190 Psyche [December 



Type 1 {subventral tympanuvi). Pericopidae. 



Geometrida? only. Herminiini and Epizeuxis of 



the Noctuidae. 

 Type 2 (subdorsal bulla). 

 Lymantriidse. ^^^'^ '^ {parapleural fovea.) 



Arctiidse. Notodontidse. 



Nolidse Noctuidne. 



Syntomidse. Agaristidse. 



It should be noted that Briies and Melander, in their key to 

 the families of Lepidoptera erroneously refer to this organ in the 

 Pericopidae and Geometridoe, as the spiracle; it is perfectly distinct 

 from the spiracle, lying above it in the former, and below in the 

 latter case. So far as I know the spiracle itself is always small and 

 normal, and varies very little in position. It should be noted, 

 however, that the "tympanic opening" of the Dioptidae (Psyche, 

 XXI, 59, alternative 32) is not homologous with thatof the Geomet- 

 ridae, but is the rudimentary structure noted in Apatelodes, and 

 w^hich becomes enlarged in the Agaristidse. In the case of the 

 Pericopidae (alternative 48) the separation from the Arctiidae is 

 slender, while the majority of Noctuidae of course completely lack 

 the bulla. The key would really be cleared up by transferring 

 Haploa to the Pericopidae, where it would not be as wholly out of 

 place as one might think. The Pericopidae certainly belong here 

 rather than in the neighborhood of the Notodontidae or Bomby- 

 cidae. 



Summary and Conclusions. 



A study of the base of the abdomen in certain Lepidoptera sug- 

 gests: 



1. The Notodontidae are better placed near the Noctuidae than 

 near the Geometridae and Bombycidae. 



2. Apatelodes is not Notodontid, but may be Eupterotid, agree- 

 ing in general with the Bombycid series. 



3. The Lymantriidae, Arctiidae, Nolidae, Pericopidae and Syn- 

 tomidae are undoubted close relatives, agreeing in a unique 

 specialization. 



4. The Noctuidae, Agaristidae and Notodontidae are similarly 



