1015.] J. STEPHENSON : Indian Oligochaeta. Ti 
The ovaries are in xiii; the oviducts and pore or pores were not seen. 
The spermathecae are single in each of the segments viii and ix. The ampulla 
of each is ovoid with the longer axis transverse, or perhaps spherical; the sections 
being longitudinal, a calculation from the number of sections in which the sperma- 
theca appears and the thickness of each section gives 5 mm. as the breadth, while 
direct measurement gives ‘4 mm. as the height of the ampulla. The duct is thick, 
in length equal to the ampulla; in one case the duct makes a considerable bend, in 
the other it is straighter and so shorter. The ducts are placed on opposite sides of 
the nerve cord, the anterior on the left, the posterior on the right; each passes 
underneath the cord to end in the mid-ventral line. A small diverticulum is given 
off from the duct near its junction with the body-wall. 
There are no penial setae. 
Remarks.—The combination of lumbricine setae, rudimentary gizzard, coexist- 
ence of mega- and micronephridia (with a characteristic distribution), ‘ Pheretima- 
prostate’, and unpaired spermathecae gives Comarodrilus a somewhat isolated posi- 
tion in the subfamily ; and apparently the genus is without any very near relations. © 
The single series of spermathecae immediately recalls Fletcherodrilus, an Aus- 
tralian genus; but the coexistence of micro- with the meganephridia in the present 
form, and the compact, not tubular, prostate, divide the two genera rather widely. 
In these two characters Fletcherodrilus shows the more primitive, Comarodrilus the 
derived condition; /letcherodrilus however cannot be the ancestor of Comarodrilus, 
since in the former the male pores also have fused in the middle line, while they are 
separate in the latter. 
The present worm recalls Pontodrilus in the limitation of the meganephridia to 
the segments xiii onwards, and in the reduction of the gizzard. It would, however, 
seem to be impossible to derive Comarodrilus from Pontodrilus ; nephridia have dis- 
appeared altogether from the anterior region of the latter, and according to current 
views of evolution they can hardly ,—whether of the same or of a different type,— 
reappear in a descendant. There can of course, in view of the unpaired sperma- 
thecae and compact prostate of the present genus, be no question of deriving Ponto- 
drilus from Comarodrilus. | 
On the whole it seems to me that Woodwardia must be taken as the connecting 
link between this form and the rest of the subfamily. Woodwardia inhabits Aus- 
tralia, Tasmania, Java, Burma (one species) and Ceylon (one species); it is charac- 
terized by a lumbricine arrangement of setae, presence of a gizzard, ‘ Pheretima- 
prostate’, meganephridia only, and, of course, the usual double series of sperma- 
thecae. In those points in which it differs from Comarodrilus, therefore, it shows 
the primitive, Comarodrilus the derived condition ; in other words, Comarodrilus may 
be supposed to have arisen from Woodwardia by the substitution of micro- for mega- 
nephridia in the anterior part of the body, the reduction of the spermathecae to a 
single series, and some degree of degeneration of the gizzard. Megascolides is less 
suitable as a starting point, since the nephridial system is already in that genus more 
broken up than it is in Comarodrilus. 
