them will, in time, make a sea, if one does not already
exist.
But it is when we examine the zoology of these countries
that we find what we most require—evidence of a very
striking character that these great islands must have once
formed a part of the continent, and could only have been
separated at a very recent geological epoch. The elephant
and tapir of Sumatra and Borneo, the rhinoceros of
Sumatra and the allied species of Java, the wild cattle
of Borneo and the kind long supposed to be peculiar to
Java, are now all known to inhabit some part or other
of Southern Asia. None of these large animals could
possibly have passed over the arms of the sea which now
separate these countries, and their presence plainly indicates
that a land communication must have existed since
the origin of the species. Among the smaller mammals
a considerable portion are common to each island and the
continent; but the vast physical changes that must have
occurred during the breaking up and subsidence of such
extensive regions have led to the extinction of some in
one or more of the islands, and in some cases there seems
also to have been time for a change of species to have
taken place. Birds and insects illustrate the same view,
for every family, and almost every genus of these
groups found in any of the islands, occurs also on the
Asiatic continent, and in a great number of cases the
species are exactly identical. Birds offer us one of the
best means of determining the law of distribution; for
though at first sight it would appear that the watery
boundaries which keep out the land quadrupeds could be
easily passed over by birds, yet practically it is not so ;
for if we leave out the aquatic tribes which are preeminently
wanderers, it is found that the others (and
especially the Passeres, or true perching-birds, which form
the vast majority) are generally as strictly limited by
straits and arms of the sea as are quadrupeds themselves.
As an instance, among the islands of which I am now
speaking, it is a remarkable fact that Java possesses
numerous birds which never pass over to Sumatra, though
they are separated by a strait only fifteen miles wide, and
with islands in mid-channel. Java, in fact, possesses more
birds and insects peculiar to itself than either Sumatra
or Borneo, and this would indicate that it was earliest
separated from the continent; next in organic individuality
is Borneo, while Sumatra is so nearly identical
in all its animal forms with the peninsula of Malacca,
that we may safely conclude it to have been the most
recently dismembered island.
The general result therefore at which we arrive is, that
the great islands of Java, Sumatra, and Borneo resemble
in their natural productions the adjacent parts of the
continent, almost as much as such widely-separated
c 2