464 Observations on Habits and Parasites of Common Flies 
sun and in the screen, the minimum temperatures in the screen and 
on the ground, and the velocity and direction of the wind. 
These charts failed to indicate any combination of weather conditions 
invariably associated with a high death-rate, but it was noticed that 
numbers of flies were often found dead on the days following heavy or 
continuous rain, especially during cold and windy weather. On the 
other hand heavy rain of short duration in mild weather was not usually 
associated with a high mortahty. 
The greatest mortahty occurred on Sept. 15 and 16, very hot, 
oppressive and sultry days, when the flies were very lethargic, clung 
tenaciously to the sides of the cage and to the clothes of any one 
entering it, and hardly attempted to fly. Somewhat similar weather 
conditions prevailed on Sept. 8 and 9, and the flies behaved in the same 
manner, dying in great numbers. It was frequently noticed that very 
hot sunshine, as opposed to oppressive heat, was fatal to flies confined 
in traps. The insects attempted to take advantage of the shade 
afforded by the labels, and other parts of the traps, but many became 
unconscious and considerable numbers died. In the experimental cage 
this cause of death probably did not operate, as ample protection was 
afforded by the leaves of plants. 
On many occasions when large numbers of flies had died the bodies 
were carefully examined. Among the dead were large and small, old 
and young specimens. The old specimens are recognisable by their 
frayed wings and hard chitin, and the young by their perfect appearance 
and soft bodies. The high death-rate which apparently occurs regularly 
amongst young flies, before they are sexually mature, must be an im¬ 
portant factor in limiting the numbers of flies. 
It is difficult to draw any conclusions in regard to the effects of 
cold, apart from rain, from the observations in the autumn of 1915 since 
there were few days in which some rain did not fall, and throughout the 
weather might be described as damp. The first cold days occurred 
between Sept. 26 and Oct. 4, followed by a warmer period between 
Oct. 5-17, but the chart continues to exhibit the same features as during 
the previous months. The period from Oct. 18 to the end of November 
was very cold and wet and the flies rapidly died off. It should be 
noticed however that the curve in its general shape resembles part 
of the curve for the second generation in July when the weather 
was warm. In view of the facts to be mentioned in the next section 
relating to the resistance of flies to cold in winter, it is open to doubt 
whether the late autumn mortahty was not due partly to deaths from 
