SOME BEES FROM FORMOSA. 13 



clypeus normal not keeled ; mandibles red subapically ; front, vertex, 

 mesothorax, and scutellum granular from minute dense punctures ; 

 basal sulcus of metathorax with fine transverse ridges ; posterior face 

 of metathorax dull and rough, with much hair ; tubercles, upper 

 border of prothorax and postscutellum densely covered with hair, but 

 surface of mesothorax not hidden by hair ; tegute clear ferruginous ; 

 wings dusky, nervures and stigma pale ferruginous ; b. n. not quite 

 reaching t. m. ; abdomen very finely punctured ; hind margins of 

 segments testaceous, with heavy broad fulvous hair-bands, widely 

 interrupted on first segment, thin in the middle on second ; apical 

 ventral segment broadly rounded. 



Hob. Koroton, Formosa, September 8th, 1907, two females. 

 N. takauensis, Friese, was based on males from Takao and Ku-sia, 

 Formosa ; it appears to resemble N. mediorufa, having the same 

 fulvous pubescence, but the abdomen is not red at base, and the 

 mesothorax and scutellum are much more densely hairy. I do 

 not think takauensis can be the male of mediofura. In Bingham's 

 table N. mediorufa runs near N. floralis, Smith, but the clypeus 

 is produced at apex into a short truncate lobe, not toothed as in 

 floralis ; the hair of face is pale fulvous, not fuscous as in 

 floralis ; and the scutellum is closely punctured, while in floralis 

 it has scattered punctures. N. maciditarsis, Cam., may also 

 be compared, but that species has smooth silvery white apical 

 margins to the abdominal segments. 



Nomia mediorufa gyammensis, n. 



5 . Like N. mediorufa, but abdomen not at all red basally ; legs 

 darker, hind tibise with dark, fuscous hair on outer side ; pubescence 

 paler, the abdominal bands very pale ochreous ; flagellum black 

 above except at apex. 



Hah. Gyamma, Formosa, April 6th, 1907. This is almost 

 exactly like the Australian N. halictella, Ckll., so much so that 

 without very close scrutiny they would be considered identical. 

 N. halictella has the wings distinctly darker, and is a larger insect, 

 but the resemblance is most extraordinary, extending to every 

 microscopical detail. It is a question whether N. gyammensis 

 and medionfa should not be regarded as races of halictella, in 

 spite of the wide geographical separation. Among the Indian 

 species N. gyammensis may especially be compared with N. 

 kangrcB, Nurse, in which, however, the pubescence hides the 

 sculpture on front, and there are other differences. It is easily 

 known from N. opposita, Sm., by the smaller size and light 

 nervures, and from N. fruhstorferi, Per., by the yellowish hair- 

 bands on abdomen, &c. In Bingham's tables it runs nearest to 

 N. aurifrons, Sm., which is very different. 



