H. L. Duke 
361 
A cyclically-infected fly has only to plunge its proboscis into its victim 
to bring about infection, and in this way the intimate contact between the 
insect and man demanded by the direct-transmission theory would be favour¬ 
able to the spread of trypanosomes by cyclically-infected flies. On the other- 
hand, full feeding is presumably favourable to the establishment of trypano¬ 
somes in the fly, and in this respect conditions favouring direct transmission 
are against a high rate of cyclical infectivity. 
We have no data on the degree to which G. palpalis depended on man 
for blood in the old days in Uganda, and for even a 0*3 per cent, infectivity 
rate to be alone responsible for the terrible death roll of the epidemic, this 
dependence must have been very intimate. 
Another factor is the effect of cyclical development on the parasite. It 
is generally held that the processes involved in this phenomenon serve as a 
kind of filter through which secondary variations of the trypanosome cannot 
pass. In other words, we should expect that individual variations in virulence 
would be steadied down by passage through the fly, at any rate until the 
characteristic had become firmly established in the course of generations of 
selection. The extreme virulence of the epidemic in the early days thus seems 
to require some additional explanation, over and above the part assignable 
to cyclically-infected flies. 
The most commonly advanced explanation of the origin of the Uganda 
epidemic seems to be the introduction of trypanosomes from outside, rather 
than the sudden acquisition of increased virulence by a previously existing 
organism. 
Whatever the origin, the two factors which challenge our attention are 
the remarkable outburst and spread of a virulent disease and its equally re¬ 
markable dying down, which was first manifested by the death-returns, years 
before any steps were taken to remove the populations. Nature adjusted 
matters for British East Africa where the populations have been increasing 
for years, but where cases of chronic trypanosomiasis are still to be found. 
The direct-transmission hypothesis supposes the development of a strain 
of enhanced virulence as the result of continued transmission unaccompanied 
by the stabilizing influence of cyclical development in Glossina. As a means 
of spread the direct method is more rapid than the normal one, as the trypano¬ 
some makes less demand on the tsetse. Removal of the intimate contact 
between fly and man necessary for the free operation of direct transmission, 
would lead to the disappearance of this hypothetical virulent strain. 
As will be seen below it is probable that direct transmission depends for 
its success on the presence of trypanosomes in fair numbers in the blood, and 
this phenomenon is usually a manifestation of a virulent disease which will 
thus be especially suited to this method of propagation. Chronic disease, on 
the other hand, with few parasites in the blood, will not lend itself to direct 
transmission; but, if we may judge by analogy with antelope infections, 
such cases may be well suited to cyclical development in the fly. 
24—2 
