THE CANADIAN ENTOMOLOGIST. 



not more than a stone's throw from the tide-water mark. The railway is 

 here bordered by thickets of brush that fence both sides, and make an open 

 lane not much wider than the roadbed. The sun was just sinking behind 

 the heights of Arlington, and the air was decidedly cool as I reached the 

 track. In spite of the chill, or more properly, I suppose, because of it, 

 the iane through the underbrush was occupied by dancing swarms of 

 Trichocera that hovered at various h.eigiats from three feet to ten, each 

 swarm maintaining itself in a fairly constant position, except when a whiff 

 of air blew it about. Occasionally these assemblies would coalesce or 

 subdivide, but not often. The swarms were of all sizes, from a dozen or 

 two individuals up to hundreds. 



It required several minutes of close attention to get an intelligent idea 

 of the individual movement within each collection of dancers. The first 

 impression was of chaotic activity, a sort of delirious motion without order 

 or purpose. And it was only by singling out and following an individual 

 thai the riddle was read. Each fly went through three movements, and 

 repeated these continually, a slow curving rise for ten or fifteen mches, a 

 rapid perpendicular fall, and a peculiar swaying flight that affected the 

 exact position of the .swarm in the air. Even after the movement had 

 been analyzed, a look through the swarm at an object beyond gave the 

 former effect of whirling atoms and rapid motion. 



A sweep of the net through a swarm revealed, as I had expected, that 

 only males were performing the airy incantation. But the "canto" was 

 unheard, probably because of the much slower wing-motion of the 

 Tipulidse as compared with the smaller and more active midges, with their 

 high-keyed song. At any rate, I was unable to hear any sound from even 

 the largest swarm. 



Creeping cautiously beneath a well-defined body of dancers, I was 

 able to watch them clearly outlined against the fast-darkening sky and see 

 every movement. No females were observed to fly into the swarm, yet 

 before I left for home I was able to distinguish a difference in the swing 

 of the flies, that indicated the presence of a female. Just what the differ- 

 ence was is hard to describe, but somehow the dancers, instead of neatly 

 avoiding each other as before, would interfere, the lines of flight seemed 

 to be more angular and less graceful, a series of tackles could be 

 distinguished as if a number of small fights were in progress, until finally 

 a pair would drop from the swarm, clumsily steering for the grass and 

 bushes that bordered the open. 



