391 SEXUAL SELECTION. [Part IL 



details. On the Avhole, I cannot perceive that an ine- 

 quality in the numbers of the two sexes would influence 

 in any marked manner the effects of ordinary selection on 

 the character of the offspring. 



Female Lepidoptera require, as Mr. "Wallace insists, 

 some days to deposit their fertilized ova and to search 

 for a i^roper place ; during this period (while the life of 

 the male was of no importance) the brighter-colored fe- 

 males would be exposed to danger and would be liable to 

 be destroyed. The duller-colored females, on the other 

 hand, would survive, and thus would influence, it might be 

 thought, in a marked manner the character of the species 

 — either of both sexes or of one sex, accoi'ding to which 

 form of inheritance prevailed. But it must not be for- 

 gotten that the males emerge from the cocoon-state some 

 days before the females, and during this period, while 

 the unborn females were safe, the brighter-colored males 

 would be exposed to danger; so that ultimately both 

 sexes would probably be exposed during a nearly equal 

 length of time to danger, and the elimination of conspicu- 

 ous colors would not be much more effective in the one 

 than tlie other sex. 



It is a more important consideration tliat female Le- 

 pidoptera, as Mr. AVallace remarks, and as is knov»'n to 

 every collector, are generally slower flyers than the males. 

 Consequently the latter, if exposed to greater danger 

 from being conspicuously colored, might be able to escape 

 from their enemies, while the similarly-colored females 

 v.'ould be destroyed ; and thus the females would have the 

 most influence in modifying the color of their progeny. 



There is one other consideration : bright colors, as far 

 as sexual selection is concerned, are commonly of no ser- 

 vice to the females ; so that if the latter varied in bright- 

 ness, and the variations were sexually limited in their 

 transmission, it Avould depend on mere chance whether 



