MEMOIRS OF THE XATIOlSfAL ACADEMY OF SCIENCES. 139 



Expanse of wiugs. i , 30 nun.; length of body, i , 14 nun. 



Tliis not nncoraniou spt'cies differs from all the others in the distinct S-shaped ]iortion of the 

 fourth or outer line, situated between the costal edge and the cubital vein, the outer curves of 

 the S being tilled in with reddish brown. Also the basal line is not dislocated, both tiiis an<l the 

 second line crossing the entire wing aud being ])arallel. 



Var. sjiecifica Dyar. Under this name Dr. Dyar briefly mentions a form of this species 

 cajitured at Manitou, Colo., May 2, ''whit'h differs from the type l)y its jiuuh ]ialer color." 

 Colorado (French). 



J'JiJII- — '"Of a heuiispheric form and dark brown, with a wide glaucous gray ring on the outer 

 margin." (Fitch, p. 275.) 



Larra. — The following description is drawn up from Mr. Bridghaui's colored drawings of the 

 three last stages and an alcDholic specimen of the mature larva. It occurred on the i)oplar, July 9 

 to 13, those in the three last stages occurring at these dates. Other specimens were reared by 

 Mr. Bridgham and the moths obtained from them. (For stage II see Appendix A.) 



Larra in Stage III. — Length, 20 mm. Head as wide as the body, reddish. The body reddish 

 on the sides, and green along the back, interrupted by a reddish patch on the first and one on the 

 eighth abdominal segments, each of which incloses a median tubercle. The greenback incloses 

 three parallel daik green, indistinct, interrui)ted lines. There are two greenish tubercles on the 

 side of the body, one above and the other below the spiracle. 



Stuf/e IV. — Length, 30 mm. The hair is still reddish, but the body has now lost its green 

 shade on the back, which is pale, with three darker parallel dorsal lines. The two median 

 tubercles are now as well developed as in the last stage. The side of the body is pale reddish, 

 with dark lateral tubercles on the thoracic and first abdominal segments, those on the succeeding 

 segments being yellowish, as on the abdominal legs, including the anal pair and suianal i)hite. 

 The thoracic legs are pale. 



Full-fed larra. — Length, 30 mm. Head hardly as wide as the body, black, with a y-sha])cd, 

 pale brown line in front, formed of a median line extending down from the vertex to the apex of the 

 clypeus, aud then dividing so as to extend down ou each .side, ending before reaching the anteniue. 

 The head is flattened and densely covered with grayish hairs. The three thoracic segments bear 

 each six lateral, rather large, yellowish warts, the lowest one the largest, ea -h bearing about six or 

 seven hairs of unequal length. There is a high median finger-shaped, tleshy nutaut black tubercle 

 on the first abdominal segment, bearing numerous short, unequal hairs; it is rather high, finger- 

 shaped, and bent over backward. Ou the eighth segment is a shorter, smaller, paler one. It is 

 evidently of double origin, its longest diameter being transverse to the bod}', and somewhnt 

 wedge-shaped; the end is somewhat swollen on each side, with a slight valley between the 

 swellings, showing that it was originally formed of two separate tubercles, and this is also 

 suggested by the fact that each swelling bears eight or ten short uuequal hairs. The thoracic legs 

 are black; the abdominal legs are dark, especially toward the planta. 



Colors (described fiom Bridghnm's figure): Body straw-yellow, with three dorsal, more or less 

 interrupted grayish or pearly j)ale brown lines and a broad lateral stripe, below which the 

 tubercles are yellow ocherous. The suranal jilate is flattened, rounded in outline, and hairy, with 

 the surface rather rough and hairy. In my single alcoholic sjiecimeu there is no sign of a 

 prothoracic shield or plate. 



Although the imago of /. apicalis is very near that of I. inclusa in markings, the larva is verj- 

 different, there being no mediau dorsal tubercle on the first abdominal segment. In the lack of 

 these tubercles /. strUjnua resembles /. apicalis. On the other hand, the larva of I. alboxiyma, in 

 respect to the presence of the two ilorsal abdominal tubercles, approaches that of J. inclu.sa. 

 These two species, then, as larviv, belong to the same genus, while the two other species (apicalis and 

 stri(jo.'<a), as respects the larva-, differ generically from inclusa and albosigma, though the nuiths 

 are congeneric. It is evident that the larvaM)frt^)/«(//.s' and strir/osa are more generalized, since they 

 lack the rather highly specialized dorsal tubercles so prominent in the two other species of the 

 genus. If we regarded the moths alone we might erroneously consider that apicalis and inclusa were 

 both coeval, whereas apicalis must be a much older, more generalized form: hence, speculations on 

 the phylogcny of Lepidoptcra based ou the imagines alone may often be uucertaiu. 



