The Crane-Flies of New York — Part II 719 



various chitinized jaws, legs, heads, and other insect remains, the dis- 

 tended proventricular regions usually show a considerable amount of 

 sand particles and much plant tissue. 



The pupa 



When ready to pupate, the larva ceases feeding and becomes much 

 contracted and sluggish. The pupa is formed within the last larval skin, 

 which is then shed completely except in certain Cyhndrotominae and a 

 few scattered genera m other tribes, in which cases the larval skin adheres 

 to the posterior end of the abdomen. The pupal existence is spent in or 

 near the haunts of the larva. 



In the strictly aquatic genus Antocha the pupa lives in water in a silken 

 case, respiration being accomplished by means of the many-branched 

 breathing horns. The species of Elliptera and certain Dicranomyia 

 (simulans, for example) approach this aquatic condition. The other 

 species of crane-flies with aquatic larvae known to the writer go to the 

 soil in order to pupate, this category including Eriocera, Hexatoma, Tip- 

 ula caloptera, T. ahdominalis, and many others. 



The pupae of some, at least, of the Cyhndrotominae attach themselves 

 to plant stems for the purpose of pupation. The leaf -mining Dicrano- 

 myia foliocuniculator pupates within the larval passages. The majority 

 of the limnobiine forms spend the pupal existence in silken cases to which 

 pebbles and particles of debris or plant tissues adhere. 



The pupae of the Ptychopteridae have one of the two breathing horns 

 enormously elongated, the tip of this being projected above the water 

 level into the air for respiration. Certain tipuline crane-flies have a some- 

 what similar development of the breathing horns, discussed later. 



The duration of the pupal existence is remarkably uniform thruout the 

 group, averaging from six to eight days. The following table illustrates 

 this for the more representative genera and species. Records which 

 have not been determined sufficiently close, and which as stated are 

 probably too long, are indicated by an asterisk. 



