306 
THE VOYAGE OF H.M.S. CHALLENGEE. 
over the surface (PL XXXIII. fig. 10). Just as in the case of Didemnum aurantiacum, 
the colour is confined to the superficial layer of the colony, where it forms a thin crust. 
In the present species, however, it is not the matrix alone that is coloured, as the spicules 
have a slight but distinct yellow tint. 
The test is firm and nearly opaque even in the lower, almost colourless part of the 
colony. The spicules are very numerous in the upper part of the test (see PI. XXXIII. 
fig. 11, sp.), and less abundant in the deeper parts. They are very regular in shape and 
size (PI. XXXIII. fig. 12). 
The Ascidiozooids, as seen in a surface view of the colony, are of elliptical or ovate 
shape (PI. XXXIII. figs. 10, 11), and are inclined at an angle to the surface, so that the 
greater part of the ventral edge of the thorax, as well as the anterior end, is visible. 
The branchial aperture is usually distinctly seen (PI. XXXIII. fig. 10, hr) surrounded 
by a well-marked sphincter, and the short wide straight endostyle (PI. XXXIII. 
figs. 10, 11, en.) is always a conspicuous object under a low power. The darker spaces 
between the bodies of the Ascidiozooids are formed of test containing closely-packed 
.spicules (PI. XXXIII. fig. 11). 
The branchial sac is larger and the stigmata are more distinct (PL XXXIII. figs. 13, 14) 
than in most allied forms. The transverse vessels are regular, and the stigmata are 
numerous and of very fair size (PL XXXIII. fig. 14, sg.). The ciliated cells are distinct. 
Figure 13 on Plate XXXIII. shows part of the sac of a young Ascidiozooid where the 
stigmata are shorter and more rounded. In the specimen from which this figure was 
drawn there were no muscle fibres in the transverse vessels. 
The dorsal and the ventral tentacles are longer than any of the others. The 
median lateral tentacles come next in size (see PL XXXIII. fig. 15). In each interspace 
l)etween these four primary tentacles four shorter ones are usually present ; these 
vary in length amongst themselves, and are not arranged with perfect regularity 
(PL XXXIII. fig. 15, tn.). It is interesting to find that the two largest tentacles in this 
Compound Ascidian are not in the same position as those which E. van Beneden and 
Julin have shown to be the first developed in the young Simple Ascidian,^ and that the 
next largest pair (the medio-lateral) correspond to the pair which are developed first. 
The order of development in the Compound Ascidian is not known. In von Drasche’s 
Polycijclus violaceus the medio-lateral tentacles are the largest. 
The prebranchial zone is rudely circular in outline, the peripharyngeal bands are 
distinctly marked, and there is no peri tubercular area (PL XXXIII. fig. 15). The 
anterior extremity of the large endostyle encroaches upon the ventral part of the pre- 
branchial zone, carrying the inner peripharyngeal band forwards in front of it. The 
Ijrancliial spliinctcr is well developed (PL XXXIII. fig. 15), and a number of dehcate 
radiating muscle bands start from its edge and run outwards and backwards over the 
* Rtichcrches sur le d(5veloppement post-einbryoimaire d’une Phallusie, Archives de Biologie, t. v. p. 611, 1884.,^ 
I 
