14 BULLETIN 647, U. S. DEPARTMENT OF AGRICULTURE. 
aphids, immature stages of the white fly, and adult aphid and scale 
parasites, but so rarely that this activity is unimportant. The more 
important relationship of the ant as an enemy of the white fly in 
the adult stage is discussed on pages 38-40. 
Insect Excretions or Honeydew as Ant Food. 
The most dependable, if not the most abundant, supply of food of 
animal origin utilized by the ants in the orange groves is the honey- 
dew excreted by the several species of soft scales, plant-lice, and tree- 
hoppers which it attends. 
METHOD OF ORTAINING HONEYDEW FROM THE SOFT SCALES, APHIDS, AND TREE- 
HOPPERS. 
The ants can be best observed obtaining sweet excretions from their 
host insects on the warmer days of winter, as fewer ants are run- 
ning at such times and they can be observed more closely without 
disturbing them. The process is essentially the same with one species 
of host as with another. Taking the black scale, for example, the ant 
approaches a mature or immature but settled insect and strokes the 
body with one antenna after the other, rapidly and rhythmically. 
If no liquid appears after 15 or 20 strokes, the ant usually passes on 
to another scale or rests motionless by the first. Unless the scales are 
very numerous a proportion of the ants always are waiting, and the 
principal function of the small shelter structures found over scale 
groups is believed to be to protect the waiting ants from light, 
breezes, and, sometimes, the too copiously falling honeydew and its 
attendant mold. When the scale is ready to excrete the anal plates 
open slowly outward and from between them is extruded a tubular 
organ, at the extremity of which appears a droplet of colorless fluid. 
This the ant takes and swallows at once. The tube is then retracted 
and the anal plates close. The whole operation requires only a few 
seconds, not allowing time for closer examination of the mechanism. 
The extreme lightness of the antennal stroke suggests the possi- 
bility of the presence of minute sense hairs on the body of the scale, 
which, if they occur, probably are distributed over the entire surface, 
as the stroking is not confined to the immediate region of the excre- 
tory pore. Attempts to induce excretion by stroking with a hair in 
imitation of the ants failed. From scales under the microscope there 
was no response to palpation with hairs of various stiffness. When 
the shell was pierced with a needle the anal plates half opened re- 
flexly, but not far enough for further observation. 
The process is very similar with the mealybug, as the following 
observation will illustrate: The droplet of mealybug excretion is 
considerably larger in proportion to the size of the individual insect 
than that of either the black or the soft brown scale. Two ants were 
