Feb. io, 1923 
Genetics of Bunt Resistance in Wheat 
465 
and Alaska, the parent strains, produced a small amount of bunt each 
year. 
An exhaustive test of Turkey X Red Russian was attempted in 1917, 
but the season was so unfavorable that the grain winter-killed badly 
and much of it did not germinate till spring, during which time the 
bunt spores with which the seeds hadbeen inoculated probably perished, 
for out of 513 rows planted, but 2 produced as much as 10 per cent 
of bunted heads. Random selection of 42 plants from this F 3 generation 
gave a preponderance of resistant type in the F 4 generation but was 
not carried further. 
There are five families in which the numbers are sufficiently large 
and the seasons sufficiently favorable to give results that are dependable 
for purposes of comparison. The distribution of the different classes 
has been refigured on the basis of the number of each class per 1,000 
rows and the result put in the form of curves in figure 1 in which the 
ordinates represent the number in each class and the abscissas represent 
the different classes. There are four distinct types of curves, one in 
which the 5 per cent class predominates, one in which the 95 per cent 
class predominates, one in which the 35 per cent class predominates, 
and two in which the major predominating class is 55 or 65 per cent 
and a minor class predominating on the resistant end of the curve, 
and a low center at the 25 per cent class. These two similar crosses 
are not so similar in reality as they appear from the position of the 
curves, for the data from which the curve representing the F 3 generation 
of Turkey X Hybrid 128 was taken was obtained in 1918, while the 
other curve representing the data of the F 3 generation of Fortyfold X 
Turkey was obtained in 1920. Reference to Table III shows that 
the three parent wheats produced an average of 13 per cent more bunt 
in 1920 than in 1918. If the curves were corrected for this seasonal 
difference, the F 3 generation of Turkey X Forty fold would show a parallel 
curve that would be from 10 to 20 per cent more resistant than the F 3 
generation of Hybrid 128 X Turkey. Turkey is one of the parents 
in both crosses, so the difference in the two curves must be due to the 
difference in the inheritance of the other two parents, Fortyfold and 
Hybrid 128. It seems more than coincidence that the 5-year average 
of the two wheats (Table III) should show 17 per cent less bunt pro¬ 
duced by Fortyfold than by Hybrid 128. In fact the F 3 generation 
of Fortyfold X Turkey shows a mean of 2.3 per cent less bunt than the 
mean of the other F 3 generation notwithstanding the more favorable 
season for bunt in which it was grown. Any combination of the parents 
in question shows from 10 to 15 per cent more bunt in 1920 than in 1918. 
The seasonal differences of each parent is, for Turkey, 7 per cent, for 
Hybrid 128, 14,6 per cent, and for Fortyfold, 15.6 per cent. It seems, 
therefore, that the curves in question should be separated, on account 
of seasonal differences, by at least one class, or somewhere between 
7 and 15.6 per cent. In other words if the F 3 generation of Turkey X 
Hybrid 128 had been tested in 1920 instead of 1918 the 194 rows tested 
would have produced between 7 and 14 per cent more bunt on the 
average than they did. This would have been between 9.3 and 16.9 
per cent more than the F 3 generation of Fortyfold X Turkey actually 
did produce in 1920. The picture of the inheritance of these two crosses 
is one in which susceptibility is dominant, with the susceptible segregates 
fluctuating around the mean of the susceptible parent in each case. 
A study of the material in the field revealed an important difference 
