384 
Mammalian Trypanosomes of Africa 
( c ) Failure to find flagellates in flies held to he infective. 
An important point in experiments with laboratory-bred flies is whether 
an infective fly .can, under any circumstances, become cleaned of its flagellates, 
either in gut or glands or both. Miss Robertson obtained no evidence that a 
trypanosome infection once established in a fly is ever got rid of, meaning, 
of course, under normal feeding conditions. Experiments at Mpumu on the 
effect of arsenic-containing blood on positive flies, showed that flies fed on 
a monkey sufficiently soon after the dose of arsenic lost their gut infection 
while the trypanosomes of the glands were unaffected (Duke, 1913 a) . Similar 
results were obtained by Roubaud with proboscis-and-gut infections. 
Bruce and his co-workers at Mpumu and in Nyasaland record two inter¬ 
esting observations. Twelve palpalis remaining in positive Exp. 975 were 
dissected and found negative to flagellates; they were then pooled and injected 
into a healthy monkey wffiich became infected. The inference drawm from this 
experiment w T as that an infective fly may escape detection by the microscope 
(Bruce, 1910 6). 
The Nyasaland instance is that of a positive experiment of morsitans in 
which an infective fly had been isolated in a glass tube and had, alone, in¬ 
fected a mouse and a rabbit. The fly remained ah we in the tube for 13 days, 
and on dissection proved to be free from trypanosomes throughout: another 
example of the same kind w T as subsequently observed. The inference drawm 
was that it must, therefore, be held as probable that an infective fly, with 
presumably both salivary glands and alimentary tract swmrming with trypano¬ 
somes, can lose all these flagellates and become non-infective (Bruce, 1914 a). 
It wms thought possible that by feeding positive flies upon animals wffiich 
are naturally resistant or immune to infection by the trypanosome, a complete 
or partial clearance of the fly might be effected. That this is not the case 
with baboon blood and either the human or the buck strain, w r as showm by 
the presence of flagellates swarming in both gut and salivary glands of the 
positive flies of Exps. 109 (human) and 113 (buck), both of wffiich boxes had 
fed well for two successive days on a baboon. In Exps. 85 and 86, also, the 
positive fly contained quantities of Varanus corpuscles, but the flagellates in 
both glands and gut were unaffected. 
When confronted by the absence of flagellates from flies wffiich are pre¬ 
sumed to have infected a clean animal, the first thought should be that a 
contamination infection of the clean animal has occurred, as, for instance, 
the direct transmission by w r ild biting flies from a neighbouring infected 
animal. A second possibility is that the gut has by some means been cleaned, 
as happened in the arsenic experiments above referred to, the glands remaining 
infective. 
These two contingencies must be excluded before an explanation is sought 
in some obscure and hitherto unsuspected process of sterilisation, or in the 
more simple conclusion that the positive fly has been overlooked. 
