A MIXTURE OF GOOD AND BAD POLLEN 
Section of a pollen sack which may represent any one of a very large number of forms of Cra- 
taegus. 
Part of the grains are perfectly formed, the rest are shriveled or empty. 
This 
is the usual condition of pollen in hybrid plants, and is one indication that many of the 
North American hawthorns are hybrids. 
Dunbar believed in the Darwinian view 
of gradual change in evolution. 
In the midst of so much conflicting 
opinion there seems to be one point 
on which all the authorities agree, and 
that is the extreme variability of the 
genus. Sargent alone during the last 
twenty years has numbered his new 
forms up into the thousands, and the 
process of multiplication appears to be 
still going on. 
Unusual variability in plants is gener- 
ally considered good evidence of hybrid- 
ism, and it is more than probable that 
the hawthorns of the United States 
share with the European species in 
particular, and with the other Rosaceae 
in general, a marked disposition to 
hybridize. Luther Burbank has, as it 
were, specialized in producing variations 
by means of crossing; and his work, 
done on such a wholesale scale, 1s 
268 
(Fig. 8.) 
based on the principle that hybridism 
breaks up the continuity of inherited 
characteristics and results in the appear- 
ance of variations. 
EVIDENCE FROM STERILITY 
For further and more exact evidence 
we can turn to the morphological 
peculiarities of hybrids. Partial or 
complete sterility is and has long been 
recognized as an important basis of 
distinction between crosses and genetic- 
ally pure species. To be sure, when the 
parent forms show a considerable degree 
of compatibility, the fertility of the 
offspring may be practically normal or 
even entirely so. Then again sterility 
may be largely eliminated by selection; 
but it is not present except when 
crossing is possible, it does not occur in 
monotypic species, and it is absent in 
genera that are isolated either 
geo- 
