186 PROCEEDINGS OF THE NATIONAL MUSEUM. tol. 55. 



fihort of transverse median a distance at least as great as length of 

 transverse median, but a lesser distance in the OedisceUs. The 

 females of the two species agree in having little hair on hind legs. 

 flylaeosoma ashmeadi female has second submarginal cell large, with 

 first recurrent nervure falling only a little short of it, and basal 

 nervure does not fall far short of transverse median. 



O. styliventris male has long antennae, and first recurrent nervure 

 falling short of second submarginal cell. O. inermis male has simi- 

 lar characters, but hind femora angled below. The face marks of 

 male inermis and styliventris are very different. 



It thus appears that Oediscelis contains quite diverse elements, 

 and so far as external characters go it is difficult to separate it from 

 Hylaeosoma. It has not been possible to compare the mouth parts or 

 genitalia. Various generic names have already been proposed (Pseu- 

 discelis for O. rostrata Friese, Protodiscelis for O. -ftebrigi Brethes, 

 Prosopoides for 0. paradoxa Ducke) for species of Oediscalis taken 

 in the broad sense, and if all these bees are regarded as belonging to 

 a single genus it will be hard to exclude Hylaeosoma. 



Ducke cites Chilicola Spinola (1851) as doubtfully identical with 

 Oediscelis. If the two are identical, Chilicola has priority. I have 

 a pair of bees which I received as Chilicola pleheja Spinola, but they 

 do not agree with Spinola's diagnosis of the genus, and are, in fact, 

 Panurgids, closely allied (at least) to Panurginus herbsti Friese. 



If Hylasosoma is regarded as a subgenus of Oediscelis it will ap- 

 parently be necessary to restrict it to the West Indian H. longiceps. 



CENTRIS HAEMORKHOmALIS (Fabricins). 



St. Andrew, Jamaica, April 1898 (C. B. Taylor). 



CENTRIS ELEGANS Smith. 



Male. — Windward side, St. Vincent, West Indies (H. H. Smith). 

 Smith described the female only. The male is like that of C. haemor- 

 rhoidalis, with the same face markings, but has the hair on and about 

 the postscutellum reddish ochreous, the face markings (perhaps al- 

 tered by cyanide) more decidedly yellow, the stripe on anterior tibiae 

 reduced to two spots (representing the ends), the marks at sides of 

 second abdominal segment transversely elongate, the apex of abdo- 

 ;men not red, and the hair on hind basitarsus all black. 



CENTRIS ELEGANS GRENADENSIS, new subspeciei. 



Male. — Grenada, West Indies (H. H. Smith). Differs from the 

 St. Vincent form thus : Face marks creamy white ; stripe on anterior 

 tibiae well developed, except basally, where it fails except for a basal 

 spot; scutellum ferruginous; much fulvo-ochraceous hair behind 

 wings; tegulae ferruginous; hair on hind tibiae and tarsi dark red- 



