DESCRIPTIONS OF GENERA AND SPECIES 579 



which I have myself examined, there is a cup-shaped layer of cells which seem to be epidermic; 

 they are not, however, continuous with the epidermis, so that there may be some mistake in the 

 interpretation which I advance here and have advanced before (84) ; this layer of cells does not, as 

 does the corresponding layer in Paradnlm, immediately surround the lumen of the large spermathecal 

 sac ; the cells of the latter, which seem from their resemblance to the cells in the spermathecal sac of 

 Hyperiodriltis, &c., to indicate a mesodermic origin for the sac in question, are continued right on 

 to the external pore, forcing their way between the epidermic cells, which they have pushed aside. 

 The last stage is, perhaps, shown by the genera Libyndrilus and Polytoreutus ; in the former, at any 

 rate, there seems to be but the merest trace, if any, of an epidermic invagination to meet the 

 spermathecal sac ; the latter I have shown to be of mesodermal origin ; in sections through a non- 

 mature individual of Lilywli-ihts I found the spermathecal sac, as it were, burrowing its way through 

 the muscular layers of the body-wall on its way to reach the exterior. In Polytoreutus also there 

 can be but little of the spermathecal sac which is of epidermic origin. We have, therefore, a series 

 showing the gradual decrease in importance of the spermathecae and their replacement by a large 

 coelomic sac which serves the same purpose. 



MICHAELSEN has divided the Eudrilidae into two subfamilies Eudrilini and 

 Teleudrilini. The latter he defines as follows : 



' Die Teleudrilinen sind meganephridische, mit 4 Borstenpaar-Reihen ausge- 

 stattete Terricolen, die eine einzige, ventral-mediane mannliche Geschlechtsoffnung auf 

 oder am 17. Segment und eine einzige ventral-mediane Samentaschen-Offnung binter 

 der Intersegmentalfurche 10/11 besitzen.' 



The only distinguishing mark of the subfamily is therefore the unpaired character 

 of the sexual apertures ; in the other points mentioned there are no differences from 

 the Eudrilini (of which MICHAELSEN gives no definition). Our knowledge of other 

 Oligochaeta seems to me to render it unwise to attempt to make a wide distinction 

 on the grounds of the paired or unpaired condition of the generative apertures. The 

 genus Fletcherodrilus would hardly be constituted a representative of a distinct sub- 

 family of the Cryptodrilidae by reason of its unpaired orifices. To speak of the 

 Teleudrilini as a ' zweifellos " natlirliche " Gruppe ' is to my mind far too strong an 

 expression. The unpaired genital orifices would be a matter of greater importance, 

 if they were invariably associated with other anatomical differences. This is, however, 

 by no means the case. A glance at the following tabular statement of the main 

 structural features of the Eudrilidae indicates a better line of division. 



I have already advanced this view of the affinities of the different genera (84). 

 In the genus Eudrilus there are a number of anatomical peculiarities which distinguish 

 it from other earthworms ; in the first place, of course, the remarkable structure of 

 the female reproductive system ; secondly, the opening of the sperrn-clucts into the 

 glandular part of the spermiducal gland a relation met with in no ' earthworm ' 

 except in the Moniligastridae. Apart from these characters which EudrUus shares 



4 E 2 



