FACTORS INFLUENCING FLOWER SIZE IN NICOTIANA 345 
in 1 91 2, and five of them were transferred to the greenhouse when the 
plants were cut back. Fi H38 and Fi H44 were cut back on approxi- 
mately the same dates. 
Table Vb 
Frequency Distributions for Length of Corolla in Fi Ha 
Class 
Centers 
in 
Ml 
llir 
neters 
Designation 
No. 
Mean 
Coef. Var. 
46 
47 
48 
49 
50 
51 
52 
53 
54 
55 
56 
57 
58 
59 
Means for totals, 1913.. . . 
I 
I 
I 
2 
5 
I 
3 
5 
19 
53.42 ±.37 
4.44i.49 
Means for totals of same 
plants, 1912 
Means for totals, 1912.. . . 
2 
2 
I 
I 
5 
I 
3 
2 
I 
I 
19 
52.l6i.46 
6.07 i.66 
2 
2 
I 
2 
6 
I 
3 
3 
3 
I 
24 
52.54rb.42 
5-84±.57 
Means for first dates, 1913 
I 
I 
2 
I 
4 
8 
I 
I 
19 
54.05 ±.26 
3-i5±-34 
Means for mid dates, 1913 
I 
4 
3 
2 
6 
2 
19 
54.53 ±.28 
3.33=^.36 
Means for last dates, 1913 
Means for different dates. 
I 
2 
I 
3 
2 
3 
3 
3 
I 
19 
52.84i.49 
5.99i.66 
1913 
I 
I 
I 
3 
I 
3 
6 
3 
19 
53.26i.43 
5-27±.58 
The 1913 measurements reported in these tables were taken with 
a different purpose in view and without a proper appreciation of the 
variation in flower size which has been shown in the previous tables. 
The measurements in this case were taken as the plants were just 
coming into flower. The method of procedure was to measure all the 
fully open flowers which were shedding pollen on the dates on which 
measurements were taken. In this way measurements were taken 
beginning at the first plant and working through the whole series 
one after the other. This process was repeated several times, the 
object being to obtain at least 100 measurements on each plant. As 
soon as this number had been exceeded on some date for some indi- 
vidual plant, no further measurements were taken on that particular 
plant during the period. Accordingly the measurements on these 
plants cover a rather short period from less than 3 weeks to slightly 
over 9 weeks, and represent a period during which the plants were 
coming into full flower as shown by the progressive increase in the 
number of flowers measured on individual plants with the advance- 
ment of the season, from as low as 4 at the beginning of the period to 
as high as 186 on one day at the end of the period. Naturally only a 
slight fluctuation in the mean would be expected during this period, 
particularly in view of the semi-sterile nature of the plants involved, 
and this is exactly the result obtained. That these plants, however, 
can show a wide variation in such respects is plainly evident from a 
comparison with the distributions for 191 2. 
