LIV
3005

P

Utverpool fIDaiine Biology Committee.

L.M.B.C. MEMOIRS

ON Typical British Marine Plants & Animals

EDITED BY W. A. HERDMAN, D.Sc, F.R.S.

I.

ASCIDIA

BY

W. A. HERDMAN, D.Sc, F.R.S.

Rrofessor of Naliiral History in University College. Liverpool

Hon. Director of the Port Erin

Bioloncal Station

(With 5 Plates)

Price Eighteenpence

LIVERPOOL

T. DoBB & Co., Printers, 229 Brownlow Hii.l

October, 1899

HARVARD UNIVERSITY.

LIBKARY

MUSEUM OF COMPARATIVE ZOOLOGY.

JLtJLyWolybv- ^^V^VCi

L.M.B.C. MEMOIRS

I.
ASCIDIA

NOTICE.

The Committee desire to intimate that no copies of these
Memoirs will be presented or exchanged, as the prices
have been fixed so low that most of the copies will have
to be sold to meet the cost of production.

The Memoirs may be obtained, post free at the nett prices
stated, from the Hon. Treasurer, Mr. I. C. Thompson, 53,
Croxteth Road, Liverpool ; Prof. Herdman, University
College, Liverpool; or the Curator, Biological Station,
Port Erin, Isle of Man.

Memoir I. Ascidia — now ready, 60 pp. and five plates,
price 1/6.
,, II. Cardium — in a few weeks, price 2/-.
,, III. Echinus — at Christmas, price 1/-.

Liverpool marine Bioloc^y Committee,
L.M.B.C. MEMOIRS

ON Typical British Marine Plants &- Animals

EDITED BY IV. A. HERD MAN, D.Sc, F.R.S.

I.
ASCIDIA

BY

W. A. HERDMAN, D.Sc, F.R.S.

Professor of Nalmal History in University College, Liverpool

Hon. Direclor of the Port Erin

Biological Station

(With 5 Plates)
Price Eighteenpence

1
^ LIVERPOOL

T, DoBH & Co., Printkrs, 229 BuowNLOvv Hill

October, 1899

EDITOE'S PKEFACE.

The Liverpool Marine Biology Committee was constituted
in 1886, with the object of investigating the Fauna and
Flora of the Irish Sea.

The dredging, trawling, and other collecting expeditions
organised by the Connnittee have been carried on inter-
mittently since that time, and a considerable amount
of material, both published and unpublished, has been
accumulated. Thirteen Annual Eeports of the Committee
and four volumes deahng with tlie "Fauna and Flora"
have been issued. At an early stage of the investigations
it became evident that a Biological Station or Laboratory
on the sea-shore nearer the usual collecting grounds than
Liverpool would be a material assistance in the work.
Consequently the Committee, in 1887, established the
Puffin Island Biological Station on the North Coast of
Anglesey, and later on, in 1892, moved to the more
commodious and convenient Station at Port Erin in the
centre of the rich collecting grounds of the south end of
the Isle of Man.

In our twelve years experience of a Biological Station
(five years at Puffin Island and seven at Port Erin), where
College students and young amateurs formed a large
proportion of the workers, the want has been constantly
felt of a series of detailed descriptions of the structure
of certain common typical animals and plants, chosen
as representatives of their groups, and dealt with by
specialists. The same want has probably been felt in
other similar institutions and in many College labora-
tories,

IV.

The objects of our Committee and of the workers at our
Biological Station have hitherto been chiefly faunistic and
speciographic. The work must necessarily be so at first
when opening up a new district. Some of our workers
have published papers on morphological points, or on
embryology and observations on life-histories and habits ;
but the majority of the papers in our volumes on the
"Fauna and Flora of Liverpool Bay" have been, as was
intended from the first, occupied with the names and
characteristics and distribution of the many different kinds
of marine plants and animals in our district. And this
faunistic work will still go on. It is far from finished, and
the Committee hope in the future to add greatly to the
records of the Fauna and Flora. But the papers in the
present series are quite distinct from these previous
publications in name, in treatment, and in purpose. They
will be called the "L.M.B.C. Memoirs," each will treat
of one type, and they will be issued separately as they are
ready, and will be obtainable Memoir by Memoir as they
appear, or later bound up in convenient volumes. It is
hoped that such a series of special studies, written by
those who are thoroughly familiar with the forms of which
they treat, will be found of value by students of Biology
in our laboratories and in Marine Stations, and will be
welcomed by many others working privately at Marine
Natural History.

