J. CHESTKK BKADI>KY. 133 



Tlie wings of the subfamily are of special interest, inasmuch as 

 we find in them a very complete series showing specialization by 

 atrophy, from the condition found in Evania as the most generalized 

 to that found in Evaniellus at the other end of the group. The wing 

 of Evania appendigaster (Fig. 76) is as generalized as any I know 

 of in the subfamily, with the exception of a few which show Rj 

 more distinctly (see Fig, 75). C and Sc + R + M are separated as 

 in Aulacinse and Gasteruptioninse, forming a distinct cell C; R3 

 after separating from R4 bends either obtusely or acutely or at right 

 angles upward, and reaches the margin from one-third to two-thirds 

 the distance from the stigma to the apex, but the angle at which it 

 bends seems not to be entirely constant within a species. As has 

 been stated R4 is present in the most generalized type (Fig. 75), but 

 it is always partly atrophied, and usually there is only a stump or 

 no trace at all of it left. R5 is never present. M separates from R 

 within or close to the stigma, running almost backward till it 

 reaches m-cu when it turns at right angles; m-cu is thus very long, 

 and the cell R -f- 1st Ri is diamond shaped. But in Zeuxevania 

 (Figs. 80-81) M separates from Sc + R about two thirds of the 

 way from the base of the wing to the stigma, or the base of M is 

 entirely wanting, if present it runs backward a short distance and 

 joins m-cu, this vein being much shorter than in Evania, and con- 

 tinuing in an unbroken curve with the longitudinal part of M. 

 After R4 the next vein to become atrophie<l is Mi^o 'iu<l Mi at its 

 base, and later along its whole length together with m and the longi- 

 tudinal part of M2 (Figs. 76-80). A trace of tliese may generally 

 be seen except in Evaniellus, Hyptia, Semceodogader and Semceo- 

 myia (Figs. 84-87) ; in Evaniscus (Fig. 82) the longitudinal part of 

 the base of M (from m cu to separation of M3-1-4) is lost; Zeuxevania 

 (Figs. 80 and 81), which has been described, seems to be the next 

 modification, and along a diflferent line, the only instance which 

 involves modification of the position rather than atrophy of the 

 nervures. In Sema^odogaster and Semceomyia Rg, all of INI beyond 

 m cu except a longitudinal vein representing a portion of M4 and 

 Ms, and except M4^,Cui, and M4+CU1 + 1st -|- 2nd -f 3rd A are 

 lost; so that we have only three closed cells left, namely C, ^I and 

 Cu + Cui; the next step is in Hyptia (Fig. 86), where only C, Sc 

 -j- R -f M, the stigma, a trace of R and Rg, Cu and ^Ii remain. 

 The forewings are also broader than in the other genera, except 



TRANS. AM. ENT. SOC. XXXIV. APRIL. 1908. 



