REPORT ON THE TUNICATA. 
311 
The Alimentary Canal is of large size, and forms a wide loop. The stomach is 
globular and smooth-walled. 
Locality . — Off the Arrou Islands, September 18, 1874. 
One specimen of this interesting species was obtained off the Arrou Islands, to the 
south-west of New Guinea, from shallow water. It is a moderately large colony, 
attached slightly by part of one edge to some gravel and shell fragments, and expanding 
upwards into an irregularly rounded and lobed form (PI. XLII. fig. 5). 
The specimen is a good deal creased and corrugated on the surface, but that is, 1 
believe, to a large extent the result of preservation in alcohol ; in all probability when it 
was living and all the Ascidiozooids in the colony had their branchial sacs expanded 
with water, it was larger and more rounded, and the surface was more regular than is 
shown in the figure. From its great softness the specimen is readily deformed, and it is 
probably now abnormally flat from having lain on one surface in a bottle for more than 
ten years. 
The colour is the same dull white all over the surface. It is due mainly to the 
superficial layer of test containing spicules. The anterior ends of the Ascidiozooids show 
as small, rounded, more transparent dots on the surface. They are closely placed, and 
are equally numerous over the whole colony. The common cloacal apertures are not 
numerous, and they are not conspicuous. They are of slit-like form. 
The test is gelatinous, and is perfectly clear and transparent in the interior of the 
colony. The outer layer, however, is slightly firmer, and is of a whiter colour and less 
transparent. It contains numerous small spicules, most of which are spherical, with or 
without numerous minute spines projecting from the surface. Some, however, are 
stellate, with many ver}^ delicate, sharp-pointed rays (PI. XLII. fig. 12), and intermediate 
forms between the stellate and the spherical spicules occur. The calcareous sj)icules in 
this colony are absolutely confined to the superficial layer, and it is this which causes the 
difference between the surface layer of the test and the deeper parts. The test cells are 
especially abundant where there are no spicules. 
The mantle allows the branchial sac and the other organs to Ire seen distinctly through it. 
There are two strong muscle bands which run longitudinally down the dorsal edge of the 
thorax. The branchial siphon is large. It is lined by a layer of test, in which calcareous 
spicules are found. These spicules are usually aggregated in three distinct clumps which 
lie just inside the sphincter. The margin of the aperture is not lobed. The branchial 
sphincter is moderately developed (PI. XLII. fig. 13). 
The superficial layer of test along with the Ascidiozooids may be stripped off from 
the rest of the colony (see PI. XLII. fig. 6, which represents such a piece removed from the 
colony, and seen from the lower surface) as the deeper layers of test are very soft and 
easily torn, besides being much cut up by the cavities and passages by means of which the 
