176 PROBLEMS OF EVOLUTION 



wing power. Had there been constant fresh arrivals of birds of 

 his species, intercrossing would have maintained his power of 

 flight and enabled him to escape from the conscienceless sailors 

 who exterminated him. Among the higher animals isolation is, 

 probably, very often due to preference. Some difference in 

 colour or marking arises by which sympathy is bred, and then 

 those in whom this marking is found, herd together and form a 

 race that may in time develop into a species, kept apart by what 

 Clannish we may call clannish isolation. Thus recognition marks are of 



isolation t j ie utmost importance, and those who have made light of them 

 cannot have realised the importance of isolation. Call notes, the 

 cries peculiar to a species, answer the same purpose as recogni- 

 tion marks, and are often alternative to them. Of this kind of 

 isolation I shall give some further account, and also of another 

 which operates very generally in the vegetable world and pro- 

 inter- bably among the lower animals. This latter form is sterility 



sterility b e t ween species, due to differences in the reproductive cells. 

 We must regard it as arising simultaneously with beneficial 

 variations and protecting them from the effects of intercrossing. 

 Among the higher classes it is probable that clannish isolation 

 often replaces it, but without numbers of experiments it is im- 

 possible to speak positively. And experiments are very difficult ; 

 if animals of two different species are crossed, there may be per- 

 fectly healthy offspring, and yet the species may be intersterile. 

 For the progeny may be mules and incapable of continuing their 

 race. It is, however, beyond a doubt that intersterility among 

 wild species is by no means invariable. Our domestic cattle are 

 descended from various wild stocks. At the Zoological Gardens 

 some success has been attained in crossing the bison with the 

 yak and the gayal, and it is much to be regretted that such ex- 

 periments have not been more frequently made and carried out 

 more systematically. 



There must be isolation in some form. This is true though 

 prepotency is an undoubted and very important fact though one 

 parent is often able to transmit his or her characteristics while 

 those of the other are but little represented in the offspring. 

 Continual intercrossing with ordinary members of the species 



