MEMOIIIS OF THE NATIONAL ACADEMY OF SCIENCES. 267 



spotted with still paler patches. Tlie protlioiacic segment is still large and broad, but the lateral 

 projections are much shorter, and now the tubercles of the itreccding stages are represented by 

 sunken pits, from the bottom of which arise small hairs.' The hairs on the body are minute, only 

 being visible with a lens. The suranal plate is smooth, the papilliform tubercles much thicker 

 and shorter in proportion than before, and the bristles arising from them slenderer and niore 

 flexible. The spinules on the stemapods are much slenderer and smaller than before, but it is to 

 be noticed that by this time they are larger on the underside, i. e., that side now almost constantly 

 held up and thus more exposed to external stimuli, Ilian those on the upperside of the filaments. 



The colors of the body are nearly the same as in vStage II, but the brown is tinged with lilac 

 and reddish, with greenish patches (ui the upper side of the second to lifth abdominal segments. 



8t((ge IV. — September l(i. Length of body, 20 mm. ; of stemapods 15-1(3 mm. The larva is still 

 much paler in hue than before, with more decided lilac blotches on the back. The thoracic dorsal 

 hump is now very marked, while the lateral projections of the prothoracic segment have nearly 

 disajipeared. The front edge of this segment is vermilion red. 



Full-fed Iiirra. — Length, 45 mm., exclusive of the tails, which are about 15-20 mm. The head 

 is pale reddish or mahogany-brown; about one third as .wide as the body at the third thoracic 

 segment. The prothoracic segment is veiy broad above, square in front on the sides, and not 

 ending in a distinct tubercle, but simply a low projection. The body is pale greenish yellow (the 

 colors somewhat failed in my three sijecimens, as they are about to transform). A dorsal median 

 reddish browu band beginning at each angle of the prothoracic segment and narrowing on the 

 second and third thoracic segments; it begins to widen on the first abdominal segment, becoming 

 widest on the fourth, and extending down on each side to near the base of the abdominal legs, and 

 contractmg aud becoming narrowest on the eufl of the seventh abdominal segment, and widening 

 a little on the ninth. The anal plate is triangular, rounded at the end; the " tails" are brown, 

 with three paler rings ou the outer half. The thoracis legs are deep red; the abdominal ones 

 pale, with browu xilanta\ 



It differs from ('. multiscripfa in that the dorsal reddish patch in the middle of the body does 

 not descend so far down on the side; otherwise it is like it in general shape and appearance. (Dr. 

 Dyar tells me that the dorsal patch is very differently colored in C. multinvripia; it is a white 

 jjatch, whereas borcalis has a brown one.) 



When at rest the head is retracted and sunken between the lateral fleshy conical projections 

 of the prothoracic segment, which are temporarily improvised or ]>ushed out by the larva when 

 at rest. The thoracic legs are held close to the body and directed forward, the tail extended out 

 behind, with the tips slightly curved up, the flagella being retracted. But when irritated or 

 teased, and probably when visited or stung, by an ichneumon, the tails are jerked up and the 

 flagella iirotruded, the head, with the thoracic segment, also being jerked up. The colors at this 

 time are precisely those of a cherry leaf partly turned yellow and partly brown. 



The caterpillar described below occurred in August and September at Brunswick. Me., on the 

 aspen. It apparently differs from those of G. occidcntalis aud cinerva in the longer spines on 

 the prothoracic segment in the young larva, aud in the smooth slight rounded i)rojections which 

 replace them in the full-grown caterpillar. 



Larva hrfore last molt. — Length to base of caudal appendages, 11mm. Head large, full, 

 rounded, dark lilac- brown, speckled with yellow, slightly wider than the body except the front part 

 of the prothoracic segment; the latter very broad, over twice as broad as long, the front edge 

 laterally produced, aud at each angle bearing a large, long, spiny tubercle three-fourths as long as 

 the segment itself, the tubercle bearing about twelve setiferous spines; across the posterior edge 

 of the segment is a row of four setiferous spines. Ou the back of the other segments are four 

 short tubercles arranged in a short trapezium, and on each side of the segments are two smaller 

 sharp tubercles. The dorsal tubercles ou the mesothoracic segment are larger than those behind; 

 those ou the metathoracic segment smaller than those ou any other segment. The body tapers 

 gradually to the end; the supraanal iilate longer than broad, rounded, bearing two long, large, 



' Dr. Dyar writes me that he has uever oljserved any species of Cerura to lose the tubercles on the cervical horns 



till the last staire. 



