292 THE VARIATION OF ANIMALS IN NATURE 



important part of those used in distinguishing species. In 

 many groups of insects, for instance, the dichotomic identifica- 

 tion keys have to be constructed separately for each sex, because 

 of the great use made of secondary sexual characters. 



Sex-limited specific characters may be roughly divided into 

 four 1 groups, viz. : (a) colours or structures apparently of an 

 ornamental nature or probably used in fighting for mates ; 

 (b) apparatus for holding the sexes together during mating 

 (apart from the genitalia) ; (c) small differences in colour or 

 structure of no apparent significance ; (d) differences in the 

 male and female genitalia. All these categories intergrade, 

 but it is easy to find examples which appear to belong definitely 

 to one or another. 



(a) Typical examples are the bright colours and ornamental 

 excrescences of many male birds and butterflies, sound-pro- 

 ducing organs in many insects, horns and antlers in various 

 mammals, and enlarged chelae in some Crustacea. Sexual 

 selection, in its original meaning, was a process by which 

 certain individuals of a species were favoured at the ex- 

 pense of the remainder ; the selection was supposed to 

 be purely intraspecific and not beneficial to the species as a 

 whole, except in so far as it might lead to a reduction of the 

 period elapsing between sexual maturity and successful mating. 

 In recent years the tendency has been to lay stress on the latter 

 function and less on the supposed advantage to individuals 

 (see Sturtevant, 1915 ; Huxley, 1923 ; Richards, 1927a). As 

 Fisher (1930, p. 138) has pointed out, even with a relatively 

 low death-rate per week, a distinct advantage would accrue 

 to individuals mating earliest. Some of the ornaments and 

 weapons found in the animal kingdom are probably of use to 

 their possessors and may have been largely evolved under the 

 influence of some form of sexual selection, though we can 

 hardly claim that there has been as yet sufficient experiment 

 to put the matter on a very sound basis. The problem of the 

 great specific diversity exhibited in ornaments is not nearly so 

 difficult as in the case of the diversity of cryptic patterns 

 (p. 280). Our knowledge of the emotional life of animals 

 is extraordinarily small; but it appears legitimate to assume 

 that any colour or structure which ' caught the eye ' of the 



1 In a number of species the female is modified in connection with her 

 maternal duties, giving a fifth type of secondary sexual difference. 



