188 T. D. A. COCKERELL. 



nearly in a straight line with them. The tongue is quite as in the 

 long-tongued bees, but only moderately long. The apex of the <? 

 abdomen is bispinose, recalling Oxim. The $ antennae are strongly 

 clavate, but those of the 9 would do very well for an Andrena. The 

 marginal cell is obliquely truncate; the first recurrent nervure 

 meets the second transverso-cubital as in Protoxcea. The female 

 abdomen looks like that of an Andrena. All of this beautifully 

 connects the Anthophorid bees with the Andrenoid and other primi- 

 tive types, which we concluded to be derived from ancestors allied 

 to the Scoliidse. 



STEGANOMUS Ritsema and CTENOI'LECTRA Smith. 

 These do not belong to the Megachilidse according to their 

 authors, Mr. Vachal remarks. Smith thought Ctenoplectra near to 

 Macropis, remarking : "The posterior legs have a dense clothing or 

 pollen brush as in that genus." Steganomus was separated by Rit- 

 sema on a % , "he said that his genus was closely allied to Nomia, 

 that is, a Nomia with two cubital cells" (Vachal). Smith put Oya- 

 thocera (== Steganomus) in the Andrenidae near Nomia, and said of 

 the 9 that the posterior legs have the tibiae and basal joint of the 

 tarsi furnished with a dense scopa. Mr. Vachal thinks Mr. Ash- 

 mead was misled by what appears to be a typographical error in 

 the table in Bingham's work on the Hymenoptera of India, whereby 

 these genera appear to go with those having an abdominal scopa. 

 Mr. Vachal adds that the species of Ctenoplectra from Africa (C 

 antinorii Gribodo) which he has before him has the three last seg- 

 ments of the abdomen fringed. 



EUASPIS Gerstaecker. 

 Mr. Vachal has the 9 of the two species of this genus, and says 

 both are without scopa, ventral or tibial, so the genus should go in 

 the Stelidinae. 



ALLOD4PE Lepeletier. 



Mr. Vachal states that this has a tibial scopa in the 9 ; he thinks 

 it belongs next to Ceratina. 



MACROPIS Panzer 

 This is not a Panurgid, according to Mr. Vachal ; he says it is an 

 isolated genus of uncertain affinities. It appears to me to be a 

 modified Andrenid, but it wholly lacks the lateral facial depressions 



