2 Myron Harmon Szvenk 



Stelidae {Transactions American Entomological Society, XXVI, 

 p. 79), in which family he also included the subfamily Coelioxynae 

 which is referred by the writer to Megachilidae. Cockerell refers 

 both the Stelidinae and Anthidiinae, as here considered, as sub- 

 families of Megachilidae along with the subfamilies Megachilinae. 

 Osmiinae, Coelioxynae and Dioxynae {University of Colorado 

 Studies, VII, pp. 184-185), but the writer would restrict the 

 Megachilidae to the four last named subfamilies. 



The subfamily Stelidinae is composed of the parasitic bees 

 referable to the old genus Stelis. Our American forms of this 

 family fall into two groups, based on the venation, in one of 

 which the second submarginal cell receives both recurrent nerv- 

 ures while in the other the second recurrent nervure is inter- 

 stitial with or received beyond the second transverse cubital nerv- 

 ure. The first type of venation is confined to Nearctic species, 

 while the second type is common to Pal?earctic and some Nearctic 

 species. The chief objection to its employment as a generic char- 

 acter is that in some species, which usually have the second re- 

 current nervure before the second transverse cubital nervure, 

 specimens occur in which these nervures are opposite each other 

 {e. g., foederalis), while other species have these nervures either 

 opposite or the second recurrent nervure beyond the second trans- 

 verse cubital nervure {e. g., lateralis), so that there is no marked 

 defining line between the two groups, and they had apparently 

 better stand as subgenera. The first type of venation occurs in 

 Robertson's alleged genus Stelidium (type S. trypetinnm), and 

 characterized Ashmead's genus Melanostelis {Psyche, VIII, p. 

 283, Nov. 1898), the type of which, M. hetheli Ashmead, is a 

 synonym of Stelis rubi Ckll. {Entomologist, July 1898, pp. 167- 

 168) . In 1888 Provancher proposed the monobasic genus Chelynia 

 {Additions a la faune Hymcnoptcrologique du Canada, p. 321), 

 and Titus, after examining Provancher's type of Chelynia labiata, 

 found that it was the same as Cresson's species nitida, which he 

 doubtfully referred to Stelis {Transactions American Entomol- 

 ogical Society, VII, pp. 92-93, 1878). S. nitida Cresson and 

 5'. rubi Ckll. both have the first type of venation and are prob- 

 ably congeneric, although rubi shows a unique form of pygidium 



2 



