304 BROOKLYN BOTANIC GARDEN MEMOIRS 
tion by reason of their coincidence of flowering periods, is actually the 
result of previous specific crossing. This view of the matter is strongly 
confirmed by the fact that the investigations of Brainerd and others 
on that genus have revealed a large number of natural hybrids. 
For a parallel case we may now turn to the genus Ranunculus. 
If any of the species which flower in the early summer are examined, 
such as R. acris, R. repens, R. aquatilis, R. Cymbalaria, a considerable 
proportion of imperfect pollen is usually present and frequently the 
grains vary extremely in size. This situation is shown for R. acris 
in Fig. 6, Plate V. Obviously there is a great range of size in the 
grains and some are imperfect. This condition is most naturally 
interpreted as a consequence of previous hybridization. Fig. 5, Plate 
V, illustrates the condition of the pollen in R. rhomboideus, a species 
which flowers in the very early spring long before the other species 
of the genus have opened their blossoms. The numerous grains 
shown in the illustration are clearly well developed and somewhat 
uniform in size. In the species under consideration as well as in R. 
odoratus, perfection in development of pollen is clearly correlated with 
isolation from possibility of contamination with other species. 
In view of the facts supplied in the above instances, which might 
be almost indefinitely multiplied in representatives of other angio- 
spermous families, it appears clear that the obvious interpretation 
of pollen sterility is to be adopted, namely as an indication (where 
it occurs under normal conditions of growth) not of mutability but 
of previous hybridization. The large situation which is so _ briefly 
illustrated by the accompanying photographs, indicates the necessity 
of bringing morphological criteria relating to the cytology and develop- 
ment of the gametic cells (pollen and embryo sacs) into court, as well 
as the data derived from genetical behavior, in speculations in regard 
to the origin of species. 
It seems clear from the evidence supplied on the systematic and 
phytogeographical sides on the one hand and that from the standpoint 
of morphology on the other, that the crossing of species in nature is 
an extremely common cause of the multiplication of species. It is 
further obvious that physiological and genetical criteria must not be 
given greater weight than the more reliable ones supplied by actual 
history and by morphological structure, in the all-important biological 
question of the origin of the species. It is finally apparent that the 
genetical status of the Oenotheras is so dubious that they cannot be 
brought into court to furnish decisive evidence in favor of the muta- 
tion hypothesis of De Vries. It may be added in conclusion that the 
multiplication of species by hybridization does not by any means 
invalidate the Darwinian hypothesis but merely supplies an additional 
