178 ABDOMINAL PATTERN IN TABANIDAE (dIPTERa) 



which, while the dorsal side is of a clearly derived and much 

 simplified type, the underside shows a spotted pattern which 

 may well be compared with that of a Chrysops. However, there 

 are only three rows of spots, the median ones not much larger 

 than the lateral ones. I notice that while in most Tabanids the 

 underside of the abdomen is of a uniform coloration, a number of 

 species, as T. trimaculatus audits allies, showabroad, black, median, 

 longitudinal band. If our views are correct, this band represents 

 the last trace of an underside pattern in process of disappearance. 

 The banded, wasp-like patterns are quite frequent among the 

 South American forms, and Eisenbeckia mattogrossensis (fig. 28) 

 is only one of the less striking examples. If these views are cor- 

 rect, they have to be considered as derived, not primitive types. 

 The same will hold good for the Pongonia group. Here we sel- 

 dom encounter anything like a primitive pattern. Of the Brazil- 

 ian forms, Bombylomyia splendens (fig. 29) still shows an ill- 

 defined trace of light triangles on a yellowish-brown abdomen, 

 while the abdomen of Erephopsis nana (fig. 30) is almost uni- 

 formly fuscous, and that of Erephopsis albitaeniata (fig. 31) metal- 

 lic green with whitish borders of the segments. The almost 

 globular form of the abdomen in these bee-like species is itself an 

 indication that we have to deal with derived, not primitive forms 

 of abdominal coloration. 



