OF WASHINGTON, VOLUME XVI, 1914 31 



tubercles and with two dorsal rows of black dots, one on each seg- 

 ment. Head light ochreous with black eye spots and reddish 

 brown mouth parts. 



Cocoon ribbed, typical of the genus, pearly white, length 8-9 mm. 



BEES VISITING THURBERIA. 



BY T. D. A. COCKERELL. 



In August, 1913, Mr. W. D. Pierce collected bees from the 

 flowers of Thurberia thespesioides Gray, in Stone Cabin Canyon, 

 Santa Rita Mountains, Arizona. This plant, given in the Synop- 

 tical Flora as a synonym of Ingenhouzia triloba D. C., is so near to 

 GoNxi/])ium that it was once described under that generic name. 

 On this account any insects frequenting it are of more than ordi- 

 nary interest. The bees collected are as follows: 



Melissodes thurberiae n. sp. 



Female. Closely allied to and resembling .17. //////;//// (All., to which 

 it runs in my table in Trans. Amer. Ent. Society, 1906. It differs from 

 thelypodii by the pale hair of thorax above (which agrees in character and ar- 

 rangement with that of M. martini, except that there is no black hair) ; tin 

 wings darker and redder; the tegulse piceous, with the posterior margin 

 broadly ferruginous; scutellum with a slight median longitudinal ridge. 



The disc of mesothorax has considerably smaller and closer punctnn'- 

 than M. martini Ckll., and they run principally in transverse lines. 

 The same characters, and the dark teguhr, readily distinguish it from .17. 

 hitei Ckll.. Although the hair of thorax above is creamy white, then- i- 

 a little orange tuft on base of wings. White hair appears at extreme sides 

 of fifth abdominal segment, whereas in .17. hitci the hair in this place is 

 black. Head very broad; vertex in type with only one dark hair. Length 

 of anterior wing 11| mm. 



Type: Cat, No. 16845, U. S. N. M.. Collected OH August 26. 



Melissodes communis Cresson. 



Female. Differs from a cotype by smaller si/.e, and darker stigma and 

 nervures. The single specimen is in bad condition; probably a series, well 

 preserved, would indicate a distinct subspecies. Collected August !'.">. 



Perdita mentzeliarum Ckll. 



I cannot distinguish these from the variable species P.mentee- 

 liarum, which usually visits Xuttulln \ M, ///:> lia Aucit. I. Perhaps 

 they are strays from adjacent Xnttnlin flowers. Two female 

 specimens August 27. 



