30 BULLETIN 647, U. S. DEPARTMENT OF AGRICULTURE. 
The initial mealybug infestation, on April 20, was much greater 
on these trees than on those used in experiment No. 2, there being on 
the trees left free to ants 1,661 individual mealybugs and small 
groups, and 10 infested ripe fruits; and on those from which ants 
were excluded, 1,896 individuals and small groups and 4 infested 
ripe fruits. 
There was no appreciable increase of infestation between April 
20 and May 25, but on the latter date a few mealybugs occurred 
under the sepals of many of the little fruits, a larger percentage of 
infestation occurring on the trees from which ants were excluded 
than on those to which they had access. 
Between May 25 and July 7 the intensity of fruit infestation in- 
creased on all trees, though the percentage of fruits infested de- 
creased. On July 7 about 18 per cent more fruits and 22 per cent 
more new shoots were infested on the trees protected against ants 
than on those frequented by them; and, while several fruits on the 
latter were infested more severely than any on the former, the trees 
free from ants continued to suffer a larger amount of fruit infesta- 
tion from July 7 to August 15. 
From August 15 to the close of the experiment, on September 11, 
the infestation was slightly worse on the trees to which ants had ac- 
cess. With the exception of such minor fluctuations as those indi- 
cated, however, the amount of mealybug infestation remained practi- 
cally the same on ant-invaded trees as on those free from ants 
throughout the period from July 7 to September 11. 
The struggle of the mealybugs to find suitable spots to feed and 
avoid their natural enemies on these scale-infested trees was marked. 
Every available spot free from sooty mold was occupied by them, 
and groups often occurred under sheets of the mold where it had 
lifted from the leaf. Even on the fruits the mealybugs were crowded 
by the black scale, and the practically equal and slight infestation 
on both sets of trees was due largely to this crowding. 
Mealybug enemies were numerous on both sets of trees throughout 
the experiment, especially the green and the brown lacewings, and 
larva? of the green frequently were seen feeding upon larvae and 
cocoons of their own kind and of the brown lacewings. Bits of 
cottony secretion of the mealybugs entangling the exuviae of mealy- 
bug enemies were numerous at every examination. Others of the 
more numerous mealybug predators were the lady-beetles Hyper- 
aspis lateralis Muls. and Rkizobius ventralis Erh., 1 the predacious 
caterpillar Holcocera iceryaeella Riley, 2 the predacious fly Leucopis * 
bella Loew, and the predacious bug Zelus renardii Kolen. 
1 R. ventralis is primarily a black-scale enemy, but it also feeds upon mealybugs. 
2 Identified by Mr. Carl Heinrich. 
