1248 
Journal of Agricultural Research 
Vol. XXVIII, No. 12 
the factor x mutates readily to X during somatic development so that at game- 
togenesis the Fi hybrid of type with rogue is homozygous for the rogue factor. 
The allelomorphe of X in Mummy is called x' and differs from x of the rogue- 
producing variety in being more stable, rarely, if ever, mutating to X when in 
the homozygous x'x' condition. However, in the heterozygous Xx' combina¬ 
tion (F x and succeeding generations of rogue X Mummy) the combination, while 
more stable than the Xx germplasm of rogue X type, nevertheless is fairly un¬ 
stable, and x' mutates frequently to X. The frequent change of x or x' to X 
during development of the plant is called mass somatic mutation and the un¬ 
equal proportion of the two kinds of gametes formed, that is, x or x': X, results 
in masking the true Mendelian nature of the inheritance. 
The behavior of certain atypical Fi hybrids of the same cross indicated that 
occasionally x and x' mutate to produce intermediate forms in regard to stipule 
shape. The inheritance of these latter forms has not been studied. 
Since the assumed genetic constitution of the rogue, type and Mummy was 
derived from the fact of segregation rather than from the actual F 2 ratios ob¬ 
served, it seemed desirable to analyze further the character of the gamete pro¬ 
duction by the F! hybrid of rogue crossed with a nonrogue-producing variety. 
When out-crossing such an F hybrid with Gradus type, a rogue gamete uniting 
with a type gamete would produce a rogue, and a nonrogue gamete uniting with 
a type gamete would produce a nonrogue plant. This paper is a preliminary 
report of the results of such an analysis. 
The Fi plants of rogue X Mummy are very late in flowering and therefore 
difficult to cross with the early-flowering Gradus type. For this reason the F t 
plants of rogue X Rice's 330 (a nonrogue-producing variety) were substituted. 
Rice's 330 is an early-maturing sort with stipules intermediate in size and shape 
between Gradus type and Gradus rogue. The mean ratio of of stipu l e as 
determined by measurements of four stipules from the upper part of mature 
plants for the three parental types was as follows: Gradus type, 1.693±0.0107; 
Gradus rogue, 2.339±0.0096; and Rice's 330, 1.991 ±0.0086 (Table I). 
In 1919*, 36 Fi hybrids of Gradus type X Rice's 330 were grown and found to 
have a mean stipule ratio of 1.825±0.0098, i. e., intermediate between the two 
parents (Table I). The stipules were larger than those of Rice's 330 and narrower 
than Gradus type, but had the emarginate apices typical of Gradus rather than 
the Dpore acutely pointed stipules of Rice's 330. 
During the summer of 1920, the writer was unable to have stipule measure¬ 
ments taken until very late in the season, at which time the foliage of only a few 
of the late flowering segregates was still green enough to measure. In all 87 
plants from 21 families were measured. The mean, standard deviation, and 
coefficient of variation are given in Table I. 
The F 2 segregates were not classified by inspection, and since the data are 
meager, no attempt is made to analyze the factorial differences between the two 
parents. The data are given at this time merely for comparison with that which 
follows. 
The F 2 hybrids of Rice's 330 X Gradus rogue and reciprocal, 32 in all, grown in 
1919, were intermediate between Rice's 330 and Gradus rogue in appearance, 
with a mean stipule ratio of 2.185±0.0213 (Table I). 
The F 2 generation consisted of 47 families, from which the mean stipule ratios 
of 247 plants were obtained. The range of variation includes that of both parents 
and indicates segregation into rogue and nonrogue forms (Table I). No data are 
available as to the exact segregation, since the plants were not classified by stipule 
shape. 
