568 THE PANORPOID COMPLEX, iii., 



Psychopsidae, I turned to that family for further evidence. The 

 result was the discovery, both in Megapsychops illidgei and in 

 Psychopsis elegans, of the vein M 5 in its original position, but 

 not strongly connected with Cu lt This condition is shown in 

 Text-fig. 45«. The cubito-niedian Y-vein is not fully formed, in 

 so far as M 5 weakens distal]}' as it approaches Ci^. M 5 is also 

 more parallel to Ci^ than in the normal formation, and is some- 

 times connected with it by one or more cross-veins. 



It is not possible to decide definitely whether the condition 

 found in the Planipennia is a reduction from a more archaic 

 stage than that seen in Belmontia (Text-fig. 41b), or whether 

 it is due to the effect of a large addition of branches of Rs 

 crowding the older veins together, so that M 5 and Cu r though 

 originally forming a Y-vein, have become pressed together in 

 the manner now shown. The evidence from the hindwings 

 would incline us towards the latter supposition, but the point 

 is by no means certain. As the Megaloptera and Planipennia 

 are more archaic than Belmontia in certain other venational 

 characters, such as the complete system of costal veinlets, it 

 may well be that the condition of M 5 in the forewing is really 

 due to debasement of this vein from an older stage in which 

 it ran freely between M 1—4 and Cu 1( without having attained 

 any definite fusion with the latter vein such as we find in 

 Belmontia. 



An examination of the hindwings of various archaic types in 

 the Order Planipennia shows that M 5 is generally present, and 

 easily recognised owing to its oblique position, and the very 

 evident forking of M which gives rise to it. Thus, we may 

 take as an example Osmylus (family Osmylidae), in the hind- 

 wing of which (Text-fig. 45, d) the median fork is particularly 

 well developed, with M 5 as a very strong oblique vein descend- 

 ing on to Cu 1% In the Psychopsidae, the trachea supplying 

 M 5 persists in the imago, and passes some distance along Cuj, 

 thus offering an interesting parallel to the case of the Lepi- 

 doptera already discussed. In the American genus Polystoe- 

 chotes (family Polystoechotidae) the trachea traverses the base 

 of 1V1 only, then leaves it, and runs freely in the membrane 

 between M and Cu for a considerable distance (Text-fig. 45, c). 

 This is an interesting survival, in view of the condition found 

 in the forewing of Psychopsidae, dealt witli above. 



In the Planipennia in general, the chitinisation of the base 



