found in Madagascar, sufficiently proves that it is no land-connection
that has brought about this small infusion of Indian birds and bats; while
we have sufficiently shown, that, when we go back to remote geological
times no land-connection in this direction was necessary to explain the
phenomena of the distribution of the Lemurs and Insectivora. A
land-connection with _some_ continent was undoubtedly necessary, or there
would have been no mammalia at all in Madagascar; and the nature of its
fauna on the whole, no less than the moderate depth of the intervening
strait and the comparative approximation of the opposite shores, clearly
indicate that the connection was with Africa.

_Concluding Remarks on "Lemuria."_--I have gone into this question in some
detail, because Dr. Hartlaub's criticism on my views has been reproduced in
a scientific periodical,[101] and the supposed Lemurian continent is
constantly referred to by quasi-scientific writers, as well as by
naturalists and geologists, as if its existence had been demonstrated by
facts, or as if it were absolutely necessary to postulate such a land in
order to account for the entire series of phenomena connected with the
Madagascar fauna, and especially with the distribution of the
Lemuridae.[102] I {427} think I have now shown, on the other hand, that it
was essentially a provisional hypothesis, very useful in calling attention
to a remarkable series of problems in geographical distribution, but not
affording the true solution of those problems, any more than the hypothesis
of an Atlantis solved the problems presented by the Atlantic Islands and
the relations of the European and North American flora and fauna. The
Atlantis is now rarely introduced seriously except by the absolutely
unscientific, having received its death-blow by the chapter on Oceanic
Islands in the _Origin of Species_, and the researches of Professor Asa
Gray on the affinities of the North American and Asiatic floras. But
"Lemuria" still keeps its place--a good example of the survival of a
provisional hypothesis which offers what seems an easy solution of a
difficult problem, and has received an appropriate and easily remembered
name, long after it has been proved to be untenable.

It is now more than fifteen years since I first showed, by a careful
examination of all the facts to be accounted for, that the hypothesis of a
Lemurian continent was alike unnecessary to explain one portion of the
facts,