ISOLATION 185 



Jess than five months difference between first and 

 last. 



As these birds all live on the same kind of food, 

 and all have the same simple habits when they are 

 at sea, we cannot suppose that their distinctive speci- 

 fic characters are due to natural selection : for that 

 which would favour one would favour all. Neither 

 can we sruppose that they are due to the action of 

 external conditions, because what would affect one 

 would affect all. Neither can we suppose them to 

 be recognition-marks, for when the breeding time is 

 drawing near each bird goes separately to its old 

 nest before courtship begins. It cannot, therefore, 

 be that the species of southern Albatrosses were 

 formed by competition on the ocean, and subse- 

 quently chose separate breeding grounds. We must 

 believe that isolation preceded the development of 

 the specific characters. Now it is not difficult to 

 imagine that those birds in wEich the breeding im- 

 pulse came on first would retire to their breeding 

 ground and there mate. Those in which the impulse 

 was delayed might find the old breeding ground fully 

 occupied and would have to choose another. Thus 

 owing to physiological difference, physical isolation 

 would be brought about. 



It is evident that in an equable climate, where the 

 exact time of breeding is not important, many varia- 

 tions could thus be preserved ; while in more rigor- 

 ous climates, where the breeding season must be 

 short, this kind of physiological isolation could not 

 take place. And we may perhaps in this way 

 account for the great number of species in tropical 



