156 PROCEEDINGS OF THE AMERICAN ACADEMY 



middle line, and be buttressed by a multitude of minute, crowded, lus- 

 treless scales ; or there may be at one point a sort of whirlpool of large 

 party-colored scales, imbricated, in the most regular fashion, like the 

 normal scales, and, beyond them again, a multitude of the minute, 

 crowded, lustreless fcales. These peculiarities, however, must be 

 studied with a glass ; the naked eye may indeed discern that the patch 

 differs in different insect?, but the general effect in all alike is a vari- 

 ously formed velvety patch or olilitpie streak of black. 



It may be remarked, in passing, that wherever antigeny, colorational 

 or structural, manifests itself in the wings of butterflies, the differences 

 between the sexes almost invariably occur upon the upper surface, and 

 generally upon the front wing only ; it occasionally happens that there 

 is a slight difference in the general tone of color on the under surface of 

 both sexes, corresponding to what appears above, as in Semrwpsyche 

 Diana ; but it rarely affects the markings of the wings. The differ- 

 ences upon the upper surface, however, and especially upon the fore 

 wings, are, as we have seen, often conspicuous and very curious. One 

 can scarcely doubt tliat tliis is in direct relation with the general 

 absence of all ornamentation from the lower surface of both wings, and 

 usually also from the upper surface of the hind wings, of moths. 



Sexual dimorphism in the legs shows itself in the proportional length 

 of the different pairs in the two sexes, in the special development 

 of certain joints, in the appendages, and in the clothing. It ap|)ears 

 remarkably in the appendages of the two higher families of butter- 

 flies, Nijmphales and Rnrules, and especially in the latter family, where 

 the terminal appendages of the fore legs are nearly or quite lost in the 

 males, and are as cons|)icuous as on the other legs in the female. I have 

 not discovered that the differences in the length of tlie leg-joints follow 

 any general law, although tliere are few of our butterflies whose sexes do 

 not vary in this particular ; this form of antigeny is also most conspic- 

 uous in the Rurales. The males of certain VilUcantes ( Chrysophaniis, 

 Epidemia, Heodes, Feniseca) also present another curious feature in a 

 tumid swelling of the basal joint of the middle and hind tarsi. Finally, 

 the fore legs of the males of Nijmphales are frequently furnished with a 

 spreading brush of hairs ; or, in other butterflies, the thighs and shanks 

 of the middle and hind legs are supplied with curious pencils or fringes 

 of stiff hair, which appear to have the same sigaificauce as similar 

 adornments in higher animals. 



Darwin supposes that these various male appurtenances, which 

 occur throughout the aniratd kingdom, have all arisen by natural selec- 

 tion, — that one of rival males being selected as a mate whose outward 



