1912.] Trypanosoma garnbiense in Glossina palpalis. 245 



currently, and shared the same weather and other external conditions, and 

 were similarly fed. Here, as all through this paper, results of flies dissected 

 before and on the oth day are excluded. In the experiments just mentioned 

 the flies were dissected at different periods, but all after the 20th day. One 

 is forced to the conclusion that, although the monkey showed trypano- 

 somes on all the four days in cpiestion at the time of feeding the flies, the 

 blood was only very slightly infective on the 23rd, probably not at all to fed 

 flies; that on the 24th and 25th it was infective, producing (adding the 

 results for the two days) seven infected individuals out of 98 flies, i.e., 

 7"1 per cent. ; and that it was once more non-infective on the 26th. Under 

 conditions so similar it is impossible to consider a difference of 7"1 per cent, 

 to be due to the individual variations of the two laboratory-hatched sets of 

 flies. The onus of this discrepancy must obviously be borne by the trypano- 

 somes derived from the monkey. 



It is interesting to note in this connection the result of microscopic 

 examination of the live blood of this monkey over these and the adjacent days. 



For the sake of completeness I add the percentage of infected flies 

 produced over all the boxes containing infected flies obtained from 

 Monkey 113. This includes some results from dates ear her than those of the 

 foregoing table. Starvation results and flies dissected before the fifth day are 

 as usual excluded. 





Flies. 



Infected. 



Experiment 25 , , 



69 



2 



29 



60 



1 



32 



71 



1 



•A3 



32 



3 



45 



13 





46 



53 



2 



54 



107 



2 



64 



104 



2 



68 



70 



1 



75 



20 



2 



Total 



599 



18 



The percentage of infected flies is equal to 3 per cent. I do not add 

 the percentage obtained by including all the boxes in which the result was 

 negative. The number so obtained would be a mere numerical curiosity, as, 

 with our present knowledge, it is purely a matter of chance from the point of 

 view of the experimenter, how often he happens to strike a non-infective 

 period. It is obvious that a quite uncontrollable factor is introduced 



