REPOKT ON THE THNICATA. 
161 
as wide as the length of the adjacent stigmata, are produced inwards to form series of 
hollow papillse of various sizes and shapes. These papillse are arranged with a certain 
amount of regularity. They form a single row only on each transverse vessel, and they are 
placed at equidistant points, but they are much more developed in some parts of the sac 
than in others, while in some regions they are altogether absent (PI. XXII, fig. 8). 
Where they are large their free ends show a tendency to split into two or three branches 
(PI. XXII, figs. 7, 10, p.), and this as well as their equidistant arrangement gives them very 
much the appearance of the connecting ducts and rudimentary internal longitudinal bars 
often seen on the transverse vessels of some Simple Ascidians {e.g., Corella parallelo- 
gramma, Corella japonica, and Asciclia tnangularis ^). In some parts of the sac where 
the papillge are rather wide, flattened, and triangular in shape, vdth the ends unbranched, 
they resemble the large triangnilar connecting ducts of some Clavelinidse {Ecteinascidia 
crassa and Ecteinascidia fusca),^ and I think there can be no doubt that they really are 
connecting ducts in a rudimentary condition. The only question is whether this species 
should be regarded as a form in which internal longitudinal bars are beginning to develop, 
or one in which, as the result of degeneration, these structures have been lost, with the 
exception of the papdlse representing the rudimentary connecting ducts. The answer to 
this question involves a discussion of the probable phylogeny of the Compound Ascidians, 
and may therefore be deferred to the general summary at the conclusion of this Report. 
The languets representing the dorsal lamina are long and narrow, and are not 
flattened antero-posteriorly. They have much more of the shape of a tentacle than is 
usual, and closely resemble the languets of Ecteinascidia turhinata ^ amongst Simple 
Ascidians. They obviously correspond to the papillse on the transverse vessels with which 
they are in series (PI. XXII. fig. 12, 1.), and from which they only differ in size. This 
adds to the arguments which I have already brought forward ^ in favour of the languets 
being homologous with the connecting ducts, and not with the papillse which project 
from the internal longitudinal bars in many Simple Ascidians, 
The tentacles are numerous and closely placed (PL XXII. fig. 13, t7\). Most of them 
are long. Some are shorter,. but they do not occur in any definite order. The endostyle 
undulates considerably from side to side (PI, XXII. fig. 6). 
The periphar 5 mgeal band is well marked (PI. XXII. fig. IS, p.p.). It curves pos- 
teriorly in the dorsal region to form a moderately deep peritubercular area in which 
the large but simple dorsal tubercle lies. The opening of this organ is a narrow 
ovate slit, with the wider end posterior (PI. XXII. fig. 13, d.t.). The margin is broad 
and conspicuous. The tubercle touches the peripharyngeal band with its posterior 
1 See Herdman, part i. of tliis Eeport (vol. vL 1882), p. 192, pi. xxvi. fig. 8 ; and Notes on British Tunicata, 
Journ. Linn. Soc. ZooL, vol. xv. p. 283, 1880. 
2 Part i. of this Eeport (vol. vi. 1882), pp. 240, 242, pi. xxxvi. figs. 8, 13, 14. 
3 Loc. cit, p. 243, pi. xxxvi. fig. 6. 
* Loc. cit., p. 284. 
(ZOOL. CHALL. EXP. — PAKT XXXVni. — 1885.) 
Pp 21 
