1922] Hymenopterous Genus Harpagocryptus and its Allies. 107 



on the basis of the simihirity of the genitaHa of the male and the 

 number of antennal joints in the two sexes which is the same as 

 that prevaihng in Rhopalosoma and most aculeate Hymenoptera. 

 There is also a curious similarity in the habits of Rhopalosoma and 

 Harpagocryptus. Hood ('13) has shown that the larva of the 

 former lives as an external parasite on the jumping tree cricket, 

 Orocharis, while Bridwell's Harpagocryptus was reared from an 

 Australian cricket of the family Trigonidiidse on which the larva 

 forms a sac like that of certain Dryinidse.' This habit would, 

 however, not give anj^ reason to associate Harpagocryptus with 

 Rhopalosoma rather than with the Drynidse. 



I am unable to reconcile the differences between Rhopal- 

 osoma and the genera here discussed sufficiently to assign them 

 to the same family. The head in both sexes of Rhopalosoma is 

 thin and strongly transverse, the eyes and ocelli very large and 

 the front is not produced anteriorly. The thorax has the pro- 

 notum very short and collar-like and absoluteh^ different from 

 that of Algoa, et al. The propodeum is elongate-oval, not 

 truncate nor sharply declivous behind; the abdomen has an 

 extremely long petiole; the femora are only slightl}^ thickened 

 and the middle coxse are approximate (widely separated by the- 

 mesosternum in Algoa). Such divergence, particularly in the 

 form of the prothorax, head and propodeum is certainly of great 

 importance, although the reduction of the eyes, mesothorax and 

 scutellum is usually encountered in wingless or subapterous 

 Hymenoptera. 



Ampulicimorpha Ashmead, referred by him to the Embo- 

 leminse does not show any great similarity to Algoa except in 

 the general form of the head and thorax and the external male 

 genitalia which resemble those of Olixon as described by Kieffer. 

 On the other hand, the peculiar genus Sierolomorpha (placed by 

 Ashmead ('03) in the family Cosilidse), resembles Algoa quite 

 closely in abdominal structure, in the general form of the thorax 

 and head, thickened legs and antennEe(cf 13-jointed, 9 12jointed)>. 



'This insect is evidently the undescribed Embolemid mentioned by Perlvins '05 (footnote, 

 p. 27) as having been reared from "small cricketes of the genus Trogonidium or allied forms." 



'Ashmead knew only the male, but several years later (Brues '05) the present writer found 

 the female of this interesting insect). 



