THE ARGENTINE ANT IN RELATION TO CITRUS GROVES. 39 
was shared by the writer, was that the ants fostered the white fly 
for its excretions, and when many ants were seen carrying adult 
white flies down an orange tree it was taken as evidence that they 
were transferring this pest to other trees and colonizing them thereon. 
As a principal occupation of the ants on nearly every tree was carry- 
ing white flies, however, and the reason for carrying only adults was 
not clear, the opportunity was taken to observe this work more 
closely. 
It was noticed that comparatively few ants were carrying white 
flies up the trunk, but that a very large majority, certainly over 95 
per cent, were carrying them down only. If the ants were estab- 
lishing the white flies on other trees, it seemed that at least there 
should be somewhere near the same proportion carrying them up 
as were carrying them down the trees. Many of the ants therefore 
were traced as they carried the insects up and doAvn the trees. In- 
variably those going up trees were traced to some cranny, where 
they poised in the dark for a rest or to avoid a breeze, or they would 
go up a short distance and then turn and go down again. In- 
variably those going down the tree were traced to the entrance to 
an underground nest, where they disappeared from view. Some of 
these entrances were directly at the base of the tree, but digging out 
such tunnels proved that the ants were not nesting about the roots 
of the trees or other plants on which the white fly might feed. It 
also disclosed the complete absence of underground colonies of living 
white flies and the presence of piles of dead remains of adults in the 
ant tunnels. 
The next step was to examine white flies carried by the ants to de- 
termine whether they were living or dead. Some of them were liv- 
ing, and a good many more were dead, but the most important dis- 
covery was that a very large majority still had their wings crumpled, 
as they are immediately after emergence from the pupa case, show- 
ing that they were captured just as they emerged. 
The percentage of white flies which the ants destroy must vary 
widely in the various groves at different times, and is probably never 
high enough to be of great economic importance. In a series of ten 
examinations to determine what proportion of the ants descending 
orange and privet trees with forage had captured white flies, the fol- 
lowing data were gathered on the subject: 
All the ants passing a point on the trunk going down the tree in 
a certain length of time, usually from 10 minutes to a half hour, 
were counted and classified as to whether or not they carried forage. 
Those carrying liquid forage could be distinguished by the distended 
gaster. The kinds of insects carried were noted without disturb- 
ing the ants where possible; otherwise the prey was collected. In 
