152 THE VARIATION OF ANIMALS IN NATURE 



genitalia (as often happens in insects) differ sharply in charac- 

 ters whose degree of variation is not enough to make them 

 overlap, in a species in which most or even all other structures 

 intergrade from species to species, it is tempting to assume that 

 we see here the actual agency for permanent isolation in these 

 forms. The essence of this theory, the well-known ' lock-and- 

 key ' theory of L. Dufour (cf. Perez, 1894), is that the females 

 should also differ in some way from each other ; differentia- 

 tion in the male alone would not be effective. Whether the 

 females do differ and whether the male armature really is 

 effective in isolation have for many years been matters of 

 controversy. The argument has chiefly lain amongst the 

 entomologists, and a decision for the insects would probably 

 also be valid in the case of many parasitic worms, Crustacea 

 and Arachnida. 



The chief supporter of the ' lock-and-key ' theory has been 

 Jordan (1896, 1905). Boulange (1924) has reviewed the 

 subject and takes the opposite view. Jordan, in his first 

 paper, dealing with the swallow-tail butterflies {Papilio), showed 

 that the differences in the male genitalia are quite manifest, 

 sometimes in geographical races. The females sometimes 

 differ markedly in their genitalia, though they were much less 

 thoroughly investigated. The actual proof that the male 

 structures coincided so accurately with those of the female that 

 copulation between different species would be difficult or 

 impossible was not very convincing, and the evidence put 

 forward was derived from a few species only. In his second 

 paper the correlation between differences in genitalia and in 

 other characters is examined. His main thesis is that local 

 and seasonal * polymorphism in colour and wing-shape is 

 quite independent of variation in male genitalia. In one geo- 

 graphical area the genitalia vary only slightly and at random, 

 but as soon as a distinct geographical race becomes recognisable 

 the variation in the genitalia tends to be correlated with the 

 size and colour characters defining the race. It may be 

 admitted that the male genitalia are easily modified in the 

 evolution of species, but it is much more uncertain what part 

 they actually play in that process. 



In the Lepidoptera as a whole interspecific crosses are not 



1 Mercier (1929) claims to have demonstrated seasonal variation of the male 

 genital organs in the fly, Cynomyia. Jordan also records one case in Papilio. 



