XVII] BIONOMICS, ETC. 327 



colour-pattern goes, on p. 255. We may offer on p. 328 an attempt 

 at a classification of the different habitats of Dragonfly larvae. 

 In this table, the letter A indicates the commonest habitat for 

 the group, while B, C, D indicate less usual habitats, in descending 

 order. 



As the table is based chiefly on observations made on Australian 

 species, it is to be hoped that observers in other regions will compare 

 it with the case of their local fauna. 



For the question of the mode of progression and method of 

 stalking the prey, the reader is referred to the account of the 

 .rectal and caudal gills in chap, ix, and to the account of the 

 labial mask in chap. iv. 



The larvae of Dragonflies are not without means of defence 

 against their enemies, though they rely chiefly on concealment, 

 both for avoiding their enemies and for the capture of their prey. 

 If a leg or a caudal gill be seized, the breaking- joint at the base 

 of the organ enables it to be cast off at once. The sharp dorsal 

 and lateral spines on some larvae, and the acutely pointed anal 

 pyramid in the Brachytronini, may be used as a means of defence 

 if the larva is seized. These same spines are probably of great 

 use to the larvae in the case of a sudden rush of water during 

 flood-time. 



Food. 



Little definite is known as to the food of Dragonflies, beyond 

 the fact that it is captured on the wing, and consists exclusively 

 of live insects. Poulton [125] and Campion [35] have collected to- 

 gether all the definite records, which include representatives of the 

 Plectoptera, Trichoptera, Lepidoptera, Hymenoptera, Coleoptera 

 and Diptera, besides, of course, smaller Odonata. As definite 

 observations concerning gnats and mosquitoes being captured for 

 food seem to be few or lacking, I may add here that I have 

 repeatedly seen Aeschninae feeding on these pestilent insects in 

 the late afternoon. A specimen of Telephlebia godeffroyi was once 

 observed flying round and round a small bush about 7 p.m., when 

 the mosquitoes were particularly troublesome. After ten minutes, 

 it was captured. I found its mouth so full of mosquitoes that it 

 was unable to shut it. There must have been over a hundred, 



