164 AUSTRALIAN WOODBORING COSSIDAE 



is considerably more than the length of the ventral surface of 

 the prothorax ; the produced posterior area of the scutellum is 

 provided with numerous spicules (Zeuzera, Endoxyla). The 

 prothoracic scutellum of Cossinae is smooth, not produced 

 posteriorly, being confined to normal and proper limits, the 

 posterior margin parallel to the anterior margin (Cossus, 

 Culama). 



An important feature in Zeuzerinae is the presence on the 

 dorsum of the abdominal segments of minute tubercles (one 

 •each side) with seta, in front of the typical anterior trapezoidal 

 tubercles (Zeuzera, Endoxyla), but which are not observable in 

 Cossinae. These tubercles are probably homologous with the 

 thoracic intersegmental tubercles, observed also in Cossinae, and 

 in other groups of Lepidoptera. Mr, A. Bacot, in a letter some 

 years ago, drew our attention to these (abdominal) tubercles on 

 the larva of Zeu era pyrini (Europe), he having also observed 

 those of the thorax in several isolated groups of Lepidoptera 

 e.g. Lyctenidae, Psychidae. We have noted such thoracic 

 intersegmental tubercles with setae in newly-hatched and adult 

 Hepialidae*, in Lysiphragma (Tineina) without setaf ; they are 

 also present in Tortricina (Cacaecia) and others. Bacot believes 

 this to be a once common character, now generally lost in 

 Lepidoptera, and had not then detected the setae which are 

 present on the thoracic and abdominal tubercles in question of 

 Endoxyla. We have been unable to find whether Dr. Dyer is 

 aware of these extra abdominal tubercles. 



We are no"; inclined to insist upon minor differences ; that 

 Zeuzerin larvae are circular or nearly so in transverse section ; 

 Cossinae, being flatter, are barely more than semi-circular ; that 

 Zeuzerinae have middorsal spiculate abdominal humps, and 

 some other features, as young larvae in either group may be 

 more alike. 



A consideration of the papal structure affords additional 

 support to the conclusions arrived at from larval characters as to 

 the separation of Hepialidae from Cossidae,:^ and the subdivision 

 of Cossidae. The pupae bear a superficial resemblance to each 



* Trans. Ento. Soc, Lon., 1900 (Quail). 



t Trans. New Zealand Institute, 1900 (Quail). 



I We wish to emphasize that these groups cannot be associated, which, 

 of course, is generally admitted, but ai'e sometimes treated of or referred to 

 as if they were. Their primitive ancestors were probably neither Hepialid 

 nor Cossid, but had some of the characters now found in each group. 



