216 PEOr. HUXLEY OK THE 



nikoff and partly by Agassiz himself, to be the larval form of 

 Balanoglossus. In JBalanoglossus, as in tbe Echinoderms, saccular 

 diverticula of tbe intestine appear to give rise to the perivisceral 

 cavity and its walls. In the Chaetognatha, Echinodermata, and 

 Enteropneusta, therefore, tbe perivisceral cavity is a portion of 

 the alimentary cavity shut off from the rest ; and in contradistinc- 

 tion to the Schizocoela, in which the perivisceral cavity is pro- 

 duced by a splitting of the mesoblast, they may be said to be 

 Enteroecela. 



The Epiccela. — In the Ascidians, the investigations of Kowa- 

 lewsky, now confirmed in all essential points by Kupffer, have 

 shown that the alimentary cavity is formed by the invagination 

 of the vesicular Morula, that the blood-channels answer to the 

 blastocoele, that the central nervous system is produced by inva- 

 gination of the epiblast, as in the Vertebrata, and that, in most, 

 the mesoblast of a caudal prolongation gives rise to an axial 

 column flanked by paired myotomes, which are comparable to the 

 notochord and myotomes of the vertebrate embryo *. 



In the simplest Ascidians (the Appendicular iw) the modified 

 pharynx, which constitutes the branchial sac, is perforated by 

 only two apertures, which open on the haemal or ventral face of 

 the body, and there is no atrial chamber. But in all other Asci- 

 dians an invagination of the epiblast takes place on each side of 

 the anus, and, extending alongside the branchial sac nearly as far 

 as the endostyle, give rise to a spacious chamber, lined by the so- 

 called atrial or "third" tunic. In many Ascidians the chamber 

 extends much further, so that even the alimentary canal and the 

 generative organs are situated between the atrial tunic and the 

 ectoderm. In this manner a kind of "perivisceral cavity" is 

 formed, which is of a totally diiferent nature from the " schizo- 

 coele " of the Annelid, and from the " enteroccele " of the Echino- 

 derm, and which may be termed an epicoele. 



The resemblance of the simplest of vertebrated animals, the 

 Lancelet {Ampliioxus lanceolatus), to the Tunicata was first in- 

 dicated, though, it must be admitted, very vaguely, by Goodsirf. 



* It is with great diffidence that I venture to express my dissent from the 

 views of my venerated friend Von Baer, from whose works I first gathered 

 soimd principles of morphological science, and whose authority in such a matter 

 as this has no equal ; but I cannot think that the doubts he has expressed re- 

 specting the fundamental similarity between the Ascidians and the Vertebrata 

 are warranted. 



t "On the Anatomy of Amphioxus lanceolaius." Eead before the Eoyal 



