xxxiii, '22] ENTOMOLOGICAL NEWS 261 



during" August and September no larvae of an intermediate 

 body length were found, although those of 10 mm. and those of 

 13 mm. were quite plentiful. The wings of this instar were 

 0.4 to 0.45 mm. long, the tracheation in them was much less 

 variable than in those of the preceding instar. The additional 

 tracheae, often so prominent in the previous instar, were less 

 frequently observed or were at least comparatively smaller and 

 of little importance. Other tracheal branches, especially those 

 of R and M, heretofore simple, are in this stage composed of 

 two or more fine branches which tend to cling together. The 

 radius is, in all cases, at least tw r o-branched, a posterior branch 

 Rs crosses over the two anterior branches of M. In all of the 

 twelve individuals of this instar which were examined the 

 trachea Rs always behaved in this way. A single exception is 

 shown in fig. 7. This wing was from the same larva as the 

 wing in fig. 6, which represents the conditions found in all of 

 the other three wings of this larva. The wing in fig. 7 may 

 therefore be considered as a variation having no special sig- 

 nificance and not by any means as representing a normal occur- 

 rence. We again note that compared with the wing repre- 

 sented by Needham's fig. 1, B, the wings of Ana.v. although only 

 0.4 mm. in length, have outstripped, in the specialization of 

 tracheal paths as regards Rs, wings of Goniplnts which were 

 (tcstc Needham) 3 mm. in length, whilst in regard to many 

 other features the .-Ina.v wings are far behind the Gomphus 

 wings of Needham. Not only is the wing in Needham's figure 

 much larger and more elongated ( its length being greater than 

 its width at base) than the wings in my figures, but the tracheae 

 themselves, with the exception of Rs. speak of a more advanced 

 stage of development. All the tracheae, excepting C, are com- 

 paratively closer together at their origin and along their par- 

 :i.'!t'l courses; Cu and .-1 have taken on quite decidedly the 

 characteristic paths which they assume in anticipation of the 

 formation of the triangle; the nodus and the stigma are already 

 indicated ; and linally, the tracheal trunk supplying the win^ 

 tracheae describes an arc of a comparatively shorter radius, a 

 condition more typical of later instars. 



Fig. 8 represents a wing rudiment of the next succeeding in- 



