MEMOIRS OF THE NATIONAL ACADEMY OF SCIENCES. 63 



The geuus^ as shown l>y tlu» larval rliaracters. is more giMicrali/.cil than Auisota. since the 

 spines on the prothoracic and abdoininal soo-nients arc lai'gci-, and the sphingiform imao'oes or 

 moths are rather more primiti\e and ancestral. 



The species of the oeniis ditl'er very considerably, and may he divitlcd into several groups or 

 what may be perhaps regarded as incipient genera. The most sphingiform group of species and 

 that nearest Syssphinx is the Hrazilian ^1. xi(h((iHji(lutu. A.f<(Ua,v l}oisd.,aud A. hiieetu^ while the 

 rounder-winged species ^-1. afho//iit'a(a, A. Iiell>(/l>rodtli, and A. bicolm' avc later, more degenei'ate, 

 or modified forms. A. hicolor. with pectinated 9 antenna', shows a remarkable divergence from 

 the normal 9 form, the front of the head being wider, while the tongue is not visible. We 

 consider those species with a well-developed tongue as the more primitive forms. 



Druce states: " Adelocephala I think will have to be divided, the typical species having 

 pectinated antennre in both sexes, those of the A. alholineatu group being only pectinated in the 

 males. I have not sufBcient material, however, to be certain how far this character holds good." 

 It appears, however, that only ^1. hicolor has pectinated antenna^ in the female. (See fig. 8.) 



The South American and our ^4. hisecta are apparently the more primitive forms, both as 

 regards the sphingoid shape of the bod\- and wings and the more fully developed maxilla;. As 

 the type found its wa\- into North America it seems to have undergone a reduction of the 

 maxilla^ until they practically became aborted, while the outlines of the w'ings became less 

 distinctive and less specialized, as seen especially in^-1. hicolor, which in this respect approaches 

 Anisota. It is interesting to observe that all the South American species (except A. anthonilis) 

 have sharp fore wings and subtriangular hind wings, while in several of the Central American 

 and North American species the wings tend to become broad and rounded at the apex. 



The most aberrant species is A. aJhoIineata. Its maxilhe are comparatively well developed. 

 I have not been able to examine the male antenna% having no examples of that sex. It differs in 

 venation from all the other species in the much wider discal cell of both wings, and in vein III; 

 of the fore wings, which arising between the anterior discal vein and III.,, springs off very near 

 the origin of the latter. The two discal veins taken together form a straight line, not being 

 curved as in most of the other species; the subcosto-radial (II) vein is much curved, so that the cell 

 between this and III is much wider than in any other species of the genus. If the genus is to 

 be subdivided A. aJhoUneata should be generically sepai'ated from all the other species, but it is 

 better to wait until we know the larva and its history. 



E(nj. — Flattened, oval-cylindrical, disk-like, each end alike. Shell smooth under a low power, 

 but when highlj- magnitied seen to be ornamented with faint polygonal areas, with a swollen 

 nucleus in the center. 



Liirra. — Head siibconical. narrowing above. Bodj- cylindrical, inclined to Ije slightly com- 

 pressed; in the more primitive species all the segments, except the prothoracic, bearing long, 

 high, saber-shaped dorsal spines as long as the body is thick; tubercles it nearly atrophied; in 

 the more specialized species (^1. hicolor. etc.) the four thoracic and caudal horns much longer than 

 the abdominal ones; of the latter those on each, or on the second, fourth, and sixth segments, 

 are stout, conical, smooth spines, conspicuously tinged with silver and rose red; no median spine 

 on the ninth abdominal segnnent; suranal plate coarsely tuberculated. but with no specialized 

 spines. 



Young larva, stage I. — Armed with four thoracic horns, three-fourths as long as the body, 

 and ending' in bulbs bearing two dark rods; caudal horn spinulated. large, as long as the body 

 is thick, divided deeply into two setiferous lobes; tubercles // present on abdominal segments 

 1 to 9; suranal plate with three lateral and two terminal setiferous tubercles; the boch' marked 

 with conspicuous longitudinal sti-ipes. 



Pupa. — Body moderateh' stout, head rounded; maxilhe short, taken together forming a 

 nearly equilateral triangle; a group of three stout spines at base of and between the antenna?, 

 as in Anisota; surface of the integument more rugose than in Anisota, and the spines a little 

 stouter; the cremaster is long, slender, narrow, rather deeply divided at the end, but not differ- 

 ing from that of Anisota. 



Vol. 9—0.5 b 



