THE SUBFAMILY LIBELLULDOE. 251 



The Libellulidas are easily to be distinguished from all the other families of Odonata 

 by the large contiguous eyes and by the difference in shape between the triangles of 

 the fore and hind wings ; the apex of the triangle of the fore wings is generally pointed 

 downwards, and that of the hind wings outwards. It is sometimes harder to distin- 

 guish between the subfamilies Corduliinse and Libellulinse, the characters of which are 

 not always very salient. In the Corduliinse, however, the eyes have a slight horny 

 tubercle behind, the triangles are generally wider, less regular, and often free, the 

 sectors of the arculus are often completely separated, the last antenodal cross-nervure 

 is always continuous ; there is only a small number of postnodal nervures ; the nodal 

 sector is never undulated beyond the middle, and the males usually have the anal angle 

 of the wings angulated, and the anal appendages of the abdomen more prominent. But 

 notwithstanding all this, the Libellulinse and Corduliinse are more easily distinguished 

 by their general facies than by any decisive characters which have yet been laid down ; 

 and 1 doubt whether the boundaries between the two subfamilies have yet been very 

 clearly defined. I have therefore restricted myself to the discussion of genera which 

 appear to be undoubtedly Libellulinse, more especially as the Corduliinse are rather 

 poorly represented in the British Museum at present. 



a. Head. 



Eyes. — These are very large in the Libellulinse, covering the sides of the head. 

 They are very rarely separated (Diastatops), but in many genera they only touch each 

 other on the crown of the head. In Zyxomma they are connected by a suture for 

 nearly their whole length ; and in a few other instances they are connected by a 

 shorter, but still a considerable, portion of their length (Pantala, Tramea, &c). 



Frontal tubercle. — Between the eyes and the ocelli rises a protuberauce, which may 

 be either convex, truncated, concave, or even bifid on the summit. Its shape is 

 frequently of generic value. 



b. Thorax. 

 Dr. Brauer lays stress on the shape of the hinder lobes of the prothorax, a character 

 which I have not used in the present paper. In one or two genera (JSTephepeltia and 

 Rapliismia) we find simple or forked appendages springing from the metasternum. 



c. Legs. 

 I have employed the claws of the tarsi — which are sometimes bifid (Macrothemis) or 

 sometimes simple {Onychothemis), but more frequently toothed distinctly below the tip 

 — as an occasional character. The spines of the hind tarsi are sometimes important ; 

 thus in Lepthemis and Mesothemis they are armed with a double row of a few very 

 strong spines, instead of a larger number of more slender ones. 



2 p 2 



