Contributions to a Knoidedgc of South African Oligochaeta — Part I. 237 
that of Phreodrilus. In this genus, however, a most important modifica- 
tion is wrought in that the spermathecal ducts open, not to the exterior, 
but into a thin-walled sac — an ovisac — from which the oviduct leads 
definitely to the exterior, opening in the position or segment characteristic 
of Phreodrilus. The structure of the spermiducal gland and male 
apparatus resembles fundamentally that of Phreodrilus. The communi- 
cation between such a " true " spermatheca and the female duct is unique. 
In Gondwanaedrilus we find definite paired spermathecae which 
occupy the typical position but which communicate with the male 
chamber. This is an exceedingly interesting counter-condition to that 
in Astacopsidrilus. 
As has been pointed out previously in this paper in considering the 
structure and histology of this organ its wall differs very much from that 
of Phreodrilus and Astacopsidrilus. The thinness of the wall posteriorly 
suggests at first that here the spermathecae are of the same nature and 
origin as in the Eudrilidae. This view we cannot accept, however, since 
the antero-ventral portion of each chamber near its junction with the 
spermathecal duct is lined with thick pear-shaped cells recalling similar 
cells in connection with the spermiducal glands to which histologically 
there is a general resemblance in Oligochaeta in general. The significance 
and importance of this view will be seen when considering Phreodri- 
loides. 
Lastly, the extension of the penial chamber upwards and backwards 
beyond the entrance of the penis to meet the spermathecal duct is impor- 
tant since this extension undoubtedly represents the muscular sac or 
" autospermatheca " of Phreodriloides. 
In Tasmaniaedrilus one of us (E. J. G.) noted a series of unpaired 
and disconnected chambers with thin walls, and recognised such as the 
remains of spermathecal structures. The structure of the wall in the 
spermathecae of Gondwanaedrilus no longer leaves doubt concerning 
the correctness of this view. In all probability in Tasmaniaedrilus we 
see the remains of an evanescent spermatheca devoid of ducts. 
In Phreodriloides it is now perfectly clear that the true spermathecae 
have disappeared, and that tbe so-called " autospermatheca." corresponds 
to the extension of the penial chamber in Gondwanaedrilus. 
The various modifications of the spermathecae are shown in the 
table below : — 
Phreodrilus Spermathecae and pores distinct. 
Wall distinctly glandular. 
Astacopsidrilus Spermathecae distinct, but ducts communicating 
with female duct. 
Wall distinctly glandular. 
