28 



In text-fig. 5 there are six probable stages indicated in the 

 phylogeny of the nodus. No. 1 shows the archaic state in which 

 there was no nodus, but the subcosta ended up on the costal border at 

 N, a little proximally to one of the cross-veins. In No. 2, the costal 

 break or joint is formed and the cross-vein is just about to move out of 

 the perpendicular to form a support. In No. 3 the support is formed 

 and we have the Mesophlebia-st&ge. In No. 4 M2 begins to move 

 backwards along Ml-2 towards the nodus level. In No. 5 one of the 

 cross-veins between R and Ml-2 begins to turn obliquely so as to 

 support M 2, and the lower part of the nodal cross-vein begins to 

 follow it. Finally, in No. 6, the complete cross-support from nodus to 

 M2 is achieved, and the subnodus definitely established, while the 

 fusion of parts around the nodal joint has reached its maximum. 

 This gives us the nodus as seen in the dragon-fly wing of to-day. 



Fig. 6. 



a 



Phylogeny of the Costal Joint in Mesophlebia. 



The only difficulty to be surmounted now is an explanation of 

 how the costal joint arose, and this is explained by a reference to 

 text-fig. 6. To understand the illustration it is necessary to remember 

 that the so-called " costa" of the dragon-fly wing is not in reality 

 one of the main veins of the wing, i.e., it is not formed about a pre- 

 cedent trachea in the larval wing-sheath. The costal trachea is 

 vestigial and the " costa" is simply the immensely strengthened 

 anterior border of the wing. Its structure is therefore quite different 

 from that of a main vein. Instead of being a uniform longitudinal 

 rod, it is seen to be very distinctly ribbed by a series of spines or 

 teeth. It seems almost certain that each portion of the costa bearing 

 a single tooth is the product of a single pair of hypoderm cells of 



