THE CITRUS THRIPS. 13 
divicluals occur. The eyes are dull red to dark purple. The average 
length varies from 0.45 mm. (about 0.017 inch) at the beginning of 
the instar to 0.90 mm. (about 0.035 inch) just before molting to the 
propupa. 
FEEDING HABITS. 
The larva becomes active immediately after emergence, moving 
about over the leaf or stem, swinging the antennae up and down, and 
surveying its surroundings for a place to feed. When feeding, the 
citrus thrips does not insert the mouth-cone itself into the plant 
tissues, and thus does not cause holes and raggedness in the leaves, 
as the pear thrips does, for example. The feeding puncture of the 
citrus thrips is made by bringing the tip of the cone into contact with 
the surface of the plant and moving the head quickly downward and 
slightly backward. This movement results in a slight folding, or 
elbowing of the cone at a point near its juncture with the head and 
in some way exerts a leverage upon the lancets, which are thus pushed 
into the tissue, starting the flow of cell sap, which is sucked up 
through the tubular interior of the cone. 
DUKATION OF LAKVAL FEEDING STAGE. 
The first of the two instars or stages of growth of the larva ranged 
from a minimum of 1 day to a maximum of 13 days, with an average 
of 3.7 days, in specimens in which the actual casting of the skin 
was observed. The second instar of these specimens ranged in dura- 
tion from 2 to 9 days, with an average of 4 days. An abnormal 
maximum of 29 days was reached by one specimen in November, the 
insect ultimately dying from exposure to prolonged low temperatures, 
without changing to the propupa. The two larval instars may be 
said to be of about equal duration. 
The duration of the complete or feeding stage of the larva ranged 
from 3 to 21 days, the average for the entire period from March 25 
to November 5 being 7.8 days. Great individual variations in the 
length of the immature stages are to be expected and do not have 
any particular significance in a practical way, but a knowledge of the 
average variation in different periods of the season and in different 
seasons, due largely, as they are, to climatic factors, are of value in pre- 
dicting what the severity of a given infestation will be. The principal 
seasonal variations in the duration of the larval stage, with a record 
of the prevailing temperature, which is the principal factor affecting 
the length of the stage, are given in Table IV. It will be seen that 
in seasons such as 1911 the larval stage will range from about 4 J 
to 7 or 8 days in the period from about the middle of May to the 
middle of September, and in the cooler weather of March and April 
