HENRY J. FRANKLIN. 183 



Bombidce Lovell, Ent. News, XVIII, 1907, p. 196. 

 Swenk, Ent. News, XVIII, 1907, p. 294. 



Cockerell, Ann. and Magaz. Nat. Hist., Ser. 8, Vol. II, 1908, 

 pp. 325 and 326. 



Dalla Torre, inhis "Catalogus Hymenopterorum," classifies 

 the bees or Anthophila (flower pollenizers) under fourteen 

 subfamilies of the great family Apidas. This arrangement 

 is the same as that proposed by Friese, and is, in most re- 

 spects, similar to the one previously proposed by Schmiede- 

 knecht and to the one followed later by Ashmead, except 

 that Schmiedeknecht and Ashmead placed the family Apidse 

 (in the large sense) in the position of a superfamily, and 

 made the subdivisions to rank as families. In Friese's and 

 Dalla Torre's arrangement, the genus Bombus was placed in 

 the subfamily Bombinae and the genus Psithyrus in the dis- 

 tinct subfamily Psithyrinse (established by Shenck in 1859). 

 Lepeletier considered Bombus as belonging to a family dis- 

 tinct from Psithyrus. Greene described species of Bombus 

 and Psithyrus indiscriminately as Bombus. Schmiedeknecht, 

 in his earlier work, and Ashmead looked upon Bombtis and 

 Psithyrus as belonging to separate families of coordinate 

 rank with the other families of the Apoidea as recognized by 

 them — Bombidae and Psithyridse. Robertson placed Bombus 

 and Psithyrus together in the subfamily Bombinae, Schmie- 

 deknecht, in his last comprehensive work on the Hymenop- 

 tera, placed Bombus and Psithyrus in the separate and distinct 

 subfamilies, Bombinae and Psithyrinae of the great family 

 Apidae. Viereck, followed later by Lovell, Swenk and Cock- 

 erell, placed Bombus and Psithyrus together in the family 

 Bombidae, but did not apparently recognize them as forming 

 separate groups of subfamily rank. After careful study of 

 the characters of the two genera and a comparative study of 

 the characters and relationships of other bee genera, I am 

 convinced that Viereck's arrangement, of all those thus far 

 published, most nearly expresses the true relationship of the 

 two genera, and that we must look upon them as being simply 

 two comparatively closely related genera of the family Bom- 

 bidae, if we are to adopt Ashmead's general classification of 



TRANS. AM. BNT. SOC, XXXVIII. 



