HOUSE-FLY INVESTIGATIONS. 499 



to house-flies enclosed in glass cylinders with muslin tops (height 

 8 inches, diameter 9 inches), which stood on the bench. When 

 flies were needed for the cylinders, either they were taken from 

 those caught in balloon traps set over the breeding-saucers, or 

 pupje were placed inside the cylinders to hatch out, etc. 

 In each experiment the following points were noted :— 



(«) the date and time of starting ; 



(6) the source of the flies ; 



(c) whether they came to feed at once ; 



{d) the number of flies (if any) lying apparently dead on 



the bench, inside the cylinders, (1) after the first hour, 



(2) each morning and evening ; and 

 (e) whether any were feeding at those times. 



The experiments usually lasted four days, as any substance 

 which failed to kill by that time was considered useless. At the 

 end of the experiment a count was made of the total number of 

 males and females which had died, and also which had lived to 

 the end of the experiment, when they were etherized before 

 being counted. 



The baits used were mixtures of casein, sugar, banana, and 

 water ; that most generally employed consisting of the following 

 proportions : — ■ 



25 c.c. casein -f- 25 c.c. brown sugar 4- 25 c.c. banana. 



{a) If the poison was a liquid, it w^as simply mixed in with 

 the casein etc. 



(b) If it was soluble, it was dissolved in 25 c.c. of water 



and added to the casein etc. 



(c) If it was insoluble, it was mixed with 25 c.c. of water 



and added to the casein etc. 



Controls with clean baits — i. e., 25 c.c. casein -f 25 c.c. sugar 

 + 25 c.c. banana -f- 25 c.c, water- — were set up for each series 

 of experiments: usually oue control for all the experiments 

 made on the same day. 



Details of these experiments, with percentage of deaths etc., 

 are given in Table II. 



The results obtained from the experiments on poisons, when 

 tested on house-flies enclosed in cylinders, were not very decisive 

 (Table II.). No substance was found which killed a really 

 large proportion of the flies. Tables III. and lY. respectively 

 give the poison-mixtures having the largest percentage of deaths 

 on the flrst day, and on the second day in cases when none had 

 died before that. Substances which killed later than this, or 

 where the death-rate was less than 10 per cent, on the first or 

 second day, were not considered of much practical value. The 

 highest percentage of deaths on the first day was only 31, 

 {Text continued on p. 51Jf. at end of TcMes.) 



35* 



