266 MEMOIliS OF THE NATIONAL ACADEMY OF SCIENCES. 



occiured at PiDvidciioe on tlie wild cbeny in Sei)teniber. Hellins states that the eggs of 

 C. rinttla are 1.0 to almost 2 inin. in diameter, and that tjie larva at its tirst molt is not more 

 than 7 mm. long. Possibly the first stage was not observed by ^Iv. Bridgham, and the following 

 description should ajjply to the second. Compare also Dyar's detailed description : 



Egif. — "Less than hemispherical — obtusely conoidal. the base flat; minutely shagreened, color 

 black; a little lustrous, but not shiny. Laid singly on either surface of the leaf." (Dj-ar.) 



Cocoon. — Oval elliptical flattened, but central area •well rounded; the edges brpad and thin; 

 spun of light drab silk; attached to side of breeding box or to bark of tree, and thus easily 

 mistaken for an excrescence on the tree; being a case of protective mimicry. 



Larva: Stage II. — Length in all, 1~> mm. September 4. Head only as wide as the body behind 

 the middle. The fllamental anal legs, or stemapods,' as we may designate them, are now more 

 than .slightly half as long as the body. The hornlike tubercles on the prothoracic segment are 

 slightly longer than in the second stage. The head and body are dark reddish brown above, the 

 fllamental anal legs with two broad, pale, greeni.sh rings. All the other abdominal legs are 

 green ; the green patch extends from the underside of the tir.st abdominal segment back over the 

 third to eighth pair of spiracles, and underneath to the end of the body. 



Stac/e II. — Length of body, 14 mm.; of stemapods, 7-S mm., and of tiagella, •'! mm. September 

 11. The head is rough and warty, the small warts bearing tine hairs. On tin? front toward the 

 vertex are four papilliforni, piliferon.-; warts of the same size and shape as those on the prothoracic 

 l)rojections, and coucolorous with the dark brown head. These spines are represented in the 

 other species (C. occiclentaUs) from the willow only by very minute warts, bearing long, tapering 

 bristles. The prothoracic segment is very wide and large, the well-defined cervical shield very 

 broad, and ending on each side in a large, stout tuberculated horn, bearing about twelve 

 piliferous, papilliform tubercles, there being a rude whorl of spines in the middle of the horn, the 

 others growing out at the end. There are four coarse piliferous warts on the hinder edge of 

 the cervical shield. 



Along the body are scattered coar.se ijiliferous warts, the dorsal four being arranged in a 

 trapezoid. The stemapods are coarsely spined (more so than in C. occidentaiis). 



A peculiarity of the genus is the pair of very long papilliform infraanal tubercles, situated 

 under the suranal plate, and ending in two long, stiff, sharp bristles.- The suranal plate is long 

 and narri>w, well rounded, and the surface is provided with high papilliform, piliferous warts. 



In this species the head and the iirothoracic horns above and beneath are reddish brown, the 

 latter in C. occidentalis being yellowish beneath, the two species by this mark being easily 

 separated. 



The body is now more green on the sides, the green hue encroaching on the back and nearly 

 meeting on the third thoracic segment. Only the fourth abdominal segment is wholly dark seen 

 from above, and the green approximates high up on the sides of the sixth and seventh segments. 



Stage III. — September 17. Length of body, 19 mm. and of stemapods, 12 mm. The body is 

 now much thicker than before. The head is now smooth, with no traces of piliferous warts or of 

 hairs representing them. The head is now larger in proportion to the body and paler red, 



' The term " tails" or caudal filaments is too vague for these highly modified anal legs ; hence we propose the 

 term stemapoda or stemapods for those of Cerura and Heterocampa. The dprivation is Gr. nrrjua, filament ; -nii;, 

 TToiSdc, leg or foot. Mr. J. Hellins, referring to these organs in Ijuckler's Larva) of the British Butterllies and 

 Moths (Roy Soc, ii, 138), remarks "but now through Dr. T. A. Cliapman's good teaehiug I regard them as dorsal 

 appendages, somewhat after the fashi(m of the anal spines of the larv:e of the Satyrid:e." This, I am saiisfied, is an 

 ■error. After repeated comparisons of the filamental anal legs of Cerura with those of Heterocampa marthesia. and 

 comparing these with the greatly elongated anal legs of young H. tiiiiculoi- as figured by Pojieuoe, and taking into 

 account the structure and homologies of the suranal and preanal flaps, one can scarcely doubt that those of Cerura 

 are modified anal legs. 



It should be also remarked that this was the view of Latreille (Gen. Crust., et Insect., 1809. p. 21!l), who defines 

 the genus thus : Eruca }>e(Ubiis analihim in eatidain fiin-ulam transformath. 



- The use of these I find explained by Mr. Hellins in his description of the larva of C. liijlda in Buckler's Larv.-e 

 of British Butterflies and Moths, ii, p. 142. as follows : " At the tip of the anal fiap are two sharp points, aud another 

 pair underneath, which are used to throw the pellets of frass to a distance." Similar duugforks are very generally 

 present in geometrid larv.^e, the infraanal papilliform tubercles being well devcloi>ed. though we have not seen 

 them in use. (See also Dyar.) 



