258 STUDIES IN THK LIFE-HISTORIKS OF AUSTRALIAN ODONATA, 



ing difficulty of capture and copulation on the part of some 

 dragonflies is evidence of special precautions developed by Nature 

 to prevent hybridisation. In those families where one form and 

 one colour-pattern predominate, many species can scarcely be 

 distinguished except by means of the parts used in coition, and 

 these may be so diverse in closely allied species as to prevent the 

 accidental copulation of the male of one species with the female 

 of another, captured by mistake. I have myself seen a male of 

 Austrogomphus collaris Selys, seize a female and vainly attempt 

 copulation. After a considerable time the attempt was 

 abandoned, whereupon a .second male seized the same female 

 and effected connection almost immediately. Here there \va» 

 probably some local deformity in the first male, which prevented 

 him either from holding the female securely or from completing 

 the act of coition. 



The method of oviposition apparently determines the shape of 

 the egg, and often the habits of the larva. In all species where 

 the egg is deposited in tissue by means of the sharp ovipositor of 

 the female, it is of an elongated oval shape, evidently a suitable 

 form for inserting into the narrow opening prepared for it. 

 Where the Qg^ is dropped into the water straightway, it is 

 usually broadly elliptical or almost round, and the vulva of the 

 female possesses no true boring apparatus. Again, the larvae of 

 nearly all the former species — those having elongated ova — are 

 clean-living insects, inhabiting floating masses of waterweed; 

 while those of the latter crawl on the trashy bottoms or burrow 

 in the sand or mud, and are often very hairy and covered with 

 mud. 



The life-histories of all those species peculiar to Australia — 

 with the exception of Lestes leda — are quite unknown. Included 

 amongst these are many ancient and isolated forms, true collective 

 types, with no very near allies in other parts of the world'; such for 

 example as Cordulephya^ Synthetnis, Petalura, Diphlebia. The 

 larvae of these four are amongst the greatest desiderata of present- 

 day systematists in Odonatology, and much lightmay be expected te 

 be thrown on the present methods of classification by their dis- 



