SPRING NUMBER 57 
shown at another house. The back porch is latticed, but not 
screened, and mosquitoes are plentiful. A crock set in various 
places on and about the porch gave almost negative results, as 
most of the mosquitoes settled in a large dark cupboard in a 
corner of the porch. This being larger and equally as dark 
as the jar seemed more attractive. In a bedroom at 203 West 
Ninth Street, where the furniture and walls are light colored, 
four or five, and at once time a dozen mosquitoes were caught 
when they were not numerous enough to be troublesome, 
while in the dormitory, where the woodwork is dark and there 
are closets and bookcases for them to hide in, never more than 
three or four, and often none, would be caught, even when they 
were too numerous for comfort. 
No data were secured on the relation of the direction of the 
wind to the number caught, but the catch was always greater 
on a still than on a windy night. But as a high wind was 
usually accompanied by a drop in temperature, this may 
account for most of the difference. The effect of temperature 
was noticed throughout the winter, a high catch always coming 
with a rise in temperature. The curve in Figure 32 shows 
the temperature recorded and the number caught in a green 
cloth-lined box on the porch at 2300 West Hernando Street 
during February, and the close correlation between the two. 
The temperature of February was the coldest and most varia- 
ble of any month of the year. 
Various substances such as apples, bananas, guavas, raw 
beef, urine, water, banana oil, etc., were placed in the traps as 
attractions, but none caused any appreciable increase in the 
number caught. Very definite results were secured, however, 
with repellants. The method employed was to place a small 
vial, or to pour a little of the substance to be tested in the 
bottom of the trap, and to have a similar trap about a foot 
away for a control, or check. The percentage of efficiency, as 
repellants, of three proprietary compounds, Bombay Vapor, oil 
of citronella and oil of tar, varied from 92.8% to 82% in the 
order named. Traps of this nature should prove useful in 
testing the efficiency of repellants because of the ease in which 
a control can be secured. 
A daily record was kept of the position and the catch of the 
individual traps and the mosquitoes placed in vials or pill 
boxes for future study. Some of the specimens were destroy- 
ed by breakage, loss, destruction by ants, etc., but during the 
