CH. IV] DRAGONFLIES 107 



certainly often functional and probably the insects can 

 and do breathe at will either through them or by the 

 apparatus just described. At these periods they not un- 

 frequently come to the surface or even climb out of the 

 water a short distance and thus expose the spiracles 

 directly to the air. If the water is foul, and presumably 

 deficient in oxygen, an ^Eschnid larva will come to the 

 surface and take air into the rectum by protruding the 

 extremity of the abdomen. 



At the close of larval life the animal ceases feeding for 

 a time and climbs up a reed or post or other upright 

 object projecting from the water, or may even travel a few 

 yards over dry land until it meets with a suitable support 

 up which to crawl. Having gained a secure foothold by 

 means of its claws the larva rests with its head uppermost 

 for some hours without undergoing any visible change. 

 The larger species when resting on rushes resemble the 

 brown clusters of the flowers of these plants and are thus 

 protected during the period of exposure and helplessness. 

 The first noticeable alteration of appearance takes place 

 in the eyes which, from being dull and more or less 

 opaque, become bright and clear in consequence of the 

 underlying brilliant eyes of the imago pushing against 

 the larval covering. Shortly afterwards a crack appears 

 along the mid-dorsal line of the thorax, and the thorax of 

 the imago begins to protrude, extending the cleft forward 

 on to the head and causing lateral splits in the region of 

 the eyes. The thorax (with the, at present, diminutive 

 wings) and head are freed first from the larval skin and 

 then by a strong bend dorsalwards the legs are extracted. 



