THE ARGENTINE ANT IN RELATION TO CITRUS GROVES. 
19 
sound and parasitized scales, the presence or absence of scale enemies, 
and the activities of the ants. This experiment was started on April 
28 and concluded October 24, 1914. There was a large colony of the 
ants about the base of the nonbanded tree throughout the experiment, 
but the ants did not visit the tree, except to keep it patrolled by 
scouts, until several soft brown scales became established there, and 
at no time were they discovered paying the slightest attention to 
the purple scales. No scale enemies of any consequence were seen 
on either tree, and there was never any evidence of parasitism. The 
results of this experiment are summarized in Table I. 
Table I. — Experiment to discover the effect of ants upon the armored scales of 
citrus. Louisiana, 1914- 
Ants present. 
Ants excluded. 
Date. 
Number of sound scales 
present. 
Number 
of scales 
showing 
parasit- 
ism. 
Number and activities 
of ants on trees. 
Number of sound scales 
present. 
Number 
of scales 
showing 
parasit- 
ism. 
May 7 
June 3 
97.... 







Only 3 scouts in tree . . 
283 

125 
591 

June 19 
198 
do 
8 ants capturing white 
flies. 
591 

July 17 
Aug. 13 
Sept. 25 
276 
530 

1,130 
1,372 

5,700 (estimated) 
10 ants attending soft 
brown scale only. 
50 ants, all attending 
soft brown scale only. 
7,200 (estimated) 
Trunk and main 
branches literally 
covered. 

Oct. 24 
Trunk and main branches 
literally covered. 

Eeference to Table I will show that on May 7 there were 97 scales 
on the ant-invaded tree and 283 on the tree from which ants were ex- 
cluded. The number gradually increased on each tree from June 
to October, except that there was a slight and unaccountable decrease 
on the tree from which ants were excluded during June and July. 
On September 25 it was estimated that there were 5,700 scales on the 
ant-invaded and 7,200 on the ant-free tree. By October 24 the trunk 
and main branches of both were literally covered with the scales, 
and it was impossible to distinguish between the two as to infesta- 
tion. The scales had increased at approximately the same rate on 
both trees. The health of the trees remained good throughout, except 
for a few yellow spots made on the leaves by the feeding of scale 
groups. 
In another experiment the ants were excluded from a block of 
* more than 200 bearing orange trees for several months, while an 
equal number of trees adjoining were left untreated as checks. The 
color of the trees in the treated block showed improvement over 
