5fi THE VOYAGE OF H.M.S. CHALLENGER. 



The body is elongated and more or less fusiform, with the branchial and atrial 

 apertures at the opposite ends of the l)ody. These openings are not surrounded by a 

 number of lobes as in many other Tunicata, but are usually tubular or bilabiate. Each 

 leads into a large space, the branchial sac and the peribranchial cavity, and these are in 

 free communication at the sides of the median dorsal lamina or "gill" (Fig- 7, d.I.), 

 which is a narrow vascular baud running obliquely from the dorsal and anterior to the 

 ventral and posterior end of the cavity. The water therefore has a free passage through 

 the body of the animal — in at the branchial aperture through the branchial sac, past the 

 sides of the dorsal lamina into the peribranchial cavity, and ont by the atrial aperture. 



The test is well developed, and in most species is thick but soft and transparent. 

 It may become thicker and firmer in some parts, and usually there is a protecting shield 

 of this nature over the visceral mass near the posterior end of the ventral surface 

 (Fig. 7, 0- 



The mantle adheres closely to the inner surfece of the test. It has its musculature 

 in the form of a variable number (generally from six to eight) of muscle bands which run 

 transversely across the dorsal surface of the body and die away on the sides, so that the 

 ventral part of the mantle has no musculature. Thus the muscle bands do not form 

 complete rings as they do in the genus Doliolum, but are merely curved bands partly 

 encircling the body. In many cases certain of the bands join or closely approach one 

 another (see PL VI. fig. 5), generally in the median dorsal line ; and all such arrange- 

 ments of the muscle bands as well as their number are constant in the species and in the 

 aggregated and solitary forms, and constitute one of the most important characters in the 

 identification of specimens. 



The branchial sac is a large cavity, but its lateral walls are entirely absent. The 

 endostyle (Fig. 7, end.) indicates its ventral edge, and the dorsal lamina {d.I.) or " gill," 

 its dorsal, so that in a lateral view, such as Fig. 7 represents, its lioundaries are per- 

 fectly definite ; but there are no side walls joining the endostyle to the dorsal lamina 

 and separating the branchial sac from the lateral parts of the peribranchial cavity. 

 There are therefore in connection with this branchial sac no transverse or longitudinal 

 vessels, and no stigmata. It is exactly as if in an ordinary Ascidian all the stigmata 

 (m each side of the sac had coalesced to form a single large aperture. The so-called 

 "gill" of Scdjjo, then, evidently corresponds to the dorsal lamina of the Ascidian 

 with the large dorsal blood-sinus which lies behind it. 



The anterior end of the " gill " or dorsal lamina is prolonged ventrally to form a 

 prominent tentacular organ, the languet, which projects freely into the anterior end of 

 the branchial sac. It is probaltly a sense organ (Fig. 7, /). The nerve ganglion is 

 placed near the front of the body in the median dorsal line. It is short and rounded, 

 and gives off nerves from the sides as well as from its anterior and posterior ends (see 

 PI. VIII. fig. 15) ; an otocyst (?) and a pigment spot are found in connection with the 



