e we have to do
with terrestrial organisms, the most effectual barriers are wide reaches
of ocean; and, accordingly, we find that these exercise an enormous
influence on the modification of terrestrial types. Moreover, we find
that the _more_ terrestrial an organism, or the _greater_ the difficulty
it has in traversing a wide reach of ocean, the _greater_ is the
modifying influence of such a barrier upon that type. In oceanic
islands, for example, many of the plants and aquatic birds usually
belong to the same species as those which occur on the nearest
mainlands, and where there are any specific differences, these but
rarely run up to generic differences. But the land-birds, insects, and
reptiles which are found on such islands are nearly always specifically,
and very often generically, distinct from those on the nearest
mainland--although invariably allied with sufficient closeness to leave
no manner of doubt as to their affinities with the fauna of that
mainland. Lastly, no amphibians and no mammals (except bats) are ever
found on any oceanic islands. Yet, as we have seen, on the theory of
special creation, these islands must all be taken to have been the
theatres of the most extraordinary creative activity, so that on only
three of them we found no less than 1258 unique species, whereof 657
were unique species of land animals, to be set against one single
species known to occur elsewhere. Nevertheless, notwithstanding this
prodigious expenditure of creative energy in the case of land-birds,
land-shells, insects, and reptiles, no single new amphibian, or no
single new mammal, has been created on any single oceanic island, if we
except the only kind of mammal that is able to fly, and the ancestors of
which, like those of the land-birds and insects, might therefore have
reached the islands ages ago. Moreover, with regard to mammals, even in
cases where allied forms occur on either side of a sea-channel, it is
found to be a general rule that if the channel is shallow, the species
on either side of it are much more closely related than if it be
deep--and this irrespective of its width. Therefore we can only
conclude, in the words of Darwin--"As the amount of modification which
animals of all kinds undergo partly depends on lapse of time, and as the
islands which are separated from each other or from the mainland by
shallow channels are more likely to have been continuously united within
a recent period than islands separa