REPORT OF THE STATE ENTOMOLOGIST I907 227 



may become two branched. And, if the two branched sector 

 at c in figure 14, and the one at e, exactly hke it, be descended 

 as indicated, they are less closely related to each other than 

 either is to the three branched sector that stands before it in 

 the diagram. Yet when the two types of branching are hover- 

 ing about the parting of the ways, as they seem to be in the 

 cases just cited, we see that even differences of kind are in 

 their beginnings merely differences of degree. 



This marked variability in a character that is elsewhere so 

 constant agd so important surely indicates that the crane flies 

 are a very generalized group. In this, and in other characters 

 as well, the main characters of the venation of the Diptera 

 were not firmly fixed. The crane flies stand near the foot of 

 the series, where venational experiments abound, and where the 

 main trends of specialization are not everywhere fully estab- 

 lished. 



A further reduction of the branches of the sector has occurred 

 in many crane flies [see pi. 2y, 28]. The two branched con- 

 dition has been reached, apparently by the fusion from the 

 base outward to the wing margin of R^ and R^, as illustrated 

 by Gonomyia [pi. 24, fig. 4, 5] and Cladolipes : also by the 

 atrophy of vein R^, as indicated by Lipsothrix [pi. 20, fig. 5] 

 and some Tipulinae. Apparently R^ and R^ have both atrophied 

 in Toxorrhina [pi. 29, fig. 5]. There has also occurred a 

 noteworthy, fusion from the tips backward of R^ and R^+^ in the 

 Cylindrotomini [pi. 15, fig. 4, 5, 6]. Certain other fusions involv- 

 ing also the cross veins, will be considered after the cross veins 

 have been discussed. 



The median vein is typically twice dichotomously branched, 

 like the radial sector. Yet even more rarely does it show the 

 full complement of branches. These were apparently fully de- 

 veloped in the fossil crane fly Rhabdinobrochus [pi. 17, fig. 2] 

 from Florissant, and all four are represented in a living unde- 

 termined Tipuline shown in [pi. 17, fig. i], although M^ is repre- 

 sented only by a spur of a vein. Usually there are but three, 

 or two branches : rarely there remains but one [as in Diotrepha, 

 pi. 29, fig. 6]. The branches of the two main forks of media 

 tend to be reduced by different methods ; those of the anterior 

 fork tend to fuse from the base to the wing margin; those of 

 the lower fork to disappear by atrophy. I have not found any 

 evidence of atrophy in the anterior fork, nor of fusion in the 

 posterior one. . A number of the larger genera, like Limnophila, 



