loo Aiinals Entomological Society of America [Vol. Ill, 



The possibly primitive state of the proleg can perhaps be 

 reconstructed from Adela and Incurvaria. Each segment bore 

 on the ventral side two chitinous sclerites, between which was a 

 fold of meml)rane, which was rough with minute conical granu- 

 lations like the rest of the skin. In this membrane there may 

 have developed a concave fold, which lost its granulations, while 

 those at each edge of it were enlarged, and became regularly 

 imbricated, gradualh' becoming like the unmodified granules as 

 >'0u go away from the fold toward the two sclerites. This is the 

 condition of Adela, (Fig. 6), except that the sclerites still pre- 

 served in Incur\-aria, are lost. On segment A6 the front side 

 of the fold only develops hooks, and the folds of the remaining 

 abdominal segments (i, 2, 7, 8, and 9) are traceable, but entirely 

 without hooks. There are no anal legs. 



In Incurvaria (Fig. 7) the posterior rows of hooks have disap- 

 peared on all the segments, so that A6 is no longer different from 

 A3 to 5, and the hooks are reduced to a single row. 



Hepialus has kept two rows of hooks, and in part, three, but 

 the teiTninal row is much more highly developed, while the others 

 are quite rudimentary. The two bands have fused on the inner 

 side, and approach each other on the outer side, thus enclosing 

 the fold, which becomes the planta. In many higher micros the 

 outer ends have also fused, making the circle (ellipse) complete. 

 In all above the Tineidae there have developed hooks on the 

 last segment, but there is never a posterior series, or a complete 

 circle; in some there is a straight band, which becomes a loop, 

 but is always open posteriorly. It is nearest a circle in the Psy- 

 chidae and lower butterflies. In the higher forms (butterflies and 

 Noctuidae for instance) , it has again become straight, but longi- 

 tudinal, by the disappearance of the outer hooks. 



The ventral legs have a similar evolution. The circle which 

 serves as a type for all the higher species may be broken on the 

 outer side {Pyralididac) , on the inner side {Psychidac) , or on both 

 (^geriid(c). It may be broken in front and back {Papilionidae, 

 Thyatiridae, Lihythcidac,) and in that case the outer half becomes 

 weaker, and then disappears, lea\-ing an inner band only {Xocini- 

 )ia, Bombycina, higher butterflies, etc.) 



In the Tinciua the last ventral legs may entirely disappear 

 (as has been mentioned they are only half develo|)ed in the very 

 primitive type Adela) while in several higher families the first 

 ventrals are lost (Xoctiiidae, Geomctridac, Nolidac). In the 



