210 EK. A. MINCHIN. 
an infection; the same result was attained in another case 
(Experiment 222, loc. cit., p. 5) with eight flies after only 
three feeds ; with twenty-four hours’ interval (Experiment 291, 
p- 6), nine flies produced an infection after thirty-eight days, 
and again, with the same interval (Hxperiment 232, p. 6), 
twelve flies infected a dog in the same number of days; with 
forty-eight hours’ interval (Experiment 317, loc. cit., p. 7) 
seventy flies fed in batches over thirty-one days were required, 
but longer intervals were not successful. With freshly- 
caught flies, one hundred and twenty-nine flies, fed in batches, 
infected a horse after twenty-four days (Hxperiment 225, loc. 
cit., p. 15), and ninety-eight flies infected a dog after 
nineteen days (Experiment 236, p. 16). If these experiments 
favour the view that the infection is a direct one, they at 
least indicate that infection of this type is much more easily 
obtained in this way with T. bruciithan with T. gambiense. 
There is, however, in one of Bruce’s experiments a circumstance 
which, taken in connection with an experimental result of ours, 
seems to me to show clearly that the infection of T. brucil 
was not always of the direct type. Bruce’s Experiment 225 
(loc. cit., p. 15), to which reference has been made already, 
was carried out to answer the question, “ Is the tsetse-fly capable 
of giving rise to the disease if taken out of the fly country 
into a healthy locality ?” In the account given we read that 
“the method of carrying out this experiment was to go down 
to the fly country in the early morning, catch the flies, return 
to the top of the Ubombo, and straightway place them on the 
animal under experiment. The greatest care was taken 
that the flies were caught on a perfectly healthy 
animal, as to have allowed them to puncture one 
already affected by the disease would naturally 
vitiate the experiment.” Now in our experiments 
on direct transmission, already recorded (p. 244), we found 
that if the fly, after feeding on an infected animal, were 
fed on two healthy animals in succession, only the first 
healthy animal became infected, not the second—that is to 
say, that by puncturing the skin of a healthy animal the 
