have been accidentally imported by ships upon some of these islands, and
even already it is observed that their descendants have undergone a
slight change of appearance, so as to constitute them what naturalists
call local varieties. The change, of course, is but slight, because the
time allowed for it has been so short. But the longer the time that a
colony of a species is thus completely isolated under changed conditions
of life the greater, according to the evolution theory, should we expect
the change to become. Therefore, in all cases where we happen to know,
from independent evidence of a geological kind, that an oceanic island
is of very ancient formation, the evolution theory would expect to
encounter a great wealth of peculiar species. On the other hand, as I
have just observed, the special-creation theory can have no reason to
suppose that there should be any correlation between the age of an
oceanic island and the number of peculiar species which it may be found
to contain.

Therefore, having considered the principles of geographical distribution
from the widest or most general point of view, we shall pass to the
opposite extreme, and consider exhaustively, or in the utmost possible
detail, the facts of such distribution where the conditions are best
suited to this purpose--that is, as I have already said, upon oceanic
islands, which may be metaphorically regarded as having been formed by
nature for the particular purpose of supplying naturalists with a
crucial test between the theories of creation and evolution. The
material upon which my analysis is to be based will be derived from the
most recent works upon geographical distribution--especially from the
magnificent contributions to this department of science which we owe to
the labours of Mr. Wallace. Indeed, all that follows may be regarded as
a condensed filtrate of the facts which he has collected. Even as thus
restricted, however, our subject-matter would be too extensive to be
dealt with on the present occasion, were we to attempt an exhaustive
analysis of the floras and faunas of all oceanic islands upon the face
of the globe. Therefore, what I propose to do is to select for such
exhaustive analysis a few of what may be termed the most oceanic of
oceanic islands--that is to say, those oceanic islands which are most
widely separated from mainlands, and which, therefore, furnish the most
unquestionable of test cases as between the theories of special c