550 THE PAXORPOID COMPLEX, iiL, 



It follows, from the above, that the subeosta may be branched 

 either basally or distally. The basal branching is complete in 

 Archi pernor pa, where Sc consists of two long concave veins, 

 running side by side between R x and the anterior border of the 

 wing. If this condition be altered, by reduction of the upper 

 branch, we arrive at the stage shown in Aristopsyche, in which 

 this branch is clearly becoming reduced to a mere veinlet. A 

 comparison of Text-figs. 65, 85, will show clearly that the final 

 fate of this branch is that it becomes the humeral veinlet (hm). 



Besides this basal branch, Sc is frequently forked distally, as 

 in Text-figs. 63, 76. This is certainly an archaic condition, and 

 1 have accepted it as such. As the usual notation for the distal 

 forking is Scj and Sc 2 , I have suggested the use of Sc' for the 

 more anterior basal branch in Archipanorpa (Text-fig. 64) and in 

 the Archetype of the Complex (Text-fig. 38) . The archetypic 

 condition, in which both Sc' and the distal forking are present 

 together, has not been found so far in any single known type; 

 but the condition in which the distal forking remains, while the 

 basal branch is reduced to the humeral veinlet, is quite common. 



The Radius and Radial Sector. 



The main stem of the radius (R) divides, in all archaic types, 

 not far from the base of the wing, into a fairly simple upper 

 branch (Rj) and a more complexly branched lower branch, called 

 the radial sector (Rs) . The point at which this division takes 

 place is the primary radial fork (rf). 



The most archaic condition of Rx is clearly that in which it 

 is a straight vein, with a definite distal forking into R la and 

 R lb as seen in Text-fig's. 63, 76. By loss of this terminal fork, 

 it usually becomes a simple, straight vein. 



Comstock has clearly shown (15) that the original condition of 

 branching of Rs was dichotomic. His diagram showing how 

 the pectinate type of branching of this vein can be derived from 

 the older dichotomic type (15, fig. 138) unfortunately assumes 

 that Rs was originally only four-branched. This is not borne 

 out by the palaeontological evidence, at any rate as far as the 

 Panorpoid Complex is concerned. The number of branches of 

 Rs in the more archaic representatives of each Order is fully 

 worked out in those Sections devoted to the venations of the 

 separate Orders (Sections viii.-xvi. ) which should be referred to 

 for details. It will suffice to state here that the evidence is 

 overwhelmingly in favour of Rs having had more than four 



