ORIGIN OF METAMBRTO SEGMENTATION. 89 



blast), which was present at the hind end of the blastopore in 

 the earliest stage (fig. 1), has become marked with a groove 

 (fig. 4). In the paper referred to, the question — Do these two 

 openings become the mouth and anus of the adult? — ^was left 

 open. I am now in a position to state that they do become 

 the mouth and anus of an embryo of an age equal to the oldest 

 stage described by Moseley^ in his original paper, so that I 

 think there can be no doubt that they do become the mouth 

 and anus of the adult. 



Thus, then, we have two undoubted facts : 



1. That the mouth of the Actinozoa is differentiated into 

 one portion for the exit and another for the entrance of 

 matter, and that this differentiation is carried so far as to 

 give rise to two separate openings (Peachia). 



2. In the development of Peripatus capensis the single 

 opening of the gastrula elongates, then divides into two parts, 

 an anterior part which becomes the mouth, and a posterior 

 which becomes the anus of the adult. 



The argument may here be briefly summarised : 



1. The blastopore of Annelida, Arthropoda, Mollusca, and 

 the mouth of Coelenterata are homologous because (a) of the 

 development {b} of the anatomical relations in each case. 



2. The structure of the mouth of Actinia and the position of 

 the mouth and anus within the primitive nerve-ring, which is 

 supposed to be homologous with the circumoral nervous dif- 

 fusion of Actinia, obviously suggests the derivation of the 

 mouth and anus from a single opening like the mouth of 

 Actinia by the completion of the fusion which is there 

 beginning. 



3. The blastopore of Peripatus, which by hypothesis is homo- 

 logous with the Coelenterate mouth and with other blastopores, 

 actually passes through the Actinia phase. 



Is this development primitive ? If it is primitive, then as 

 the mouth and anus of Peripatus are homologous with those of 

 Annelids, my point is gained and we shall have to take the 

 second alternative (p. 86), and suppose that the peculiar 



1 ' Phil. Trans.,' 1874. 



