1 
33 
Probable error 
12.4 j 
This deviation of 33 ±12.4 is not 
significantly different from the ex¬ 
pected ratio as it is less than three 
times the probable error. 
At St. Paul the F 2 plants were shown 
t} have a greater percentage of white 
glumed plants than occurred at either 
Davis or Mandan. The Kota-Hard 
Federation cross particularly segre¬ 
gated in that manner, due possibly to 
maternal influence, although the envi¬ 
ronmental conditions at St. Paul were 
less favorable than at the other points 
for developing and maintaining natural 
glume colors. It is quite probable, 
therefore, that there was a greater 
error in the classification of the St. 
Paul material, which, together with the 
possible maternal influence, may ac¬ 
count for the differences. 
Of the 28 F 2 white-glumed families at 
Mandan which segregated for white and 
brown glumes, 19 produced more white- 
glumed plants than brown. The aver¬ 
age for the 28 families was 13 white to 
10 brown. There were also 7 additional 
families included among the true- 
breeding white which produced 1 to 
4 questionable brown-glumed plants. 
The occurrence of these questionable 
plants or of more white-glumed plants 
than brown can not be satisfactorily 
explained and may be due to natural 
crossing. At Davis the 9 families 
classed as white-glumed in F 2 and which 
broke up into brown-glumed and white- 
glumed in F 3 segregated in the expected 
3:1 ratio, the deviation being 4±4.26. 
There were no families which had more 
white-glumed plants than brown, al¬ 
though one family had an equal number 
of each. 
The number of F 2 families of the 
brown-glumed class at Mandan separ¬ 
ated into homozygous and heterozy¬ 
gous families in the F 3 closer to a 1:3 
ratio than to the expected 1:2. At 
Davis, however, a close agreement to 
the 1:2 ratio was obtained, the devia¬ 
tion being 7 ± 3.24, which is not sig¬ 
nificantly different from the expected. 
Of the heterozygous brown-glumed 
F 2 families at Mandan the F 3 plants 
segregated in a ratio very close to 1:2 
rather than 1:3. This is very similar 
to the F 2 results and is what would be 
expected to reoccur. At Davis, al¬ 
though the ratio is nearer the 1:3, the 
deviation is 46 ± 13.24, which also is 
significantly different. 
In spite of the several unexpected 
segregations and ratios obtained in both 
F 2 and F 3 there is no reliable evidence 
of more than a single genetic factor 
involved in the color of glumes in this 
cross. The frequent lack of a significant 
fit to the 3:1 ratio apparently is due 
to environment, but may be due to 
natural crossing and possibly maternal 
influence. 
COLOR OF THE KERNEL 
The inheritance of kernel color has 
been explained by one, two, and three 
Mendelian factors. Biffen (4) found 
that red was dominant to white in Fi and 
segregated in a 3:1 ratio in F 2 . Nilsson- 
Ehle (24) was the first to report crosses 
which in F 2 gave 15:1 and 63:1 ratios 
of red-kerneled and white-kerneled 
plants, proving the presence of two 
and three factors. Howard and Howard 
(22) of India, and Gaines (15) in Wash¬ 
ington have since obtained similar 
ratios. 
In the cross here reported red proved 
dominant as usual in Fi, and the data 
on segregation in F 2 are given in Table 
XI. 
