124 
PSYCHE 
[October 
each other with a kind of weaving motion — a downward and forward plunge and 
back again, performed witlioiit unison or regularity. The movement.s were suffi- 
ciently slow to allow the plumed antennae of the males to be clearly distinguished 
and it a]>peared that the swarm t\as composed wholly of males. Once a female 
was seen to dash into the midst of the swarm and emerge on the other side united 
with a male. The insect-net was swept through the swarm a number of times. At 
each stroke the mosquitoes would disjierse somewhat but returned at once to their 
former position and continued their dance, and there was no perceptible diminution 
of their numbers. An examination of the captures showed 897 males and four females. 
This count includes the female observed passing through the swarm, as narrated 
above, and it is obvious that the other three females in the capture are not to be con- 
sidered as members of the swarm. They may have entered the swarm unnoticed 
in the manner described, or they may even have been hcjvcring about outside of it. 
The observations of T. II. Taylor (in Miall & Hammond, The Harlequin Fly) show 
a like condition in Chironomus. On a still evening his captures from a swarm were 
700 males, no females. On a windy evening, when the swarm was thrown more or 
less into disorder by the breeze, a capture of 4300 specimens included 22 females. 
Upon turning to leave the mosquito-swarm another one was discovered close by, 
hovering over and about a corn-stook. The swarm, extended about half way down 
the sitle of the stook and kept on the south side of it, the mosquitoes all facing north- 
ward. Although there was no perceptible breeze it was tliought that the attitude 
of the mosquitoes was in response to a current of air and subsequent observations 
confirmed this supposition. It w'as but seldom that one of the mos(juitoes alighted 
on the corn, and as in the cloud first observed, all appeared to be males. A round 
of the field showed that each corn-stook had its swarm of moscpiitoes, and further- 
more, single stalks that remained standing had small swarms dancing over them — 
sometimes of only six or eight individuals — .md the bushes and small trees on the 
edge of the field had their swarms. In every case the mosquitoes faced northward 
ami the swarm kept on the south side of the object of attraction. Always the mos- 
quitoes gathered over some }>rominent object such as a tree or a jirojecting branch, 
a bush, a corn-stook or a person. In this last case the swarm would move with the 
person and the only way to get rid of it was by passing under some taller object where 
the sw'arm would then remain. 
On the following evening at five o’clock the field was again visited. Upon ap- 
jiroaching the region of the creek swarms of moscpiitoes were noted over every tall 
object — at the tops of telephone-poles, orchard-trees and shrubbery. On a very 
tall elm, standing alone in a pasture, a swaim was dancing before a projecting branch 
