404 



DR ARTHUR T. MASTERMAN ON THE 



and of Semon brings us no nearer any other phyla of the animal kingdom ; and if the 

 structure of an early ancestor does not fulfil this condition, its credentials are open to 

 question, and its construction does not seem even to possess the virtue of utility. In 

 fact, the Pentactcea (quite apart from other objections urged by Bury, Clark, Ludwig 

 and others) appears to me an instance of precocious segregation, in which the typical 

 echinoderm axo-sym metric character of a hydroccele ring has been grafted on to the 

 simple character of an early bilateral ancestor. Such an anachronism occurs in 

 ontogeny, but not in phylogeny. 



The equivalence of the hydroccele and epigastric coelom in Cribrella leads us to 

 regard their parent organs, the left and right lateral cceloms, as being comparable to 

 the two collar-cavities of Balanoglossus, and the right and left posterior cceloms as 

 the homologues of the two trunk-cavities. Going further afield, we may recognise in 

 the pre-oral ccelom, the lateral and posterior cceloms, — the protoccele, the mesocceles, 



A B 



^ire-oral ccelom 



ityoiroport..^ 



prs-OrCcl JO.C 



_, central ccelom '■jrolmUJim 



right latera. 

 caslom 



prc-oraZ ceelon. 



colla.'r cxla-rn 



tru-nk ccelom. 



3, — Diagrams showing relationships of Cribrella larva (stage D) to Balanoglossus. (A) Dorsal view of Cribrella larva (stage D), 

 with the parts named. The water-pore and the central coelom are supposed to be pushed back into their true position. 

 (B) A dorsal view of a young Balanoglossus, with the parts named. (C) Diagrammatic coronal section of larva of 

 Cribrella at stage B. (D) Diagrammatic coronal section through larva of Balanoglossus. The hypoblast is indicated 

 black in each case. 



and the opisthocceles respectively of the archi-ccelomate, the central ccelom being also 

 comparable to the archimeric heart.* Our bilateral ancestor, as derived from stage D, 

 therefore takes us to the junction of echinoderms with numerous triploblastic Metazoa, 

 with archimeric segmentation. The 'ancestor' as here outlined would unquestionably 

 take its place in this group, and probably also would be placed beside Balanoglossus in 

 the Archi-Chorda (Hemichorda). 



It will be seen that this ' bilateral ' ancestor differs from that of such authors as 



* Proc. Boy. Soc. Edin., Dec. 1898. 



