XI. D, 5 Cockerell: Ceratinid Bees 305 



Ceratina philippinensis nigTolateralis subsp. nov. 



Female. — Rather small; length, 5 to 6.5 millimeters; yellow 

 markings smaller than in C. philippinensis, no yellow spot on 

 pleura behind tubercles. The two specimens received are pos- 

 sibly of different species, but probably indicate varieties only; 

 they differ thus: 



(a) nigrolateralis, type. Length, 6.5 millimeters ; lateral face . 

 marks continuous; mesothorax with four short linear marks; 

 second submarginal cell broad; first abdominal segment black 

 with three rather small yellow marks {Baker 3837) . 



(b) humilior var. nov. Length, 5 millimeters; lateral face 

 marks divided in middle, but upper division elongated (broad 

 median clypeal band as in philippinensis) ; mesothorax entirely 

 black ; second submarginal cell very narrow ; first abdominal seg- 

 ment yellow with a black mark on each side {Baker 3836) . 



Palawan, Puerto Princesa. 



This species is close to C. moratvitzii Sickmann, from Formosa, 

 agreeing in the sculpture of the mesothorax. It differs from 

 morawitzii by having, in the typical nigrolateralis form, only 

 the fifth abdominal band entire, and in the humilior form the 

 segments beyond the third without bands. I find, however, that 

 morawitzii varies in the abdominal banding, so that, apart from 

 geographical considerations, I would be inclined to call the Pala- 

 wan insects varieties of morawitzii rather than of philippinensis. 

 If the philippinensis type of Ceratina is derived from moi-awitzii 

 and reached the Islands via the Malay Archipelago and Borneo, 

 it is not surprising to find an intermediate form on Palawan. By 

 the sculpture of the mesothorax, typical philippinensis is to be 

 associated with C. morawitzii and not with C. hieroglyphica. 



In this whole series we have to do with variabl|k characters 

 which here and there reach conditions of relative stability, per- 

 mitting us to separate local races or species ; but it may be that 

 large series will show that the ranges of variation overlap or 

 that the normal characters of one form are represented by occa- 

 sional varieties of another. An easy way out of the difficulty 

 would be to call all these insects by one name and simply remark 

 that the species is variable; but this would obscure the very 

 facts which we are most anxious to investigate. An intensive 

 study of such a series may show that in the formation of allied 

 races or species there has been no modification of the determiners 

 or genes whatsoever, but that they have simply been sorted out 

 in various combinations for heterozygous forms. Thus, sup- 

 posing that nigrolateralis and humilior, described above, are 



