27] LEPIDOPTEROUS LARVAE— FRACKER 



thorax. They are somewhat more dorsal in position and not so close 

 together, but otherwise clearly represent the same structures. The two 

 lateral setae are in the same position as the Kappa group. The caudal one 

 is clearly theta, for it was not found on the newly hatched larva; but 

 the other may be either kappa or eta. I have labeled it kappa, altho 

 there is no evidence in this one species that it may not have developed 

 from eta instead. We shall see, however, that eta, when present on 

 the thorax, as in most of the Frenatae, takes a position quite distinct 

 from that shown here, while kappa is usually found in this place. The 

 homology as given, is therefore undoubtedly correct. 



Laterad of the leg and close to the coxa is a single seta, often borne 

 on a chitinized plate. This is pi, the only representative of the Pi group 

 ever found on the mesothorax except in a few cases. The Tau group and 

 sigma are also present. 



Abdomen. The relation of the type of arrangement of the setae of 

 the abdomen (Fig. 13) to that of the thorax can not be determined from 

 a study of the mature larva alone. This is due to the addition of several 

 new setae at the first molt and the changed position of others. As in the 

 first instar, there are four setae above the level of the spiracle, but these 

 do not so clearly represent the two transverse rows as before. The 

 ventral seta of the caudal row, rho, had migrated cephalad to the middle 

 of the segment even before the first molt and is now found very close 

 to epsilon and associated witli it above the spiracle. The fact that this 

 is actually rho and that it has come from the caudal part of the seg- 

 ment can not be doubted after seeing figures of the newly hatched larva 

 and studying the record of this seta throughout other members of 

 the order. 



In the spiracular region are three setae in a diagonal line, theta 

 and kappa caudad of the spiracle and eta some distance ventrad. Kappa 

 and eta were noted in the first instar but theta was missing. The latter 's 

 relative position on the abdomen is the same as that of its homotype on 

 the thorax. 



Below eta the maximum number of setae on any segment except 

 the last is five. These are arranged differently as we pass caudad. In 

 Hepialus hamuli all are present on the first six abdominal segments, 

 but one of those on the first segment is much smaller than the others. 

 It is entirely missing from this segment of H. lupulinus and H. hectus. 

 On the second segment the arrangement is more typical. Two setae, pi 

 and nu, are close together near the middle of the segment and only 

 slightly more ventrad than their homotypes on the thorax. Sigma is 

 present as usual near the ventromeson. This leaves only two setae, 



