20 Britisli Dragoiiflics. 



But this portion of the subject is involved in much 

 doubt, and, hke the identification of many of the n}'mphs, 

 requires close observation and careful, steady work. 



As with the larva; of Lcpidoptci-a, changes of skin 

 (ecd)'ses) occur, and after one of these we find that the 

 nymph, which in Fig. 6 is seen to be quite destitute of 

 wings, has developed on the thorax little processes, 

 which, as successive changes of skin take place, 

 gradually assume the appearance of wings ; but even 

 during the last n)-mph stage the}- are very small indeed 

 compared with their size in the perfect insect. On 

 account of the wings appearing in this way, and the 

 further fact that the n}'mph is active and feeds in all 

 stages, the imperfect Dragonfly gradually approaches in 

 appearance the imago ; and so those startling meta- 

 morphoses, with which we are so familiar in butterflies 

 and moths, are absent to a great extent in these insects. 

 How often ecdyses occur it is difficult to say, but 

 probabh' a considerable number of times. The period 

 between the last two corresponds perhaps most nearl}- 

 to the pupal stage of Lcpidoptcrcx, Co/eoptcra, &c., the 

 period before — that of growth — to the larval condition of 

 those insects. 



When a nymph is full grown, and the appearance of 

 the perfect Dragonflv' is near at hand, it apj^ears to 

 cease feeding and to take little interest in life. Perhaps 

 the former is impossible, for great changes must be 

 taking place inside the skin. Some impulse now causes 

 it to .seek the u[)per air, and the hitherto aquatic and 

 ugly, sombre-tinted nymjih ascends the stem of a reed 

 or rush, or crawls up an\'thing, in fact, wliich will enable 

 it to leave its water}- home. Up this reed, or |)Ossibl\- 



