is94 ON PHYSIOLOGICAL SELECTION 369 



tion. Where the isolation is due to physical barriers 

 (as at Malta) there is no need for any sexual differen- 

 tiation to originate a species. But on common areas, 

 sexual differentiation is the only means of securing 

 the isolation. Therefore (I say) we can see why 

 Jordan's French varieties all prove sterile with their 

 parent forms, and I should expect your Malta varie- 

 ties to prove fertile with theirs elsewhere. 



G. J. E. 



. Costebelle : April 15, 1894. 



Dear Mr. Henslow, — Yes, please write when you 

 get back, suggesting any time you may find con- 

 venient for spending a day or two with us at 

 94 St. Aldate's, Oxford (immediately opposite Christ 

 Church). I cannot talk long at a time, but I think 

 the meeting will be of use to both. 



Of course ' Isolation produces segregation of type,' 

 is only a short-hand expression, meaning — indis- 

 criminate variation being supposed — isolation supplies 

 a necessary condition to segregation of type by up- 

 setting the previous stability that was due to free 

 inter-crossing. 



I quite agree that Darwin very greatly over- 

 estimated the benefit of inter-crossing, as I am 

 showing in my forthcoming book on ' Physiological 

 Selection.' But this is quite a different thing from 

 his having made too much of inter-crossing as a con- 

 dition to stability of type ; I do not think that this 

 can be made too much of. Indeed, how is it con- 

 ceivable that there ever can be divergence of type 

 without isolation of some kind having first occurred 



B B 



