DESCRIPTIONS OF GENERA AND SPECIES 583 



it is less certain how far it is widely spread. In Libyodrilus, as is pointed out 

 more in detail elsewhere, there is a complicated integumental network formed by 

 the branching and anastomosis of nephridial ducts. I have discovered at least traces 

 of a similar network in Pareudrilus, Eudrttoides, and Stuhlmannia. MICHAELSEN'S 

 failure to find nephridiopores in other forms may very possibly be an indication 

 of a similar state of affairs. On the other hand, there is no trace of this network 

 in E'tulrilus, or any member of the first group. If we are to accept MICHAELSEN'S 

 division of the family, Pareudrilus would have to be placed by the side of Eudrilus 

 and Nemertodrilus. This would be a very heterogeneous assemblage ; but the 

 violence which would be done to the affinities of the genera in question by this 

 unsuitable collocation will be more apparent after the mutual relationships of the 

 different genera have been discussed. 



MICHAELSEN places the genus Eudriloides at the base of the series. In this genus 

 the ovaries remain unenclosed within sacs, and there is no communication between 

 the spermathecal sac and any other part of the female apparatus ; to this might be 

 added the fact that there are occasionally at any rate dorsal pores. The develop- 

 ment of other Eudrilidae in which there is a complicated system of coelomic sacs 

 enclosing the ovaries shows that this state of affairs is secondary ; but to show this 

 it is not necessary to have proved that the simpler forms are more primitive. This, 

 however, is undoubtedly one point in which Eudriloides resembles other Oligochaeta 

 more than does any other genus of Eudrilidae. I have previously referred to the 

 gradual change in the character of the spermathecae in this group (p. 578). Sperma- 

 thecae like those of other Oligochaeta seem to have been replaced by spermathecae 

 which are coelomic sacs ; and there is a series of stages in this replacement. Here, 

 as elsewhere, it might be argued that the change has been in the reverse direction 

 to that which has been suggested. The principal argument to the contrary is, of 

 course, the vast preponderance of families in which the spermathecae are epidermic 

 invaginations ; indeed there are no other forms except the Eudrilidae in which these 

 organs are not entirely formed in this way. In this case those Eudrilidae in which 

 the replacement has not commenced, or has only just commenced, will be so far the 

 most primitive forms ; assuming this for the moment, it is clear that Heliodrilus is 

 the most primitive Eudrilid. Moreover, in this genus, as in the group of genera 

 round it, the calcifurons glands are more like those of other earthworms than are 

 the peculiar structures of Eudriloides; these latter are more easily to be derived 

 from calciferous glands of the ordinary type than are the typical calciferous glands 

 from them (see p. 62). Their number too and the segments in which they occur 

 are more like what is usually met with in other Oligochaeta. On the other hand, 



