456 JOUR^'AL, BOMBAY NATURAL HIST. SOCIETY, Vol. XXV. 



odd divisions of insects into the latter order which have not fitted 

 into other existing and more homogeneous orders. The main 

 argument for inchiding them in the Neuroptera is on account of 

 the neuration of their wings, but if we enquire closely into the 

 comparative anatomy of Dragonflies we find that they possess as close 

 an affinity to the Orthoptera as to the former order, agreeing 

 with them especially in regard to the structure of their mouth-parts 

 and to their incomplete metamorphoses. They appear to lie mid- 

 way between these tAvo orders and for this reason and by virtiie of 

 the distinctive facies possessed hj these insects which makes them 

 as compact a group as are the Lepidoptera or Coleoptera, one is 

 justified in raising them to a similar rank. 



Order — Odonata. 



The Odonata are predatory, carnivorous, mandibnlate insects, whose life- 

 cycle is split up into a larval stage spent in an aquatic environment and 

 an imaginal stage spent in the air. The head is comparatively large, the 

 greater part being formed by the eyes : the prothorax is small and incon- 

 spicuous : the meso-and meta-thorax are bulky and fused into one : the 

 abdomen is composed of 10 complete segments : the wings, four in number, 

 are membranous, the membrane supported on a framework of stout nervures 

 and a closer network of finer, secondary nervures : the legs, of which there 

 are three pairs, are usually spined or ciliated : the sexual organs of the 

 male are situated on the 2nd and 9th abdominal segments ; of the female, 

 on the ventral surface of the 9th abdominal segment : respiration is carried 

 out by means of a large thoracic spiracle and a sub-ventral, pleural mem- 

 brane. The larvte are characterised bj'^ the possession of a labium fur- 

 nished with pincers or jaws and capable, by means of two joints, of 

 propulsion forwards, by which means the larva seizes its prey ; they are 

 either direct rectal breathers by means of rectal gills or indirectly so by 

 means of laraellated processes attached to the anal end of the abdomen. 



The metamorphosis is incomplete, in that there is no distinct resting stage 

 between the larva and the imago corresponding to the pupation of Lepi- 

 doptera. Their morphology, taken in detail, is remarkably diverse but 

 their facies as a whole, is so characteristic that there is never any difficulty 

 in determining whether any particular insect is or is not a Dragonfly. In 

 their life-cycle they pass through a number of stages, beginning with the 

 egg and passing through an, as yet luiknown, number of larval stages, the 

 larva sloughing its skin between each instar. Finally the skin is shed for 

 the last time, the larva leaving its aquatic home for this purpose and 

 climbing up some hand}' reed or trunk of a tree. The newly emerged 

 imago is an almost colourless, soft but remarkably plastic creature. The 

 wings rapidly expand and the stumpy abdomen lengthens but the full 

 colouring of the insect is not attained for about four days or more, so that 

 according to the age of the specimen, a very difterent appearance is 

 presented. For this reason, newly emerged specimens are sometimes 

 spoken of as " juvenile *' and those which have attained their full colouring 

 as " adult " specimens. The former may always be recognised by the 

 extreme glossiness, amounting almost to an oily appearance of their 

 wings. 



The final instar of the larva does not appear to vary much from the 

 l)revious ones and although it is sometimes spoken of as a " nymph, " it 

 can harilly be said to be the analogue of a pupa. 



