IN THE LARV^ OF ANISOPTERID DRAGONFLIES. 177 



be seen that the pads are actually circular or nearly circular discs of fairly 

 considerable thickness. 



In the great majority of Libellulidse the lamellse are very numerous, more 

 than twenty in a hemibranch, and very closely set. In these, usually the 

 first two or three and the last four or five will be found to possess no basal 

 pads at all. Occasionally also a pad may be missing here and there on any 

 chance lamella near the middle of a hemibranch. 



In the St/nthemini, the number of lamellse in a hemibranch is only twelve, 

 or occasionally thirteen, and the lamellae stand correspondingly further apart. 

 Usually every lamella has a basal pad, but one may be absent here or there, 

 especially on the first one or two lamellae. The basal pads are comparatively 

 enormous, at least twice as large as in other Libellulidse (text-fig. 11), and 

 their diameter may be as much as one-fourth of the whole length of the 

 lamella. Their large size seems to be explicable in two ways, viz. : — 



(1) Owing to the greater space between consecutive lamellae, as com- 

 pared with those of other Libellulidae, there is a correspondingly greater 

 need of basal supports. This is met by the greater size of the basal pads. 



(2) The Syntliemini are admittedly the most archaic of all Libeilulidge. 

 If, then, we may assume that the localized basal pads of the Libellulidse are 

 derived from diff'use pads such as are found in gills of the Simplex System, 

 we should naturally expect to find less localization in the more archaic 

 Syntliemini. The large pads of the latter, then, may be regarded as inter- 

 mediate in formation between the diffuse pads in the Simplex System and 

 the smaller localized pads of the rest of the Libellulidae. It seems clear that 

 the decrease in size of the pads is correlated with the increase in number of 

 the lamellae in each hemibranch and their greater closeness to one another. 



The Hypobranchial Tissue. (Text-fig. 16 ; Plate 22. figs. 25-27.) 



In all forms of gill the space between the two gill-walls at the base of the 

 gill-fold, bordered on one or both sides by the basal pad, is more or less 

 completely filled with a mass of tissue in which the principal tracheae of the 

 gill run. This tissue was called by Chun and Faussek " connective tissue." 

 Sadones called it " adipose tissue," but admitted that this name was faulty, 

 since no trace of fat can be found in it. I propose to term it hypobranchial 

 tissue {hy) , a name which avoids any assumption as regards either structure 

 or function, both of which are somewhat problematical, and simply indicates 

 its position as lying in the space at the base of the gill. 



This tissue diff'ers from ordinary adipose tissue, not only in the absence 

 of fat, but also in its staining qualities. With haematoxylin, it stains 

 almost to an equal intensity with the basal pad ; so that, apart from its more 

 irregular shape and the diff'erence in the shape of its nuclei, it might be 

 difiicult to pick out at once in sections treated only with this stain. I find, 



