396 Mr. T. D. A. Cockeiell on new Hymenoptera 



stigma piceous. Third submarginal narrowed about one 

 half to raarginah A dark shade looking like a stump of a 

 nervure at lower outer corner of third submarginal. Pubes- 

 cence of legs wholly pale, hind legs not dilated, the white 

 hairs long on outer side, but not forming a broad dense mat, 

 as in other species. Claw-joints and basal iialf of claws 

 rufous. Claws deeply bifid. Abdomen with short greyish- 

 white pubescence, tolerably dense on first two segments, sparse 

 on the others, giving way to fuscous on 4 and 5, but the 

 hind margins with narrow thin white hair-bands. Apex 

 with white hairs. 



$ . — Length 12-14 millim. 



Very much like the female of ccesalpinice^ with the same 

 broad black brusli of hairs on the hind legs, the same black 

 abdomen, &c. The thoracic pubescence averages paler than 

 in casaJpinice. Eyes grey. Clypeus and labrum dull 

 orange-yellow, the upper corners of the clypeus more or less 

 black, frequently the whole upper margin of the clypeus 

 broadly black. Scape wholly black. ]\Iandibles rufous, with 

 black ends and base. Pubescence of abdomen very sparse 

 and black, first segment sometimes with pale pubescence, but 

 never the second. Claws small, Avith a little tooth within. 

 Wings smoky hyaline. Front broad, inner orbits parallel. 



Uab. College Farm, Mesilla Valley, N. M., at flowers of 

 CasaJjnnia Jalcaria, i\Iay 18, 18^6, 1 c? , 7 ? . 



1 had at various times swept over the flowers of Ccesalpinia 

 {IJqffma7iseffgia) falcaria, var. stricta (Benth.), without getting 

 any bees, and had wondered why this was. But on May 18 

 I observed that they were indeed visited by bees of the genus 

 Cenirt'sy which were much too agile to be caught by sweeping. 

 The C. cccsn/pinicey in its typical female, was especially iiard 

 to catch, hovering over the flowers with a curious swinging 

 motion, darting away at the least alarm. It was interesting 

 to find several species of Centris, a neotropical genus with 

 only one hitherto recorded U.S. species, at flowers of Ccesal- 

 pi7iia, a typically tropical genus of j)lants. 



Wlien I came to arrange all the specimens of Centris 

 caught in the Mesilla Valley, 1 found 1 had four different 

 males and four different females, presumably belonging to 

 one another. The females are much alike, but the males 

 present wider differences; and while 1 think I have the sexes 

 properly associated, it must be admitted that there is a certain 

 probability of error. 1'hree of the forms — ccvsalpi'nicp, rhodo- 

 j!>?<.s-, and Hofmauseggiw — are described above. Tlie tburtli is 

 C. lano.sa, Cresson, originally described in 1872 from three 

 males caught by Iklfrage in Texas. Cresson's description is 



