140 The Philippine Journal of Science i9i8 



of thorax white, on anterior part of mesothorax thin and 

 white, the tuft below tegulse slightly yellowish; antennae 

 long and slender, black; vertex densely punctured; mesothorax 

 and scutellum very densely and minutely punctured, but glisten- 

 ing; tegulae black; wings dusky; abdomen rather short, shining, 

 with feeble white hair bands, the basal ones very thin, the apical 

 only at sides; keel of sixth segment feebly developed, broadly 

 rounded, obtusely emarginate, with a depression above the emar- 

 gination. Anterior coxae without spines. 



Luzon, Laguna, Mount Maquiling (Baker). This and the 

 next are named after naturalists whose contributions to the 

 knowledge of the Philippine biota can never be forgotten. 



Megachile mcgregori sp. nov. 



Male. — Length, about 10.5 millimeters ; similar to M. merrilli, 

 but larger and also differing thus : Hair of thorax above warm 

 ocherous, without black; tegulae ferruginous; wings with a dis- 

 tinct reddish tint; punctures of mesothorax quite distinct under 

 a lens ; anterior tarsi entirely black ; keel of sixth abdominal seg- 

 ment broadly emarginate. The anterior coxte are without spines. 



Luzon, Laguna, Mount Maquiling (Baker 7Jf50). 



Genus ANDROGYNELLA Cockerell 



Female wholly without ventral scopa, but with sting well 

 developed; antennae of female 13-jointed and anterior coxae 

 spined, as in male. 



Androgynella subrixator sp. vel forma nov. 



Female. — Nearly 9 millimeters long, with the general character 

 of Megachile subrixator- from the same locality, but abdomen 

 smooth beneath, without any ventral scopa, though the form 

 of the abdomen is that of a female and the sting is present; 

 anterior parts' including head, as in male, face covered with 

 yellow hair, antennae long and 13-jointed, and anterior coxae with 

 sharp spines. A second specimen, however, has shorter antennae, 

 which are only 12-jointed. 



Mindanao, Davao (from Baker). I should regard these as 

 simple gynandromorphs of Megachile subrixator, which they 

 may be, but for the fact that a similar type {Androgynella detersa 

 Ckll.) is established as a genuine species in Australia. I have 

 discussed this subject '' and have shown that the characters in 

 the Australian form are constant in a considerable series. In 



•■Ann. & Mag. Nat. Hist. (1911), VIII, 7, 314. 



