468 
Journal of Agricultural Research 
Vol. XXIII, No. 6 
The total populations of the F 4 and F 6 generations are presented in 
figure 2 in terms of the distribution of rows in the different classes per 
thousand, similar to the F 3 distribution in figure i. It will be seen that 
the F 4 and F 5 generations very nearly parallel each other. If the two 
susceptible lines had been eliminated, the other five would show decidedly 
greater resistance than the F 4 generation in spite of the fact that resistant 
varieties generally were twice as badly smutted in 1920. (See Table 
III.) Figure 2 may be taken to fairly represent the first peak of re¬ 
sistant segregates of the F 3 curve in Figure 1. No doubt subsequent 
selections will isolate the full resistant qualities of Turkey in the progeny 
in the homozygous condition. It is not unlikely that some of the F 6 rows 
are already in that condition and can not be changed by selection. 
! 
| 
» It 25 45 65 75 65 
Ptrcfttliuti 
Fig. a .-^Graphical representation of resistant segregates in later generations. The similarity of the F< and 
F» segregates indicates the constancy of the selections. The increase on the susceptible end of the Fs 
curve is due to the selection of xo families in that direction from the F4. 
For convenience in summarizing the data in respect to the bunt re¬ 
sistance of Turkey and Hybrid 128 and the progeny of a cross between 
them, Table VII has been prepared. It should be kept in mind that the 
result of a given generation represents the constitution of the generation 
preceding. Thus the results of the F 3 generation show that the F 2 
generation reproduced the extremes of the parents with a rather indefi¬ 
nite frequency center near each parent center. The great majority 
tended toward the center of Hubrid 128, indicating dominance of suscepti¬ 
bility. Similarly the F 4 and F 6 class frequencies indicate the constitu¬ 
tional resistance of the F s and F 4 generations. Each later generation 
shows less variation than the one preceding, as would be expected in a 
selfrfertilized race when selected in the direction of the recessive class 
center. 
