282 Messrs. Minchin and Thomson. Transmission of [Jan. 7, 



in all cases chloroformed in order to get' them off the rats, it is possible that 

 some of tliem may not have recovered from the chloroform when returned to 

 Cage D, but may have succumbed ; if so, the number of fleas active in this 

 experiment would have diminished as the experiment proceeded. This and 

 the short time during which some of the experimental rats were exposed 

 may account for the break in the series of infections. Allowing for 

 differences in individual susceptibility, it may be that the chances of any 

 given rat becoming infected increase, up to a certain limit, with each bite of 

 an infective flea. It must be pointed out, further, that the fleas were exposed 

 to infection for only 22 hours. As, they had been kept without food for 

 24 hours previous to this, it is probable that at least the greater number of 

 fleas that survived must have fed on the infected rat. Comparing the 

 results of Experiment D with the almost unbroken series of infections in 

 Experiment C and with the positive results of Experiment A, and allowing 

 for all possible errors, we may conclude that the chances of the flea becoming 

 infective increase, up to a certain limit, with the time during wliicli it is 

 exposed to infection. It may be that only certain individual trypanosomes 

 in the blood are capable of producing a permanent infection in the flea. 



From Experiments C and D, it would appear that bred clean fleas that 

 have fed on a rat heavily infected with T. leioisi, and showing many 

 trypanosomes in its blood, are not infective until at least six days, or more, 

 have elapsed from the time of their having first fed on infected blood ; but 

 that from the seventh day onwards they may retain the infection, as is most 

 clearly seen in Experiment C, so as to be capable of infecting a series of 

 ctean, healthy rats, without themselves being exposed to re-infection. It is 

 possible, even probable, that with less heavily infected rats and with varying 

 conditions of tempei'ature, season, etc., the incubation-period in the flea may 

 vary somewhat, and that six days may be about the minimum period 

 required for the completion of the cycle in the flea. In Experiments A 

 and B this period seems to have been exceeded, although these two 

 experiments deal, strictly speaking, with the nietliod of infection without 

 special attention to the lengtli of the incubation-period in the flea. In none 

 of our experiments, neither in those recorded here nor in any others we have 

 performed, has direct infection taken place. In the case of T. lewisi, it is very 

 doubtful if infection by the direct method ever does take place in nature, 

 and even when conditions aie i'avouial)ly arranged in experiments, infection 

 by llio direct method has not l)ecn proved. Nuttall's oxperimciits with fleas, 

 to wliich reference has been made, were concerned with tlie /ml, of trans- 

 mission only, and weie not designed in such a way as to allow of any 

 conclusion being drawn as to the wHluxl of traiiHmis.sion. He records a 



