MOSQUITOES AND FLIES 



or buried in mud or any other soft medium, so long as it 

 keeps one end of the body out for breathing. 



The rat-tailed maggot (Fig. 172), which is the larva of 

 a large fly that looks like a drone bee, has taken a special 

 advantage of its respiratory system; for the rear end of its 

 body, bearing the posterior spiracles, is drawn out into a 

 long, slender tube. The creature, which lives in foul water 

 or in mud, can by this contrivance hide itself beneath a 

 floating object and breathe through its tail, the tip of 

 which may come to the surface of the water at a point 

 some distance away. The end of the tail is provided with 

 a circlet of radiating hairs surrounding the spiracles, 

 which keeps the tip of the tail afloat and prevents the 

 water from entering the breathing apertures. 



The great disparity of structure between the larva of a 

 fly and the adult necessarily involves much reconstruc- 

 tion during the period of transformation, and probably 

 the inner processes of metamorphosis are more intensive 

 in the more highly specialized Diptera than in any other 

 group of insects. 



The pupa of an insect, as we have seen in Chapter VIII 

 (page 254), is very evidently a preliminary stage of the 

 adult, the larval characters being usually discarded with 

 the last molt of the larva. The pupa of most flies, however, 

 while it has the general structure of the adult fly (Fig. 

 182 A, F), retains the special respiratory scheme of the 

 larva and at least a part of the larval breathing organs. 

 The fact that the larvae breathe through special spiracles 

 located on the back suggests that the primitive fly larvae 

 lived in water or in soft mud, and that it was through an 

 adaptation to such an environment that the lateral 

 spiracles were closed and the special dorsal spiracles de- 

 veloped. The retention by many fly pupae of the larval 

 method of breathing and of at least a part of the larval 

 respiratory organs, though their habitat would not seem 

 necessarily to demand it, suggests, furthermore, that the 



[3 2 7] 



