SWARMING OF MALES 123 



then slowly drift towards the ground. Once a pair in copula was seen to issue 

 from one swarm and plunge into another swarm close by. The pair made great 

 haste to extricate itself while the swarm was immediately thrown into frantic 

 excitement and the mosquitoes danced up and down at a furious pace for some 

 time, until at last the ordinary measure of speed was regained. With the grow- 

 ing darkness the excitement in the swarms increased and the movements became 

 more rapid. Few successful unions now took place. Females entering the 

 swarm would be pounced upon by two or three males, and together, tumbling 

 over each other, they would fall to the ground and there separate. Towards the 

 last no more females appeared and wdth the increasing darkness the sv/arms 

 rapidly diminished, the males flying off into the air. 



" At 5 o'clock on the following evening the swarms were found as before, 

 dancing over every object projecting above the general level. Single mosquitoes 

 were seen flying rapidly and straight. These looked larger than the dancing 

 males and when captured proved to be females. The air was again very still 

 with a current from the south and, as before, the dancing males faced towards it 

 and kept on the opposite side of the objects. The west side of the held was 

 bounded by tall trees and high up on these, at least fifty feet from the ground, 

 before projecting branches, clouds of mosquitoes could be distinguished while 

 lower down on these trees there were none. Station was taken at a corn-stook to 

 determine how long the dance would continue. As the darkness grew the num- 

 bers began to diminish, and at 5.50, when the darkness was almost complete, the 

 last male flew away. The departing males flew upward and none of them 

 alighted on the stoock. 



" On the fourth evening the field was visited nearly an hour earlier than be- 

 fore. The sun was still shining and there were no mosquitoes present. Later, 

 when the sun had disappeared behind the trees, the swarms were again present 

 just as on the previous evenings. On this evening, however, there was quite a 

 breeze blowing and the mosquitoes could not maintain their position over the 

 projecting objects and swarmed altogether on the leeward side of them. Other- 

 wise their behavior was much the same, only that the freshening wind occasion- 

 ally threw the swarms into confusion and greater activity. Rain and cold 

 weather followed the next day and put an end to further observations and pre- 

 sumably to the swarming. 



" There are many records of swarms of Culicidte and related forms, although 

 in many cases there is no exact indication of the identity of the insects in ques- 

 tion. I am convinced that all such records, in so far as they refer to swarms of 

 the nature described above, apply to Nemocera and probably in every case either 

 to Culicidfe or Chironomidfe. I believe that these swarms of dancing males, 

 congregated for sexual intercourse, are peculiar to the Nemocera. Many of the 

 records from untrained observers, called forth by the appearance of these 

 Diptera in extraordinary num.bers, though incomplete, are nevertheless of in- 

 terest. In nearly all of them the fact that the swarming leads to sexual union 

 has been entirely overlooked. 



" Moufet, in 1034, already speaks of these swarms and notes how they gather 

 at the gables of houses and over the heads of people passing over bridges. It 

 should be noted that in his chapter ' De Culicidum ' the Culicid£e and Chiro- 

 nomidas are not distinguished, as indeed has been the case with many a v/riter 

 since. 



" The oldest record of the copulation of Culex appears to be that of the 

 Spaniard Diego Eeviglias, communicated to the Leopold-Carolinian Academy in 

 a letter dated 4 March, 1728, but not published until 1737. Eeviglias observed 

 under the microscope, and described at considerable length, the sexual union 

 and the copulatory apparatus of the mosquito. He treats his discovery as a very 



