1368 THE BUTTERFLIES OF NEW ENGLAXD. 



The caterpillars are easily distinguishable by the great constriction of 

 the neck enveloped in a chitinoiis collar (a posterior extension of the hinder 

 face of the skull) and by a corneous plate on the summit of the first 

 thoracic segment. The head is almost invariably rugose, and the body 

 smooth or covered with a very delicate pile composed of short hairs, 

 vifhich, at least in the first stage of the caterpillar, are always greatly en- 

 larged at the extremity, so as to appear fungiform or wine-glass shaped. 

 Neither head nor body is e^er provided with spines, though in some 

 tropical forms (Pyrrho[)yga) the Ijody is coAcred with long hairs ; it is 

 further noticeable that head and body are alnlost invariably of very 

 distinct, sometimes almost incongruous colors, and that when the head is 

 ornamented it is usually in large patches. The caterpillars live singly in 

 nests formed of the leaves of their food plant, fastened by a few silken 

 cords, and are very cleanly in their habits. They are also very slow in 

 their every movement ; it generally takes them nearly a day to eat their 

 way out of their egg shells, and they appear to spend the major part of theu- 

 lives within their nests, when it is horizontal resting feet upward, as if 

 longing for the time when they could sleep the pupal sleep in that position. 



In the change to imago, the chrysalis skin does not part at the suture 

 between the head and prothorax so as to free the head, as is the case with 

 all other butterflies, but in a special suture near the back of the head, 

 running from one antenna to the other, so that a short mai'ginal piece of 

 the head remains attached to tlie prothorax in exuviation. This suture 

 is plainly seen in the chrysalis of the Pamphilidi, but is generally obscure, 

 though no less eftectual in the Hesperidi. 



The chrysalids are generally well rounded, presenting no prominences, 

 excepting sometimes a slight projection in front, similar to that of the 

 Pieriuae. Poulton is e\'idently unfamiliar ^vith the transformations of the 

 Hesperidae, when, apropos of Weismanu's remark that it is impossible to 

 characterize the larvae of butterflies as a whole, as it is possible with the 

 imagines, he says : *' The pupae of Rhopalocera can be formed into a 

 large group corresponding to the union of the imagines into one of the 

 two chief divisions of Lepidoptera. The characters by which these pupae 

 can \>e identified as a whole are briglit, or at any rate varied, colours ; 

 angularity of outline, especially anteriorly ; and mode of suspension. 

 The divergence from the pupae of Heterocera as u lu/iole is also accom- 

 panied by a difference in conditions ; the latter being jn'otected from light 

 in the earth or in cocoons, while the former are freely exposed to it." 



Werneburg remarks that the pupae of this group are darker or lighter 

 according to the amount of exposure, those of the Hesperidi which pupate 

 within leaves being dusky, while those of the Pamphilidi changing between 

 grass blades are greenish. 



The insects enclose themselves in a slight cocoon for transformation, 



