As I do not know in what respects you have found one part of my previous letter not to 'tally' with another, I cannot fully explain it; but I fancy that you will find they do, if, in reading the letter, you carry in your mind the simple proposition that, from the nature of the case, there can be no physiological selection except where differentiating varieties ('incipient species') occur upon common areas and identical stations. I do not see any difficulty about willows, roses, brambles, &c., since Naudin's researches on _Datura_ have shown how much variability, due to the hybridisation of any two species, may give rise to the _appearance_ of there being many species. This, you will remember, is the view that Naudin himself takes with regard to willows &c.--although, of course, without any reference to phy. sel. If you will refer to p. 405 of the paper on phy. sel. you will find that from the first I have been aware of the difficulty about discontinuous areas to which you allude. But I think the converse line of evidence (viz. that of cross-sterility between incipient species on identical stations) will alone prove sufficient to verify the theory. At the same time I look for more corroboration from the cross-fertility of well differentiated species upon discontinuous areas where these are, as you say, oceanic islands, or, still better, mountainous districts where the allied species are severally peculiar to mountain tops and isolated valleys. For in these cases there must be much doubt, as a general rule, touching the species having been differentiated by topographical isolation upon the particular areas where they are now found. Moreover, and this I think quite as important, the consideration which Darwin adduces in another connection is obviated, viz. 'that if a species was rendered sterile with some one compatriot, sterility with other species would follow as a necessary contingency.' Yours very sincerely, G. J. ROMANES. P.S.--From your first letter it would almost seem that you had supposed me to doubt the fact (or, at any rate, the frequency) of cross-fertility in general. And this after I had written the article on 'Hybridisation' in the 'Ency. Brit.'! In June Mr. Romanes took a small house for the summer months outside Oxford at Boar's Hill, a district well known to Oxford people, and it was hoped country air and quiet might do him much good. He was rather headachy, and liked to lie on the grass in the garden and have novels read to him, but he was able to go up to London one day, and even planned to take a journey to Wiesbaden in order to consult an eminent oculist.