44 OLIGOCHAETA 



In all of these worms there is no muscular sac at the points of opening on to the 

 exterior ; favourable sections are wanted in order to demonstrate the actual pores, 

 since the tubes are of narrow dimensions up to the very opening. Generally speaking, 

 the cells of the epidermis show some peculiarities at the orifices ; in Perichaeta I have 

 figured the conditions which are found ; and SPENCKR has figured them in Megascolides. 

 The epidermic cells become slightly narrower and converge round the actual orifice, 

 forming possibly a kind of constrictor for regulating the outflow of secreted matter. 

 As already mentioned, the funnels may be present or absent in the diffuse nephridia ; 

 when present they do not seem to differ at all from those of the paired nephridia, 

 except that they are generally smaller. BENHAM has carefully described the funnel 

 in Perichaeta malamaniensis ; there are eight or nine marginal cells, but no central 

 cell nor any centrifugal gutter-cells ; the intracellular duct of the tube appears to 

 open at once into the midst of these cells. I have, however, described a horseshoe- 

 shaped funnel in Perichaeta bermudensis. 



Nephvidia of this kind do not exist in any of the aquatic Annelids, nor have 

 they been met with in any other groups of Annelids excepting only the leech 

 Pontobdella, where BOURNE has stated that a network exists and is continuous from 

 segment to segment ; this statement, however, has been denied. 



Integumental nephridial network. The network that exists in certain Eudrilidae is 

 not, I believe, morphologically comparable to the diffuse nephridial system already 

 treated of. This condition of the nephridial system appears to characterise a large 

 number of Eudrilidae, and I have attempted a classification of that family partly based 

 upon the presence or absence of the network. It has been studied most thoroughly in 

 the genus Libyodrilus. That worm, like all the other genera of the family to which it 

 belongs, has paired nephridia a pair to each segment. But the 'duct leading to the 

 exterior, instead of passing at once to the exterior, branches and forms a complicated 

 network in the integument ; in this worm the peritoneal layer which lines the body- 

 wall is, in places at any rate, excessively thick ; and in this layer tubes formed by 

 the branches of the ramified external duct run ; these put into communication the 

 nephridia of successive segments. The actual details of the way in which this complex 

 network is formed are perhaps subject to some variation. It appeared to me, however, 

 that there were four principal and longitudinally running trunks, symmetrically 

 disposed two on each side of the nerve-cord (corresponding in position to the setae) ; 

 from these branches arose which ramified in every direction through the longitudinal 

 muscular layer and finally joined a circular vessel running right round the body 

 between the two muscular coats. From this latter fine branches lead to the exterior. 

 These tubes are nowhere ciliated and seem to be not comparable to the coelomic 



