552 THE PAXORPOID COMPLEX, iii.. 



Secondly, a study of the Orders Megaloptera and Planipennia 

 (Section xv., xvi.) shows us that the evolution of the pectinate 

 type of branching of Rs was probably at first confined to R 2+ 3, 

 the dichotomic condition of R 4+5 remaining in these insects 

 long after the rest of Rs had become pectinately branched.* 

 Thus we are driven back to examine R 2+3 only, and to try to 

 discover what its original condition was, before it became pec- 

 tinate . 



A study of the wings of archaic Megaloptei'a and Planipennia 

 must convince us that there were more than two original dich- 

 otomic branches of R 2+3 concerned in the formation of the 

 pectinate type. All the fossil Planipennia show a considerable 

 number of pectinate branches in this region, and the same is 

 true of most of the Cory&alidae, all the Ithonidae .and Psychop- 

 sidae, and most other archaic types. Even in the very reduced 

 Sisyridae (Text-fig. 110), it must be clear that at least three 

 branches of R 2+3 took part in this formation. Probably the 

 actual number of branches varied according to the size of the 

 species affected. As an illustration of the manner in which 

 pectination of Rs could be brought about, without loss of the 

 original dichotomy of R 4+5 , from a many-branched type, we 

 have selected the actual type of branching found in Archipan- 

 orpa (Text-fig. 39), and have shown how, by very simple 

 changes, the original dichotomies of R 2+ 3 can be altered to the 

 pectinate arrangement . 



As the point at which Rs first forks is of great importance, 

 we have called it the secondary radial fori-, with the notation rf<. 



The Media. 



The original first dichotomy of this vein was clearly that into 

 M. 1 _ i and M 5 (Section iii.). This has hitherto been over- 

 looked. The point at whieh this forking takes place is the true 

 primary median fork, and should carry the notation mf. The 

 point at which M 1-4 divides into M 1+2 and M 3+4 , hitherto 

 known as mf, must now be called the secondary median fork, 

 and must take the notation mf'. 



The evidence is clear that, in the three Orders Mecoptera, 



*Note, however, that, if R 5 itself were originally branched, as in 

 Text-fig. 38, R 4 might migrate across to E.2+3 and still leave a branched 

 vein in the original position of R4+5- It is thus impossible to decide 

 this point definitely. 



