Plant Breeding. 257 
of culture and selection may be necessary before the progeny 
will uniformly resemble the parent. 
The variations in the seedlings from two crossed varieties, 
and the kind of selection needed to fix the desired variation, 
are illustrated by the following diagram (Fig. 173). Let @ 
Fic. 173. Diagram illustrating the selection of seedlings from a cross. 
represent the seeds from two crossed flowers A and B. The 
plants from these seeds will probably be quite variable, as is 
indicated by the divergent lines. Let us suppose the varia- 
tion marked ¢ to be nearest the ideal form. The plants 
grown trom 7 will be again quite variable in the second gen- 
eration 6, but probably less so than in the first generation. 
No plauts of the second generation may be nearer the ideal 
type than those of the first generation, but we select the 
plant nearest to our ideal, and plant the seeds from this. 
Each succeeding generation may be expected to produce 
less of variability than the one before it. Bye and bye, we 
may hope to secure a form that approaches our ideal ,and 
that comes tolerably true from seed. 
445. Planting with Reference to Chance Crossings. 
Many valuable varieties of fruits have unquestionably arisen 
from accidental crosses between plants of different varieties 
that chanced to be growing in proximity. Profiting by this 
hint, varieties are sometimes planted near together to favor 
the accidental crossing of the flowers, a practice to be en- 
couraged. 
