SOME SINGULAR TlllNCiS ABOUT CATERPILLARS. 1429 



segments or sliglitly change their direetion at this point, so that often we 

 may readily distingui,<ii the thoracic from the abdominal segments without 

 looking at those parts which characterize them distinctively, such as the 

 legs or spiracles. Thus, even in the mere disposition of the spines along 

 a caterpillar's back, the future separation of the thorax and abdomen is 

 foreshadowed. This is wholly independent of the larger amount of space 

 upon the thoracic tract due to tiie absence of spiracles ; for, when 

 the ■;[)ines are well developed on the first thoracic segment, whicii bears an 

 unusually large spiracle, they are aligned with those of the otJier thoracic 

 segments and not with those of the abdominal segments. 



This, like the absence of spiracles from the second and third thoracic 

 segments, might be explained on the theory that the transformations of the 

 insects are an acquired characteristic, a development backward from the 

 imago. But tiiis will not explain another peculiarity which one observes 

 in the general arrangement of the spines and other dermal appendages on 

 the back of caterpillai-s. a feature which is extremely common, though 

 perhaps not imiversal, with all vermiform creatures. I refer to what might 

 be called the polar or antithetic arrangement of these appendages, which 

 shows itself in a multitude of ways. As a general rule the hairs, spines, 

 filaments, or what not, are highly developed upon the thoracic seg- 

 ments, sometimes increasingly so from the hindmost forward, the series 

 culminating in lofty bristles or long appendages upon the first thoracic seg- 

 ment. When this occurs, it is an almost invariable rule that a similar but 

 reversed arrangement and extension of the same class of appendages is 

 found upon the terminal abdominal segments. Or, if, as is frequently the 

 case, the second or third thoracic segment is independently enlarged or its 

 armature specially magnified, a similar but generally lesser development 

 will be found to occm- on one of the preterminal, though not the terminal, 

 abdominal segments. A case in point is easily seen in the caterpillars of the 

 genus Basilarchia, where the second and third thoracic segments are mam- 

 railate and the second is crowned by a pair of stout, thorny tubercles. So, 

 too, in a less degree, the seventh and eighth abdominal segments ai'e slight- 

 ly hunched and the corresponding tubercles at that point are noticeably 

 enlarged, especially on the eighth segment. 



Many other similar features might be pointed out even among the limited 

 series of our own caterpillars, as in all tlie young Papilioninae, and this sym- 

 metrical polarity seems (juite akin to that which we have pointed out in the 

 markings of the wings of butterflies, where corresponding ocelli are found 

 upon the wings in antithetic positions as related to the vein structure 

 beneath. Another instance of this polarity is seen in many of the cater- 

 pillars of the Satyrinae, in all of which the terminal segment ends in a 

 fork of greater or less dimensions, in some instances taking the form of 

 a long, pointed spine on either side, directed backward. When this is 



