74 Mosquitoes 



very slight development of the anal gills. The larva, 

 having its tube continually in contact with the air, 

 has no need for tracheal gills, therefore economical 

 nature does away with them. On the contrary, 

 those forms which remain much under the water 

 have greatly developed gills with many tracheae. In 

 O. dupreei and G. discolor especially, the gills are 

 enormously long and well tracheated. G. discolor 

 often remains below for hours, and O. dupreei only 

 comes to the surface to pupate. There are all sorts 

 of transitions between these extremes in the gills. 

 In W. smithii, which stays most of the time at the 

 bottom, the gills, although not long, have one pair 

 very wide, and full of tracheae. The breathing tube 

 is closed at the top by a chitinous valve, and, when 

 the larva is in the habit of staying long at the bot- 

 tom, there will occasionally escape from the tube a 

 bubble of air. From the dorsal end of the ninth 

 segment some long hairs project backward, gener- 

 ally farther than the anal gills, probably acting as 

 feelers to prevent a rear attack. The anal brush on 

 the lower side of the ninth segment is evidently 

 used as a propeller and rudder. 



It may not be amiss to say a few words here as to 

 the relation of the larval structure to habits of ovu- 

 lation. The larvae of those species making rafts 

 evidently form a natural group and possess struct- 

 ural characters peculiar to themselves. Some of 

 the single-egg depositors may have one of these 

 characters (given in the tables), but none has the 

 whole combination of characters. This natural 

 grouping is further emphasised by the fact that the 



