1020 
Journal of Agr icultural Research 
Vol. XXVIII, No. 10 
It is evident that there is considerable range in sterility of Fi hybrids. In the 
cross Marquis X Alaska the Fi plant was practically as fertile as either parent. 
The only indication of sterility was the wrinkled grain and the great range in 
size of the kernels. In F 2 the range in average sterility for the different crosses 
is less than the Fi, due partly to the greater number of individuals present. 
There seems to be little relation between sterility in Fi and sterility in F 2 . This 
result is not entirely unexpected since in the more sterile Fi individuals only the 
more favorable combinations would survive, while in such crosses as Marquis 
X Alaska many of the gametic recombinations would survive and function more 
or less favorably. In the F 2 , sterility is due not only to gametic incompatibility 
but also to differences in somatic development. Certain plants are so poorly 
developed that they set no grain, while others are extremely vigorous in vege¬ 
tative development but are sterile, due to gametic imperfections. The difference 
in sterility in F 3 is significant. In the case of Marquis X Alaska the sterility is 
greater than in the cross Amby X Kubanka. This is partly due to climatic 
conditions, since the F 3 of Marquis X Alaska was grown in Washington State 
where a greater number of the weaker types undoubtedly survived, while the other 
cross was grown in Maine and only the strongest and most vigorous segregates 
could survive the extreme climatic conditions. 
Table I.— Sterility of various species crosses as indicated by average number of 
grains per spikelet ( GS ) in F\, F 2 , and F 3 . The number (n) of individuals in¬ 
volved is indicated in each case 
Crosses 
Ft 
f 2 
♦ 
Fj 
n 
GS 
n 
GS 
n 
GS 
Marquis X Kubanka.. 
13 
1.05±0.08 
135 
0.80±0.03 
Hybrid 143 X Alaska... 
34 
1.09± .05 
293 
. 63± . 02 
Marquis X Alaska.... 
8 
2 .00d= . 16 
152 
.98± .03 
1536 
1.09±0.01 
Hybrid 143 X Polish... 
19 
. 66± .06 
82 
. 70=fc .04 
Amby X Kubanka... 
8 
. 58± .03 
52 
1.07± .06 
38 
1.59± .04 
It has been found that gametic and somatic sterility is closely associated with 
chromosome behavior. Those individuals with their chromosome number 
approaching either 14 or 21 survived more frequently than the individuals with 
an intermediate chromosome number. Thus it would be expected that where 
there was a high degree of sterility in F x the segregates with an intermediate 
chromosome number would be more rapidly eliminated, while in the relatively 
fertile combinations segregates with an intermediate number might survive 
in many cases and a relatively high degree of sterility would persist for a num¬ 
ber of generations, although complete sterility would be comparatively rare. 
This hypothesis can probably be checked by an analysis of genetic behavior. 
For instance, in the more sterile combinations one might expect to find a higher 
correlation between the various characters due to the elimination of intermediate 
types. In such a case the surviving segregates would resemble the emmer 
parent or the vulgare parent and the characteristic features of the two groups 
would be more or less closely associated. In less sterile individuals the inter¬ 
mediate type would be expected to survive to a considerable extent, so that the 
vulgare and emmer characteristics might be combined in many cases and there 
would then be little correlation between the characteristic features of the two 
groups. One might also expect the recovery of occasional segregates com¬ 
bining the desirable characters of the two species in the more fertile combina¬ 
tions, while in more sterile hybrids the only homozygous fertile segregates would 
resemble one parent or the other. The available data do not, however, indicate 
