The Crane-Flies of New York — Part II 705 



the air and far overhead. A few were in copulation on the leaves, but 

 from observations made later it seems probable that mating begins in 

 the air an.d the united pair seek a support later on. Toward twilight 

 the flies may be found in great numbers in company with many kinds of 

 caddice flies and may flies. The swarms vary in munbers from about 

 fifty individuals to those including many thousands. Some of the larger 

 of these swarms cover a vertical height of at least fifty feet, the lower 

 individuals being about fifteen feet above the water. When danger 

 approaches, the swarm either mounts into the air overhead or retreats 

 before the breeze, never advancing nor moving sidewise. 



Out over the land in the smaller swarms, copulation was observed 

 several times. The males in the swarm dart swiftly at the females and 

 seize them almost instantly. They then usually leave the swarm and go 

 sailing away, the male above doing the flying, the smaller female hanging 

 limply beneath. The antennae of the male at this time are directed 

 straight ahead and are slightly divergent, the legs hang downward, the 

 wings vibrate rapidly. The female hangs downward with the legs trailing 

 limply beneath, the wings nearly horizontal and motionless. The dead 

 weight of the female continually pulls the male toward the water, and 

 often both fall into the river. As a rule, copulation ends before the male 

 is exhausted. The male opens his forceps and the female drops straight 

 downward for a foot or so, quite like a parachute released from a balloon. 

 If the pair are near the surface at the time of separation, the released 

 female drops into the water. The male darts upward again and back 

 into the swarm. The female slowly flies away, usually upstream but 

 sometimes downstream, presumably to lay her eggs. She does not stop 

 for swarms that she may encounter, and may even make a wide detour 

 in order to avoid them. In a very few cases the female is the active 

 partner and succeeds in pidling the male where she wills altho she is much 

 smaller. No matings were observed during the hours of sunlight, and 

 it was only at twilight or just after sunset that mating took place. Often 

 a second male will seize a female already in copula and the three will 

 come tumbling down into the water together. One such pairing of thi-ee 

 individuals came down, and when they were about a foot above a board 

 in the water the male in copulation dropped the other two. These rested 

 for a moment, and then the remaining male attempted to engage the 

 female in copulation. She resisted but finally he managed to seize her 



