MOULTING HABITS OP SPIDERS. 107 



the last moult is relatively so much greater, that one reasonably comes to 

 locate the transformation within the last moulting period. In some cases, 

 however, the coloration of the sexes is more strongly differenced during the 

 stages previous to final moult. Thus the color of male Trochosas during 

 the two or three last moults is notably clearer than that of females. With 

 Attus, in the corresponding periods, one may observe the sexual peculiari- 

 ties in color distinctly appearing, and that they deviate more and more 

 intensely with each moult, until the last fixes the distinction. Like the 

 male, the female awaits the final moult for perfect development, at which 

 time the genital cleft is freely opened and the hood and scapus assume 

 those various forms which serve as valuable specific characters. 



" Leidy," one of my captive tarantulas, shed its skin several times. 

 The first moult occurred some time in August (1882). I had been absent 



on my usual summer vacation, and returning August 31st saw 

 Change m^]-jg iiuinifii lying on the soil about the middle of its aranearv, 

 las ^^^^"^ ^^^ ^®^* gathered together, looking dull, gray, and faded out, 



apparently dead. I shook the glube. No responsive motion fol- 

 lowed, and I left without more careful observation, concluding that the 

 spider was dead. I was not able to visit it again until the fifth day of 

 September following. I threw oif the cover of the globe and put my hand 

 in to take out the dead body, which lay apparently in the same position, 

 in order to preserve it in alcohol. At my touch the animal leaped to its 

 feet, and as I hastily withdrew my hand it presented itself quite changed 

 in appearance. The body was a fresh bright color, the cephalothorax a 

 clean whitish gray, the head and fangs dark brown. The abdomen was 

 black, with brown hairs covering it. The legs were black, with yellowish 

 brown hairs and spines. I at once understood that the spider when first 

 seen was in the torpid condition which usually immediately precedes the 

 act of moulting. In the interval between my visits it had cast off its skin, 

 which I found lying in a tolerably complete condition on one side of the 

 glass. Another tarantula, a male, which I received when quite young, 

 came to me a dull reddish brown, but during successive moults at last 

 appeared a bright black brown, almost black. 



VIII. 



We may thus summarize the most important moulting phenomena, as 

 above disclosed. 1. The first two moults of spiderlings occur within the 

 cocoon or on the mother's back ; several occur before entrance 

 Biologi- ^ipon independent life, sometimes as many as four. 2. The 

 ca um- jj^djyij^^j^ig Qf a brood do not all moult at once, and those moult- 

 ing first, having the greater strength, in some species feed upon 

 the younger and weaker individuals ; in these cases survival depends upon 

 priority of moult. In other spe(^s cannibalism is absent or rare, and 



