ORIGIN OF METAMERIO SEGMENTATION. 61 



(Hatschek); in this animal (in the adult) most of thenephridia 

 have been lost, the three pairs which persist (two pairs of brown 

 tubes and anal vesicles) being enlarged and modified ; the 

 gangliation of the ventral cord is lost and there are no traces 

 of the somites. 



To sum up in a few words: — The Coelenterata differ 

 from segmented animals only in the fact that the 

 alimentary or archenteric pouches (mesoblastic 

 somites) and the alimentary canal do not become 

 separate; and connected with this absence of a 

 distinct coelom is the low state of differentiation 

 of such coelomic structures as the excretory organs 

 and the absence of a separate vascular system. 



On the Origin of the Excretort Organs. 



This part of my subject is so closely connected with the 

 preceding that it is difficult to separate the two. 



I have already referred to the Hertwigs^ observations ^ on 

 the marginal pores of Medusae and the cinclidesof Actinozoa. 



Metschnikoff was, I believe, the first to observe these 

 marginal pores in Medusae, and he regarded them as excretory ; 

 in this view the Hertwigs concur. 



There is, then, this common feature in the anatomy of the 

 Medusae and Actinozoa ; they both possess peripheral pores, 

 putting the alimentary pouches in communication with the 

 exterior. 



In the Actinozoa they seem to have an irregular distribution 

 as tentacular pores and cinclides (vide Her twig). In the 

 Medusae, however, they have a definite position, one pore for 

 each radial canal. 



It seems an obvious suggestion that in the less specialised 

 ancestors of Medusse and Actinozoa these pores were distri- 

 buted more or less irregularly as in the Actinozoa : that their 

 position was determined by the habits of life and form of the 

 animal. 



• Vide Hertwig ' Organismus der Medusen,' p. 39, and Hertwig, ' Die 

 A-Ctinien.' 