It is proposed that the forms selected should, as far as
possible, be common L.M.B.C. (Irish Sea) animals and
plants, of which no adequate account ah-eady exists in any
text-book. Probably most of the specialists who have
taken part in the L.M.B.C. work in the past, will prepare
accounts of one or moi-e representatives of their groups.
The following have already promised their services, and
in some cases the Memoir is already far advanced. The

first three Memoirs will be issued before the end of 1899,
and others will follow it is hoped in rapid succession.

Memoir I. Ascidia, W. A. Herdman,
II. Cockle, J. Johnstone.
III. Echinus, H. C. Chadwick.
Dendeonotus, J. A. Clubb.
ZosTERA, K. J. Harvey Gibson.
Halidrys, C. E. Jones.
CoDiUM, E. J. H. Gibson and Helen Auld,
Diatoms, F. E. Weiss.
GiGARTiNA, 0. V. Darbishire.
Alcyonium, S. J. Hickson.
Plaice, F. J. Cole and J. Johnstone.
BoTRYLLOiDES, W. A. Herdman.
Cuttle-Fish, W. E. Hoyle.
OsTRACOD, Andrew Scott.
Patella, J. K. Ainsworth Davis.
Calanus, I. C. Thompson.
Actinia, J. A. Clubb.
POLYZOON, Laura K. Thornely.
Calcareous Sponge, E. Hanitsch.
Porpoise, A. M. Paterson.
Arenicola, J. H. Ashworth.
Oyster, W. A. Herdman.

Finally, I desire to acknowledge that a welcome
donation of £100 from Mr. F. H. Gossage of Woolton
has met the expense of preparing the plates in illustration
of the first few Memoirs, and so has enabled the
Committee to commence the publication of the series
sooner than would otherwise have been possible.

W. A. Herdman.

University College, Liverpool,
Odohcr, 1899.

The ancestor remote of Man,
Says Darwin, is th' Ascidian."

Anon. Ballad.

"For Man was once a leather bottel."

Old Song.

L.M.B.C. MEMOIRS.

No. I. ASCIDIA.

Professor W. A. Herdman, D.Sc, F.R.S.

Although the Ascidian has been much talked of and
written about during the last quarter of a century, com-
paratively few people, beyond the circle of professional
biologists, have an accurate idea of what the animal is like
in appearance and structure, or have more than a vague
notion as to what the popular impression of its relation-
ship to higher animals is based upon.

. The adult or fully-developed Ascidian is a very remark-
able animal, and presents numerous interesting problems
for the biologist to investigate ; but its reputation is due,
not to any of these, but to certain changes which the
young animal undergoes in its development from the egg ;
changes so remarkable and interesting that when the more
important of them were discovered by Kowalevsky, some
thirty years ago, they gave rise to the belief that Ascidians
were a group of animals holding a position between the
Vertebrata and the Invertebrata, and indicated a possible
line along which the former might have been evolved from
the latter. The opinions of scientific men on this point
have undergone a certain amount of modification during

lafe 3'ears, and Ascidians are now regarded as the degener-
ate descendants of a very lowly-developed group of the
early Vertebrata (or more correctly, Chordata).

But before we are in a position to understand this
important matter, it is well to have some knowledge of
the ascidian structure at several different stages in the
life-history, and we shall connnence with the last stage
of all — the full-grown, or adult condition.

Mode of Occurrence.

Ascidians are all marine, and they have been found in
all seas, from tropical to polar regions, and at depths
varying from near higli-water mark down to three and a
quarter miles or so. Round most of our own coasts they
are common, and some kinds are familiar enough, under
the name of " sea-squirts," to many sea-side visitors, as
being amongst the inhabitants of rock-pools which can be
collected and kept in aquaria. Such forms are usually
found as dome-shaped l)odies of a dull red colour, adhering
to the rock or sea-weed, and having two small openings on
their upper surface from which, when touched, they emit
delicate jets of sea-water with considerable force — thus
establishing their claim to the title of " sea-squirts " ; and
their resemblance to double-necked leather bottles, whence
the name Ascidian (from the Greek ao-Kos).

Others ngain form flat gelatinous expansions attached
to sea-weeds or stones, and symmetrically marked with
bright spots of colour in the form of circles, meandering
lines, or starlike patterns. These are really colonies in
which each spot of colour or ray of the star represents an
ascidiozocid or member of the colony equivalent to the
whole animal m the case of the solitary Ascidian.

