The Origin of Chrysanthemum Scgctiun Plenum. 189 
saved 17 seed-parents with 48-67 rays in their terminal 
inflorescence. From these I collected the seed from the 
flowers which bloomed in July separately from those 
whose flowers had opened after the selection had taken 
place. But from the former specimen I raised only 
three plants that flowered (with 41-44-47 rays in their 
terminal head), which evidently could have no effect on 
the shape of the curve and were soon removed. Fertili- 
zation in 1899 therefore was again very pure. 
Having arrived at the end of the account of our ex- 
periment, all that remains is to compare the course of 
the selective process in this case with the ordinary pro- 
cess of selection carried out in the improvement of agri- 
cultural plants. I refer the reader to FRITZ MUELLER'S 
experiment with the many-rowed maize already de- 
scribed. 1 
That the difference is essential will be clear from the 
description given. In the case of the maize the object 
was to intensify the racial character (12-14 rows) as 
much as possible by selection; in the case of the Chry- 
santhemum the object was to uncover a latent character 
and to bring this to its full development. In the first 
case a visible character that had been known for ages 
o 
had to be increased as much as possible ; in the second, 
according to current conceptions at least, a new char- 
acter had to be called forth. The 26-28 rowed ears fall 
within the range of fluctuation of the 12-14 rowed race; 
and they would without doubt have appeared within it 
without any selection, if cultures of sufficient extent, 
which could be calculated beforehand, had been grown 
(Vol. I, p. 162). 
Without any doubt my crop of 1897 would have 
1 See the pedigree in Vol. I, Fig. 18, p. 73. 
