832 
Journal of Agricultural Research 
Vol. XXIV, No. 10 
As some uncontrolled pollinations may have occurred, there is possibly 
an experimental error in the results which should be kept in mind when 
interpreting the F2 data. On this account the crosses are being repeated, 
and it is planned to grow the generation under glass for protection. It 
is believed, however, that the data here presented are of sufficient interest 
to warrant publication at tliis time and that the errors from out*crossing 
will prove minimal. 
RELATION OR THR A AND A FACTORS FOR FDOWRR COLOR 
The Fj plants of the crosses between rogue and Mummy had colored 
(purple) flowers. In Table VIII is shown the percentage of colored- 
flowered (AA + Aa), segregates to white-flowered (aa), segregates in the 
F2 generation of rogue X Mummy, and reciprocal. The observed per¬ 
centages, 76.79 per cent AA and Aa plants and 23.21 per cent aa segre¬ 
gates conform to expectation. 
RELATION OF A AND C, FACTORS FOR FLOWRR COLOR 
Gradus rogue X Mummy in the Fj generation gave approximately the 
expected 56.25 per cent purples (with factors A, B, and C); 18.75 
cent pinks (with factors A and B); and 25.00 per cent whites (with 
factors aa), the observed percentages being 56.91 per cent, 19.88 per 
cent and 23.21 per cent, respectively, the deviations being well within 
the limits of error due to random sampling (Table IX). 
Although the dovetailing of observation and theory is good for the 
three classes of color segregates as a whole, there is in the F2 in both the 
case of Mummy X Gradus and Mummy X Gradus rogue, a deficiency of 
white segregates. The deficiency of aa plants can be explained by 
differences in disease resistance as well as by variations in time of matur¬ 
ing between colored and noncolored Fg segregates. On the whole, 
plants with colored flowers are more resistant to disease than are the 
white flowered forms, so that proportionately more of the latter would 
be lost before they reached maturity. For unavoidable reasons the 
Fj data were collected late in the season after a number of the plants 
had matured. Plants whose stipules were too dried to measure were 
not classified as to flower color but their occurrence was noted. 
Tschermak (12), Hoshino (7), and others have noted a correlation 
between the presence of the A factor and the habit of late flowering. In 
all probability the same correlation existed in these cultures. On this 
account, relatively more of the white segregates than of the colored 
ones were too mature to be used in this study at the time the stipule 
measurements were made. 
RELATION OF N AND N FACTORS FOR NORMAL AND FASCIATED STFM, RRSPKCTIVELY 
The 69 Fj plants of Gradus rogue X Mummy had normal stems. No 
Fj counts were made of the normal and fasciated plants, but appar¬ 
ently there occurred the expected Mendelian segregation. 
RELATION OF Xx'" AND Yy FACTORS FOR STIPULR SHAPE 
In 1919, 69 Fj hybrids (Table XII) were grown from the seed of 13 
pollinations made the preceding year. The young plants resembled 
Mummy in regard to stipule shape and were uniform in appearance. 
