LUTHER BURBANK 
recombine them variously by hybridizing, and thus 
to develop new races from the old stock. 
When, however, the plant developer, through 
his hybridizing experiments, brings together 
groups of characters in which the old guards, so to 
speak, that have control over the fundamental 
characters are in conflict, no union is possible. 
Either fertilization wiil not take place, or the 
offspring will be sterile. Only within narrow lim- 
its, and as regards the new and relatively unes- 
sential characters, can there be diversity or, at 
most, the accentuation of old characters. 
Such an accentuation, for example, occurs, we 
may suppose, in the case of the hybrid walnuts, 
which take on gigantic growth. Both Persian wal- 
nut and California walnut have in their germ 
plasm the hereditary factors of large groups of 
remote ancestors of the Mesozoic era, when gigant- 
ism was the fashion, but these factors have for 
Jong generations been subordinated by newer ones 
born in a less favorable era. Now, however, 
hybridzation brings the two strains together, and 
the two dominant groups of factors for slow and 
relatively dwarfed growth, in some way mask or 
neutralize each other, enabling the earlier groups 
to make their influence felt. 
And here, as we have seen, the factors for 
growth that have thus been rudely disturbed as to 
[236] 
