714 



Mr. T. H. Taylor on 



identity cannot be demonstrated. The reverse seems to be 

 inferred from the following considerations. In the larva 

 the mesothoracic spiracle is connected with the spiracle- 

 chamber by an initial thread. In the pupa the same 

 ai'rangement is found, and in addition a tracheal extension 

 connects the spiracle-chamber with the gill. In the tly it 

 might have been supposed that the same arrangement as 

 in the pupa would hold with these differences : (1) the 

 tracheal extension would end blindly at the surface of the 

 body in a scar marking the former insertion of the cast gill ; 



Tliorax of Simulium-fly in side view. The pro- and meteUiorax with their apiiend- 

 ages are dotted, x 45. 



s, iiitsotlinracic spiracle; s', metathoracie do.; ap, apodenie marking junction of 

 pro- and nicsothorax. 



(2) the spiracle would be open, and in communication with 

 the spiracle-chamber by a hollow initial tube. The actual 

 arrangement is, however, quite different. The epithelium 

 of the initial thread and of the tracheal extension becomes 

 flattened out, and takes part in the formation of the body- 

 wall. The spiracle-chamber thus comes to lie close to the 

 external surface, and the apertures of the initial thread 

 and the extension coalesce to form the functional spiracle 

 of the fly. The anterior thoracic spiracle of the fly is thus 



