Life Histories in the Hepicdid GrouiJ of Lepido'ptera. 425 



Notes on piqjm. 



My material is scanty ; it comprises empty pupa cases 

 of several species, and live pupa of Porina cervinata. In 

 connection with the latter, the complete preservation of 

 setce in the exact number and position of the larval setse is 

 most interesting; it is probable that with pupae of other 

 Hepialidm, the setse are equally well represented, but 

 during the operation of emergence or from other causes, 

 they get broken, and though some are easily detected, it 

 is usually not possible to detect all the (larval) setie on 

 empty jjupa cases. 



The antennal cases are always short, and are not fused 

 to the pupal integument, so that on dehiscence they 

 separate from it, and become detached with the headpiece 

 from remainder of pupa case, leaving a hollow scar at the 

 upper part of each of tlie second pair of legs. The tips of 

 the third pair of legs are always between the apices of the 

 wing cases. Only a small portion of the outer margin of 

 hind-wing case projects beyond the fore-wing case in Porina 

 {cerviiiata, fig. 21) ; a longer portion but no perceptible 

 base, in Charagia (vircscens) ; base to half outer margin in 

 Trictena (lahTjrinthica) ; base to costal side of apices of 

 fore-wings in P/iassus [■purpuresccns). 



The 1st abdominal segment is represented by a small dorsal 

 area without spiracles, 2 has spiracles wholly above edge 

 of wing-cases — except in Phassus, where it is partly covered 

 by the extended margin of hind-wing cases. 1st and 2nd 

 abdominal segments have anterior spines in Charagia, but 

 not in Porina, Trictena, Phassus; 3 to 6 have anterior 

 and posterior ridges with spines commencing in the 

 locality of the spiracles and extending across the dorsum ; 

 4, 5, 6 have ventral spines in position of abdominal feet of 

 larvae (4, 5 only in Porina) ; 7 is the most characteristic 

 pupal segment, in Porina a ventral anterior ridge of spines 

 is strongly developed ; Trictena has the ventral anterior 

 ridge strongly developed, and also spines extending com- 

 pletely round the segmental ridge ; Charagia and Phassus 

 are similar to each other, both have weak posterior spines, 

 and stronger spines in position of abdominal feet (as on 

 4, 5, 6), these are connected midventrally by a series of 

 spines posterior to the inner (leg) seta^, in Phassus, but not 

 so in Charagia; 8, 9, 10 are smooth or nearly so in 

 Porina, Charagia, Phassus ; 7, 8, 9 in Trictena are deeply 



