574 
Journal of Agricultural Research 
Vol. XXVIII, No. 6 
F x is brought into contact with stigmas of any kind, and especially with ovules 
capable of being fertilized. Thus, the apparent greater fertility of the female 
gamete may be due merely to chance. 
With 7 bivalent and 14 univalent chromosomes present in the F x , numerous 
recombinations at the time of gamete formation would be theoretically possible. 
It is probable that the gametes formed would contain 7 bivalents in every case 
and that nearly all of the gametes would contain, in addition, one or more of the 
univalents. Assuming random assortment, one gamete out of 16,384, on the 
average, would be expected to contain no univalent, and an equal number would 
contain all 14 univalents. 
There is some basis for assuming that successful backcrosses on the F x wheat- 
rye hybrid with wheat and rye pollen are related in some way to the number of 
univalents present in the hybrid gamete, and that with a high number of univa¬ 
lents a successful backcross with wheat can result, while with none or possibly 
one univalent the rye backcross is possible. If it is assumed that a gamete with 
12, 13, or 14 univalents is required for a successful backcross with wheat pollen 
the theory would explain the results obtained. The 12, 13, or 14 univalents 
present in the gamete would appear to counteract any incompatibility that may 
exist between wheat and rye chromosomes to a sufficient degree to allow the 
setting of seed. 
Sakamura (11) found 42 somatic chromosomes in a fertile strain descended 
from a backcrossed wheat-rye hybrid. Kihara (5) confirmed this number in the 
fifth generation of this same lot of fertile hybrids. Two plants resulting from 
seed produced by an almost sterile plant in this same lot of wheat-rye hybrid 
descendants contained 38 and 42 somatic chromosomes, respectively. 
Jesenko (4) observed a self-fertile plant with dense hairiness on the sheaths 
of the lower leaves in the progeny of an F x wheat-rye hybrid backcrossed with 
wheat. This character was not observed in either the wheat or rye parent. In 
the F 3 generation of this hairy-sheathed plant there were 31 hairy-sheathed 
individuals and 23 glabrous-sheathed, or approximately a 9 : 7 ratio. On the 
assumption that 2 factors are concerned in the inheritance, there should be 2 
equal groups in the F 4 , segregating into 3 : 1 and 9 : 7 ratios, or a total of 3 
hairy plants to 2 glabrous plants. The actual ratio obtained by Jesenko was 
3 : 1.96. This is apparently an example of simple Mendelian inheritance in 
wheat-rye hybrids. 
The unusual ratios obtained on the segregation for the hairy-neck character 
can not be explained satisfactorily, as they are based upon too few plants, and 
also upon plants in the ancestry of which natural crossing has occurred at least to 
a small extent. Further investigations may-show that we are observing plants 
in which an excess of recessive factors, accumulated in the gametes through 
random assortment of wheat and rye chromosomes, results at times in a reversal 
of dominance, and at times in unusual ratios in inheritance. If this is the case 
it is possible that the three selections showing all hairy-necked individuals in 
their progeny are not fixed for this character, and that it may be impossible 
of fixation. 
The F 2 and later generations of wheat-rye hybrids show greater fertility but 
less rye resemblance than the F x . As the F 2 is usually a sesqui-hybrid with the 
parentage (T. vulgare 9 XS. cereale &) 9 XT. vulgare cT (21 haploid chromosomes 
X 7 haploid chromosomes) X 21 haploid chromosomes, the rapid disappearance of 
the rye characters in generations later than the F x appears explainable. In the 
F x generation two complete sets of chromosomes are present, one from the 
wheat parent, the other from the rye. The possibility of obtaining, in the F x , an 
egg cell containing only 7 rye chromosomes or 21 wheat chromosomes, is extremely 
