1894- EARTHWORMS AND THEIR ALLIES. 49 



cell. This accounts for the non-discovery of a nucleus within the 

 muscular sheath of the fibre which was wanting to complete its Hke- 

 ness to the fibre of a Leech. A little more fanciful, perhaps, is my 

 own comparison of the remarkable integumental network formed 

 by the ducts of the paired nephridia, which ramifies through the 

 body- wall of Lyhiodrilns and more than one other Eudrilid, to the 

 intra-integumental nephridial tubes of the Nematoidea ; particularly 

 to the coil of vessels which constitutes the Lemnisci of the Acan- 

 thocephala lately shown to exist in the true Nematoidea. In the 

 Eudrilids in question the entire integument is riddled through and 

 through with a complex system of tubes, all derived from the 

 branching of the duct, by means of which the paired nephridia 

 reach the exterior. There are, therefore, here as in PericlicBta 

 and some other forms, many external nephridial pores upon each 

 segment of the body ; but this same result is arrived at in a 

 different way in the two cases. The most natural allies of the 

 Oligochaeta would seem to be the marine Polychaeta ; and yet recent 

 researches have not tended to break down to any great extent the 

 barriers separating these two groups. One character, however, 

 hitherto held to differentiate the two has broken down. Cuvier called 

 the Oligochaeta as we nov/ know them " Annelides abranches setigeres." 

 In 1855, Grube described, unfortunately too insufficiently, a curious 

 Annelid from the Nile which he called Alma nilotica ; the same species 

 has been lately re-discovered by Levinsen (12), who renamed it 

 Digitibranchus niloticns. The worm has a series of tufted gills on the 

 posterior segments. But it is not yet clear that the worm is not really 

 a Polychaete. Apart from this dubious instance undoubted Oligo- 

 chaeta are now known to possess gills. Bourne was the first to 

 discover this important fact ; he found at Calcutta a Naid having a 

 series of paired processes of the body-wall in the anterior segments. 

 Later still I found in the " Victoria regia Tank " at the Botanical 

 Gardens a worm allied to the common Ttihifex of our streams and lakes 

 with a series of dorsal and ventral gill-like processes upon the hinder 

 segments; and more recently still (5) another Tubificid, belonging to 

 a new genus Hesperodrilus, in which exactly the same kind of gills are 

 present but lateral in position. Beyond these instances no Oligo- 

 chaete has been discovered which places the Oligochaeta nearer to 

 the Polychaeta than they were at the time when Claparede wrote. 

 The phylogenetic arrangement of the Oligochaeta must, therefore, 

 still remain the aim of future investigation. The difficulties which 

 surround the solution of the problem are not so much due to missing 

 links in the chain as the inability to decide at which end of the chain 

 was the beginning. Recent discoveries have rendered it in some cases 

 almost an impossibility to distinguish family from family by any salient 

 characters. 



We may turn now to matters of special anatomical interest not 

 having so great a bearing upon the inter-relationships and phylogeny 



£ 



