THE ARGENTINE ANT IN RELATION TO CITRUS GROVES. 11 
almost exclusively on isolated trees, where the number of blossoms 
and of host insects of the ant are low in comparison with the number 
of visiting ants. 
The following points have been noted as being generally true where 
the ants do use the mandibles on the blossoms: The parts attacked 
are usually the petals and stamens of open and presumably pol- 
lenized blossoms, and in most cases there is no evidence that the fruit 
is injured thereby. The attack usually begins in a wound made by 
other insects, and the work of destruction proceeds slowly. As many 
ants as could be accommodated by the blossoms have been observed 
to work steadily for one-half day without being able to destroy two 
petals completely. The ants never have been detected carrying away 
particles of the blossom tissue; evidently they desire only the juice. 
The mandibles are used to squeeze the juice out of a portion of the 
petal or stamen, that it may be lapped up by the tongue. The work 
of other insects often may be mistaken for that of the Argentine ant 
in the orange groves of southern Louisiana. Thus the blossoms of 
both the orange and the loquat may be found badly chewed and 
ragged, with tunnels cut into the unopened buds, while all are cov- 
ered with ants inside and out, seeming to make a positive case against 
the ant. When such cases have been examined with a determination 
either to see the ants cutting the holes or to discover what did cut 
them, the real culprit always turned out to be a bud moth, 1 Uranotes 
tnelinus Htibn., an unidentified case-bearing lepidopterous insect, or 
katydids. 
A few of the flowers other than citrus more commonly visited by 
the ants in the Louisiana orange groves are those of the loquat or 
so-called Japanese plum (Eriobotrya japonica Lindl.), peach, cow- 
peas, clovers, dock, goldenrod, and aster. 
INJURY TO ROOTS. 
The possibility of the ant causing direct injury to plant parts other 
than the blossoms and fruit, and particularly to the roots, was in- 
vestigated. In the orange groves the ants habitually nest in the 
ground near the base of the trees, and often the entrance to the nest 
will be found directly against the trunk. Many nests in these situa- 
tions were examined, and both the underground tunnels of the ants 
and some of the roots of the trees traced for a considerable distance. 
Dead and dying trees which were said to have been injured or killed 
by the ants and healthy but heavily infested trees were selected for 
these examinations. 
The principal facts brought to light were as follows: The ants 
never were found nesting directly in the root clusters of young 
* Identified by Dr. Harrison G. Dyar. 
