FLIES IN CAPTIVITY 93 



be caught in a few minutes. After a sufficient number of flies 

 had been captured they were placed in one of the experimental 

 cages. These consisted of cylindrical glass chimneys about 

 three inches in diameter and nine inches in length. One end of 

 the chimney was closed by gauze kept in place by a piece of 

 thin paper gummed round the chimney over the gauze. The 

 other end was open and when in use rested on a clean quarter- 

 plate negative glass (Plate XIX, fig. i). Other cages of the 

 same kind but smaller (one and a half by six inches) were also 

 made use of The transference of the flies from one such cage 

 to another can be very easily accomplished. The fresh cage is 

 placed on the bench with its open end upwards and the full cage 

 with the negative glass still in place is placed on top of it. The 

 negative glass is then slowly withdrawn leaving the two cages 

 in free communication. By taking up the two cages in this 

 position and holding the fresh one in the direction of the light 

 most of the flies can be induced to pass into it. If any difficulty 

 occurs they can be blown from the old cage into the other. 

 A fresh glass plate is then inserted between the cages, 



" Flies have been kept alive in such cages, with daily transfers 

 to fresh cages, for more than three weeks. It was very rare for 

 a fly to escape or to be injured during the process of transference 

 from cage to cage. 



" The flies were usually fed once daily. The liquid food (syrup, 

 milk, sputum, etc.) was deposited in separate drops on a clean 

 negative glass, which was placed in contact with the one on 

 which the cage stood. The cage was then slightly tilted and 

 slipped into position over the food on the new glass. Infected 

 food was given in the same way. 



" In order to obtain a few flies from a cage for cultural purposes 

 the following plan was adopted. A piece of wood about six 

 inches square, in which a round hole, slightly larger than the 

 diameter of a cage, had been cut, was lined with cloth so as to 

 closely grip the sides of the cage when the latter was placed in 

 the hole. To one edge of the cloth gauze was sewn to form a 

 conical bag about six inches in length and about two inches in 

 diameter at its free end which was open (Plate XIX, fig. 2). 

 When in use the cage is slipped from the negative glass over the 



