232 NEW YORK STATE MUSEUM 



of Other forms. The one within the discal cell occurs in Hoplolabis, 

 Plusiomyia [pi. 17, fig. 2], Ula [pi. 25, fig. 5] and as a spur in 

 others. The one in cell M (the second basal cell, of some systematic 

 dipterologists) occurs in Ephelia, Polyangaeus, etc. The one in 

 cell M3 occurs in Idioplasta. The one in cell A^ occurs in Discobola. 

 The probable significance of all these supernumeraries has already 

 been discussed. I consider them lone survivors of a more abundant 

 venation. But their widely scattered recurrence inclines me to 

 believe in a very remote origin of genera and groups of genera in 

 this family. 



And there is among them one that is very peculiar. Apparently 

 it is not in the way of disappearing, but of redevelopment. It is the 

 one that occurs in Hoplolabis armatain cell 1st M^. There 

 is here a curious atrophy of the base of M^+^, just beyond the first 

 fork of the median vein and a creeping of all the vein tips around 

 toward the wing apex, and a most curious conformation of the discal 

 cell [pi. 2^, fig. 5, 6]. It acquires a reentrant angle from which a 

 spur starts. These things are foreshadowed in Acyphona [pi. 23, 

 fig. 4]. I studied the variability of this spur in some 50 wings of 

 this species, and found it to exhibit all degrees of length from a 

 complete cross vein down to a spur one third as long, it being usually 

 about half as long. This seems to be a relatively new acquisition 

 that accompanies the reentrant angle that probably meets some new 

 need due to the movements of veins and shift of stresses. 



The median cross vein rarely disappears by the fusion upon it 

 of the veins it connects [as in Phalacrocera pi. 11, fig. 2, and in 

 Idiophlebia figured by Griinberg in Zool. Aug. 1903. 26:525]. It 

 gets curiously reduced in length and reversed in position in Palaeo- 

 poecilostola [pi. 20, fig. 4]. 



Principal readjustments of venation in the Tipulidae 

 We come now to note the correlated behavior of veins and cross 

 veins in this family. The principal shifts of veins that clearly show 

 correlated movements of many elements of wing structure are two: 

 (i) the formation of the cord and (2) the upshift of veins Cu^ 

 and M^^. Let us consider these somewhat in detail. 



Formation of the cord. As already defined, the cord is the 

 principal line of transverse joinings of the veins that traverse the 

 wing disk. It is always composed of at least three forks and two 

 intervening cross veins. These are the first fork of the radial sector, 



