218 PROCEEDINGS OF THE ACADEMY OF [March, 



scape long, flagellura brownish beneath and delicately pubescent; 

 tegulffi very dark brown; base of metathorax with oblique radi- 

 ating strife; Avings hyaline, the costal margin, including the 

 marginal cell, very broadly dark fuliginous (as in some Tachi- 

 nidaj); legs very dark reddish-brown; abdomen with long white 

 hairs beneath; puuctures of first dorsal segment of abdomen 

 stronger and less dense than on second, which has the punctures 

 minute and as close as is possible; hind spur of hind tibia pecti- 

 nate, with three large teeth. 



Hah. — Chapada, March and April; five specimens. Very close 

 to the Mexican C discolor (Smith), thus adding another to the 

 now rather numerous instances of Brazilian bees representative of, 

 but not identical with, those of Mexico or the adjacent parts of 

 the United States. 



This might be held to differ subgenerically from the type of 

 Corynura, but in that case Cacosoma Smith is not available, be- 

 cause of the prior Cacosoma Felder, 1»74. 



C. atromarg Inata is one of five species flying at Chapada, all 

 having the first recurrent nervure interstitial with the second 

 transverso-cubital, and the mesothorax very densely punctured. 



These species are readily distinguished as follows : 



Costa fuliginous from base to marginal cell, 



atromarginata n. sp., 9. 



Costa not fuliginous from the base, or not at all, .... 1. 



1. Abdomen long and narrow, clavate, like Baccha clavata, . 2. 

 Abdomen oval or suboval, hardly or not clavate, ... 3. 



2. Head and thorax bright golden-green, pseudobaccha n. sp., cf. 

 Head and thorax dark, suliused with olive-green, 



jucmida (Smith), cf. 



3. Marginal cell fuliginous, .... semimarginatan. s])., ?. 

 Tip of wing only fuliginous, .... chapadicola u. sp., 9. 



Corynura jucunda (Sinith'i; n. syn. C. senigma (Gribodo). 



Hah. — Chapada, December; 2 cJ*. So far as I can make out, 

 C. jucunda and cenigma are the same and identical with the insect 

 now before me. This gives it a range from S. Paulo and Cha- 

 pada, Brazil, to Rioja, Argentine Kcpublic, the extremes being 

 about 1,250 miles apart. In view of what Sichel states about the 

 two sexes of Corynura, it seems likely that C. semimarg inata or 

 C. chapadicola may prove to be the 9 of C. jucunda, but as I 

 have no proof of this, I treat both for the present as distinct. 



