1917] Wheeler—Marriage Flights of Some Sonoran Ants 179 
and fro about the branches of a few small lote bushes (Zizyphus 
lycioides) in the open desert, in precisely the same manner as I have 
seen the males of Prenolepis imparis dancing about the Japanese 
barberry bushes in April in the Arnold Arboretum, near Boston. 
The winged females of imberbiculus were far from numerous and 
were found running on the ground near the lote bushes. Mating 
was not observed. 
5. Atta (Moellerius) versicolor Pergande. 
The marriage flight of this fungus-growing ant is very different 
from that of any ant I have observed, and was witnessed under 
unusually favorable conditions. We had left our camp July 30 
about 30 miles north of Florence, Ariz., and were crossing the 
desert cn our way to Phoenix. The air was very still and clear 
after a heavy rain on the preceding day. At 5.50 a. m., just after 
sunrise, we entered a region several miles in extent where the mar- 
riage flight of Atta versicolor was in full swing. The ants were 
aggregated in numerous sharply defined swarms, each of which 
was egg-shaped or elliptical, about six to ten feet long and three 
to four feet broad, stationary some twenty to thirty feet above 
and with its long axis perpendicular to the surface of the earth. 
In some places the swarms were only about forty or fifty feet apart 
but more frequently the distance between them was fully a hundred 
feet or as many yards. As far as the eye could see over the desert 
similar swarms could be discerned. Within each swarm the large 
dark brown males and females were darting about in vertiginous, 
zigzag flight. Closer examination showed that each swarm was 
constantly receiving single males and females flying straight to it 
from a distance, but it did not grow in size because pairs of ants 
in copula were constantly raining down to the ground from its 
lower extremity, so that under each swarm there was a dense 
layer, often a yard or more in diameter, of writhing and struggling 
ants. One of the swarms happened to be poised above a puddle 
of water so that the surface of the latter became black with the 
fallen pairs. We rode for fully half an hour through these swarms, 
which must have comprised hundreds of thousands of ants. The 
activity of the insects was truly surprising, for the workers of 
versicolor are sedate and slow-moving like all other Attii. The 
whole phenomenon was rendered remarkably clear and striking 
