178 Observation of the Origin of I'ariehcs. 
partial if not to complete predominance. The success 
of the experiment of course depends on factors still 
unknown to us, for it is by no means always successful. 
My belief in these principles, which DARWIN himself 
often refers to, led me to pay special attention, from the 
very outset of my experiment, to part-curves, i. e., to 
curves derived from the lateral flowers of the single 
plants (seep. 173). It is useless to give the numerous cases 
which afforded no indication of a latent character, and 
so I will proceed at once to that plant which was the 
first to do so. It was a specimen of the 21 -rayed race 
of 1896, which had 21 ray-florets in its terminal in- 
florescences and gave the following part-curve on the 
12th of August : 
L. F. 14 15 16 17 18 19 20 21 22 
No. Ic 1 1 2 2 2 3 3 4 
I refer to this plant as No. Ic 1 in order to indicate 
that it belonged to the same culture as Nos. la and Ib 
whose part-curves were given on page 175. It agrees 
with those two plants in the fact that there is not a trace 
of a maximum at 13; but it differs from them and from 
all the other plants that were examined on the same bed, 
by the possession of four flowers with 22 rays. On no 
other plant was there a single lateral flower with more 
than 21 rays. 
This indication was no doubt pretty small. It would 
not have been discovered but for the counting of the 
ligulate florets. Without this statistical method of in- 
vestigation it would certainly never have been grasped, 
for the plant If grew in a culture of about 1500 speci- 
mens. It was noted first, along with 500 others, as hav- 
1 Berichte d. d. lot. Ges., Vol. XVII, p. 91, where No ir is given 
as No. 12 in the series. 
