252 PEECY SLADEN TEUST EXPEDITION. 



oblique septum on eitlier side from the anterior portion of the splanchnocoele. The 

 latter, termed by Van Wijhe the " stomocoele," has become further subdivided and 

 forms a complex system of c-avities in relation to the mouth, for details of which we 

 must refer to his description (1902): here we need only note its complete division by a 

 ventral septum into asymmetrical right and left halves — a division of which tliere is no 

 trace in the larva, — and the fact that the only communication with the hinder parts of 

 the splanchnocoele is tliat witli the hypobranchial ccelom by the ccelomic canals of the 

 first pair of gill-liars. Further, the pterygocoeles, whose connection with the splanchno- 

 coele in AmpMoxides is at best doubtful, in the adult open freely round the anterior ends 

 of the pterygial muscles into the right and left " cava epipterygia " — the posterior 

 portions of the stomocoele. All trace of a connection between splanchnocoele and cavity 

 of the 1st myotome has disappeared. 



How are \^e to account for the separation of anterior and posterior portions of 

 the splanchnocoele, found in the adult Branchiostoma, but not in AmpMoxides ? 

 Van Wijhe (1906) sees in it the primitive separation of "mesosoma" and " metasoma " ; 

 but in view of the fact that there is no trace of any division in Ampliioxides, this 

 supposition can hardly be correct. It gains some support, it is true, from Macbride's 

 account of the development of ^raJicAios/owm (1898), according towhicli the ventral part 

 of the " collar cavity " would appear to remain separate from the rest of the splanchno- 

 coele, at any rate for a great part of larval life *. But I have little doubt that a 

 re-investigation of the Branchiostoma larva will show it to possess a simple undivided 

 splanchnocoele similar to that of AmpMoxides ^ : the identity of tlie anterior part of the 

 latter with Macbride's " collar cavity " is clearly shown by the wide communication of 

 both witli the cavity of the 1st right myotome. 



Van Wijhe's own demonstration that tlie adult position of the mouth is secondary 

 seems to point the way to a truer view of the nature of the stomocoele, viz., that its 

 separation and subdivision are purely secondary, and connected with the new position 

 and complex functions of the mouth. These events should, I believe, be interpreted 

 physiologically rather than morphologically. 



One other i)oint in Macbiide's description may be noticed here, viz., tlie great backward 

 extensions of the collar-cavities in the ventro-lateral angles of the body of the young 

 larva, originally supposed by him to give rise to the metapleural cavities. He has since 

 abandoned this view, and it is difficult to see how it can be correct, since the metapleural 

 folds, according to Laukestcr and AVilley's account, develojj from liehind forwards. 



* Even if the ventral part of the " collar cavity" does remain separate, however, Van ^Yijhe's complete itleiiti- 

 fication of the adult stomocoele therewith does not seem to be justified. Arcordiiig to llarbride's account, the 

 "collar cavity " only gives rise to a single pair of myotomes, while the stomoca4e must, on Van Wijhe's showing, 

 include the ventral portions of at least two segments. 



t It may be permissible to suggest that the apparent separation of the ventral part of the " collar cavily" from 

 the rest of the splanchnocoele, in Macbride's larva;, was due to the collapse of the walls of the latter cavity in the 

 anterior pharyngeal region, where he only saw it as a solid mass of cells. Aviphioxidt-s, owing to the greater 

 resistance of its tissues, is a much more favourable object fur the study of the coelomic cavities than the 

 Branchiostoma larva, and it is hardly probable that the cavities diti'er materially in the two forms. 



