226 NEW YORK STATE MUSEUM 



very puzzling; for such characters as this have been found 

 elsewhere in the Diptera most reliable. But it is to be noted 

 that in all these cases of departure from the assumed Tipulid 

 type of forking of the sector, the branches are all long and 

 the forks lie close together, and a slight migration of either of 

 the three branches would change the relations of the forks to 

 each other. The accompanying diagram illustrates these dif- 

 ferences of behavior of the radial sector. There are some 

 reasons for believing that these peculiar forms are only apparent 

 exceptions to the rule of Tipulidae. For example, the position 

 of the r-m cross vein in Pedicia, Amalopis and Rhicnoptila, indi- 

 cates that in these forms at least, after the complete fusion of R* 

 and R^ a further fusion of R'*+^ with the part of the sector 

 immediately before it, brought about the apparent reversion of 

 the fork; for, elsewhere when the sector is three branched, this 

 cross vein touches the posterior branch after its separation from 

 the middle branch but here it touches before this separation [fig. 

 14^]. In other words, the r-m cross vein remains in its original 

 position, while fusion has carried the fopk past it. A reason for 

 that further fusion may possibly have lain in the arcuation of the 

 wing apex in these genera. 



In Gynophstia [pi. 20, fig. i] the fusion has proceeded only 

 to the level of the cross vein, and the two forks are of equal 

 depth. In M o 1 o p h i 1 u s (not Erioptera) comatus 

 Doane [N. Y. Ent. Soc. Jour. vol. 8, pi. 7, fig. 20] the forks 

 are symmetrically arranged about the middle branch [like / in 

 fig. 14]. Obviously this condition, although intermediate, is not 

 primitive, but secondary, and has come about through migra- 

 tion of one or the other or both of the forks. 



To the other cases in which this exceptional mode of branch- 

 ing of the sector occurs, Conosia [pi. 21, fig. 5] and Molophilus 

 [pi. 22, fig. 6] the same reasons for this second fusion will not 

 apply : the wing tip is not arcuated, but straight. But in both 

 of thes^ the r-m cross vein is located unusually far out from 

 the base of the wing, especially in the former, and this second- 

 ary fusion may here be an accompaniment of the upward 

 skewing of the sector, and the unusual relations borne by the 

 elements of the cord [fig. 14/i]. 



Professor Comstock pointed out that the two modes of 

 branching of the sector are differences of kind, and he showed 

 that by further reduction either kind of three branched sector 



