472 
Journal of Agricultural Research voi. xxiii, no. 6 
of from 15 to 30 per cent when planted in the fall under conditions favor¬ 
ing maximum infection. I11 1919 Hybrid 128 produced 98 per cent, 
Marquis produced 75 per cent, and the F 3 varied from 75 to 100 per cent. 
In 1920 the parent varieties produced 30.9 and 74.3 per cent respectively, 
an F 2 progeny produced 56 per cent of bunt, and the F 4 selections varied 
from 46 to 85 per cent. Five plants, each selected from the lowest and 
highest F 4 row, produced an average of 78 and 91 per cent bunt in the 
F 6 generation in 1921. These figures include only the rows that were 
planted together in the same part of the field, and on the same date. 
There is much work to be done to establish the inheritance of the resist¬ 
ance of Marquis, especially the part played by the winter rest period, 
the segregation of the units that are not so affected, and the complex 
of its resistance for spring planting. This is being done and will be re¬ 
ported later. 
There remains one other cross to discuss, that of Fortyfold X Red 
Russian. The average susceptibility of the parent varieties and their 
F s offspring in 1920 was 36.7, 49.3, and 37.5 per cent, respectively. The 
F 3 segregates varied from 8.1 to 73.6 per cent. Reference to figure 1 
shows the greatest class frequency of the 181 F 3 rows to be between 30 
and 40 per cent, with rapidly diminishing class frequencies above and 
below. The most striking thing about this cross is that the average of 
the F 3 generation is less than the average of the parents. The fact that 
the parents were duplicated so that there were nearly 200 plants counted 
precludes the possibility of a chance error. In the other four crosses 
charted in figure 1 the average of the parents lie below the average 
susceptibility of their respective F 3 ’s, showing dominance of suscepti¬ 
bility. In this cross there seems to be a slight dominance of resistance, 
the parents producing on the average 5.6 per cent more bunt than the 
F 3 progeny. 
The lower limit is decidedly lower than any recorded example of either 
parent under similar conditions. This shows that Fortyfold and Red 
Russian (which are known to be intermediate in respect to resistance 
of bunt, when compared with wheats like Turkey, Alaska, and Florence 
on the resistant side, and Hybrid 128 and Jones Winter Fife on the sus¬ 
ceptible side) possess different kinds of resistance which are cumulative 
in effect, when brought together by crossing. These resistances have 
a value somewhere between 10 and 20 per cent in reducing the amount 
of bunt that would be produced on the extremely susceptible varieties 
when planted under conditions favoring maximum infection. By cross¬ 
ing them, segregates are obtained that are from 10 to 20 per cent more 
resistant than either parent. It is to be expected, and the evidence 
points to the probability, that segregates also occur which are from 10 
to 20 per cent more susceptible than either parent. The difficulty of 
isolating the susceptible segregates, on account of the fungus destroying 
the reproductive organs, was pointed out in another place. The segre¬ 
gate dilute or feeble resistance of Fortyfold shown here is in accord with 
that found in the cross between it and Turkey. The feeble resistance 
of Red Russian has not as yet been definitely corroborated in other 
crosses. 
DISCUSSION AND CONCLUSIONS 
The investigation described in this paper forms a part of the general 
work of the Washington Experiment Station looking toward the discovery 
of methods for reducing the destructiveness of bunt or stinking-smut in 
