!N[r. Pt. E. Turner on Fossorlal Ilijmenoptera. 293 



cJ , Maudihlcs broad, with five teeth at the apex ; the 

 cly|)eus almost flat, Avith a very low carina. Head puuctured- 

 rngosc, the hind anj^les beneath prodnced into stout tnbercles, 

 somewhat longer than broad, slightly narrowed posteriorly, 

 the eyes separated from the posterior margin of the head by 

 a little more than their own length. Pronotum transversely 

 rugose, arched anteriorly, a little shorter than the nieso- 

 notum ; mesopleur^e rugose, dorsulum and scutelluni coarsely 

 punctured. ]\Iedian segment rugose, with a nari'ow longi- 

 tudinally striated space at the base, three longitudinal carinse 

 rather near together running from the base to the apex, the 

 space between them transversely striated. Abdomen shining, 

 almost smooth. Tarsal langues bifid, with a blunt lobe at 

 the base. Neuration as in subviolacea. The sides of the 

 median segment are strongly striated. 



Hub. ]\ilanje, Nyasaland ; January (S'. A. Neave). 



This species may be distinguished by the tubercles on the 

 head. The tarsal ungues resemble those of K. nyassica, but 

 the basal lobe is more strongly developed. The recurrent 

 nervure is received well before the first transverse cubital 

 nervure. The pronotum is distinctly transversely depressed 

 on the posterior margin. The wings are hyaline at the baf?e, 

 a fuscous cloud spreads itself apically from the stigma. The 

 species is near decemdentata, Enderl. (Arch, f, Naturg. 

 p. 213, 1901), but in that species the tegulfe and mouth- 

 parts are yellow-brown, the five apical joints of the antennae, 

 although much thinner than the preceding joints in tuher- 

 culata, are no longer, as is the case in decemdentata. The 

 sculpture is also different. Enderlein does not mention the 

 shape of the head in his species, but the shape of the man- 

 dibles is very similar. A specimen from Calabar in the 

 British Museum, which I identify as decemdentata with some 

 doubt, is only 10 mm. in length and has a much smaller head, 

 rounded at the hind angles and without tubercles, the pro- 

 notum is also without a transverse sulcus at the apex ; so 

 that if I am right in my identification, decemdentata is a 

 Kathejjyris. 



Mangesia bi'evicornis, sp. n. 



J. Niger, albo-pilosus; alis fusco-violaceis. 

 Long. 13 mm. 



J, Mandibles tridentate ; antennae stout and short, not 

 more than half as long again as the head, the second joint of 

 the flagellum as long as the first and third combined, joints 

 4-7 broader than long, the four apical joints very much 



