102 MEMOIRS NATIONAL ACADEMY OF SCIENCES. vol. xn, 



little shorter than those on the first thoracic segment. The two dorsal spines on the third thoracic segment are, in size 

 and spinulation, now exactly like those on the abdominal segments 1-9. The median double one on the eighth abdom- 

 inal segment is thicker than the single ones in front, also higher, and ends in two spines; the lateral spines are much 

 more numerous than those in front. The spines of the subdorsal series are alike on both the thoracic and abdominal 

 segments. 



Last stage: The shape of the dorsal spines of the larva in its final stage is represented by figure 6. I, a prothoracic 

 dorsal spine, ending in two equal terminal piliferous spinules, with seven or eight just below it, while at and near the 

 base are the long, pale spines, each ending in a sharp black point; II, one of the second thoracic dorsal spines, the base 

 short, bulbous, with very numerous radiating spines, and a single terminal, central piliferous spine, with a smaller one 

 near it; III, a bush-like dorsal spine of the third thoracic segment, no piliferous spinules present. The abdominal 

 dorsal spines are all on the same type. 1 The median spine on the eighth abdominal segment is about twice the size of 

 the other dorsal single ones in front, though no higher, and it spreads more, having about twice as many spines on the 

 sides. On the ninth segment are two dorsal and two subdorsal ones, and behind these four on the same segment is a 

 median one. These types are already attained in stage II, though the spinules are fewer in number. 



It is to be noticed that the characters of the full-fed larva appear in large part in stage II, and are almost fully 

 developed in stage III. 



Figure 7 represents the spiracle and lateral oversible gland of the full-fed larva; <7, the eversible gland; sp, spiracle; 

 g / , an eversible gland, enlarged. 



In the large dark (in alcohol) larva of Hyperchiria, or perhaps of a Gamelia, referred to by me in Proc. Bost. Soc. Nat. 

 Hist., XXV, 91, the dorsal spines of the three thoracic segments are represented by figure 8. I, a prothoracic; II, one 

 from the second thoracic; III, one from the third thoracic; VII, one from the seventh abdominal segment. It will be 

 seen that the spines of this species are rather more generalized than in the mature A. io, and approximate those of the 

 second stage of that species ; the dorsal spine of the third thoracic segment ending in three piliferous spinules, there being 

 no piliferous spinules at all in the homologous spines of A. io; the abdominal spines also (VII) ending in three piliferous 

 spinules, though the other spinules are much (about one-half) less numerous. 



Intermediate between those of A, io and the Mexican species is the Heniileuca artemis, from Las Cruces, N. Mex. 

 (fig. 9), in which the prothoracic dorsal spine is like the Mexican form, the second thoracic dorsal spine like the pro- 

 thoracic ones of ^4. io, and the third thoracic dorsal tuft like the second dorsal one of Hemileuca yavapai from Arizona. 

 In this last species the dorsal tufts of the body, as a whole, are intermediate between H. maia and A. io, but as regards 

 the second and third dorsal and the dorsal abdominal ones, it approaches much nearer to A. io, as will be seen by 

 an examination of the figures, the second and third thoracic spines being alike in shape. Hence the most gener- 

 alized or primitive form, as regards its larval armature, appears to be the genus Hemileuca, and H. maia is the 

 most like the young larva of Aulomeris io; then succeeds the Cordova larva, then the New Mexican larva, while 

 Hemileuca yavapai is more modified, Aulomeris io being the most so of any under consideration and this may have been 

 the last to be evolved. 



A0TOMERIS LILITH (Strecker). 



Plate LX, figs. 5, 6. 



[Hyperchiria lilith, Strecker, Lep. Rhop Het. (1878), p. 139; pi. 15, fig. 17; Georgia.] 



The o* is not yellow, but with brown fore wings; the 9 is like that of A. io; when the s 

 is yellow on the fore wings it is much deeper yellow than in A. io. 



Larva. — Stage I: The freshly-hatched larvae differ from A. io in being bright green, not 

 brown as in the stem or io form. After the first molt they become more yellowish. 



[The following appeared in Proc. Amer. Phil. Soc, XXXI, p. 171 :] 



About a dozen living specimens of these interesting caterpillars were kindly presented to 

 me by Mrs. Annie Trumbull Slosson, who had collected them at Punta Gorda, Fla., where they 

 were found in March feeding on the mangrove. They were described April 6. Mrs. Slosson 

 considered them as belonging to Strecker's var. lilith. 



Length 20 to 25 mm. The body is yellowish green all over. The lateral broad, reddish, 

 spiracular band is as in northern specimens of H. io; it is broadly and distinctly bordered below 

 with white. The head and all the legs, both thoracic and abdominal, are straw-yellow. The 

 spines in general are bright, straw-yellow, more yellow than the body; the ends of the dorsal 

 ones on the prothoracic segment are black, while the ends of the long spinules in general are 

 more or less black, some merely tipped with black. 



This is apparently a case of acceleration of development as the larva in its second (or third) 

 stage resembles in coloration the full-grown northern form offie larva of H. io, the markings, in- 

 cluding the lateral reddish and white spiracular line, being as in the full-fed normal larva of 



i The spines have been somewhat flattened, but have been drawn with the camera. 



