REPORT ON THE TUNIC ATA. G5 



more gradually into the larger middle part. The body is rounded anteriorly, while 

 posteriorly it is truncated. This difference is due to the shapes of the apertures. The 

 branchial is bilabiate, and the anterior parts of the lips arc rounded, — the dorsal 

 projecting slightly farther forward than the ventral. The atrial aperture, on the other 

 hand, has no distinct lips, but is merely the circular termination of the atrial siphon 

 (PI. VL figs. 1 and 3). 



The dimensions given above are those of one of the specimens obtained on April 12, 

 1876. The other is 3 '2 cm. in length and nearly 1 cm. in breadth. The two specimens 

 collected on April 13, 1876, measure 3"2 cm. and 2"8 cm. in length, while their greatest 

 breadth is about 1 "3 cm. All the specimens are from the Atlantic ; they are in rather 

 bad condition. 



All the specimens of this species are of an opaque yellowish-grey colour ; but this is 

 due more, I believe, to the strongly-developed musculature of the mantle than to the test. 

 In the narrow intervals between the muscle bauds, the test appears clear and transparent. 

 The longitudinal ridges on the test are very sbghtly thicker than the rest of the surface. 

 They are, however, due more to folding than to thickening. They are not echinated. 



The musculature is very remarkable (PI. VI. figs. 1, 3, and 4). It is in a condition 

 paralleled only by the solitary form of Salj^a hexagona. The muscle bands are so wide 

 that they almost touch their neighbours, and here and there they do touch and fuse 

 with one another. There are nine or ten bands on each side, of which the first is single 

 in the dorsal middle line (PI. VI. figs. 3 and 4), but breaks up as it runs outwards to 

 the sides into three branches, one of which continues its course transversely, while the 

 other two turn forwards towards the branchial aperture (PI. VI. fig. 4). If these two 

 anterior branches be regarded as distinct muscle bands, then the total number would be 

 twelve, of which the first, second, and third fuse dorsally. At the posterior end of the 

 body it becomes difficult to distinguish the exact number of bands as they lie close 

 together and partially fuse at intervals. In Salpa hexagona, solitary form, there 

 are eleven bands, of which the posterior is partially divided, while the arrangement 

 at the anterior end of the body is quite different from that seen in the present species. 



The endostyle is large (PI. VI. fig. 3. en.), and extends farther forwards than the 

 anterior end of the dorsal lamina. The peripharyngeal bands run dorsally and 

 posteriorly. The dorsal lamina is attached to the mantle as far back as the third 

 muscle band (PL VI. fig. 4, d.l). Its ribbed portion is narrower than usual, and the 

 transverse ridges are slight and irregular (PI. VL fig. 4). The dorsal tubercle scarcely 

 projects from the surface of the triangular peritubercular area (PL VL figs. 3 and 4), 

 and the languet is apparently absent. The ganglion is placed on the first, or at the 

 junction of the first and second muscle bands. It is farther back than in most species 

 of Salpa, but not nearly so far back as in Salpa hexagona, where, according to 

 Traustedt, it lies on the fifth muscle band. 



(ZOOL. CHALL. EXr.— PART LXXVI. — 1SS3.) Gggg 9 



