June, iS/n] DyaR : On THE OrIGIN OF PeRICOPID/E. 69 



As to the origin of this family, three alternatives seem possible : 

 (i) from the Arctiidas (2) from the NoctuidK (3) from the Dioptidai. 



(I). This alternative seemed to me at first sight most plausible. 

 However, the subcostal vein of secondaries in the Pericopidai is free 

 from the radius and the wart formation of the larva is of a simple type. 

 Therefore, if of Arctiid origin, they are a branch representing a more 

 generalized type than the present Arctiidte. But no confirmatory evi- 

 dence for this possible origin occurs to me. 



(2). By pushing the origin a little further back, we reach a deriva- 

 tion from the Noctuida^. This involves an independent wart formation 

 from that in the Arctiidse ; that is that we may suppose that the Arcti- 

 idae and Pericopidse arose simultaneously, or nearly so, from Nociuid 

 ancestors, but independently and with separate larval wart formation, 

 the Pericopidse remaining the more generalized. I do not think of 

 any confirmatory evidence for this view. 



(3). The Pericopidae and Dioptidge differ in that in the former 

 cubitus is four-branched while three-branched in the latter. This dis- 

 tinction is not sharp, since in Gonora heliconiata ( a Pericopid ) vein 5 

 arises only a short distance below the fold representing media. In Di- 

 optis pheloides, Fhryganidia calif ornica and Tithraiistes haemoti (Diop- 

 tidae ), vein 5 is a distinct continuation of medial fold. The families 

 almost intergrade, as the confusion in the genera in Kirby's catalogue 

 bears witness. 



The larvae of the Dioptidce ( only known to me in Phryganidia ) 

 have single Noctuidous setce. To derive the Pericopidce from them, 

 we must suppose an independent wart formation analogous to that of the 

 Apatelidse from the Noctuidse or the Eupterotidte from the Notodonti- 

 d£e, but simpler than either. 



Now as confirmatory evidence, we have in the moths the marked 

 similarity in build and pattern of coloration ; also a tendency in both 

 families to the coalescence of veins 3 and 4, not seen in the Noctuid^ 

 and Arctiidae. The Noctuid habitus of the larva is also equally con- 

 sonant with an origin from the generalized Noctuiform Dioptid larva as 

 from any existing type of the Noctuidae. 



I think that we may conclude, therefore, that the Pericopidse are 

 a family representing a specialized form of Dioptidae, with independent 

 moving of vein 5 toward 4 and wart formation in the larvae. Their 

 geographical distribution is also in favor of this conclusion, since, being 

 confined to x^merica, their origin from the strictly American Dioptids 

 is more reasonable than from either of the world wide Noctuidce or 

 Arctiidae. 



