25] LEPIDOPTEROUS LARVAE — FRACKER 25 



spiracle. Above it are only five setae, two in a cephalic and three in 

 a caudal row. The homology is perfectly evident, gamma, the middle 

 of the three setae of the prothorax, being the only absentee, and 

 alpha, beta, delta, rho, and epsilon having the same position as before. 

 Below kappa is located pi, the caudal seta of the Pi group. Tau occupies 

 the same position as before. 



We thus find that all the thoracic segments in this instar are arrang- 

 ed on the same plan but that the prothorax bears three more setae, gam- 

 ma, eta, and nu, than the other two segments. Whether these setae have 

 been lost from the second and third segments or added to the first will 

 be discussed in a succeeding paragraph. 



On the abdomen (Fig. 4) above the spiracle there are only four 

 setae, two in each transverse row. The cephalic row, exactly as in the 

 metathorax, consists of alpha and epsilon, but in the caudal row delta 

 is wanting. Four of the six dorsal prothoracic setae are then retained 

 on the abdomen. Of the lateral setae, kappa and eta are both present as 

 on the prothorax but they are widely separated. It will be shown later 

 that this separation is not of great importance in showing the ances- 

 tral condition or the homotypy, for on the abdomen of the Microlepid- 

 optera, kappa and eta are approximate as they are on the prothorax. 

 Unquestionably they have been derived from the same source on all the 

 body segments. 



At the base of the proleg are foTind two cephalolateral setae and 

 one cephalic seta. The first two are almost directly behind pi on the 

 metathorax and clearly represent pi and nu. As the other is in front of 

 the proleg and bears the same relation to it that tau does to the thoracic 

 leg, it doubtless represents that seta. We may therefore conclude that 

 the arrangement of the abdominal setae is homotypic with that of the 

 prothoracic. 



Later Instars 



[Figs. 5, 6, 13, 14] 



The mature larva of Hepialus differs in some important particulars 

 from the condition in the first stage. Most of these differences were 

 pointed out by Dyar and their bearing on phylogeny was discussed. Our 

 point of view is somewhat different from his, for we are considering homo- 

 typy, a field which he did not enter. 



The most striking change brought about at the first molt is the 

 appearance of a certain seta, theta, on each segment. This is always 

 caudad of kappa and dorsad of it on all but the prothoracic segment. 

 It is the best established subprimary seta in the whole order, for it sud- 



