to monkeys and other animals by means of tse-tse flies caught in the jungle. It is manifest, however, that before these tse-tse flies could have fed on the experimental animals an interval of many hours must have elapsed since their previous blood-feed. Therefore, in view of Minchin's failure to communicate the disease by insects that had been kept for only a short time after being fed on trypanosoma-infected animals, it is reasonable to conclude that there could have been no direct inoculation by these jungle-caught flies.
A more probable interpretation of the experiments, and one based upon what we know of the transmission of other hsemoprotozoa, is to the effect that glossina serves as an alternative host in a truly biological sense, and not as a simple mechanical transmitter; that the trypanosome, after entering the intestinal canal of the insect (Fig. 47), undergoes developmental changes requiring a considerable space of time for their completion, developmental changes which enable it subsequently, when the opportunity occurs, to effect a lodgment in the human or other vertebrate host. That this view of the role of the tse-tse fly is correct has been established by Kleine, who, after feeding G. palpalis on animals infected with T. brucei, set the same flies to bite fresh animals at various intervals. Up to twenty (possibly sixteen) days the flies failed to convey the infection, but from that period onwards to the forty-seventh day, when the experiment concluded, they communicated the trypanosome to eight animals. Bruce and others have confirmed Kleine's results and have ascertained that they apply to T. gambiense and other trypanosomes. It is estimated that 0-03-0-34 per cent, of wild G. palpalis in Uganda are infective; but in captivity, if fed on trypanosome-infected blood, 5-6 per cent, of flies become infective, and, according to Miss M. Robertson, a still greater proportion (21 per cent.) if the flies are starved after feeding. More recently Bruce and his co-workers have found that in Nyasaland 1 -35 per cent, of all wild tse-tse flies are infected with one or other of
