8 
BULLETIN 1403, U. S. DEPARTMENT OF AGRICULTURE 
consistent or significant difference between reciprocal crosses, as the 
percentages alternate slightly at the two stations. 
The study of normal and dwarf segregation was continued in the 
F 3 generation at Bozeman in 1924, where F 3 families were grown 
from 224 normal and 25 dwarf F 2 plants. The data obtained from 
normal plants are given in Table 4. 
Table 4. — Breeding behavior of 18,897 F z plants grown from seed of normal F 2 
plants of crosses between Marquis and Hard Federation wheats, at Bozeman, 
Mont., in 1924 
Num- 
ber of 
F 2 fam- 
ihes 
Expect- 
ed on 
7:4:2 
ratio ' 
Percentage of— 
Number of F 3 plants 
Devia- 
tion 
from 
ratio 
indi- 
cated 
Prob- 
able 
error 
Cor- 
rected 
Breeding behavior 
in the F3 genera- 
tion 
Normal 
plants 
Stand 
Dwarfs 
clumps 
Total 
per- 
centage 
of 
total 
F 2 
Normals breeding 
true... .. 
127 
64 
33 
121 56. 7 
85.2 
85.9 
85.6 
i 
7,570 
3, 177 673 
7,570 
3,850 
1,977 
45.3 
Normals breeding 
13 normals to 3 
69 
34 
28.6 
14.7 
49 
33 
16.30 
13.01 
22. & 
Normals breeding 
3 normals to 1 
1,450 
527 
11.7 
Total 
224 
224 
100.0 
85.4 
12, 197 
1,200 
13, 397 
79.8- 
1 P=0.72. 
The data show that of the thirteen-sixteenth s which were normal 
plants in the F 2 generation, approximately seven parts bred true for 
normal plants and six parts segregated for normal and dwarfs in 
the F 3 . The families throwing dwarfs were divided according to 
whether they were below or above 21.9 per cent for the 13:3 and 
3:1 ratios accordingly. Of the six segregating parts, approximately 
four parts bred as in F 2 with 13 normals to 3 dwarfs, and two parts 
segregated in a 3 : 1 ratio of normals to dwarfs. A very close fit to 
the 7:4:2 ratio was obtained, P equaling 0.72. Of the four parts 
segregating in a 13 : 3 ratio, the deviation is greater than that in 
F 2 . This is evidently due to the overlapping with the"3:l ratio. 
The deviation of 49 ±16.30, however, is no larger than would be 
expected by chance in 1 out of every 22 trials. The two parts 
segregating in a 3:1 ratio have a deviation of 33 ±13.01, which is 
close to the expected ratio, as the deviation is but 2.5 times its 
probable error. 
As shown, the percentage of stand for the three types was prac- 
tically 85 per cent for each, so that it appears certain that the seed 
of the dwarfs is not less viable than those of the normals. 
Table 5. — Breeding behavior of 934 F z plants grown from dwarf F 2 plants of crosses 
between Marquis and Hard Federation wheats at Bozeman, Mont., in 1924 
Breeding behavior in the 
F3 generation 
Num- 
ber of 
F 2 
families 
1 
Percentage of— ; Number of F3 plants 
Devia- 
tion 
from 
ratio 
in- 
dicated 
Prob- 
able 
error 
Cor- 
rected 
percent- 
age of 
total 
F 2 
Dwarfs 
F 2 
Normal 
Dwarfs 
Total 
Dwarfs breeding 3 dwarfs to 
21 
4 
84.0 
16.0 
63.0 
12.0 
176 
577 
181 
753 
181 
12 8.01 

12. & 
Dwarfs breeding true 
2.4 
Total. .. 
25 
100.0 
75.0 
758 
934 
15.2 
