166 Observation of the Oriyin of J'arieties. 
to be avoided on account of the sterility of the most 
highly modified individuals. 
My cultures embraced, as a rule, a hundred individ- 
uals each, but sometimes a few hundreds. There can 
hardly be a doubt that if I had carried out more extensive 
sowings I should have attained my object at least one 
year earlier. But the more stringent the selection is, the 
smaller are both the harvest and consequently the next 
year's crop. 
Of course the reader will ask, has this transition been 
a gradual or a sudden one ? I consider it sudden ; but 
much depends on the meaning that we attach to the 
words. At any rate the change did not occupy centuries, 
as is commonly supposed by the current theory of selec- 
tion ; it did not even require one decade. Three years 
were sufficient in a culture of no more than a few square 
meters in extent. 
I now come to the details of the experiment and shall 
first give a short description of the original wild species. 
The species does not grow around Amsterdam. The 
herbarium material collected by me in various parts of 
the Netherlands points to the general occurrence of a 
mean number of 13 rays. HEINSIUS plotted curves from 
plants from two localities in the province of North Bra- 
bant, and obtained the following numbers. The first row 
relates to plants which were collected near Vucht, the 
second to a collection from Hintham. 1 
NUMBER OF LIGULATE FLORETS (L. F.) IN THE NETHERLANDS. 
I- F. 6789 10 11 12 13 14 15 16 17 18 19 20 21 
Vucht 1 13 5 3 8 18 78 37 22 11 17 2 33 
Hintham 10 99 8 15 14 33 9 4 1 1 00 
1 Be r. d. d. bot. Gcs., Vol. XVII, p. 87. I have already exhibited 
the variation in both localities united into a single curve in Vol. I 
(See p. 152, Fig. 32). 
