1916} FORSAITH—ONAGRACEAE 469 
and which present impotent microspores as hybrid segregates, or 
hybrid species. 
In view of this great quantity of evidence in favor of the coexist- 
ence of hybridism and sterility, together with a habitual lack of 
contradictory proof, it seems safe to assume that these conditions 
illustrate a general biological correlation. Consequently, whenever 
there appears a large percentage of impotent pollen on the anther, 
the natural conclusion is that this sterile state has been brought 
about by previous ancestral crossing. 
These facts naturally lead to the question of the stability of 
hybrids in nature, which the morphological evidence of pollen 
sterility reveals as such. This question can likewise be answered 
in the same manner as the first, namely, by reference to cited 
examples. BATESON (3) in his discussion of this subject mentions 
_@ number of instances where hybrids have remained constant, 
instead of following the usual Mendelian principle. In regard to 
this phenomena he states as follows: ‘Literature of hybridization 
and heredity abounds with examples of hybrids which are said to 
have bred true, or, as we should say, without segregation”; Dr 
Vrirs (7) in a similar discussion of Oenothera biennisX muricata 
mentions constant hybrids which resemble one parent more than 
the other; and Davis (6) in speaking of crosses of Oenothera 
biennia and O. grandiflora says that hybrids were produced which 
resembled O. Lamarckiana to such a degree as to justify the con- 
sideration as a working hypothesis that O. Lamarckiana arose as 
a hybrid, and (to quote more directly) “it will be apparent to 
the reader that if the evidence should finally indicate Lamarckiana 
to be of hybrid origin, as a number of writers have already sug- 
gested, many of the so-called mutants are likely to be snterpayted 
as segregates splitting off according to Mendelian expectations.” 
Thus it is apparent that there are numerous quite constant hybrids 
which may be considered as segregated species. 
Having discussed the morphological indications of the relation 
of pollen sterility to that of hybridization, and the proof of the 
Stability of hybrids thus formed, it is evident that these two so 
well established principles must have some bearing upon the origin 
of certain races of plants. At present there are two conceptions 
