H. L. Duke 
377 
few days. In the Tables, Nos. V, VI and VII, column 4 records the day upon 
which, after the first infecting feed, dissection was commenced. For the first 
few days of the experiment the dead flies are not dissected, as the discovery 
of trypanosomes in such early flies does not necessarily imply a developing 
infection. The dates appearing in the second column of the Tables are given 
to show the stage of the disease in the infecting mammal. 
(a) Meteorological conditions at Entebbe and Mpumu. 
Records are available of the wet and dry bulb readings inside the laboratory 
and at the Entebbe meteorological station in the open air. 
The laboratory readings are on the whole higher than the official figures, 
and as compared to the figures obtained from Mpumu the Entebbe laboratory 
showed a generally higher temperature throughout the year. 
(b) Tables of fly experiments and discussion of same. 
The human, wild-fly and Damba antelope strains are dealt with in Tables 
V, VI and VII respectively. The percentage of infected and infective flies 
obtained with each strain is given at the end of the corresponding table. Tn 
none of the flies dissected were any flagellates found in the proboscis. 
Tables V, VI and VII show that: 
(a) With the limited number of flies employed, the average duration of 
the full developmental cycle, with the human strain, was 32 days (maximum 
45, minimum 25 days); with the fly strain, average 29 days (maximum 31, 
minimum 26 days); with the buck strain, average 28 days (maximum 40, 
minimum 19 days). 
(b) Fully formed and apparently healthy larvae can be produced by flies 
fed on a purely reptilian diet. 
(c) All three strains of trypanosome can develop in flies which are 
nourished solely on reptilian blood. This matter will be further dealt with 
in subsection (d) below. 
Discussion of the fly-dissection Tables. 
It is much to be regretted that the salivary gland dissections were not 
more carefully performed in the experiments recorded in the above Tables. 
Press of other work made it impossible for me to undertake all the fly dis¬ 
sections myself, and the routine method practised by the native assistant 
consists in pulling out the gut, after snipping off the terminal part of the 
abdomen, and teasing it up in saline. In this way the glands usually come 
out with the intestines and are chopped up therewith, so that only fragments 
are available for scrutiny. 
The destruction of two of the experimental boxes by ants was due to these 
boxes, after having been selected for dissection, being left on the table over¬ 
night instead of being replaced on the water dishes. The tiny ant responsible 
for these devastations appears with magical suddenness wherever anything 
