34 ILLINOIS BIOLOGICAL MONOGRAPHS [34 



prothorax. In the Tortricidae (Fig. 39), Aegeriidae, and Yponomeu- 

 tidae (Fig. 35), this seta remains constant in position while rho moves 

 forward toward it. This is distinctly noticeable in Plutella while in 

 Atteva aurea (Fig. 36) it has scarcely been begun. The opposite move- 

 ment is to be noted in all Macrolepidoptera. The newly hatched larva 

 of Feltia shows rho and epsilon in their normal positions distant from 

 each other. In the mature larva, however, epsilon has migrated back to 

 rho, leaving only alpha and gamma remaining in the cephalic subdorsal 

 group. 



On the mesothorax and metathorax of Frenatae, epsilon and rho 

 are again found associated. Like alpha and beta they have yielded to 

 the mobility of this part of the body and been crowded to the middle 

 of the segments. The first stage of Hepialus and Feltia show that epsi- 

 lon is the upper of the two and rho the lower. Both prothoracic and 

 abdominal segments of Hepialus and Cossus cossus prove that epsilon 

 was originally above the level of rho and that its present position ventrad 

 of that seta on the abdomen of many of the species we know today has 

 come about through migration. 



On the abdomen of Hepialus, the presence and large size of epsilon 

 in the first as well as later instars creates a serious problem if 

 that seta be considered missing from its usual position in the first stage 

 of Frenatae. I have been unable to detect it in the newly hatched 

 Feltia larva but according to Quail (1904), who quotes A. Bacot, it is 

 of general occurrence in all instars of Frenatae. The probability is, 

 therefore, that it is disappearing from the abdomen and is retained in 

 its former vigor only by Hepialus and Cossus cossus. There seems to 

 be no evidence that it is in any sense subprimary. 



Rho. The principal migrations of rho have been described in the 

 discussion of epsilon. It is present in all instars on all segments of all 

 caterpillars except when obscured in later stages by a secondary arma- 

 ture. Usually it is associated with epsilon when that seta is present. 

 In species bearing verrucae, epsilon never forms one distinct from that 

 of rho, except sometimes on the mesothorax and metathorax. 



Theta. The dorsocaudal seta of the Kappa group needs considera- 

 tion merely to prevent its confusion with other setae. No question can 

 be raised as to its homotypy on the segments of Hepialus, where it is 

 subprimary throughout, in all cases being dorsad and eaudad of the 

 primary kappa. The same is true on all thoracic segments of the tineoid 

 series of Microlepidoptera and on the meso- and metathorax of Pyralid- 

 idae and Macrolepidoptera, where it is present and subprimary. 



Some confusion might arise concerning it on the abdomen of 

 Frenatae in those cases where kappa is well toward the upper border 



