376 WILLIAM T. 31. FORBES 



rollers. The larva 1 (figs. 245, 284-287) never have conspicuous mark- 

 ings, except for the shields and tubercles. Head always exposed and 

 fully chitinized, with normal mouth-parts (as in all the higher groups). 

 Mandibles tending to point more forward than those of external feed- 

 ers, but not different in structure. Front extending from one-fourth to 

 three-fourths way to vertex ; adfrontals almost always touching the 

 vertex, sometimes very large; ocelli normal. Prespiracular wart with 

 three seta 1 . Abdomen with setas i and ii well separated, and iv and v 

 approximate usually on one tubercle; prolegs with hooks in a com- 

 plete circle, which is often broken in. front and back, except in a 

 couple of (/ossida 1 ; often with tubercles ii of ninth segment of abdomen 

 approximated or united in mid-dorsal line, and often with a special- 

 ized multiple supra-anal spine. Many species are injurious as borers 

 or external feeders, probably a larger proportion than of any other 

 superfamily. 



Pupa of a normal incomplete type, with segments 3 to 6, and 7 of 

 male, movable, with two rows of spines on each segment, as in the 

 higher Tineoidea and ^Egeriidaj. Head often with a cocoon-breaker ; 

 prothorax convex and wider than dorsal headpiece; maxilla distinct, 

 not divergent as in the Hepialidas, but sometimes almost as small, and 

 completely separated by the labial palpi, which are always exposed. 

 Mentum well developed. Maxillary palpi distinct, separated by a 

 suture ; but in Zeuzera, at least, remaining with maxilla? on dehiscence. 



Antennae varying, following their character in the adult; rarely if 

 ever as long as wings. Fore femora exposed. Cremaster of various 

 types. 



The four families are quite clean-cut ; in fact, are united mostly by the 

 combination of specialized micro larva with typical incomplete pupa. 

 The Eucosmina3 (Olethreutina?) are often made a separate family, but 

 have no clean-cut characters to distinguish them from the Tortricina?, 

 though certainly a well-marked group in the imago. Several workers 

 separate the Zeuzerina? also as a family Zeuzeridre. In this case inter- 

 mediate forms are rarer if they exist, and the separation is very pos- 

 sibly justified; it has not generally been made by American workers. 



Family. 28. TORTRICID^ 



(With Eucosmidce, Grapholithidce, Epibleniidce, Olethreutidce) 

 Ocelli present (so far as looked for) ; antenna? rarely pectinate (never 

 in our species) ; the scaling confined to dorsum of antenna, but the outer 

 row stronger and longer than the inner ; ventral surface pubescent, more 

 strongly so in male ; palpus moderate, upturned to middle of front, or 

 rough and porrect, often triangular; tongue present, usually rather 



