BY R. J. TILLY ARD. 419 



also. A great deal of struggling followed, and then the other 

 legs were freed, together with the thorax and wings. After this 

 exertion the insect rested about ten minutes. Then suddenly, 

 with a violent bending motion, it worked itself to and fro, 

 wriggling its abdomen bit by bit out of the tightly fitting 

 nymphal shell. Four or five segments were thus liberated, and 

 then the insect hung backwards helpless for about another ten 

 minutes. During this time the legs appeared to have gained 

 some muscular strength and hardness, for the insect soon after- 

 wards began waving them wiluly about, and finally fixed them 

 on the mosquito netting. JSTow, using the leverage, the insect 

 easily extricated the remaining segments of its abdomen, and 

 climbing up the netting got clear of the nymphal case. 

 The dragonfly was now full in the sun's rays, and a few 

 minutes' rest was sufficient to stir it to further action. Develop 

 ment of the wings, which so far had appeared like a set of short 

 flabby greenish bags, began to take place immediately. By 

 observing this process through a strong lens I was enabled to 

 note exactly how it went on. There was very little time for 

 observation, for so rapidly did the wings develop that in ten 

 minutes' time they had reached their full size. Under the lens, 

 I could see the regular pulsations of the thorax, and the quick 

 growth of the wings as the fluids were forced out into their 

 baggy folds. The development was remarkable, and of sufficient 

 interest to be given in detail. 



The wings at first lay in a series of folds, or rather rucks, 

 parallel to, and I suppose corresponding with, the main nervures 

 of the completely developed wing. These rucks gradually 

 smoothed out into more or less regular curves. Distension 

 occurred first in the parts before the nodus, so that the basal 

 portion of the wing was fairly well formed, while the rest still 

 remained flabby and shapeless. A large transverse ruck at the 

 nodus itself took a great deal longer to smooth itself out. 

 Development of that portion of the wing lying between the nodus 

 and pterostigma followed later, and began along the postcostal 

 margin from the base outwards. This part of the development 



