338 
Journal of Agricultural Research 
Vol. XXIX, No. 7 
f 2 progeny 
In the fall of 1922 the kernels from 
one average head of each of the F x 
plants were sown at the University 
Farm, Davis, Calif. The kernels were 
spaced 2 inches apart in the row. The 
date of emergence of the first head on 
each plant was noted. Seven hun¬ 
dred and .ninety two hybrid plants, 
representing 23 families, were tagged 
for first heading. A frequency table 
of these data, showing the variation 
in time of heading, is given in Table I, 
and a frequency polygon is shown in 
Figure 3. The frequency curve shows 
a so-called “heaping up” of individ¬ 
uals into a large group of fairly early 
frequency class in this part of the 
curve, and half of the 8 individuals 
therein are assigned to each group, 
there will be 609 plants in the early 
group and 183 plants in the late group, 
which is a ratio of 3.07 : 0.93. This 
indicates a one-factor difference for 
earliness in the two varieties. The 
deviation from the 3:1 ratio probably 
is not significant, for it is less than 
twice the probable error, ±0.04. 
Figure 3 gives the curve of the 
maximum daily temperature during 
the heading period of the cross. The 
maximum temperature gives a fairly 
good measure of the prevailing daily 
temperatures at this time. These 
Fig. 3.—Curves showing the frequency distribution of the F 2 hybrid population, the Sunset parent, and 
the Marquis parent, by dates of heading, and the maximum daily temperature for the heading period 
in degrees Fahrenheit, at the University Farm, Davis, Calif., in 1923 
plants and a smaller group of rather 
late plants, with a distribution cover¬ 
ing almost the entire range of the par¬ 
ents. A few plants were earlier than 
the earliest plants of the early parent, 
while only one was as late as the latest 
plants of the late parent. Consider¬ 
ing the family progeny in each row, 
in practically every case there was a 
definite segregation into an early group 
and a late group, with a distribution 
of each closely approximating a normal 
frequency curve. A combination of 
all families represented shows that 
there is some overlapping of the two 
distributions. If an arbitrary division 
between early and late groups is made 
at April 28, which is the minimum 
data are included to show that low 
temperatures were not the cause of 
the low frequency of first heading on 
April 28, and the consequent bimodal 
form of the curve. 
f 3 progeny 
The sowings for the F 3 progeny 
were made in the same way as those 
for the F 2 . One head from each of 
the F 2 plants was harvested and the 
kernels were sown 2 inches apart in 
rows 5 feet long and 16 inches apart. 
The index for earliness, however, was 
obtained in a different manner. The 
date of first heading for each individ¬ 
ual plant in each row was not noted, 
but the appearance of the first head 
