f 
422 SEMI-CENTENNIAL OF TORREY BOTANICAL CLUB 
variation, but the plants grown in all three years were quite similar 
in general vigor and habit of growth, although there was more or 
less variation in the number of branches and the length of the 
flowering period. 
Data on flower number in this variety have been obtained from 
two parent plants of 1913, and from three generations (1914, 
1915, and 1916 crops) derived by further intravarietal breeding. 
Until the summer of 1915 all plants (with the exception of one 
feebly self-fertile plant in 1915) were found to be self-sterile. 
The variety was kept in culture by crossing plants (usually im- 
mediate sister plants) that were cross-compatible. 
As a result of the vigorous and robust growth in the lines 
grown of this variety as many as 2,000 to 3,500 flower heads per 
plant were produced. At the climax of development 100 or even 
150 heads opened on a single plant in a single day. The period 
of bloom was also somewhat extended, but data were not used 
unless a plant completed its bloom before the first autumn freeze. 
TABLE 41 presents data for one of these plants. For this 
plant the period of bloom extended from July 14 to October 21. 
There is a decided decrease in both daily range and average. 
The partial variability seen in the entire range from 31 to 7 was 
the greatest in both directions seen in any plant of this variety. 
Of the crop of 1913, grown from commercial seed, there are 
data for the two plants which were crossed in obtaining seed from 
which all cultures were derived. For one of these the values ob- 
tained were а = 19.5, b = — 0.084, and the range was 23-12; 
for the other plant a = 17.4, b = — 0.081, and the range was 
19-13. With respect to values of a, these two parent plants were 
somewhat diverse. 
For the 1914 crop, data were obtained from nine plants: these 
gave values of a ranging from 18.6 to 20.8, with average of 19.8 
0.67. As shown in TABLE 42, the values of these plants are quite 
- well distributed about the value of the parent having highest value 
of a. 
In 1915, offspring weie grown from crosses involving four 
sister plants (of the 1914 crop) for which the values of a were 
almost identical and were quite of the average. The average 
values for each line of descent according to immediate parentage 
