ause they show us, firstly, that the physiological {184} test of
lessened fertility, both in first crosses and in hybrids, is no safe
criterion of specific distinction; secondly, because we may conclude that
there must be some unknown law or bond connecting the infertility of
illegitimate unions with that of their illegitimate offspring, and we are
thus led to extend this view to first crosses and hybrids; thirdly, because
we find, and this seems to me of especial importance, that with trimorphic
plants three forms of the same species exist, which when crossed in a
particular manner are infertile, and yet these forms differ in no respect
from each other, except in their reproductive organs,--as in the relative
length of the stamens and pistils, in the size, form, and colour of the
pollen-grains, in the structure of the stigma, and in, the number and size
of the seeds. With these differences and no others, either in organisation
or constitution, we find that the illegitimate unions and the illegitimate
progeny of these three forms are more or less sterile, and closely resemble
in a whole series of relations the first unions and hybrid offspring of
distinct species. From this we may infer that the sterility of species when
crossed and of their hybrid progeny is likewise in all probability
exclusively due to differences confined to the reproductive system. We have
indeed been brought to a similar conclusion by observing that the sterility
of crossed species does not strictly coincide with their systematic
affinity, that is, with the sum of their external resemblances; nor does it
coincide with their similarity in general constitution. But we are more
especially led to this same conclusion by considering reciprocal crosses,
in which the male of one species cannot be united, or can be united with
extreme difficulty, with the female of a second species, whilst the
converse cross can be effected with perfect facility; for this difference
in the facility of making reciprocal crosses, and in the fertility of their
offspring, must be attributed either to the male or female element in the
first species having been differentiated with reference to the sexual
element of the second species in a higher degree than in the converse case.
In so complex a subject as Hybridism it is of considerable importance thus
to arrive at a definitive conclusion, namely, that the sterility which
almost invariably follows the union of distinct {185} specie