253 The Control of Mosquitoes 



flies. When all flies were killed, no mosquitoes re- 

 mained alive. An hour longer was required to kill 

 flies. Cages containing both were hung near the 

 windows, close to the glass, that they might be ob- 

 served through the fumigation period, and it was 

 realized that the pyrethrum on the local market 

 varied so much in quality as to be useless for prac- 

 tical fumigation, and samples for experimental 

 purposes were obtained from the United States. 



On receiving a satisfactory quality of powder, 

 larger quantities were purchased, with the privilege 

 of rejection if there was variation in quality as to 

 generation of fumes and culicidal powers. The 

 contractors kept the quality up to standard, and 

 good restdts on a large scale were obtained. 



The experiments determined the fact that in a 

 room with air-tight walls, floor, and ceiling, one 

 pound of this particular grade of pyrethrum pow- 

 der per thousand cubic feet and a three-hour 

 exposure to the fumes was sufficient to kill all mos- 

 quitoes in the room. Subsequent experiments 

 proved that with any possible chance of leakage 

 of the fumes, the quantity of pyrethrum used must 

 be increased to allow for the leakage. Where the 

 room was not reasonably air-tight and the leak- 

 age was large, pyrethrum could not be relied upon, 

 unless all walls, ceilings, roofs, etc., were covered 



