ISOLATION 151 



is little developed, and it is probable that the dances are also 

 stimulatory in their effect. The female behaviour has two 

 phases, an amatory and an aggressive one, and when the 

 former holds sway she is much less likely to attack the male. 

 Thus courtship dances, besides giving the female a chance to 

 recognise her mate, also put the female into a state in which 

 attack is unlikely. After copulation, when the aggressive 

 phase reasserts itself, she may devour the male, though she can 

 scarcely be said not to recognise him. 



Against these examples we may set others in which the 

 epigamic characters are not yet known to play a part in 

 isolating species. Among birds, as Huxley (1923) and others 

 have pointed out, the exhibition of coloured parts and the 

 performance of special antics, flights and songs take place 

 usually after the birds are already mated up for the season. 

 The displays are supposed to have a purely stimulatory effect. 

 It is possible that male epigamic characters may play a minor 

 part as recognition marks, though on this point we have no 

 evidence. The stimulatory function seems likely to be impor- 

 tant in many groups. Species which hybridise naturally also 

 provide important evidence, since they show that no single 

 element in the isolationary complex is necessarily and always 

 competent to produce its normal effect. 



II (iii). Differences in the mechanical relations of the copulatory 

 apparatus. 



We may, in the first place, mention a rather exceptional 

 example amongst the fish. In the genus Anableps (Nor- 

 man, 1 93 1, p. 296) the male genital orifice is prolonged into 

 a tube. The genital aperture of the female is covered by a 

 special scale, free on one side only. The opening may be on 

 either the right or the left and the males may have the intro- 

 mittent organ turned in either direction. Copulation takes 

 place sideways and a right-sided male always pairs with a 

 left-sided female and vice versa. 



The whole problem becomes much more complex when we 

 consider the more usual type of specific differences in the 

 genitalia, which are so often found, especially in the males, 

 in a number of groups (see p. 296). It is important to dis- 

 cover how far these elaborate structures act as a mechanical 

 means of isolating allied species. When we find that the male 



