The Origin of Chrysanthemum Scgctum Plenum. 175 
the other maximum, without however being symmetrical. 
Even in this group the race was therefore still far from 
being pure. Below I give the curve for the offspring 
of the single best seed-parent together with the sum of 
the curves representing the offspring of the five remain- 
ing seed-parents (Nos. 2-6). These curves therefore 
refer to the initial culture of the 21-rayed race (1896). 
L. F. 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 
No. 1 1 3 9 15 15 22 30 33 36 64 123 15 5 
Nos. 2-6 1 5 11 12 70 84 69 92 79 77 114 150 416 46 3 1 
All in all 370 plants were recorded for No. 1 and 
1220 for Nos. 2-6. 
Only the first named group, that is to say the off- 
spring of the plant numbered 1 in 1895, was used for 
the continuation of the experiment, and from it the best 
seed-parents for the purification of the race were selected 
on the basis of an examination of their lateral branches. 
* 
These were two plants the lateral flowerheads of which 
gave the following curves (1896) : 
L. F. 12 13 14 15 16 17 18 19 20 21 22 
No. la 01143220330 
No. I b 000000203 14 
Of the two, No. \b most obviously belongs to the 
race I was looking for. 
I harvested only the seeds of these two plants, and 
sowed them separately in the following year. In har- 
vesting this seed I confined myself to flowers which had 
bloomed after the other plants had been removed and 
had therefore been pollinated with their own or similar 
pollen. The result corresponded with my expectation, 
for in the following summer the race was pure on both 
beds. 
