BY R. J. TILLYARD. 



697 



Ria ..Rib 



"3A ?A LA cm 





R, 



Text-Fig. 107. 

 Wings of Raphidia maculicollis Steph., (fam." Raphidiidae) . (x 7|), 

 Lettering as on p. 535. The cross-veins are indicated by dotted lines. 



additions to the original venational scheme; and it is probable 

 that the same cause has operated to produce the extra forks of 

 M. As against these specialisations, we find Sc and R x remain- 

 ing unfused distally, and no sign of any tendency towards ex- 

 pansion of the anal area of the hindwing, such as is very ob- 

 vious in all the Sialoidea. 



We should also note that, in the Sialoidea, there is no fusion of 

 1A with Cu 2 in the hindwing, though the two veins approach 

 very close to one another. This fusion is completed in the 

 Raphidioidea, as already mentioned in Section v. 



As regards the costal veinlets, there is a complete system pre- 

 sent in all Megaloptera, and hence also in the Archetype. 



The cross-vein system can be seen to be closely similar in 

 Raphidia, the Sialidae and ArchichauUodes, but is most primi- 

 tive in the last-named, where there is little tendency to any 

 definite arrangement. A constant character is the presence of 

 only three cross-veins between R x and Rs. The addition of 

 numerous cross-veins between these two main veins in Proto- 

 hermes (Text-fig. 106) is clearly not an archaic character, as 



