214 WING-VEKATION IN ZYGOPTEROUS DRAGONFLIES, 



as likely to yield good results. Asa result of two days' search- 

 ing, six fine larvse of D. lestdides were taken, clinging close to 

 the undersides of rocks. Two were in the last instar, two in the 

 penultimate, and two in the antepenultimate. One of the first 

 was apparently approaching metamorphosis. I, therefore, killed 

 it at once, and made camera-lucida drawings of the wing-cases. 

 These drawings led to an interesting discovery, but were not 

 reliable guides to the problems on hand. On my return home, 

 a week later, the other larva in the last instar was so ad- 

 vanced that it had to be dealt with at once. The wing-cases 

 were exceedingly hard and opaque, so that it was most difficult 

 to obtain any photographic results of value. The results are 

 shown in PL xxxiii., figs.1-2. Of the other four larvse, one in 

 each stage was dealt with at once. The wings were fairly hard, 

 and yielded only moderately good results (PL xxxii., tig. 3). The 

 other two were kept alive, in the hope that they might undergo 

 ecdysis. They both gradually grew weaker, but one managed 

 just to survive an ecdysis, and yielded fairly good results (PL 

 xxxii., figs.1-2: PL xxxiii., fig.l). 



In the broad outlines of venation, the EpallagincB are more 

 archaic than the Calopterygince. I shall, therefore, deal with 

 Diphlehia first, and then indicate the chief points in which 

 Calopteryx either agrees with, or difiers from it. 



The Wing-Tracheation of Diphlehia lesto'ides (Text- figs. 1-2; 

 PL xxxii., figs.1-3; and PL xxxiiL, fig.l). 



The last three instars were studied, but only the antepenulti- 

 mate and the penultimate showed the tracheation; the larvse, in 

 the last stage, had the tracheation masked by cuticularisation 

 (see below). In both instars, the following results were noted — 



(i,)R. is a straight trachea without a single branch. (The 

 waviness in the photographs is due to unavoidable displacement 

 in cutting off the wings). 



(ii.) The basal contact between R and M is not very close. 



(iii ) M is a stronger trachea than R, and gives off four pos- 

 terior branches. Three of these come off close together near the 

 level of the bifurcation of Cu. The fourth comes off under the 

 subnodus (level of N in figures'). 



