160 THE WINGS OF NEUROPTERA 



four definitive branches are doubtless veins R2, R3, R4, and R5, as indicated 

 in the figure. Vein R5 is spHt back nearly to the base of the sector and the 

 remainder of the sector, after the separation of vein R5, anastomoses with 

 vein Ri. (Figure 154 represents this part of the wing greatly magnified.) 



A study of the tracheation of the hind wing of a pupa of this species 

 (Fig. 155) confirms the interpretation of the venation of this part of the 

 wing given in Figures 153 and 154. The trachea of the radial sector is 

 typical except that trachea R5 is split back from trachea R4; and this 

 splitting has been carried so far that the undivided stem of the radial sector 

 is very short; the tracheation of the pupal wing and the venation of the 

 adult wing correspond very closely in this respect. But in the pupal wing 

 there is no indication of the anastomosis of veins Ri and R2+.3+4. It is 

 evident, however, that these two tracheas come together at the point 

 marked a in Figure 155 in the course of the later development of the 

 wing. 



This anastomosis of vein R2+3+4 with vein Ri is carried much farther 

 in the fore wing, where the two veins coalesce for a considerable distance, 

 with the result that two of the branches of the radial sector are stranded on 

 the main stem of the radius before the separation of the remainder of the 

 radial sector from vein Ri. The first of these two stranded branches is 

 doubtless vein R5, as it is obviously serially homologous with vein R5 of the 

 hind wing, in which there are no definitive accessory veins to cast a doubt 

 on the determination of the homologies of the branches of radius. From 

 this it follows that the extra branch of the radial sector in this wing has not 

 been added to the proximal end of the series but to vein Ro, and is the one 

 labeled R^a in the figure. 



According to the view that the accessory veins are added to the proxirnal 

 end of the series we must believe that the first fork of the radial sector has 

 been pushed out to near the apex of the wing by the development of the 

 accessory veins, and that the three short veins next to vein Ro are veins R3, 

 R4, and R5. If this be tine, either the size of the wing has been greatly 

 increased or veins R3, R4, and R5 have been greatly shortened during the 

 specialization of the wing. Either of these alternatives may be true, but 

 neither seems probable. 



In many cases the branches at the tip of the radial sector are merely 

 short twigs, which differ in size, number, or an^angement in the wings of the 

 two sides of the same insect; in such cases it is obviously impracticable to 

 begin the nimibering at the apex of the sector. The fact is, these tenninal 

 branches in cases of this kind are merely marginal accessory veins, that have 

 not attained a definitive form and position. 



Finally, while there are many cases, like those of Chauliodes and 

 Corydalus, in which there appear to be taking place the development of 

 additional branches at the distal end of the series of branches of the radial 



