1893 EVIDENCE FOE PHYSIOLOGICAL SELECTION 309 



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 dif- 

 ferentiated 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 

 j-endered sterile with some one compatriot, sterility 

 with other species would follow as a necessary con- 

 tingency.' 



Yours very sincerely, 



Gr. J. EOMANBS. 



P.S. — Prom your first letter it would almost seem 

 iihat 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 ' Hybridi- 

 sation ' in the ' Ency. Brit.' ! 



