506 PROCEEDINGS OF THE NATIONAL MUSEUM vol. no 



toothless mandibles mth a corresponding development of the epistoma 

 and pleurostoma. In the Metopiinae, Anomalinae, and Ichneumoninae 

 the hypostoma is also well developed. In the Pimplini the hypostoma 

 is reduced, but the hypostomal spur is well developed. The hypostomal 

 spur is absent in the Metopiinae, Anomalinae, and Ichneumoninae. 

 The labial and stipital sclerites are, except in the Pimplini, relatively 

 poorly developed. The reduction of these sclerites might be corre- 

 lated with the relatively slight cocoons spun by these endoparasites 

 (see section on Anomalinae) . 



In the species examined of the Metopiinae, Anomalinae,^^ Ichneu- 

 moninae, and Pimplini, development takes place in, or is completed in, 

 the pupa of a lepidopterous host. The Campoplegini, which are endo- 

 parasites of larvae, have a different type of head structure with the 

 epistoma absent and the labial and stipital sclerite relatively well 

 developed. With regard to the degree of development of these sclerites 

 it is significant that many Campoplegini emerge from the host larva to 

 spin a well developed cocoon, although some species of Hyposoter and 

 Benjaminia pupate within the host larva. 



It cannot be claimed that all Ichneumonidae which emerge from the 

 pupa of the host, or even from the pupa of a lepidopterous host, will 

 have a larval head of similar form. The Cryptinae have a head struc- 

 tm-e which is very different from that of the Metopiinae, Anomalinae, 

 Ichneumoninae, and Pimplini, Also, larval habits are not constant. 

 Mr. J. F. Perkins (in litt.) has directed my attention to some reared 

 specimens of Trichomma occisor Habermehl with a quite well developed 

 cocoon and to an Ophion in which the cocoon is spun within the host 

 pupa, which remains adhering to the outside of the cocoon. Thorpe 

 (1930) has observed that the ectoparasite Ephialtes rujicollis (Graven- 

 horst) spins a light and irregular cocoon, yet when reared in a gelatin 

 capsule usually does not spin a cocoon. But it does appear that in most 

 Metopiinae, Anomalinae, Ichneumoninae, and Pimplini there is some 

 connection between the form of the larval head and the habits of the 

 larva. It is thus possible that convergence in larval characters asso- 

 ciated with a similarity in larval habits might explain the similarity of 

 the Metopiinae and Anomalinae. 



From the standpoint of larval characters alone, the following inter- 

 relationships are suggested among the subfamilies of the Ichneu- 

 monidae. The Metopiinae and Anomalinae are related. The Ophi- 

 oninae, Mesoleiinae, and Lissonotinae are also related subfamilies. 

 Similarly, but to a lesser extent, there appears to be some degree of 

 relationship between the Tryphoninae, Cryptinae, and the tribes of 

 the Pimplinae other than the Pimplini and Acaenitini. The Ichneu- 

 moninae show some likeness to the Orthopelmatinae. The Adelogna- 

 thinae, Collyriinae, Plectiscinae, Diplazoninae, and Mesochorinae are 



" Except Anomalon, which pupates within the larvae of Elateridac. 



