86 
ZOOLOGY OF THE FAR EAvST. 
(2) I propose to discuss here the "coelomic sac" of Kawamuria, which, as 
also in Branchiuya sowerbyi, surrounds the terminal part of the male deferent 
apparatus. 
The sac is ovoid, completely closed, planted on the ventral body-wall over the 
situation of the penis; it possesses a muscular wall of some thickness. Its probable 
function, which I take to be the extrusion of the penis, will be understood from fig. 2. 
The penis is a pear-shaped bladder-like organ, the space between the axial ejacu- 
latory duct and its outer wall being free from the strands which, in other genera with 
a penis, frequently form a connection between the two. Its extroversion would be 
effectively brought about by the contraction of the walls of the coelomic sac and the 
forcing of fluid into this space. 
It is difficult to guess any other function for the sac; and it seems not unlikely 
that sac and penis have evolved together as a connected whole. 
But among Tubificids a similar sac occurs elsewhere only in Branchiura; and 
there can be no doubt, on other grounds, that Kawamuria and Branchiura are closely 
related; Kawamuria might be described as a Branchiura without gills and with a 
penis. But if, as just implied, Branchiura has the sac but no penis, what must be 
the nature of the relationship? It can only be that Branchiura has lost a penis that 
it formerly possessed ; since (taking for granted the function of the sac as explained 
above) we cannot credit Branchiura with having produced a sac in the expecta- 
tion of a penis to follow. In other words the sac in Branchiura is a vestigial 
organ. 
There are possibly a few indications of this in its structure ; it appears to be 
somewhat less ample than in Kawamuria (cf. figs, in Michaelsen, 13, and Stephenson, 
17); the wall is also not particularly strong (seine Wandung ist diinn, mit zarten 
Muskelstrangen ausgestattet) , and is certainly thinner than in Kawamuria. 
If the above reasoning is correct, Branchiura would be a direct descendant of 
Kawamuria.^ by loss of penis and development of gills. (On the systematic aspect of 
the connection between the two forms v. inf.). 
(3) Criodrilus bathybates is remarkable on account of the depth (180 ft.) at which 
it is found. A certain number of the Eumbricidae, as well as scattered examples in 
other terrestrial families of the Oligochaeta, may be found regularly or occasionally 
in marshes, ponds and streams, — usually on the banks or near the shore, — while 
more than one subfamily of the Geoscolecidae are partly, and the Criodrilinae wholly, 
limnic; but even these latter seem to occur for the most part in shallow water only. 
The occurrence of one of the "Megadrili" at the depth of 180 ft. is probably very 
exceptional, though without going through the literature of all the limnic forms I 
could not say that it was unique. But it is noteworthy that in L. Baikal, where a 
systematic investigation of the whole fauna was carried out, no Megadrili at all were 
obtained, though Microdrili, — Tubificids, Lumbriculids and Haplotaxis, — were pres- 
ent at depths even below 1,000 metres. (Michaelsen, 11). 
With this deep-water habitat, in the case of C. bathybates, is probably to be cor- 
related the absence of a gizzard. While this organ is present, though rudimentary, 
