^.1331. DRAGON-FLY WING VENATION— NEEDHAM. 705 



>ranches of a single main tractiea, some were shown to be incorporated 

 nto convex and some into concave veins. Thus disappeared the fun- 

 lamental difference which had been assmncd to exist between the two 

 iLinds of veins. 



i Brauer and Iledtenbacher also affirmed that the homology of veins 

 n remotely related insects is only to be determined by the study of 

 f,heir developuient — a suggestion which has until (juite recently been 

 [generally commended in theory and more generally disregarded in 

 practice. 



i Brogniart, in his Etude sur la nervulation des ailes des insectes/' 

 figures both the tracheation and the venation in careful detail, and 

 ipoints out the close correspondence between the two. He fails to 

 ^•ecognize the individuality of the veins and trachea% however, desig- 

 nating mere branches as independent veins, and he confuses the 

 identity of one branch (Bs) by giving it a different designation in the 



cwo stages. 



\ The foregoing papers have shown (without expressly stating) the 

 following points: (1) The precedence of the trachea and the subse- 

 jquent development of the veins about them, through hypodermal 

 selevations becoming paired troughs, which by fusion become hollow 

 ,tu^es inclosing the tracheiv; (2) the difference in kind, and (3) the 

 ^correspondence in arrangement between the two structures: and (4) 

 the fact that simpler conditions are found in the earlier stages. 



I have traced the development of the venation through a series of 

 nymphal stages and have published recently, in collaboration with 

 Professor Comstock. a preliminary account of it.'^ but nuist for present 

 purposes review the matter more at length. 



I ONTOGENY IN (JOMPHUS DESCKIPTUS. 



For the present I pass l)y all points of histological structure, not as 

 being uninteresting or unimportant, but as being uness(>ntial to the 

 specific problem now in hand. 1 shall deal in this paper with facts 

 that may be observc^l without the aid of sections, and, for the most 

 part, with no special preparation whatever.'' 



Six principal trache?^ traverse the wing of a dragon-tiy in all stages. 

 These arise verv earlv in the l)udding wing, springing separately from 

 a longitudinalthoracic tracheal trunk which makes a slight bend 

 outward to meet the base of t he wing. These trachea' and their cor- 



" Recherches sur lea insectes fossiles, etc., Paris, 1894. See especially pp. 204-208, 



and pi. vni. -v-x-x-ir 



'' Comstock, J. H., and Needhani, J. G., The Wings of Insects, Anier. >at., VXXli 



and XXXIII, 1898 and 1899; Art. Odonata, XXXII, pp. 903-911, 9 figs. 

 '•The only preparation necessary to demonstrate the facts cited in the folloxvmg 



pages is that of the wings of nymphs, which should he removed fresh, mounted 



quickly in glycerin jelly, and cooU-d su.ldenly (to retain the air in the trachea- ) , and 



are then ready for observation. 



