PLANT-BREEDING 147 
but in the majority of instances the plants are not 
merely conserved, but cultivated; cultivation has led 
to selection of the best varieties; and continued selec- 
tion has resulted in the production of forms often 
very different in appearance from the wild plants from 
which they originated. We cannot create new forms; 
but by taking advantage of the innate tendency to 
vary which all plants display—some to a much greater 
degree than others—and by raising, generation after 
generation, the seeds of those individuals in which a 
certain abnormal feature is best displayed, we can 
produce an artificial race in which the selected 
character may -be developed to an extraordinary 
degree. But we have not by this means produced a 
new species. Seedlings of such plants will tend to 
“throw back” towards the original form; we can 
preserve or improve the special characters only by 
continued selection; if allowed to grow and seed 
unchecked, most of such plants will revert to the 
natural type in a few generations. Often this rever- 
sion is so rapid that seeds are useless for cultural 
purposes, and it is only by cuttings or graftings—that 
is, by growing parts of the original possessor of the 
required characters—that constancy can be main- 
tained ; this is what is usually done in the case of fruit- 
' trees, Roses, Pansies, and so on. 
Equally efficient in the hands of the cultivator has 
been another method of producing new forms— 
namely, hybridization. If the pollen of a plant be 
transferred to the stigma of a related species, offspring 
is often produced; and the product is a batch of plants 
intermediate in characters between the two parents, 
and generally uniform in appearance. Should these 
