252 Transactions of the Boyal Society of South Africa. 
ment due to the pouch. Otherwise the male ducts agree with those of 
other Glossiphoniidae. 
The ovary is an unpaired hollow organ, and differs to a remarkable 
extent from the large sac-like ovaries of the other Glossiphoniidae, in 
which the ovaries occupy a large portion of the body. In many, for 
example, they stretch from somites xiii-xviii inclusive, and constitute 
a large portion of a transverse section in this region. In Marsupiobdella 
the ovary is limited to somite xiii, and would thus correspond to the 
anterior portions of the ovaries of other genera. It is enclosed in a layer 
of follicle cells which are ellipsoidal in places. The ovary is further 
peculiar in that the oesophagus pierces its dorsal wall, and passing through 
the cavity of the organ emerges through the posterior wall. The folli- 
cular cells of the ovary pass upwards for some distance around the 
oesophagus from its point of entrance into and exit from the ovary, and 
become ellipsoidal. 
The external limiting layer is otherwise composed of flattened cells 
essentially similar to those limiting the brood pouch internally, and this 
may be taken to support the suggestion that the latter represents the 
posterior portion of the typical Glossiphonid ovary. 
The tracing of the female ducts has been very difficult and unsatis- 
factory. The female pore, and two indistinct structures with convoluted 
walls which undoubtedly represents the oviducts were made out. The 
ducts pass from the ovary downwards towards the female aperture, but 
although there is every indication that they would communicate with the 
exterior, in a study of serial sections it is seen that the ducts turn near 
the pore upwards and backwards towards the antero -dorsal surface of the 
brood pouch. This portion directed upwards and backwards really 
represents part of the ovary of other Glossiphoniidae. No distinct open- 
ing into the brood pouch could be made out. There can be no doubt, 
however, that such communications do exist, and that the eggs pass 
directly into the brood pouch, and in all probability are fertilised in 
that sac. 
The failure to find definite openings of the oviducts into the pouch is 
readily explained by the presence of well-advanced embryos or segmenting 
ova in the pouch, and by the small size and immature nature of the ova in 
the ovary. The specimens were all taken apparently after the breeding 
season, or, rather, towards the close of the same. Such a closure of the 
female ducts, in other Hirudinea has frequently been noted by one of us 
during the resting season. 
The abundance of spermatozoa in the dorsal sinus and other parts of 
the body indicates that in all probability impregnation is hypodermic as 
in species of Glossiphonia, and this would be greatly facilitated by the 
delicate nature of the body wall in the brood pouch region. 
