504 
Journal of Agricultural Research 
Vol. XXVII, No. 7 
lations are not fairly comparable. This objection applies to all com¬ 
parisons of recessive F 3 progenies with recessive parental populations in 
Table Xll, the former having had, in every case, a higher rate of early 
flowering than the latter. This factor of relative earliness of the plant 
also affects the comparisons between recessive F 3 progenies which had 
the same F t grandparent. Of the two pairs of recessive F 3 ’s mentioned 
on page 499 as showing significant differences in mean grade of petal 
spot (28-32 and 28-49, 21-16 and 21-22), in each case the progeny 
which gave the lower mean for petal spot had the higher rate of early 
flowering. 
On the other hand, the evidence of the occurrence of modifying factors 
from comparisons involving dominant F 3 progenies is not vitiated by 
differences in earliness, since in the dominant populations a retarded 
condition of the plants apparently did not result in a more pronounced 
expression of the petal spot. Of the two comparisons of dominants in 
Table XII, one (F s 24-46) gave a significantly higher and the other 
(F 3 20-14) gave a significantly lower mean grade of petal spot than the 
corresponding dominant parental population, yet the rate of early flow¬ 
ering in both F 3 ’s had resembled the parental rate. Comparison with 
one another of the dominant F 3 progenies 20-14 an d 20-15, both of 
which had the same F 2 grandparent, shows that the latter gave a sig¬ 
nificantly higher mean for petal spot grade, although it had much the 
heavier rate of early flowering. The relations of earliness and develop¬ 
ment of the petal spot are in this case the reverse of what was noted in 
the recessive populations. 
Returning to the indications of segregation in the first generation 
afforded by a correspondence between the differences in F x and F 3 (Table 
IX) it may be said that, in this case, the factor of relative earliness did 
not operate consistently in the recessive populations. In the cross 
1-3-12-14X 12 the recessive F 3 descendants of F t plant 28 gave a very 
significantly higher mean for petal spot than the recessive F 3 descendants 
of F x plant 20; and in the cross 3-2-4-1 X 13 the recessive F 3 descendants 
of F x plant 24 gave a very significantly higher mean for petal spot than 
the recessive F 3 descendants of F x plant 21. Rut the heavier rate of 
early flowering was shown in the first cross by the population which 
gave the higher mean for petal spot and in the second cross by the popu¬ 
lation which gave the lower mean for petal spot. 
The position as regards modifying factors for petal spot in this material 
may be summed up in the statement that there are indications, but not 
cionclusive proof, of the existence of such factors which, by their segrega¬ 
tion and recombination, have brought about a slight degree of differen- 
tation in the hybrids. 
VARIATION IN PETAL SPOT ON THE INDIVIDUAL PLANT 
The frequency distributions based upon averages of from 3 to 10 
flowers per plant (Table VI) show a gap, amounting to 2.5 grades, be¬ 
tween the plants which gave the lowest averages in the dominant parental 
and F 3 populations and the plants which gave the highest averages in 
the corresponding recessive populations. The gap is bridged, however, 
if individual flowers be considered. This is shown by the data presented 
in Table XIII, which gives the grades of the individual flowers 
grading lowest in the 12 dominant and highest in the 12 recessive parental 
and F 3 progenies of 1923. 
