BY R. J. TILLYARD. 239 



certainly breed out from these larvae, still remains a matter for 

 conjecture. 



Bearing these facts in mind, we may confidently expect that 

 new species will continue to be found in all parts of Australia 

 for many years to come; nor is it likely that we shall reach a 

 complete knowledge of this genus until more workers in many 

 localities are induced to study these insects. 



De Selys described in his four works on the Gomphince (Synopsis 

 des Gomphines, 1854; Additions, 1859; Seconde et 3°^° Additions, 

 1869, 1873) ten species of the grand genus Gomphus from 

 Australia. For the reception of these he proposed the subgenera 

 Austrogomplius and Hemigomp)hus, relegating one species only to 

 the Old-World subgenus Onychogomphiis, viz., 0. praeruptus. 

 The two former subgenera were, however, purely artificial, and 

 in 1901, M. Rene Martin placed all ten species in the single genus 

 Austrogomplius, and added two new species from Queensland. lb 

 seems best to keep to this arrangement for the present, until a 

 careful study of both larval and imaginal forms may reveal some 

 distinct lines of separation. As regards wing-venation, all the 

 species are practically alike. Very little variation in this respect 

 can be found throughout the whole subfamily, and this makes the 

 work of subdivision very difficult. Natural selection seems to 

 have mainly busied itself, in this subfamily, with providing each 

 distinct form with appendages differing widely from those of its 

 nearest allies. In this we see a wise provision for the preserva- 

 tion of the species; for amongst insects so seldom on the wing 

 and so widely scattered, every chance meeting of a male and 

 female must be made use of; the capture of the latter by the 

 former must be swift and certain; and to this end, specialization 

 has proceeded to great lengths in the form of the male abdominal 

 appendages; and also, as a complementary result, in the structure 

 of the occiput of the female, which becomes modified to suit the 

 grasp of the male. 



Throughout the subfamily protective colouration is found. 

 Corresponding to their forest habitat, we find a general scheme 

 of colouration, in which the markings are always some shade of 



