156 



THE WINGS OF NEUROPTERA 



with the tips of the branches of media in which there is a similar sphtting 

 of the tips of veins. 



Finally, excepting those cases in which the development of a pectinate 

 radial sector has not progressed very far, as in Sisyra (Fig. 139), the most 



Fig. 150.— Wings of Protohermes davidi (After van der Weele). 



striking feature of this vein, when its dichotomy has been suppressed and a 

 pectinate form attained, is the presence of a greater or less number of super- 

 numerary veins. These are accessory veins; and it has been shown that 

 veins of this type are produced and their number increased by the splitting 

 of veins. 



It is evident, therefore, that the force that makes for the development 

 of the pectinate type of radial sector acts by the splitting of veins ; and it is 

 not probable that this force should cause veins R4 and R5 to coalesce in 

 the Sialidae, instead of splitting apart as we know they do in the Hemero- 

 biida:;. 



The mimhering of the branches of the radial sector when it is peciinately 

 branched. — In most of the Neuroptera in which the radial sector is pectin- 

 ately branched, the number of the branches of this vein is more than four, 

 the number of branches of the typical radial sector. It is evident that when 

 this is the case, some of the branches are accessory veins. The question 

 arises, therefore, which of the branches are the primitive ones and which are 



