BY R. J. TILLYARD. 719 



comparative depth of the crenations of the inner margin of the 

 lateral lobes in the nymphal labium. Ranging from a deeply den- 

 tate margin in some forms, to the more prevalent erenate form of 

 the Corduliince s. str., (in which the depth of the erenate lobe is 

 roughly equal to the width of the depression between them), we 

 find this as the distinguishing mark of the larvae of this subfamily. 

 The only exceptions are the most caenogenetic forms, such as 

 Hemicordulia, Tetragoneuria, etc., which show the more shallow 

 Lihelluline crenations; while, in the Lihellulince,, the only nymph 

 having moderately deep crenations, is that of Pantala flavescens. 

 This latter being one of the most recent and highly developed 

 Libelhduiff, we must suppose this Corduline character of the 

 labium to be secondarily developed. 



It will be seen, at once, that the larva of Nannophlehia possesses 

 this Corduline character. This is clear evidence of the proximity 

 of Group i. to the Corduline stem, and, therefore, of their being 

 archaic, as compared with the general body of the Libellnlince. 



(2) Coupled with the above character, but not so distinctive, is 

 the smooth unspotted surface of the lateral lobes, and their fuller 

 and more rounded outline in the Corduliince^ together with a 

 smaller development of spines on the crenations. The tendency of 

 the Lihelluline labium seems to me to be the development of a 

 broad subtriangular mentum, the compression basally of the lateral 

 lobes to subtriangular form, and a general decrease of the broad 

 curvature of the inner margin. With this comes the increase in 

 the number (as distinct from the size) of the crenulations, an often 

 corresponding increase in lateral setae, and a strong tendency to 

 the production of black spots and warts on the surface of the lobe. 



The larger number of setae is not usually, however, found in any 

 small nymphs, and need not therefore be looked for in Nanno- 

 phlehia. Judged, however, by the general form of the lateral lobe, 

 and by the appearance of black warty spots on it, we must regard 

 tlie labium as showing distinct Lihelluline characters. 



On the whole, then, the evidence of the labium is that Nanno- 

 phlehia is an archaic Lihelluline which branched off long ago from 

 tlie old Corduline stock. 



