MEMOIRS OF THE NATIONAL ACADEMY OF SCIENCES. 25 



Colomt'oii of Si/ssj>/ii».r. — The larva of the last stage has a thick green body, with the horns 

 and spines much reduced, while the lateral stripe is ^-ellow. Hence this form is protected simply 

 bv its green color. Its young, freshly hatched larva, judging from alcoholic examples, is plain 

 green, without stripes, but with long, large horn.s, much as in .1. Vtcolor. 



Cohmdion of Eiu-le!< hnpii'kdh. — We now have an entirely different style of markings in 

 this genus. In the freshly hatched young of this polyphagous form there are no longitudinal 

 stripes on the pale sienna-hi'own body, but the abdominal segments 1 to 7 have each two narrow 

 regular parallel dark brown bands across the back, with either one or two short ones on the side 

 reaching up to or a little beyond the spiracular region. The body is verj' spinose. the four 

 thoracic and the caudal horn very long and deeply forked, and pinkish. From what ancestral 

 form this unusual style of markings has been derived is difhcult to conjecture. The hue of the 

 body is similar to that of the sheathing base of pine needles. 



This style of markings is retained in the next stage, but the spines are somewhat reduced, 

 and are now black. At the second molting the larva enters on its third stage with no markings, 

 while the horns arc pale, black at tips. After a third ecdysis the ))ody is green (sometimes 

 reddish brown); the horns are reddish and the spines are yellow. as are the suranal plate and anal 

 leo'S and thoracic legs; the head is partly banded with yellow, while the midabdoniinal legs end 

 in yellow. The spiracles are in stage I^' noticeable from the rich dark green ring in tiiem. the 

 color of the pine needles. 



These hues and markings are apparently protective, the catei'pillars. both the green and 

 l>rown forms, in the Northern States being most conunon on the white pine. 



In a larva of ^. ^^tv/.Av*, living in the high mountains of the interior of Brazil, Peters figures 

 a form somewhat like K. hiij>, rial Ik, ]>ut with two violet dorsal l)ands bordered with white. This 

 kind of marking is exceptional and apparently unique in this subfamily. It lives on a melasto- 

 maceous plant and also on the guava (Pxidhimpomifevinn); neither of these plants is spiny. 



It is questionable whether the caterpillar of E. imj'>erUdisoi-\g\\vA\\\ lived on the pine, though 

 when feeding among the needles it is not readilv detected. 



Coloration in Citheronia. — Unlike Eacles, the two species of the present genus, the only ones 

 whose earliest larval stages are known, are not marked witli bands either longitudinal or trans- 

 verse, or any spots. In C. regidix and fiejju?cndii< the body at birth is either dusky pale on the 

 upper side of abdominal segments four to six {C. reycdix), or pale yellowish brown and dusky on 

 the dorsal side of abdominal segments live to eight ( ('. xej>idcr(dlx). 



The markings, or rather their absence in this stage, throw no light on the relationships of 

 this genus, whatever may be said of the armature. 



After the first molt ('. rcijidix is reddish and C. >« pidcndix yellowish. Now, some very 

 interesting stripes appear in ( '. /■,<jiilis. On each side of the back of abdominal seg'ments one to 

 eio-ht are three short, dark, irregular longitudinal bands, the lowe><t of v'hich is directed a litth> 

 dovnward and extends midi r tin- spiracle, this hecoming the ohlique hand, a characteristic color- 

 mark of thi.s genus, not usually present in the Ceratocampidte. Unlike the oblique stripe.s 

 in the majority of the Si)hingid;e, these stripes pass from the front edge of the segments, and 

 also do not overlap on the succeeding segment. This is also the case with the corresponding 

 markings in the Saturniidie (Telea. etc.). The two shortei' upj)er bands are not retained, but 

 disappear with the second casting of the skin. 



In stage HI the seveti pairs of oblique stripes are dark aliove and pale beneath. They 

 probably harmonize with the dark twigs of the food tree. 



Those in C. sepidcratis arise in the same wa}', there being at first three, though moreoljscure 

 and irregular, and in this species two are retained, only one being eliminated; there being on 

 each side a di.stinct straight dorsal stripe, liesides the oblique one situated in front of the spiracles 

 which is not quite so long as in C. regaHs. We thus see that coloi-ntional features in these two 

 species appear to be more stable than the armature, since the two sjjecies under discussion belong, 

 as regards the armature, to two quite difl'erent sections of the genus. 



