318 BULLETIN OF THE 



lids and those of Vertebrates differ conspicuously, viz. in the mode in 

 which the nephridia terminate. It is a familiar fact, tliat in Annelids 

 each nephridium opens separately on the surface of the body ; while in 

 Vertebrates the nephrostomal tubules all connect, with a pair of longi- 

 tudinal ducts opening into the cloaca. In commenting upon this fea- 

 ture of difference, it is important to note in the first place that the 

 contrast is not of universal application. It has been recently shown by 

 E. Meyer ('87, pp. 618-625) and Cunningham ('87% 87% pp. 248-253) 

 that in Lanice conchilega, a terebelluid Annelid, certain nephridia open 

 into a longitudinal trunk, and only secondarily communicate with the 

 exterior. On the other hand, it is probable that Amphioxus possesses 

 nephridia which open to the exterior (atria cavity) without the inter- 

 vention of a longitudinal duct. If such differences can occur among the 

 members of either group, it seems to me that it would be unjust to deny 

 the homology of the other portions of the system in consequence of 

 the fact that Vertebrates in general possess a longitudinal duct, while 

 Annelids in general do not. It appears to me, moreover, that the con- 

 dition of the nephridia in Lanice conchilega and the ontogeny of Verte- 

 brates both serve to indicate the manner in which the duct may have 

 secondarily arisen. In Lanice conchilega there is no doubt tliat the 

 nephridial duct is a secondary growth, and it is highly probable that the 

 channel is formed by outgrowths extending from each of the nephridial 

 tubes backward and communicating with the next following nephrid- 

 ium. Two groups of nephridia can be distinguished in Lanice con- 

 chilega. The more anterior of tliese consists of a short longitudinal duct 

 which bears three nephrostomal tubules, and terminates at its posterior 

 end by a single pore. In the posterior set, the longitudinal duct is 

 merely a canal which connects the several nephridia, while these con- 

 tinue to retain their external orifices, I have already pointed out 

 that the ontogeny of Vertebrates presents a similar process in the devel- 

 opment of the longitudinal canal of the pronephros, and have shown 

 that such changes may likewise have taken place in the region of the 

 mesonephros. I by no means wish to imply by this comparison a belief 

 that the ordinary mode of development in Vertebrates is to be directly 

 derived from that presented by Lanice conchilega, nor to assume a close 

 genetic relation between Vertebrates and genera presenting this condi- 

 tion. I merely wished to emphasize the fact, tliat in Lanice conchilega 

 we have an instance of a species which, primitively possessing discrete 

 nephridia, such as may have been pi'esent in the ancestor of Vertebrates, 

 has acquired a longitudinal excretory canal by a process of transforma- 



