70 ME. R. J. TILLYARD : LIFE-HISTORIES AND 



Va r i e ti e s. — Individual imagines vary much in the clearness and details 

 of the beautifully mottled pattern of the abdomen. The amount of dark 

 brown and pale yellow on 9 and 10 varies greatly, some specimens appearing 

 yellow-tipped when flying, some almost black. Variation occurs also in the 

 shape of the male appendages, both superior and inferior. It is possible that 

 a long series of Western Australian specimens might prove to be constant, 

 the form that I have noted in my short series from that State — viz., that the 

 superior appendages are broader and more leaf-like, and the inferior slightly 

 longer than in those from Eastern Australia. 



C. GYNACANTHA SERIES. 



M2 arching strongly up at or near level of pterostigma. Ms unsymmetri- 

 cally forked well before the level of the pterostigma. Mspl curved concavely 

 to the lower branch of the fork, so as to enclose several rows of cells between. 

 Mspl similarly concave to Mi. Membranule small. Hind-wing of male 

 angulated; narrower near base than towards nodus; anal triangle either 

 short and broad or long and narrow. Anal appendages long or verj^ long in 

 both sexes. Females with a remarkable fork projecting downwards under 

 segment 10. 



Three genera are represented in Australia, and may be distinguished as 

 follows : — 



-■ I Basilar space reticulated Hkli^schna. 



I Basilar space free 2. 



j Large insects with short, rather wide, three-celled, right- 

 angled anal angle in male. M2 much arched upwards 



J near pterostigma , . . . . Gynacantha. 



2. I Medium-sized insects with long, narrow, four-celled anal 



I triangle in male. M.^ less arched near pterostigma, more 



[ regularly curved ; Austrogynacantha. 



All three genera agree in having a hijid fork under segment 10 of the 

 female. This distinguishes them from the closely-allied genera Tria- 

 cantliagijna (trifid fork), Tetracantliagyna (four-pronged fork), Platacantlia 

 ( ? without fork, but with three fine points under 10), and Cornacantha 

 ( ? without fork, but with two strong horns under 10). Probably these are 

 all scarcely more than subgenera of Gynacantha, but may be retained for 

 systematic purposes as genera. 



Genus 1. GrYNACANTHA, Ramhvr. 



Characters as given (see above). 



Type : G. gracilis (Burm.) Handb. Entom. ii. p. 837 (1839). 



I am unable to find anywhere that any described larva has been definitely 



