LEPIDOPTERA OF NEW YORK AND NEIGHBORING STATES 



15 



an undoubtedly early type; and the lowest butterflies are but little 

 removed from the highest of the Cossid type (the tropical American 

 and Australian Castniidae). In certain ways (as in the thoracic 

 sclerites) even the highest butterflies show their primitive character. 

 For such reasons some recent authors have inserted the butterflies next 

 to the Cossidae, and have put the Euchromiidae or Noctuidae at the 

 pinnacle of the order. It should be remembered, in any case, that each 

 is the terminal of a long line of separate descent: and a true arrange- 

 ment would put them side by side rather than one above the other. 



The diagram (fig. 1) indicates the writer's impression of the probable 

 relationship of the families. The Jugatae are undoubtedly an early 

 type and approaching extinction. The Frenatae can scarcely have 

 descended from either of the three Jugate families, but must have come 

 from a stock with the haustellum, ovipositor, and tibial spurs of the 

 Eriocraniidae, and the larva of the Hepialidae. The earliest type of 

 Frenate must have been much like Incur varia, but doubtless larger, 

 and with characters that have been lost in all the aculeate genera, but 

 survive in the lower Tineidae, such as Acrolophus and Scardia. From 

 such a type the Tineine superfamilies have arisen as separate branches, 

 the Tortricoidea perhaps as another branch, or in association with the 

 Yponomeutoidea, and the Pyraloids and higher Frenatae doubtless 

 from a common stem with the Yponomeutoids, but diverging very early. 



In a similar way, the individual superfamilies of the Macrofrenatae 

 must have separated very early; but the Drepanoidea, Uranioidea, 

 Saturnioidea, and Geometroidea may have hung together a short time 

 after the Sphingidae and the Noctuoidea had become distinct. 



As to the internal evolution of the superfamilies, little is clear in 



most Cases, Save that the author has No/idae /Imatidae Ayanstidoe 



aimed to arrange the families in 

 order, from the more generalized to 

 the specialized. In the noctuid 

 group the relationship is a little 

 clearer (fig. 2), the connection be- 

 tween Notodontidse, Liparidae, and the 

 remaining families being through the 

 exotic families Hypsidae and Peri- 

 copidae. The Hypsidae have the 

 primitive venation of the Liparidae 

 but have preserved the ocelli and the 

 tongue. The Dioptidae have the tri- 

 fid venation and the free Sc of the 

 Notodontidae, but the tympanum is of a primitive type from which the 

 others may be derived. 



In the butterflies, the line of descent through Hesperiidae and Papili- 



Lipartdae 



Coct/h 



A/ofodonhdae 



P/opfidae 



FlG. 2. GENEALOGY OF THE FAMILIES 

 BELATED TO THE TSOCTUIDM 



