IV] THE LARVA OR NYMPH 91 



male and female larvae of the subfamily Aeschninae (in which 

 the appendages are best developed). 



Three large appendages can be easily seen the same three that 

 we have already noticed in the young larva. These are the 

 appendix dorsalis (ad), mid-dorsal in position, and the two latero- 

 ventral cerci (c). When closed, the three together form the anal 

 pyramid (fig. 29 A, B, ap). These appendages are the same in 

 both male and female larvae, and are the true larval appendages. 

 In the Zygoptera they form the three caudal gills (fig. 39 H, J). 

 In most Anisoptera they are much shorter than in the Aeschninae. 



From the fourth or fifth instar onwards, a second set of appen- 

 dages develops. These are destined, at metamorphosis, to give 

 rise to the imaginal appendages. In both sexes, these consist of 

 a pair of small pointed appendages, the cercoids of Heymons, lying 

 one on either side of the appendix dorsalis, above the cerci. They 

 give rise to the "superior appendages" of the male imago, or the 

 "anal appendages" of the female. Thus, in larvae of both sexes, 

 there are five appendages, of which the three largest are truly 

 larval ; the two smallest, imaginal. Now if we examine a male 

 Aeschnine larva, we shall see, placed directly above the appendix 

 dorsalis, a short projecting piece overlying the base of the latter 

 (A-D, k). This is the developing "inferior appendage" of the 

 male imago. In most Anisoptera it is not distinctly separated 

 off from the appendix dorsalis, and simply appears as the swollen 

 base of the latter (E, k). As Heymons states that this appendage 

 is really formed from the base of the appendix dorsalis, we forbear 

 to name it here. It is, however, clear that its condition in the 

 more archaic Aeschninae (c, D, k) raises the question of its having 

 a separate identity, and gives further point to Handlirsch's remarks 

 about the relationships of the cercoids with the cerci. 



In the Zygoptera, the position is much simplified. Besides the 

 three caudal gills, which are the true larval appendages, we can 

 distinguish the two cercoids as small processes (H, c'), either blunt 

 or pointed, placed as in the Anisoptera. No other process corre- 

 sponding to the basal swelling k in male Anisoptera is to be found. 



At metamorphosis, the following changes occur : 



(i) In all cases, the appendix dorsalis is cast off. But, in the 

 males of Anisoptera, it leaves behind it, basally, the inferior 



