18 Annals Entomological Society of America [Vol. XII, 



The very great similarity of the planidium stage of Psilo- 

 gaster to that of Perilampus is extremely interesting and sug- 

 gests a close affinity between the Eucharidas and Perilampidae 

 as has already been pointed out by Smith from a comparison 

 of the planidia of Perilampus and Orasema. In fact, this stage 

 of Psilogaster fasciiventris is almost as close to that of Perilampus 

 hyalinus as the latter is to "Perilampus, species a" figured 

 by Smith. In this case it may of course be possible that 

 "species a" is not a true Perilampus as Smith did not succeed 

 in rearing the adult. The host relations are very different 

 in the case of the two families so far as is known; all bred 

 Eucharids have been found to be parasites of ants while Peri- 

 lampus has been bred from a diverse series of hosts, including 

 Lepidoptera, Coleoptera, Neuroptera and some parasitic 

 Hymenoptera and Diptera. 



The later larval and especially the pupal stages are much 

 more highly modified in the Eucharids than in Perilampus, 

 although in the larvae there seems to be much diversity in this 

 respect. In the case of the larva, that of Psilogaster is com- 

 paratively simple, without conspicuous exudate organs, although 

 it is well supplied with minute surface hairs which no doubt 

 function as secretory organs of this type. Orasema possesses 

 numerous, highly developed exudate tubercles arranged seg- 

 mentally on both the thorax and abdomen. On the other hand, 

 Perilampus shows similar conspicuous tubercles and projections, 

 although it seems improbable in view of the host relations in 

 this case that these can function as true exudate organs. In 

 the pupa, that of Perilampus is simpler, although with small 

 intersegmented ridges on the abdomen. In Orasema the pupa 

 exhibits well developed transverse welts between the abdominal 

 segments as well as some on the head. The climax appears in 

 Psilogaster where the exudate organs are extremely large and 

 elaborate as described in the present account. 



There would seem to be two possible reasons for the presence 

 of the highly developed exudate-like organs on the Perilampus 

 larva. It might be that Perilampus is derived from Eucharid- 

 like ancestors in which these organs actually [functioned for 

 the elaboration of an exudate, but this seems hiAly improbable 

 \ .-in view of the much more specialized habits and structure of the 

 ". Eucharidse. On the other hand, it may be that the integu- 

 mental modifications of Perilampus really do f uncition as exudate 



'f'-r 



