BY R. J. TILLYARD. i f 



in Australia by the two genera, Protoplectron and Pseudo/ormica- 

 leon, two genera which may be compared respectively with 

 Acanthadisis and Myrmeleon in the subfamily Myrmeleontinm. 

 The African genus Creagis would also appear to belong to this 

 ancient group. 



It appears that the position of the cubital fork has remained 

 fixed, throughout the evolutionary development of the subfamily, 

 at a point about one fourth of the wing-length from the base. 

 This is, therefore, a satisfactory Jixed level from which we can 

 observe the evolution of the radial sector in the forewing. This, 

 as might be expected, has followed the same lines as in the hind- 

 wing, i.e., gradual removal of its point of origin on R, from a 

 point close to the base of the wing to a position further and 

 further distad. Thus, just as the DendroJeontAnn' are, as a whole, 

 more archaic than the Myrmeleont'm(t>, on account of the position 

 of the origin of Rs in the hindiving, so we may now select, from 

 amongst the Dendrohontince themselves, those more archaic 

 genera in which the same holds good for the forenjing. This 

 tribe, to which I give the name Dendroleontini, includes, then, 

 all genera in which Cu., does not run parallel with Cu, -f M., and 

 having the origin of Rs in the forewing before the level of the 

 cubital fork. This tribe contains a large number of genera, of 

 which the Australian representatives are Dendroleov, Ghnoleon, 

 Froggattisca (an asthenic offshoot of Glenoleon), Peridystus, and 

 Chrysohon. 



Finally, we may group together those genera in which the 

 origin of Rs, in the forewing, has become shifted to a level either 

 exactly above, or distad from, that of the cubital fork, as the 

 tribe Disfoleontini. The connecting-link with the Dendroleon- 

 tini is not wanting, since the genus Gymnocnemia has the origin 

 of Rs exactly above the cubital fork. But as, in other respects, 

 this genus appears to belong rather to the Disfoleontini than the 

 Dendroleontini, it seems advisable to include it here rather than 

 to separate it out as an intermediate tribe. The Dendroleontini^ 

 then, also contain a large number of genera, of which the Aus- 

 tralian representatives are Gymnocnemia, Jh-achyleon, n.g.,>Steno- 

 lean, n.g., Xantholeou, n.g., Macronemnrus, Formlcaleon, Alio- 



