THE CRANE-FLIES OF NEW YORK PART II 747 



The Tipulinae show six lobes in practically all genera, the only exceptions 

 being that there are eight in a few rare cases of Tipula and five in Doli- 

 chopeza, and that lobes are indistinct or lacking in Tanyptera. 



In the Rhyphidae (Trichocerinae) the spiracular disk is very similar 

 to that in many Tipulidae, and is surrounded by four lobes. In the 

 Ptychopteridae the very reduced disk is borne at the tip of a slender, 

 retractile breathing tube. In the Tanyderidae the condition is somewhat 

 similar, but here the disk is larger and is surrounded by five lobes at the 

 tip of a long, stout, non-retractile breathing tube. 



The spiracles vary greatly in size from very large to small and vestigial, 

 or they may even be lacking in some species of Antocha. They consist 

 of an apparently uniform middle piece surrounded by a radially folded 

 margin, or ring, of various widths, called the stigmal ring. Many authors 

 (De Meijere, Mik, Mtiggenberg, Brown, Keilin, and others) hold that 

 the middle piece is an imperforate chitinized plate and that respiration 

 takes place thru the stigmal ring. Gerbig (1913), however, shows that 

 the middle piece is split across the disk, .the cleft being closed by two 

 overlapping membranes. Directly behind the spiracles the tracheae 

 enlarge into the felt chamber, whose walls are provided with long, branched,' 

 treelike structures, the branches apparently anastomosing. Surrounding 

 the felt chamber in many larvae are dense masses of air tubes, which 

 make up the tracheal lungs. These tubes are arranged in bundles, which 

 arise in special cavities of the felt chamber; thus, in Tipula paludosa, 

 there are about fifty bundles, each of about twenty tubules, making 

 a total of one thousand of these air canals (Gerbig). 



The early stages of the larva are quite different from the later develop- 

 mental stages, as Gerbig (1913:137-140), working on Tipula paludosa, 

 has well shown. The prominent six-lobed spiracular disk of the more 

 matured larva is represented in the first developmental stage by four 

 heavily chitinized projections, which bear but few bristles on their outer 

 margin. The dorsal lobes are not evident, but are replaced by eight 

 branched bristles, about equidistant from one another. The spiracles 

 are oval, not circular as in the grown larva, and project a little beyond 

 the level of the disk. The writer has noted several first-stage larvae 

 with an appearance almost as described but showing several points of 

 difference. The immature larvae of Phalacrocera are described elsewhere 

 in this work (page 963). 



