June 7 ,1924 A Genetic and Cytological Study of Wheat Hybrids 1027 
but the correlations are neither high nor consistent. In the second cross there 
are no significant correlations between fertility and number of homozygous or 
heterozygous factors. This method is not entirely satisfactory, however, since 
many factors might be in a heterozygous condition without resulting in any 
qualitative differences which can be described and tabulated. 
It would appear from these results that segregates combining the desirable 
characters of the emmer group and the vulgare group can rarely be obtained in 
the comparatively sterile combinations, but that in the more fertile crosses such 
individuals might be obtained more frequently. It has previously been found 
that there is a very high degree of association between the various characters of 
the emmer and vulgare wheat in hybrid segregates (11). It is found, for instance, 
that most of the segregates which are rust resistant also resemble the durum or 
emmer parent in morphological characters and quality of grain. However, 
several vulgare segregates which were rust resistant were found in about 20,000 
F 2 and F 3 segregates of crosses between wheats in the emmer and vulgare groups 
(7). None of these segregates was of economic value, however. By using 
species which result in relatively fertile segregates the probabilities of obtaining 
F 2 segregates which would combine the desirable characters of the emmer and 
vulgare wheats would be greatly increased, although it is not known at the 
present time how fertile and homozygous these types might be in subsequent 
generations. It is possible that ultimately all of the intermediate types would 
be eliminated, although Sevier wheat is an economic variety combining several 
durum and vulgare characters. 
In comparing the genetic behavior of certain characters in the species hybrids 
it is also desirable to know how these same characters behave in fertile varietal 
crosses. Such characters as awn length, color of awns, color of grain and pubes¬ 
cence, and in certain cases hardness of grain, have been found to be dependent 
on Mendelian factor differences in varietal crosses within the vulgare group. 
It is obvious that certain characters used in species crosses can not readily be 
analyzed in fertile crosses because the characters are more or less common to 
the entire group. Such characters, for instance, as size of keel, extremely flinty 
grain, pithy culms and very long awns are found only in the emmer group and 
in very few cases are any of these characters absent in any variety within the 
emmer series. It is true that a few beardless varieties of durum and emmer 
wheats exist, but these have been derived for the most part, if not entirely, from 
crosses between members of the emmer group with members of the vulgare 
group. There is, however, one character which is found in Alaska and in none 
of the other varieties of the emmer group. This is the branching habit of the 
spike. An analysis of the behavior of this character was obtained from data 
on a cross between Alaska X Spring emmer. The Fx was slightly branched 
although the Fx plants frequently had some heads which were normal. In F 2 
the segregates showed 294 normal to 107 branched. Only six of the F 2 plants were 
branching to the same extent as the Alaska parent. - The degree of branching 
varied all the way from a reduplication of one or two spikelets to individuals 
which closely resembled the branched parent. The fact that there is much 
greater variation in the branched segregates, which are apparently recessive, 
is not unusual, as in Chelidonium doubleness is a recessive character and yet 
the doubles vary enormously while the singles do not vary at all in regard 
to petal number. The genetic behavior of this cross is being continued into F 3 
for verification of F 2 classification. At any rate, it is evident that the genetic 
behavior in the varietal cross is entirely different than it is in the specif cross. 
In the cross between Marquis X Alaska only five individuals were classed as 
branched in a population of 152 plants. In F 3 17 families bred true for normal, 
89 segregated, and only 4 bred true for the branched condition. 
