330 
A Critical Review , etc. 
We consider this is imperative and that the conclusions Keysselitz and 
Mayer have come to after using caught flies are erroneous and mis¬ 
leading. 
Roubaud (1908) has recently devoted attention to what he considers 
to be a special development of pathogenic trypanosomes in the proboscis 
of Glossina palpalis. In two of these flies, naturally infected, he found 
flagellates fixed in tufts on the internal surface of the proboscis channel, 
walls of labrum and hypopharynx ; the parasites were slender, 
measuring from 20—22 y in length. The intestinal tracts of the flies 
also contained enormous numbers of flagellates, some of which were 
without flagella. Roubaud was unable to infect animals with any of 
these parasites. As a result of his observations he concludes that there 
are three methods of development of vertebrate trypanosomes in 
the Glossinae : 
(1) Harmless culture of trypanosomes in the posterior portion of 
the midgut in the residuum of the digested blood. This culture 
disappears as soon as the flies are allowed to starve or when they feed 
again on blood. 
(2) Special development of trypanosomes in the salivary fluid in 
the proboscis which is independent of (1) and is also as transitory; 
Roubaud considers this to be the important development in connection 
with the transmission of the pathogenic trypanosomes. 
(3) An active multiplication in the intestine which may end in 
complete infection of the gut and proboscis, the parasites behaving like 
true parasites, and which until now have only been observed in cases of 
natural infection. 
We can find nothing in Roubaud’s work proving conclusively that 
these various developmental stages have come from vertebrate try¬ 
panosomes, and in spite of what he says we cannot see any difference 
between the forms he observed in the proboscis of G. palpalis and those 
described by Keysselitz and Mayer in G. fusca. We therefore can only 
consider these various developmental forms as representing stages in the 
life-cycles of one or more natural parasites of the fly which have no 
connection with any vertebrate trypanosome. The very fact that 
Roubaud was unable to infect animals with a pathogenic trypanosome 
by inoculating them with the forms he has seen in the proboscis of 
G . palpalis is, in our opinion, conclusive evidence that they are natural 
parasites of the fly. We have no difficulty in understanding that the 
conditions in nature are much more favourable for the infection of the 
flies with these natural flagellates. The three types of infection as 
