DISTRIBUTION OF BACTERIA 99 



At intervals flies were killed and their crops dissected out. Some 

 of the fluid from the crop was drawn up to the first mark, diluted 

 and sown. Between each culture the pipette was sterilized 

 with alcohol and washed and dried. Approximately the same 

 number of colonies grew in cultures made from the emulsion and 

 from the crop contents of flies dissected one, four, five and a half 

 and eight hours after feeding. About half the number of colonies 

 grew in cultures made from the crop contents of flies dissected 

 24 and 31 hours later, which had been allowed to feed once in 

 the interval on plain syrup. 



In another similar experiment the colonies were counted. 



A fly was dissected 45 minutes after feeding and 4500 

 colonies were counted. 



A fly was dissected 75 minutes after feeding and 5490 

 colonies were counted. 



A fly was dissected 275 hours after feeding and 4900 colonies 

 were counted. 



A fly was dissected 5*5 hours after feeding and 4098 colonies 

 were counted. 



A fly was dissected 24 hours after feeding and 247 colonies 

 were counted. 



A fly was dissected 3 days after feeding and 10 colonies were 

 counted. 



These experiments seem to indicate that in the case of 

 B. prodigiosiis multiplication does not take place in the crop. 



id) The infection of agar plates by living flies. 



In each of the following experiments two or more flies which 

 had previously fed on syrup infected wath B. pi^odigiosus were 

 allowed to walk for 30 minutes over the surface of agar plates. 

 For the first few minutes the flies generally walked rapidly over 

 the surface, but subsequently they frequently stopped, and 

 applying their proboscides apparently sucked the agar. In 

 some cases, especially if the agar was not too dry, oval marks 

 were left where the proboscides were applied, and it was round 

 these marks that the prodigiosus colonies grew. The following 

 table shows the result of these experiments. 



7—2 



