366 



Mr. F. W. Ed\var.ls on 



specimens, liowevev, I was astonished to find, first that the 

 insects were only about half as large as they appeared to be 

 when on the wing, and secondly that the swarm consisted 

 exclusively of jemales of Palpomj/ia brachialis, Hal. The 

 fact that only females were present seemed sufficiently remark- 

 able, for, as is well known, these dancing swarms of Chiro- 

 nomida? and other Diptera nearly always consist of males 

 only, and are often an essential preliminary to copulation, the 

 females flying into the swarm one by one, pairing, and the 

 pairs dropping out. There was, however, something still 

 more iv'niavkable about these insects. 



By accident T squeez'^d one of the flies between my fingers, 

 and noticed that some curious reddish tubes were being 

 everted from the end of the abdomen. Subsequent examina- 

 tion of other specimens confirmed the presence of these tubes, 

 and by careful observation of the insects on the wing it was 

 asceitainei] t'lar the tubes were always fully everted daring 

 flight, the jiosition in which they were iield being made 

 out with some accuracv. Wlien the specimens were taken 

 in the Tiet the tubes were withdrawn into the bodv with 

 amazing ra)ndity,.so that l)y the time the flies could be 

 examined there was rarely a trace of the tubes visible exter- 

 nally. This, then, was the explanation of the apparent 

 discrepancy in size between the flying and the captured 

 insects. 



Palpomyia brachialis, Hal. Flying attitude. 



The accompanying diagram shows the arrangement of the 

 fully extended tubes wiiile the insect is in fligiit. The tubes 

 arise from openings in the membrane at the bases of some of 

 the abdominal segments. There is a pair at the bases of each 

 of segments 6-1 , each tube being nearly as long as three 



