eitter: octacnemus. 243 



can be no doubt about their identity. Their general character, but 

 most of all their relation to other clearly identifiable parts, leave no 

 room for hesitation. They are, as will be noted especially from Figure 

 9, Plate 3 situated on each side of the dorsal lamina, d. I., and behind 

 the ganglion and peripharyngeal band. Furthermore, as we shall see 

 presently, though far removed from the endostyle, their general relation 

 to this organ is as it should be. Absence of cilia and the peculiar struc- 

 ture of the bordering epithelium are probably associated with the fact 

 that the stigmata are no longer functional as respiratory organs. 



But while there is no doubt that these orifices are branchial stigmata, 

 and consequently that the membrane in which they occur is the strict 

 homologue of the branchial membrane of other ascidians, the fact that 

 only the inner one of the two layers above pointed out seems to consti- 

 tute the visceral wall (Figs. 3, 5, 8, and 9 in. and the layer immedi- 

 ately beneath it) is perforated by the stigmata, does raise a diflScult 

 question as to the peribranchial cavity, i. e. the cavity into which the 

 stigmata of the typical ascidian open externally. 



Something of the character of this external, unperforated layer is 

 stiggested by the facts that a series of six or eight distinct, though 

 rather delicate, muscle bands (Fig. 9, m. h' .) are situated in it, and ex- 

 tend across the median dorsal line, and are disposed at nearly regular 

 intervals from before backward ; and that the same layer extends over 

 the gonads (Fig. 3, m-). The suggestion from these facts is that the 

 layer belongs in reality to the mantle ; and its resting immediately upon 

 the branchial membrane suggests farther that the peribranchial cavity 

 Jias become ohliterated, or, more exactly, reduced to the very narrow 

 interval between the two membranes. On this interpretation the large 

 cavity beneath the " horizontal membrane " in the oral disc, which was 

 above regarded, with some doubt, as the peribranchial-atrial chamber, 

 would not be such ; at least would not be peribranchial (the condition 

 here described was the occasion for the reservation as to identification, 

 not indicated in my treatment of that subject). 



I am unable to reach entire clearness on these points. It is possible 

 that further study on additional material, particularly on young or de- 

 veloping specimens, will find that the atrial chamber is here distinctly 

 set off from the peribranchial, and that the great interior space already 

 described is atrial and not at all peribranchial; or it may be that this 

 whole space is artifact, as I have interpreted the portion above the 

 horizontal membrane to be. This, however, seems hardly probable, 

 though such a view would furnish an explanation of the apparent 



