740 



STUDIES IN AUSTRALIAN NEUROPTERA, 



(1). The course and limits of the subcosta (Sc), radius (R), 

 radial sector (Rs), and analis (lA, 2 A) are the same as in the 

 generally accepted scheme (Text-fig. 3), and need no further con- 

 sideration. 



(2). The supposedly simple median vein is really tim>-hranched^ 

 as in most Hemerohiidit^, 0.^mylid(t', and Xymphidiv. In the tra- 

 cheation, the thicker and straighter of the two branches is the 

 lower one, Mj, from which the weaker Mj arches weakly upwards, 

 to run above and neai'ly parallel to M^. The upper branch of 

 the cubitus (Cuj), which forks at a level slightly proximal to that 

 at which M forks, is also weak, and arches similarly upwards, to 

 meet M->jnst distally from the fork^ and there unite with it. 



Text-tig. 5. — Portion of Text-fig. 4 enlarged ( x 50), to show the formation 

 of the oblique vein 0, and the cubito-median,Cuj -}- M2. Compare 

 Plate Iviii., fig. 2. (Original). 



Now, in the imaginal venation, a peculiar transformation 

 occurs. The strong, straight, lower branch of the media. Mo, 

 becomes sharply bent downwards at the fork, thus forming the 

 oblique vein O, whose persistence had so puzzled me before. This 

 lower branch Mo consequently unites with the much weaker Cui 

 at the lower end of O. At that point it bends away distad 

 again, exactly aontiyiuiny the line of Cuj itself No wonder, then, 



