14 CHARLES T. BRUES. 



The majority of caterpillars occur upon their food plants in 

 exposed situations, and as they are very helpless creatures, h.i\<- 

 developed in many cases elaborate spines, tufts of bristles, or 

 poisonous hairs which are undoubtedly protective adaptations. 

 Many of them are also remarkably colored, presenting shades 

 and patterns that may be readily interpreted on the basis of the 

 theories of protective resemblance, warning color and mimicry- 

 Consequently, at first blush these larvae would not appear to 

 show any consistent characters applicable to their classification 

 on a rational basis. In i8Sf> \V. Miiller found that the position 

 of certain primary body-seta- might be utilized for the purpose 

 of classification. Dyar ('94), Forbes ('10), Tsou ('14), and most 

 recently Fracker ('15), have elaborated and modified this idea 

 and it is now applicable in a most satisfactory way throughout 

 almost the entire order. Thus we find that the arrangement of 

 these setae follows a certain plan, that the individual setae are 

 homologous, and that the modifications present in specialized 

 families may be derived from those of more primitive groups. 

 Still more interesting is the fact that very young larvae of all 

 groups tend to approach more closely to the primitive arrange- 

 ment than the larvae of later instars. We find thus quite evi- 

 dently that in ontogeny the caterpillar reproduces certain changes 

 in setal arrangement that have occurred in the phylogeny of the 

 group. Fracker has called attention to this as he found it to occur 

 quite consistently, although he fails to point out some of the inter- 

 esting conclusions to which it may lead. This is really a beau- 

 tiful case of the morphological independence of the larval state 

 in insects. These characters, which are of a purely transitory 

 nature, have never been in any sense adult structures, yet they 

 have so thoroughly impressed themselves on the species, that 

 they repeat their history in a very specific way in the develop- 

 ment of the individual caterpillar. 



On account of the small size and relatively unimportant I unc- 

 tion of the setae, their use in classification is greatly enhanced. 

 Thus the larger groups of Lepidoptera can be de-limited more 

 easily from a study of the larvae than can those of most other 

 insects, so far as our present knowledge extends. 



Until very recently the pupae of the Lrpidoptera have not 



