Studies on the Variability and Heritability of Pigmentation in Oenothera. 357 
the percentage of O. rubrinervis was larger, but there was an excellent 
opportunity for crossing with O. rubrinervis to have taken place. How 
closely the 3:1 ratio will be adhered to therefore, remains to be seen. 
The 1g10 offspring of the self-fertilized Wood’s Hole plants (therefore 
belonging to the F,) numbered 70 individuals. Unfortunately, only 
15 of these, which came into bloom, could be determined, owing to 
my absence earlier in the season. Of these, 10 were O. rubricalyx 
and 5 O. rubrinervis. 
The 71 plants of O. rubricalyx obtained in 1909 from the open- 
pollinated plants at Chicago, together with the 45 obtained from the 
self-fertilized Wood’s Hole plants, gave an ample number for observing 
the range of variability in this form. In the great majority of cases 
practically no variability in the color pattern of the buds could be 
observed. Every flower corresponded exactly with plate VI, fig. 8. 
Two plants were observed, however, in which there was a marked 
falling off in the amount of pigment in the bud color pattern on flowers 
produced late in the season. This was not noticed earlier in the season, 
and as all the plants were carefully scrutinized in every stage of 
development, it could not have occurred until observed. Both these 
cases were found in self-fertilized plants of O. rubricalyx (F2). My 
notes for these two cases read as follows: 
“No. g. Sept. 18. Very late in blooming. Buds paler than typical 
red, the red pattern not extending quite to the margin of the sepals. 
Hypanthium red, but less deep than typical”. “No. 14. Sept. 30. 
Buds, color pattern 7 with red hypanthium, green streak where sepals 
meet. Also less red on hypanthium, so that the bud as a whole has 
decidedly less than typical. Where shaded, one side of a bud is quite 
green”. This shows a tendency in the individuals to a falling off in 
pigment production toward the end of the season. These extremes 
of variation were not found in any other case. The range of variation 
in the color pattern of the buds is thus seen to be much more circum- 
scribed than in O. rubrinervis, although it is probable that the quantity 
of anthocyanin produced in the cells may vary considerably without 
producing a visible effect. The increase in pigment production in 
O. rubricalyx is greater than would be indicated by the amount of 
extension of the color pattern, for the cells of the colored area are 
evidently more densely packed with anthocyan than in O. rubrinervis. 
O. rubricalyx does not sustain the same relations to O. rubrinervis as 
the other mutants do to their parent, for it merely differs quanti- 
tatively in a single character (pigment production), while O. oblonga, 
