RV R. J. TILLYARD. 



551 



branches originally in the Orders Mecoptera, Paramecoptera, 

 Protomecoptera, Megaloptera and Planipennia; whereas, in the 

 more specialised Orders Diptera, Triehoptera, and Lepidoptera, 

 it is reduced to the four-branched condition. It will, of course, 

 be perfectly obvious that the pectinate type of branching, which 



Text-Fig.39. 

 Stages in the evolution of the pectinate type of radial sector from 

 original dichotomic type, shown in a. Lettering as on p. 535. 



distinguishes Handlirseh's Sub-class Neuropteroidea from Ins 

 Panorpoidea, is just as easily derived from a dichotomic con- 

 dition with more than four branches, as it is from one with only 

 four branches. Text-fig. 39 shows us how this could be brought 

 about, starting with a dichotomic type in which the branches of 

 Rs are arranged as in the Archetype of the Complex (Text-fig. 

 38). In this connection, I wish especially to draw attention to 

 the position of the dichotomic forkings in the fossil genus Bel- 

 montia (Text-fig. 63), and to emphasise the obvious fact that a 

 very little movement of the point of origin of R 4 from its 

 original position on R4J.5 across to K 2 +3 w ould at once produce 

 the pectinate type of branching. 



The fossil Belmontia (Text-fig. 03) may not, however, be taken 

 as the starting point of the Neuropteroid Orders, for two rea- 

 sons. Firstly, Belmontia is clearly more specialised than the 

 older Megaloptera in the reduction of the series of costal veinlets 

 and the arrangement of its cross-veins, as well as, probably, in 

 its simplified anal veins, and possibly also in the very complete 

 formation of the cubito-median Y-vein (see Section hi.) . 



