I THE DARWINIAN HYPOTHESIS 3 
described and may be identified in collections, and 
the number of separable kinds of living things is 
under-estimated at half a million, Seeing that 
most of these obvious kinds have their accidental 
varieties, and that they often shade into others 
by imperceptible degrees, it may well be 
imagined that the task of distinguishing be- 
tween what is permanent and what fleeting, 
what is a species and what a mere variety, 
is sufficiently formidable. 
But is it not possible to apply a test whereby a 
true species may be known from a mere variety ? 
Is there no criterion of species? Great authori- 
ties affirm that there is—that the unions_of 
members of the same species are always fertile, 
while those of distinct species_are_either_sterile,. 
or their offspring, called hybrids, are so, It is 
affirmed 1% only that this is an experimental 
fact, but that it is a provision for the preservation 
of the purity of species. Such a criterion as this 
would be invaluable ; but, unfortunately, not only 
is it not obvious how to apply it in the great 
majority of cases in which its aid is needed, but 
its general validity is stoutly denied. The Hon. 
and Rev. Mr. Herbert, a most trustworthy authority, 
not only asserts as the result of his own observa- 
tions and experiments that many hybrids are 
quite as fertile as the parent species, but he goes 
so far as to assert that the particular plant Crinwm 
capense is much more fertile when crossed by a 
B 2 
