BY B. ILI.IDGE AND AMBROSE QUAIL, F.E.S. 163 



until the pupa forces its anterior segments out of the burrow 

 for the imago to emerge. Young Cossus larvae feed at first 

 beneath the bark of the tree, then burrow into the wood, there 

 to spend, as do Hepialidae and Zeuzerinae a lengthy existence 

 often of three or four years. The habits of Cossinae are not so 

 exclusive as those of Hepialidae and Zeuzerinae, the external cover 

 is often absent (Culama), indeed we have frequently observed 

 the larva of Cossus cossus when full fed, expose its whole length 

 to warm sunshine, moreover it will commonly leave its larval 

 burrow to pupate elsewhere, even in the earth away from its 

 ligneous habitat. May it be that a too numerous colony 

 (Cossus and Culama are gregarious) in the same tree, by 

 breaking in up^n each others barrows, always strictly avoided 

 by Hepialidae, become a source of irritation, or even danger at 

 the critical time of changes to the pupal condition, and the 

 larva is compelled to pupate elsewhere. The normal habit 

 appears to be that the larva tunnels to the bark which it eats 

 away, leaving however a very thin surface, and it may be noted 

 these larvae frequently fill their burrows with a kind of solidified 

 sawdust (Hepialidae scrupulously eject all frass). fA cocoon of 

 silk and chips is constructed within which it pupates, in the 

 immediate vicinity of the exit ; finally the pupa forces its 

 anterior segments through the thin outer cover of the burrow 

 and the imago emerges. 



Always remembering the limited material at our command, 

 so far as it goes our observations show that Zeuzerinae and 

 Cossinae may be associated by some identical larval structures. 

 The number of scutellar setae of prothorax, the tubercle 

 arrangement (and spiracle scars) of meso and post thoracic 

 segments, the lateral thoracic intersegmental tubercle ; the 

 duplicate remote supraspiracular seta, the position of the basal 

 setae, and the hooks of the abdominal feet. These structures 

 are not peculiar to Cossidae, being also observed in other groups 

 (composing Dyer's superfamily Cossina), but they sharply and 

 distinctly separate Cossidae from Hepialidae, which cannot be 

 associated . 



Some larval features appear to afford good characters upon 

 which we may separate Zeuzerinae from Cossinae. The Zeuzerin 

 prothoracic scutellum, viewed laterally, slopes upward and 

 backward in dorsal outline, so that the length from front to back 



t Proc. Eoy. Soc. Queensland Vol. XIV. (Illidge). 



