TRYPANOSOMES IN TSETSE-FLIES AND OTHER DIPTERA. 225 
(3) The observations and experiments of Koch, Stuhlmann, 
and others, show that T. brucii goes through a develop- 
mental cycle in G. fusca. 
(4) It. is probable that IT’. gambiense has an inverte- 
brate host in which it completes its life-cycle in regions where 
it is indigenous, and it is possible that the true host may be 
Glossina fusca. 
(5) Considerations of phylogeny indicate that the lfe- 
cycles of different trypanosomes should not be expected to be 
in all cases of one invariable type. 
(6) The encystation observed in T’. grayi indicates that 
contaminative infection may occur as well as moculative. 
VY. Recorp oF OBSERVATIONS. 
The batches of flies used in the experiments here recorded 
were infected by feeding them once on an infected animal, in 
all cases a monkey. ‘Those flies which did not feed on the in- 
fected animal-were destroyed. The flies that had fed were kept 
alive by being re-fed on a “clean” animal, in almost all cases a 
guinea-pig, but sometimes a monkey. At first I used to feed 
the tsetses once every two days, but latterly I found it better 
to give the flies the chance of feeding daily. ‘The method of 
feeding in all cases was as follows: The flies were kept in 
boxes with mosquito-netting at the sides and the box was 
placed on the skin (previously shaved) of the experimental 
animal, usually on the belly, so that the flies could feed 
through the gauze. Asa rule they fed readily; indeed, care 
was necessary in handling the boxes to avoid getting one’s 
fingers bitten through the gauze. ‘The boxes containing the 
flies were kept over water, as without moisture the flies died 
quickly. ‘The monkeys used im my experiments were species 
of Cercopithecus, either pygerythrus or smithi, and 
in all cases were infected, directly or indirectly, with try- 
panosomes from human cerebro-spinal fluid. The followmg 
