1250 
Journal of Agricultural Research 
Vol. XXVIII, No. 1 
However, some idea of the character of the F 2 segregation following the cross 
between rogue and Rice’s 330 may be obtained from a classification of 262 F 2 
plants grown in 1923 from 5 Fi hybrids grown in 1922 under a wire cage to protect 
them from volunteer crossing by insects. The F 2 plants were grouped as rogues, 
intermediates with pointed stipules of greater width than in typical rogues, and as 
broads with large stipules resembling Rice’s 330 or Gradus type. The variability 
of the rogues and intermediates was such that it was not always easy to separate 
them, but the differences between the two roguelike classes and the broads or 
type-like plants were easily recognized. In all there were 217 rogues, 37 inter¬ 
mediates, and 8 broads, clearly not a typical Mendelian segregation. 
During the winter of 1922-23 there were grown 155 plants from 32 back- 
crosses of the kind [(rogue X Rice’s 330) X type], as well as 8 of the reciprocal 
back-crosses. The Fi hybrids used as either pollen or pistillate parents were the 
same 5 plants whose F 2 progeny, derived from self-fertilized seed, were grown in 
1923 and described above. 
The progeny of these back-crosses were for the most part easily distinguished 
as broads with wide, rounded stipules, or as rogues with characteristic narrow, 
pointed stipules. The stipule differences between these two classes of segregates 
is clearly shown by the data for their mean stipule ratios given in Table I. Ths 
frequency distribution and mean of 1.748±0.0065 for the broad segregates 
are very near those of the Fi hybrids of Rice’s 330 X Gradus type. The varia¬ 
bility in stipule shape among the broad segregates, as shown by the coefficient 
of variation, is also strikingly near that of the Fi generation of the latter, i. e. 
4.63+0.2664 and 5.07+0.3054, respectively. 
The narrow or roguelike segregates resemble closely the pure rogues in regard 
to their frequency distribution, except that the range of variation of the hybrids 
extends below 2.05, the lowest class of pure rogues. The mean, 2.223+0.0155, 
is significantly different from that of 2.339+0.0096 for the pure rogues, but is 
also decidedly different from that of the broad (nonrogue) segregates. 
The roguelike plants also resemble the rogue parent in the great variability 
they exhibit, caused in part by the presence of chimeralike segregates char¬ 
acterized at the middle and upper nodes by having the corresponding stipules of 
each pair unlike, that is, one stipule narrow and pointed while the other stipule 
would be broad and obtuse. In all, 13 chimeras were identified, 10 from seed 
pollinated by the type and 3 from the reciprocal cross. The following measure¬ 
ments of the stipules at successive nodes are typical of this class. The ratios for 
the pair of stipules from the same node are inclosed in brackets and the largest 
ratio, usually above 2.05, indicates in every instance the roguelike stipule. 
Plant No. 
3.100-2 
3.111-2 
3.118-1 
(1.71) 
(1.66) 
(1.86) 
(2.76) 
(1.81) 
(2. 55) 
(1.87) 
(1.63) 
(1. 55) 
(3.36) 
(2.33) 
(2.77) 
(1. 53) 
(1.82) 
(1.92) 
(2.09) 
(2.36) 
(2.66) 
Average. 
2.05 
1.93 
2.21 
In the data given in Table I the chimeras are classed as rogues, presumably 
heterozygous for the rogue factor, as shown by the presence of both the pointed 
and obtuse stipules. Whether the chimeras are merely heterozygous forms or are 
mosaics in the sense of Bateson and Pellew is not known, since each plant produced 
only two or three seeds in the greenhouse, and these too late for spring planting. 
