1004 
Journal of Agricultural Research 
Vol. XXIV, No. I* 
The Marquis seed used was from a carefully selected sample and was 
typical of the Marquis variety. There was the possibility that these 
Marquis seedlings which in the greenhouse gave type 3 infections were 
of a different genetic nature than the greater part of the Marquis seed¬ 
lings, due to mixtures or some other causes. Accordingly, seedlings 
which showed the type 3 and the type i or 2 infection in the greenhouse 
were planted in the field and grown to maturity. All plants appeared to 
be of the Marquis variety. Seeds from several plants in each group were 
saved separately and progeny of these plants tested in the greenhouse by 
inoculating with spores of Form XIX. Five out of six of the progeny of 
the seedling plants which had type 3 infections again gave seedlings 
which were placed in the 3 group, while the infection of one family was 
mainly in types i and 2. Six out of seven progeny lines which descended 
from seedlings having type 2 infections again had similar degrees of in¬ 
fection, while one family bore infection mainly of type 3. The reason, then, 
for the greater part of the type 3 infections obtained in the Marquis 
seedlings appears to be genetic differences in the Marquis material. 
To determine their reaction to biologic Form XIX, 455 F3 families were 
tested. Some proved homozygous in the moderately resistant and the 
moderately susceptible types. Other Fg families proved to be heterozy¬ 
gous. The number of susceptible types is too large to be explained on 
the basis of a single factor difference. Forty resistant families were 
obtained. 
Certain of the Fg parent plants of these various Fg families were selected 
and their progeny grown in the rust nursery in 1922. Some lines ap¬ 
peared desiraWe and individual plants within these lines were selected. 
Seedlings were grown in the greenhouse from the individually selected 
plants and were inoculated with Form XIX. Progenies of 3, 4, and 5 
plants of three different Fg families, which were classified as IS, or of the 
Kota type of infection, were grown in the greenhouse. All 12 plants 
bred true to the IS type of infection. One family from the IR group 
was tested in a similar manner. Progeny of 5 plants of this family were 
tested, and all showed type 2 infections, as was expected. 
The reactions to Form XXVII were of a more definite nature than 
those to Form XIX, as will be seen in Table VI. Of the Kota plants all 
except 33 out of a total of i ,040 were in the immune class, while Marquis 
had 9 plants in type 4 and 725 in type 2 infections. The deviations 
obtained could all be explained on the basis of natural causes, for the 
frequency of such natural crossing is in the neighborhood of 2 to 4 per 
cent in Minnesota under nursery conditions. 
A total of 462 Fg families were tested. Of these, 76 showed either no 
infection or infection of the o or immune type, 14 were resistant, and 39 
were susceptible, and there were four separate classes of heterozygous 
types. Immunity is clearly dominant over resistance or susceptibility 
as shown by the results for the heterozygous classes Hi and H4. 
The segregating classes obtained when Fg families were inoculated with 
Form XXVII could result from two independently inherited genetic 
factors for resistance and immunity each allelomorphic to a factor for 
susceptibility. Furthermore, if the factor for immunity was epistatic to 
the factor for resistance, three times as many homozygous immune fam¬ 
ilies would be obtained as homozygous for either resistance or suscep¬ 
tibility. It is possible that a much more complex genetic condition may 
be the cause of the classes of segregation obtained. The clear proof that 
