P.S.—After giving much condensed, my argument versus continental extensions, I shall append some such sentence, as that two better judges than myself have considered these arguments, and attach no weight to them.
CHARLES DARWIN TO J.D. HOOKER. Down, August 5th [1856].
...I quite agree about Lyell's letters to me, which, though to me interesting, have afforded me no new light. Your letters, under the GEOLOGICAL point of view, have been more valuable to me. You cannot imagine how earnestly I wish I could swallow continental extension, but I cannot; the more I think (and I cannot get the subject out of my head), the more difficult I find it. If there were only some half-dozen cases, I should not feel the least difficulty; but the generality of the facts of all islands (except one or two) having a considerable part of their productions in common with one or more mainlands utterly staggers me. What a wonderful case of the Epacridae! It is most vexatious, also humiliating, to me that I cannot follow and subscribe to the way in which you strikingly put your view of the case. I look at your facts (about Eucalyptus, etc.) as DAMNING against continental extension, and if you like also damning against migration, or at least of ENORMOUS difficulty. I see the ground of our difference (in a letter I must put myself on an equality in arguing) lies, in my opinion, that scarcely anything is known of means of distribution. I quite agree with A. De Candolle's (and I dare say your) opinion that it is poor work putting together the merely POSSIBLE means of distribution; but I see no other way in which the subject can be attacked, for I think that A. De Candolle's argument, that no plants have been introduced into England except by man's agency, [is] of no weight. I cannot but think that the theory of continental extension does do some little harm as stopping investigation of the means of dispersal, which, whether NEGATIVE or positive, seems to me of value; when negatived, then every one who believes in single centres will have to admit continental extensions.