NATURAL SELECTION 293 



female might be effective, so that a great variety in adorn- 

 ment might be adapted to the same end. Any bright-coloured 

 patch in the male might serve to raise sexual excitement in 

 the female and so hasten mating, and it would not be sur- 

 prising if in one species a blue patch and in another a red one 

 first gave the opportunity to selection. 



Much the same argument can be applied to the develop- 

 ment of scent-producing organs, which occur so widely in 

 insects (Richards, 1927a), are not uncommon in mammals 

 (Pocock, 1916), and also play some part in the courtship of 

 spiders (Bristowe and Locket, 1926). They appear usually 

 to be the main factor in bringing the sexes of insects together, 

 and in others also seem to be employed to stimulate sexual 

 excitement in the female. In the latter role they are exactly 

 comparable to an ornament, but when used for attraction and 

 recognition of members of the other sex the evolution of 

 specific diversity is more difficult to explain, since changes in 

 production would have to be very closely correlated with changes 

 in perception. Exactly the same difficulty has to be met with 

 in trying to explain the evolution of male genitalia (p. 299) . 



With weapons the case appears rather different, since we 

 might expect a much closer degree of correspondence between 

 the structure evolved and the needs of the animal in fighting. 

 It is very doubtful if such correspondence could at present be 

 established, but our information is very scanty on the observa- 

 tional side. Although horned mammals certainly fight to a 

 considerable extent in the breeding season, the remarkable 

 horn-like structures found in many male insects do not appear 

 to have this function, and much fuller records of the behaviour 

 in nature of animals bearing such excrescences are required 

 before we can confidently assert how far presumptive ' weapons ' 

 are really useful either to the species or to the males. The 

 occurrence of secondary sexual characters is very capricious 

 — e.g. in some Pulmonate Mollusca ' darts ' are present ; in 

 many they are absent (cf. also the contrast between rodents 

 and ruminants among mammals). 



(b) Special organs for grasping the female during copula- 

 tion are characteristic of many invertebrates, especially 

 arthropods. For our present purpose we are excluding the 

 most typical grasping organs of all, those developed in con- 

 nection with the genitalia. Almost any part of the body may 



