248 GILBERT 0. BOURNE. 



the oviducts and the anus. Posteriorly tlie atrium is pro- 

 longed backwards on the right side of the body as far as 

 the lowest limit of the bend of the intestine, but there is 

 no corresponding cavity on the left side. Anteriorly the 

 atrium extends forward between the stomach and the dorsal 

 body-wall (ectoplerome) to about the middle of the length of 

 the body, and is there interrupted by the dorsal suspensory 

 fold (fig. 34, D. sus.) ; but laterally it extends right and left 

 of the stomach and oesophagus towards the mid-ventral line, 

 to form the right and left peribranchial cavities {Qg. 13, 

 R. phr. and L. pbr.), separated from one another by the 

 ventral suspensory fold. The peribranchial cavities are pro- 

 longed forward far beyond the limits of the atrium itself; 

 as shown in fig. 12, they are completely separated from one 

 another (except for their communication by way of the 

 atrium) before they reach the level of the branchial sac. 

 In the region of the branchial sac the peribranchial cavities 

 continually diminish in size as they pass forwards (figs. 8 to 

 11), until in the most anterior branchial region they are 

 reduced to mere canals, which lie dorsally right and left 

 of the branchial sac, and eventually end blindly in front. 

 The left peribranchial cavity is deeper dorso-ventrally, and 

 extends rather further forward than the right. The limits 

 of the right peribranchial cavity are indicated by the deeper 

 shade of blue in fig. 34. 



The walls of the atrium and peribranchial cavities, where 

 they are not bulged by the gonads, intestine, and other viscera 

 embedded in the ectoplerome, are for the most part smooth 

 and devoid of ridges and folds. But in the oesophageal 

 region a thick, longitudinal ridge of the ectoplerome, covered, 

 of course, by atrial epithelium, projects into the peri- 

 branchial cavity of either side. As these ridges are highly 

 vascular — that is to say, they contain an exceptional number 

 of blood-spaces crowded with corpuscles, — they are probably 

 respiratory in function, and I regard them as the representa- 

 tives of the endocarps of other Ascidians. 



