566 THE DESCENT OF MAN. 



variations which recur so frequently that they would imme- 

 diately be fixed, if they served to allure the female. The 

 laws of variation must determine the nature of the initial 

 changes, and will have largely influenced the final result. 

 The gradations, which may be observed between the males 

 of allied species, indicate the nature of the steps through 

 which they have passed. They explain also in the most 

 interesting manner how certain characters have originated, 

 such as the indented ocelli on the tail-feathers of the pea- 

 cock, and the ball-and-socket ocelli on the wing-feathers 

 of the Argus pheasant. It is evident that the brilliant 

 colors, top-knots, fine plumes, etc., of many male birds 

 cannot have been acquired as a protection; indeed, they 

 sometimes lead to danger. That they are not due to the 

 direct and definite action of the conditions of life, we may 

 feel assured, because the females have been exposed to the 

 same conditions, and yet often differ from the males to an 

 extreme degree. Although it is probable that changed 

 conditions acting during a lengthened period have in some 

 cases produced a definite effect on both sexes, or some- 

 times on one sex alone, the more important result will have 

 been an increased tendency to vary or to present more 

 strongly marked individual differences; and such differ- 

 ences will have afforded an excellent groundwork for the 

 action of sexual selection. 



The laws of inheritance, irrespectively of selection, 

 appear to have determined whether the characters acquired 

 by the males for the sake of ornament, for producing vari- 

 ous sounds and for fighting together, have been transmitted 

 to the males alone or to both sexes either permanently or 

 periodically during certain seasons of the year. Why 

 various characters should have been transmitted sometimes 

 in one way and sometimes in another is not in most cases 

 known; but the period of variability seems often to have 

 been the determining cause. When the two sexes have 

 inherited all characters in common they necessarily resem- 

 ble each other; but as the successive variations may be 

 differently transmitted every possible gradation may be 

 found, even within the same genus, from the closest simi- 

 larity to the widest dissimilarity between the sexes. With 

 many closely allied species, following nearly the same 

 habits of life, the males have come to differ from each 

 other chiefly through the action of sexual selection; while 



