92 



THE PALEONTOLOGICAL DATA 



Fig. 80. — Nymph of a Paleo- 



dictyopteron (After 



Handlirsch). 



studied the tracheation of the wings of nymphs and of piipee of the more 



generahzed members of all of the orders of living insects of which material 



was available. 



In the case of each of the wing-veins, a comparative study was made of 



the trachea that precedes it in all of the orders in which that trachea is 

 retained; and in those orders where the trachea- 

 tion is reduced, a study was made of the wing- 

 vein in the more generalized members of the 

 order. It was upon these studies that our conclu- 

 sion regarding the probable primitive fonn of the 

 wing-vein was based. It is of interest now to 

 ascertain to what extent the validity of these con- 

 clusions is supported by the paleontological data 

 at hand. 



The evolution of the costa.^Little can be 

 said about the evolution of the costa; for in the 

 oldest of winged insects known to us it had reached 

 the form characteristic of the most specialized 

 of living insects, that of an unbranched marginal 

 vein. 



The need of a firm support of that margin of 

 the wing where the greatest stress falls in active 



flight evidently led very early to the attainment of the form of the vein 



that has persisted to the present. 



The evolution of the subcosta.^ — Among recent insects three types of 



the subcosta exist: first, that which is well-shown in Nemoura, in which 



the vein is divided into two branches (Fig. , , 



82, a); second, the type found in many insects, 



in which it is a simple unbranched vein .(Fig. ,--'' 



82, 6); and third the type that is well-illus- 

 trated by the subcosta of Corydalis, in which 



the principal vein bears many small branches 



Fig. 82, c). 



It seemed probable to Comstock and 



Needham that the second and third types 



had been derived from the first; that the 



second type was evolved by the loss of one of 



the branches existing in the first ; and that the 



third type was the result of a specialization by 



addition. 



Among the facts that led to the conclusion that the two-branched type 



was the more primitive one are the following: In Xeiioneura (Fig. 73), 



which at the time we wrote our series of articles was believed to be a 



Fig. 81. — Nymph of a Palcodi- 

 ctyopteron (After Ihindhrsch). 



