82 WHITE FLIES INJURIOUS TO CITRUS IN FLORIDA. 
observations of the past leading to the statement of three more or 
less distinct periods of emergence are correct, the number of genera- 
tions annually ranges from two to five, or, under unusually favorable 
conditions, from three to six. In figure 11 the maximum and mini- 
mum number of generations as actually known to occur in groves 
at Orlando during 1907-9 has been plotted. Figure 11 is based 
upon the development of individuals in rearing experiments. The 
generation between January and March may or may not occur, 
according to whether the winter weather is warm or cold, but when 
present is numerically insignificant. The other generations are 
more confused than can be indicated diagrammatically. As may 
(JAN. 
(2)MARCH 
(9 MAY 


(4) JUNE, JULY (Q) JUNE, JULY 
(4) AUG, SEPT. (S)AUG., SEPT. 
(5) SEPT, OCT * (6) SEPF; OCT. 

Fra. Fr. Sth. Fh. 5th. 67. ord. 
* 4DULTS EMERGING AT THIS TIME ARE FROM EGGS DEPOSITED 
BEFORE THE (Sth. OF AUGUST. 
Fig 11.—Diagram showing generations of the citrus white fly. (Original.) 
have been inferred from datain Tables XV and XVII, the variation in 
the number of generations is due almost entirely to the length of the 
pupal stage, which, as has already been pointed out, is subject to 
wide variation, the cause of which can not be traced to food, tem- 
perature, humidity, or location on the tree. The most striking 
variation in the length of life cycle, with its effect upon the possible 
number of annual generations, is found among individuals developing 
from eggs deposited in April and May. In one instance eggs depos- 
ited on April 20 produced adults on June 5, July 31, and in the 
following March. In the main, each generation has two more or 
less distinct periods of emergence, as reference to the data in Tables 
XV and XVII will show, 
