36 PECKHAM. [Vol. 1. 



difference between the sexes, the males having the legs longer, 

 more robust, and more ornamented than the females. This is 

 also true in the North American species Icius ■palmanum. Saitis 

 barbipes has the third leg of the male fringed with hairs on 

 either side, which gives it quite a plume-like appearance, while 

 that of the female is entirely plain. In an undescribed species 

 of Habrocestuvi from Arizona the first leg of the female is plain, 

 while that of the male has on the tibia a fringe of long, silky, 

 yellow hairs, mingled with which are other hairs, which are 

 enlarged and flattened at the end. These spatulate hairs also 

 appear on the first legs of Habrocestum hirsutum ; in this species 

 we have only the male. 



Walckenaer remarks that his earlier division oi Attidas into 

 Sauteuses, or short-legged, and Voltigeuses, or long-legged, is 

 vicious, since in many species of Sauteuses the males have very 

 ]oi:ig legs, and are, therefore, if we have not seen the female, 

 placed in the Voltigeuses, while the females must be put into the 

 Sauteuses} We may go further than this, and say that all these 

 modifications of the legs are sexual and of little or no import- 

 ance in taxonomy. 



The instances which we have given of secondary sexual 

 differences might have been indefinitely multiplied, but we 

 have only thought it necessary to give examples of each kind 

 or class of modification ; these will serve, we trust, to establish 

 the fact that these differences are not less numerous among 

 spiders than among birds and insects. 



MATING HABITS. 



For a number of years prior to 1888, we had been much 

 impressed by the many important differences between the 

 sexes of our jumping-spiders, and since we thought it most 

 probable that they had come about through sexual selection, 

 we had often tried to watch them during their courtship, but 

 up to that time with very little success. We had occasional 

 glimpses of their habits, but they were so incomplete as to con- 



1 Loc. clt., p. 482. 



