38 BULLETIN 640, U. S. DEPARTMENT OF AGRICULTURE. 
ground exposed to the direct sunlight in summer, but many larvae 
escape even from such fruits. An examination of 17 mangoes 
exposed over sand on shallow trays to the sun for two days in August 
revealed 17 living and 98 dead well-grown larvae in the fruit, and 103 
that succeeded in safely leaving the fruit to form pupa? in the sand 
beneath. The small brown ant, 1 so common about the lowlands of 
Hawaii, unquestionably is a factor in natural control. It is fre- 
quently found swarming over and in fallen fruits and kills many 
larvae as they leave the fruits to pupate. Ants were observed to 
remove from a fallen ball kamani nut 86 medium-sized larvae during 
a 40-minute period, but they failed to reach 34 other larvae in a flriner 
portion of the fruit. 
No natural checks upon fruit-fly increase in Hawaii, aside from 
introduced parasites, are of practical value; but in many other lands 
climatic checks are unquestionably of great value, and, as in many 
places in Spain, for instance, are the only checks that make fruit 
production possible without the use of remedial measures. 
CLIMATIC CHECKS. 
Observations made in various countries indicate that the Mediter- 
ranean fruit fly will not be a serious pest when the monthly mean 
temperature falls to or below 50° F. for from three to four consecutive 
months during the year. In Hawaii the climate is not cold enough 
throughout the fruit-growing regions to act as a serious check on 
fruit-fly increase. Development progresses most rapidly after the 
Hawaiian means reach 75° to 79° F. At a mean of 68° F. develop- 
ment requires about twice the time. A temperature of 58° to 62° F. 
may increase the period of development to three to four times the 
normal period for the warmest weather. Larval development in 
apples stored outdoors at temperatures ranging from 31° to 64° 
F. (mean, about 51° F.) was slow and was attended by no un- 
usual mortality. No development occurred at 26° to 70° F. 
(mean, 48° F.), and nearly all larvae were dead at the end 
of six weeks. Very few fruit flies can develop at 49° to 50° F., 
and none at temperatures below this point. Complete mortality 
will follow continued exposure to temperatures below 50° F. An 
exposure for two to three weeks at 32° F. will kill all stages of the 
fruit fly, but an exposure to this low temperature for four days has 
practically no effect upon the fly. Sixty-two of 248 larvae survived 
an exposure for five days to 21° to 28° F. These facts indicate that 
the Mediterranean fruit fly is a very hardy and persistent enemy in 
spite of the quickness with which it responds to checks upon its 
development resulting from the low temperatures ordinarily experi- 
enced in semitropical countries. 
1 Pheidole megacephala Fab. 
