166 OLIGOCHAETA 



from all Megascolides in a number of important points, of which the following are 

 the most important: 



(1) The spermathecae, if present, are in the neighbourhood of the female gonads ; 



they are always accompanied, and generally replaced, by coelomic sacs, 

 which apparently perform the same function. 



(2) The sperm-ducts always open into the glandular part of the spermiducal 



glands. 



(3) The spermiducal glands open by a terminal copulatory chamber. 



(4) Both paired and unpaired calciferous glands exist, or they are replaced 



by peculiar structures of probably a different function. 



(5) The ovaries are nearly always involved in coelomic sacs. 



(6) The nephridia are always paired, and sometimes form a network in the skin. 

 There seems to be no doubt that the remarkable development of coelomic spaces, 



which is so characteristic a feature in the organization of the Eudrilidae, is a secondary 

 development. But, on the other hand, the group shows certain primitive features. 

 This is particularly the case with the spermiducal glands, which show, in the 

 possession of a terminal copulatory chamber, a close agreement with the copulatory 

 glands from which they are, in my opinion, derivable. As a rule, there are no 

 traces of these glands other than the spermiducal glands in the Eudrilidae ; but 

 MICHAELSEN has described and figured in Reithrodrilus minutus something which 

 appears to be of the same nature. Mainly on account of the frequent presence of 

 such glands in the Geoscolicidae, and the absence in them of coelomic spaces 

 surrounding the genitalia, I should be disposed to regard that family as more 

 primitive than the Eudrilidae. 



The relation of the Geoscolicidae to the Megascolides is not a very near one, 

 and it is furthermore not an easy task to decide which of the three families of 

 the Megascolides has most claims to have given rise to the Geoscolicidae. The 

 Perichaetidae may, I think, be dismissed from consideration. The only positive 

 point of resemblance is in the intestinal caeca of Urobenus, which not only are 

 like those of Perickaeta, but occur in the same segment. In no Perichaetous worm 

 are the setae ornamented ; this is one of the most general characters of the 

 Geoscolicidae ; nor are there any segments in the anterior part of the body in 

 Perichaeta free from setae ; this is another character often met with in the 

 Geoscolicidae. It is among the Cryptodrilidae that all the characters distinctive of the 

 Geoscolicidae occasionally occur; in Deodrilus we have ornamented setae and an absence 

 of setae from the most anterior segments of the body. The long sperm-sacs of the 

 Geoscolicidae (such as occur in Trichochaeta, Pontoscolex, &c.) have their counterpart 



