830 
Journal of Agricultural Research voi. xxiv, no. 10 
Plates 4 A, B, and 5 A, are of the upper parts of a Graduslike type 
and of an intermediate segregate, respectively, from the F2 of Gradus X 
Mummy. The intermediate shown has stipules wider than the stipules 
of the Fj plant of the same cross shown in Plate 3, A. The difference is 
largely due to the fact that the stipules of the F^ plant are on a branch 
while those of the Fg intermediate are on the main stem. Stipules on a 
branch are smaller and narrower than those of the main stem in nearly 
all instances, irrespective of the variety or strain. 
It should be pointed out here that Y * and y are considered as the 
factors chiefly responsible for the difference for stipule shape between 
Mummy and Gradus. It is highly probable, from the increased varia¬ 
bility of the Fj generation over that of the parents, that a number of 
modifying factors interact with Y and y to affect stipule shape. 
RELATION OF FACTORS Aa FOR FLOWER COLOR AND Nfl FOR CHARACTER OF STEM 
It has been said before that stipules from the fasciated portion of the 
stem of Mummy are narrower than those of the apparently nonfasciated 
part of the same plant or from a normal plant. The Fj plants of the 
crosses of Mummy with Gradus were normal stemmed and had narrow, 
intermediate, or Mummylike stipules. No accurate counts were made 
of the fasciated F2 segregates, but apparently the expected 3 : i ratio 
of normal to fasciated plants occurred. 
Since fasciated plants appeared, on the whole, to have narrower 
stipules than normal plants, it might be suggested that linkage existed 
between n and A, and N and a, so that the existence of a factor pair Yy 
would not have to be assumed. That such is not the case is shown by 
the results of White (14), who found factors Aa and Nn to be inde¬ 
pendently inherited. The two cases are comparable as the AAnn parent 
of the hybrids studied by White was from the same strain as that used 
by the present writer. 
STATISTICAL EVIDENCE OF LINKAGE BETWEEN FACTORS AY AND ay 
In addition to the stipule measurements of the 34 plants mentioned 
in the preceding section, stipules of 41 other plants were measured from 
the F2 generation of Gradus X Mummy. Of the 75 plants measured 
(Table VII), 53 were A segregates (39 purples and 14 pinks) and 23 were 
a segregates (whites). The mean stipule ratio of ^e F2 plants as a 
whole was found to be 1.868i0.0126, for the A’s 1.905 ±0.0142, and for 
the a’s 1.710±0.0260. Considering a difference between the means of 
at least three times the probable error of the difference as being signifi¬ 
cant, it is found that the difference between the means of the A and a. 
segregates is significant, since it is five times the standard error. 
The same relation holds between the mean ratio of the colored and 
white F2 segregates, whether the A plants be considered collectively, or. 
separately as purples (with the factors A and B) and as pinks (with the 
factors A and b), the mean of the purples being 1.909 ±0.0175 and of the 
pinks i.864±o.oi 42 as compared with 1.905±0.0142 for the A segregate 
as a whole (Table VII). 
No essential difference exists between the means of Gradus and the 
white-flowered (aa) F2 segregates from the cross Mummy with Gradus. 
However, the hybrid whites show a greater variability than do the pure 
whites, the coefficient of variation of the former being 10.58 ± 1.0870 
compared with 5,78 ±0.2381 for the latter. 
