STUDIES ON CEYLON H#MATOZOA. 693 
the Limnatis, which is apparently not a true transmitting 
host. 
The work here recorded was done partly at the Govern- 
ment Museum in Colombo, Ceylon, and partly in the Zoolo- 
gical Laboratory in the University of Glasgow. Iam much 
indebted to Prof. J. Graham Kerr for kind suggestions 
during the course of the work. 
Guiasaow ; December, 1908. 
Nore.—It may be objected that no absolute proof has been 
adduced that the form in the Glossiphonia is T. vittate. 
Absolute proof was not possible from the nature of the con- 
ditions, but the evidence seems to me to be very strongly in 
favour of the identity of the parasites. The correspondence 
of the early stages in Glossiphonia with those in Limnatis 
where experimental feeding could be carried out, the final 
stage in the Glossiphonia at the end of digestion being a 
slender flagellate and not a rounded-off organism as occurs 
in the flagellate of non-bloodsucking insects, and the absence 
of the parasite in the leeches from the tortoise where no 
Trypanosomes could be found all seem to me to point to the 
stages in Glossiphonia being true developmental stages in 
the life cycle of T. vittate. 
I may add that no flagellates were ever found in the land 
leech or in Ozobranchus or in Limnatis. Ozobranchus, of 
which a very large number were examined, lives in exactly 
the same habitat as Glossiphonia but is parasitic on Nicoria 
trijuga instead of Huyda vittata. Nicoria never showed 
a ‘T'rypanosome. 
