THE ARGENTINE ANT IN RELATION TO CITRUS GROVES. 47 
all others. Ants, with scales, also were traced to underground tun- 
nels, which led neither to tree roots nor to any plant roots on which 
the scales could live. 
Aphids, too, are transported only when the infestation is very 
heavy. The largest numbers carried were upon maple and elder 
trees, on which aphis infestation persisted somewhat longer than on 
orange trees because of more rapid parasitism of the orange-infesting 
species. A great majority of the aphids carried were dead. Since 
almost all are destroyed by parasites, some undoubtedly contain 
parasites when taken by the ants, but the number thus destroyed 
is too small to reduce the effectiveness of these enemies. The destina- 
tion of the aphids carried was generally the underground nest. Only 
a very small percentage of the ants carry these insects up tree, 
and, when traced, these always have gone into one of the ant shelters 
for rest or, ultimately, retraced their steps down the tree. 
Experiments have been tried several times to induce the ants to 
remove scales and aphids from unsuitable food to a place where they 
could thrive. As an example of these experiments, about two dozen 
elderberry stems, very heavily infested with aphids, were on one 
occasion placed in the midst of thousands of ants at the base of an 
elderberry tree and examined at intervals thereafter. At the end 
of an hour aphids were leaving the stems, many were scattered about 
in the short grass, and a considerable number of others were travel- 
ing up the tree trunk. A great majority of the ants paid no atten- 
tion to these wanderers, but a few followed and stroked individual 
aphids while in motion. One such ant, becoming impatient after a 
few minutes of unrewarded effort, seized an aphid by a leg and 
pulled it about this way and that for a distance of fully a foot, when 
it let go and went its way. 
Similar experiments were performed with mealybugs, infested 
stems being cut and placed among numerous ants in pots containing 
vigorous young orange trees. The ants would attend and stroke these 
mealybugs indefinitely, but in not a single instance did one transport 
a mealybug from a dying stem to the flourishing growth of the 
young orange tree. Many of the mealybugs would wander off the 
dry stems, and some of them would find their own way sooner or later 
into the healthy tree. 
On one occasion, in California, a good opportunity was presented 
to the ants to assist mealybugs to regain trees from which they had 
been knocked by spraying with water under high pressure. The ant 
invasion was from " very heavy " to " extremely heavy " in about 70 
per cent of the trees examined, the remainder having " light " or 
" very light " trails. Some of the mealybugs hit by the water were 
27139°— 18— Bull. 647 4 
