530 Observations on Habits ancl Parasites of Common Flies 
and in the jar transferred from the “sun” to the “shade tin.” In the 
former it is 60 %, in the latter nearly 90 %. That the latter figure is 
a true index of the extent of braconid infection in the original stock 
of pupae is shown bv the fact that 9 % of the pupae in the “sun tin” 
were not parasitised. 
These facts reveal the extraordinary destruction wrought by these 
parasites, and indicate that larvae feeding in warm and sunny situations 
are more liable to attack than those living in shady places. 
It may be argued that in these experiments natural conditions were 
not simulated sufficiently closely and that the braconids enjoyed 
unusually favourable opportunities. Other evidence, however, shows 
Pig. 15. Aphaereta cephalotes Hal. $ ( x 11 and nat. size) from pupa of 
S. melanura. 
that under natural conditions a great destruction of the fly larvae 
results from braconid infection. In the autumn of 1914 very large 
numbers of blow-fly larvae were present in two experimental cages 
previously described (p. 443). Yet in the spring of 1915 only 308 and 
206 blow-flies emerged in these cages. The number of pupae was not 
ascertained, but very numerous braconids are known to have emerged. 
Another experiment carried out in the winter of 1912-13 tends to 
confirm this view. Several samples of dog excrement containing 
sarcophagid larvae were placed in jars in the autumn of 1912. The 
mouths of the jars were covered with muslin, and they were kept 
outside until the spring of 1913. The pupae in the jars were then 
