ioo6 Journal of Agricultural Research voi. xxiv, no. xa 
hypersensitive area. For this reason the seed of Kota is generally well 
filled even under rust-epidemic conditions. 
The F3 families studied in their reaction to Form XIX and Form XXVII 
in the greenhouse were the progeny of Fg plants taken at random from an 
Fj population. The seed characters of these plants were examined and 
the plants which appeared to have desirable seed from an agronomic 
standpoint were selected and their progeny grown in the rust nursery in 
1922. The various F3 lines were exposed, as has been previously noted, 
to at least 9 biologic forms of stem rust. The epidemic was a satis¬ 
factory one. Individual plants within the different Fg families were 
carefully examined and on the basis of the infection obtained the Fg 
families were placed in the following classes: 
R=Resistant; while there was considerable infection the plants all showed 
the Kota type of resistance. 
NR=Semiresistant, all plants infected in a similar way. Much more resistant 
than Marquis, but more heavily infected than Kota. 
H=Heterozygous, containing both resistant and susceptible plants. 
S=Susceptible, all plants heavily infected, with large uredinia, similar to 
the infection shown by Marquis. 
The results of the field experiment are presented in Table VII. The 
9 control plots of Kota were all clearly resistant while the 20 control 
plots of Marquis were fully susceptible. 
Table VII .—Distribution of 206 Fg families of the Marquis-Kota crossy which had 
not been tested previously under rust-epidemic conditionsy when inoculated in the rust 
nursery with a mixture of g different rust forms y in ig22 
Number of families. 
Parents or hybrids. 
Homozygous. 
Heterozy¬ 
gous. 
Resistant. 
Semi¬ 
resistant. 
Susceptible. 
Fg awnless homozygous. 
7 
18 
22 
Fg bearded homozygous. 
0 
15 
10 
35 
33 
0 
16 
22 
Fg heterozygous. 
12 
17 
Koffl, bparded . 
0 
Marnuis. awnle^y; . 
y 
20 
Of a total of 206 Fg families 28 were homozygous resistant, 61 were 
susceptible, 86 appeared semiresistant, and 31 were clearly heterozygous. 
These results can not be explained on a single factor basis, because only 
28 out of 206, which is a ratio of i to 7.36, were as resistant as Kota. 
It should be remembered that 9 biologic forms of rust were used to pro¬ 
duce the field epidemic. Apparently the Kota type of resistance to all 
9 forms depends on more than a single genetic factor. 
In 1921 an F2 generation of the Kota-Marquis cross was planted in 
the rust nursery. Several of the 9 biologic forms used in 1922 to produce 
the epidemic were available, although only three or four biologic forms 
were used to produce the 1921 epidemic. A total of 666 normal plants, 
which were carefully examined, consisted of 80 bearded susceptible, 233 
awnless susceptible, 89 bearded apparently resistant, and 264 awnless 
apparently resistant. The 353 plants which appeared resistant were 
harvested and on examination 47 appeared to have desirable seed char- 
