natural species are either absolutely infertile if crossed with
individuals of other species, or, if they give rise to hybrid
offspring, the hybrids so produced are infertile when paired together.
The horse and the ass, for instance, if so crossed, give rise to the
mule, and there is no certain evidence of offspring ever having been
produced by a male and female mule. The unions of the rock-pigeon and
the ring-pigeon appear to be equally barren of result. Here, then,
says the physiologist, we have a means of distinguishing any two true
species from any two varieties. If a male and a female, selected from
each group, produce offspring, and that offspring is fertile with others
produced in the same way, the groups are races and not species. If, on
the other hand, no result ensues, or if the offspring are infertile
with others produced in the same way, they are true physiological
species. The test would be an admirable one, if, in the first place, it
were always practicable to apply it, and if, in the second, it always
yielded results susceptible of a definite interpretation.
Unfortunately, in the great majority of cases, this touchstone for
species is wholly inapplicable.

The constitution of many wild animals is so altered by confinement that
they will not breed even with their own females, so that the negative
results obtained from crosses are of no value; and the antipathy of
wild animals of the same species for one another, or even of wild and
tame members of the same species, is ordinarily so great, that it is
hopeless to look for such unions in Nature. The hermaphrodism of most
plants, the difficulty in the way of insuring the absence of their own,
or the proper working of other pollen, are obstacles of no less
magnitude in applying the test to them. And, in both animals and
plants, is superadded the further difficulty, that experiments must be
continued over a long time for the purpose of ascertaining the