21] LEPIDOPTEROUS LARVAE — FRACKER 21 



realm of speculation. Let us consider for a moment the nature and kind 

 of these indications and their use as evidence on a correct nomenclature 

 of the setae. 



APPLICATION OP PRINCIPLES 



Similar position in all modern forms. — When a group of setae are 

 in similar positions throughout the entire series of mature caterpillars, 

 we may conclude that they are homologous with each other. If further 

 evidence does not point distinctly in some other direction, we can not go 

 back of this stand. Practically every segment of every caterpillar bears 

 one seta near the medioventral line on each side. This the writer calls 

 sigma and it is clear that in these setae we are dealing with truly homo- 

 logous organs ; that on the thoracic and abdominal segments in Hepialus 

 as well as the Frenatae, these ventral setae are homotypes. 



Similar position on certain segments of all modern forms. A nearly 

 uniform arrangement of the setae on the prothorax of practically all the 

 members of the order is excellent evidence that this arrangement is 

 ancestral and that it has not arisen through convergent development. 

 On the other hand, such a condition cannot bear on the relations of the 

 segments to one another. 



Similar arrangement on all the segments in the most generalized 

 groups. Every structure of Hepialus points to the view that this genus 

 is one of the most generalized of Lepidoptera. While distinctly in a dif- 

 ferent suborder, it bears much evidence that it is closer to the ancestral 

 type than are most of the Frenatae. When we find in it, therefore, that 

 the relations (Figs. 5, 6, 13, 14) of alpha, beta, and rho, to each other and 

 to the boundaries of the segments, are identical throughout the body, 

 we have reason to believe that they are homologous in spite of their 

 changed position in the Frenatae on the mesothorax and metathorax. It 

 simply remains to derive the condition found on these two segments 

 of Frenatae from that shown by the same two segments of Hepialus, 

 and again we have a complete series of homologues. 



Similar position on all the segments of newly hatched larvae. 

 The bearing of the setal arrangement of first-stage larvae was discussed 

 under the subject "Ontogeny". Homologizing a seta never present in 

 this stage with another that is present cannot usually be admitted as 

 justifiable. On the abdomen of Hepialus (Figs. 6, 13) there are three 

 setae, theta, kappa, and eta, in a long diagonal row caudad and ventrad of 

 the spiracle. Of these, the upper one is absent from the first stage on all 

 the segments and therefore, according to Weismann's hypothesis, may 

 be assumed to be of more recent origin. It would certainly be incorrect 

 to homologize it with any primary seta of generalized Frenatae. 



