72 MEMOIRS OF THE NATIONAL ACADEMY OF SCIENCES. 



caterpillar is at rest they are held close together in a recurved position and in the grown-up larva 

 when touched they are not moved or the body jerked in response to such.stiniulu.s. They are 

 adorned with white, blunt spines, which are often tipped with lilack. 



"The 'silver horns' on the third to the sixth abdominal segments are now one-sixteenth to 

 one-eighth of an inch long, bright pink inside and ))urnished silver externally. The numl)er of 

 these 'silver horns' varies in different larvte, some having them only on the fourth and sixth 

 abdominal segments; others have them on the fifth, seventh, and ninth segments; still others have 

 them on the second, fourth, sixth, and seventh segments." (Papilio, II, 49.) "I have now to 

 add that this j'ear I reared three larvte having these silver horns on every segment except the 

 twelfth; still the imagines from these three larva? did not differ from the ordinary form." 

 (Jewett, Papilio, II. 144.) 



The horn on the eighth abdominal segment is now only about one-fourth shorter and thicker 

 than the thoracic spines, and is of the same color and structure, the spinules being conical, 

 rounded, blunt, white, and liearing a tine bristle. 



On abdominal segments 1-7 are two dorsal rows of acutel}' conical spines, which are recurved 

 and directed backward. Those on the fourth and sixth aljdominal segments are twice or thrice 

 as large as those on the other segments (1-3 and 4 and 7), and provided with three or four blunt 

 spinules; the spines themselves are roseate on the inner side, and exteruallj^ briUiantly painted 

 with a pearly silvery white, giving off' all the colors of the rainljow during the movements of the 

 animal. The corresponding spines on the other segments are painted in the same fashion, though 

 less brilliantly. 



I find, as did Jewett, that the number of the dorsal "silver horns" varies, one larva having 

 but two pairs, one on the fourth the other on the sixth abdominal segment, while in another 

 there is a pair on the second abdominal segment (fig. 3). The degree of specialization of these 

 dorsal spines varies, but those of the first abdominal segment are always'smaller than any behind, 

 both in stages IV and V. This stage in armature differs from IV in the )nuch shorter and 

 stouter thoracic horns, and in the diff'ei'entiation of the two or three pairs of dorsal abdominal 

 silvery spines. 



On the side of the body from the third thoracic horns to the eighth abdominal spiracle is a 

 bicolored stripe; it is pure marble-white below, and above rosy purple, and is interrupted by the 

 wax-colored spiracles, which extend above the \ipper limits of the reddish line. The suranal 

 plate is vei-y large and long, deeply divided at the end, the two forks being, like the surface, 

 coarselj' granulated with stout short conical spines; the plate is green, with the edge straw- 

 yellow. There is a minute median spine on the ninth abdominal segment. Each abdominal 

 segment with two dorsal transverse rows of white, liead-like, coarse granulations. Below the 

 bicolored lateral stripe is a black, double, conical spine on each segment, and underneath on 

 abdominal segments 1. 2. 7, and 8 is a group of unequal, smaller, black, sharp spines. The body 

 beneath is granulated with white, and also on the sides, as well as above. 



The thoracic legs are black, partly greenish beneath; the abdominal legs, including the anal 

 pair, are greenish, with a group of singular black piliferous spines, while some of the spines ai'e 

 tipped with white. 



The general color of the body is of nearly the same hue as the under side of the leaves of 

 the honey-locust, and thus colored it is partly assimilated and protected by its color, while the 

 horns are in general like the spines of its food plants. On the other hand, the gleaming silvery 

 spines certainly render the creature conspicuous, as well as the lateral parti-colored ))and. 



It would appear probable that the formidalile spines of the grown-up caterpillar save it not 

 infi-equently from being ^swallowed by birds; though the horns are probably of greater use in 

 the earlier stages, when they are much longer and much more movable, in frightening away 

 ichneumons and Tachiniv. For example, even when 20 mm. in length, a larva was seen when teased 

 to spread apart its great arm-like horns, while the full-fed ones did not notice such stimulus. 



