1430 THE BUTTERFLIES OF NEW ENGLAND. 



the case, and especially when it is most developed, the head also is 

 crowned with a similar pair of pointed spines, and at rest the head is bent 

 downward, so that these spines are thrown forward and the body ends at 

 each extremity in a jiair of long pointed spines. In this instance, at least, 

 a purpose might be seen m such an armature, for it would appear as if the 

 head bearing these long pointed spines would present a formidable appear- 

 ance to some of its enemies, especially as it is able to present these organs 

 at any point with great force and rapidity ; if an enemy, alarmed at the 

 front aspect, sought to assail the creature in the rear and were to find a 

 similar pair of spines, it might well be conceived that he would presume 

 that these also could be used with equal offensiveness. Possibly this will 

 explain many other cases. 



If we examine the arrangement of the spiracles upon the sides of the 

 body, we shall find that the first thoracic and the last abdominal pair are 

 invariably much larger than the others, which are equal among themselves. 

 The explanation of this is easy. The respiratory tube of each has to feed 

 a very much larger field, the head and second thoracic segments being fed 

 by the tubes finding their outlet at the first thoracic segment, and several 

 of the hinder abdominal segments of the body being equally dependent 

 upon that of the eighth abdominal segment. The spiracle of the first 

 thoracic segment is also situated on a higher level than the ordinary ab- 

 dominal spiracle, and this is a consequence, in jjart at least, of the ordi- 

 narily smaller size of this segment ; yet it is also true in those forms in 

 which the first thoracic segment is greatly enlarged. But what is cu- 

 rious is that in certain groups, the Lycaeninae, in particular, and the 

 Hesperidae to a less extent, the spiracle of the eighth abdominal seg- 

 ment is also situated at a considerably higher level than those of the other 

 abdominal segments. This seems another instance of the polar arrange- 

 ments of parts to which we have alluded, but the explanation here is less 

 obvious, since it is a characteristic only of certain groups, and even here 

 is not invariable ; for in the caterpillar of Feniseca, one of the Lycaeninae, 

 the eighth abdominal spiracle is quite on a level with those in advance of 

 it, just as it is in the bulk of butterfly caterpillars. The only reason for 

 this elevated position in these cases would seem to be the particular form 

 of the termination of the body, for in all the Lycaeninae, excepting 

 Feniseca, and in all the Hesperidae in which this occurs, we find a flattened 

 subonisciform shape, one which, indeed, throws the spiracles of all the 

 abdominal segments a little higher relatively to the base of the body than 

 is common among caterpillars in general. 



Besides the spines, filaments, bristles, etc., which form so noticeable and 

 common a feature among butterfly caterpillars, there is another still more 

 common and of a very similar nature ; that is, the short hairs or pile with 

 which the body is provided, always supported by little papillae and dis- 



