1428 THE BUTTERFLIES OF NEW ENGLAND. 



some new and unexpected feature often presents itself and will then appear 

 in the same or in some modified form through a long series of different 

 species. Or the arrangement or disposition of parts with which he is per- 

 fectly familiar may suddenly be found to follow certain laws which he can 

 formulate but not explain, and which he had before overlooked. 



One of the most fundamental facts in the structui-e of caterpillars, as of 

 all arthropods, is the repetition of similar parts along the several rings of 

 which the body is composed. But there are in caterpillars, at least, two 

 disturbing elements which modify this law ; one, considering that the meta- 

 morphoses of insects are conceded to be an acquired characteristic, is readily 

 explained, and that is the difference between the structures found on the 

 thoracic and the abdominal segments. Thus, on account of the develop- 

 ment of the future wings, the spiracles of the seconti and third thoracic 

 segments are omitted, and the legs found upon the thoracic and abdominal 

 segments are very different in character, those of the abdominal segments 

 being a temporary expedient for the long trailing abdomen, while those of 

 the thoracic segments are more highly developed. But these exceptions, 

 which have a direct relation to the future needs of the animal, cannot explain 

 certain other features which show similar differences. In general terms, 

 the entire surface structure of the upper part of the animal is j^ractically 

 identical on the thoracic and abdominal segments, but there are certain 

 differences which appear, to which no explanation is readily given. 



For instance : By the transverse creases which simulate the incisures 

 between the segments, each segment is ordinarily divided into two or more 

 transverse sections. Now these sections, at least when there are more than one 

 or two, invariably differ upon the thoracic and abdominal segments, the ab- 

 dominal segments having a system of their own, distinct from that of the tho- 

 racic. The first thoracic segment, indeed, is the subject of great specialization 

 and differentiation, and often differs widely in its divisions, as also in the ap- 

 pendages it bears, from the other thoracic segments ; but this is not true, or is 

 true in an extremely limited sense, as regards the other thoracic segments, 

 which to all practical purposes are identical in general appearance with the ab- 

 dominal and would seem to have much the same office to fill. Yet, if we 

 examine carefully the dermal appendages of these segments, we shall note 

 some curious features distinguishing them from those of the abdominal 

 segments. Thus, the spines, bristles, filaments, or other special develop- 

 ments of the skin, are ranged in most caterpillars of butterflies in longitu- 

 dinal rows when they have any regular disposition whatever. There may 

 be one or two or more upon each segment in a single row ; all the spines 

 of one row will be found at corresponding points of the different segments, 

 either on the middle or back or front, as the case may be. Yet with scarcely 

 an exception among the caterpillars of butterflies, those series which extend 

 along the abdominal segments will either stop altogether at the thoracic 