By dredging around the coast, outside low-water mark,
many other kinds of Ascidians are found, of diverse forms

and often most gorgeous colouring. One of tlie commonest
species of our seas, Ascidia vlrglnca, 0. F. Miiller, has the
form of a short and somewhat irregular cylinder, with
rounded ends, one of which is attached to a stone or dead
shell, or some other object at the bottom, while the other
end is directed upwards into the water, and bears two
short projections, each terminated by an opening (see
fig. 1, PI. II.). This species is of a dull grey colour ; it is
usually found in from five to twenty fathoms of water, and
is in some places so abundant that the naturalist's dredge
may come up absolutely filled with it. An average size is
an inch and a half in length, an inch in breadth, and half
an inch in thickness ; but in a dredge ful one usually finds
all sizes, from a quarter of an inch to two inches in length.
Our largest British species, Ascidia mentula, 0. F. Muller
(see PI. I.), measures from three to six inches in length,
and is usually found on a muddy bottom in from ten to
fifty fathoms. Either of these two species of Ascidia will
serve very well as the type of a simple ascidian, and the
following description in nearly all details applies to both.

EXTEENAL ApPEAEANCE.

The two openings at the upper end, although they
appear very similar at first sight, are really different, and
can readily be distinguished with a little practice. One of
them {Br. in fig. 1, PI. II.), is higher, or more nearly
terminal in position than the other {At.), which may be
placed some way down one edge ; the former is the
Branchial, and the latter the Atrial aperture. A close
examination will show that the margin of the branchial
aperture is cleft into eight small projections, or lobes ;
while the atrial aperture is bounded by six lobes only (see
also PI. I.- — -as an mdividual peculiarity this specimen
has a seventh small lobe at its atrial aperture). Conse-

qnently, we can distinguish the apertures both by their
position and by the number of their lobes ; and this is a
very important matter, since it enables us to recognise
the different regions in the body of the Ascidian ; and also
because the exact number of lobes is a characteristic of
certain families and genera of Ascidians, e.g., Molgula
has six branchial and four atrial lobes, while Ascidia
has eight branchial and six atrial. Plate I. shows the
exact shape of the lobes and the apertures in Ascidia
mentula.

The branchial aperture {Br.) indicates the anterior end
of the body, the region which corresponds to the head end
of a man, a dog, or a fish ; and, consequently, the opposite
part, which is attached to the ston^e, is the posterior end.
Then, the atrial aperture {At.) is invariably placed to the
dorsal side of the branchial ; hence, that side of the body
near which the atrial aperture is situated corresponds to
the back of the man, dog, or fish ; while the opposite side
is, of course, ventral.

Now that it is known which is the anterior end, and
which is the dorsal surface, or back, it is easy, by placing
the Ascidian alongside oneself, and comparing the parts,
to determine which is the right-hand side and which is the
left. In PI. I. and in PI. II., figs. 1 and 4, it is the right
side of the specimens which is shown. The area of attach-
ment at the posterior end frequently involves a certain
amount of the left side.

The Test.

The outside of the Ascidian — all that is visible, unless
one looks into the expanded apertures — is formed by a
stout, gristly, translucent layer called the Test or Tunic
(hence " Tunicata," the name of the group), which
is notable amongst animal structures for containing

" tniiicine," a substance which appears to be identical
in composition, and in behaviour under various treatments,
v^ith "cellulose" — a characteristically vegetable substance,
entering largely into the composition of plants. The test
is a protective layer, usually about a quarter of an inch
in thickness, and may be considered as an exo-skeleton.
It is the only tissue of a skeletal nature which the adult
Ascidian possesses.

In shape the test is an oblong sac, pierced only by the
branchial and atrial apertures, and forming, at its posterior
end and left side, the place of attachment by which the
Ascidian adheres to the rock (PI. II., figs. 1 and 4). At
this point it may become greatly thickened and expanded
to form a margin, or may even grow out in the form of a
short stalk, raising the body above the surrounding surface.
Stones, sea -weeds, dead shells, and remains of other
animals may be overgrown by the test and incorporated
in its substance ; many sessile animals and various kinds
of sea-weeds may become attached to its outside ; and
some parasitic Amphipod Crustaceans (Tritceta gihhosa)
and Lamellibranch Molluscs (Modiolaria marmorata) may
inhabit cavities excavated in its thickness. Microscopic
parasitic Algae may also be present amongst the cells in
the interior of the test, and help in giving the animal its
colour.

The test is cartilaginous in appearance and consistence,
and to some extent in structure, as it consists of a clear
or slightly fibrillated matrix in which are imbedded many
cells. As the test is morphologically a cuticle, being at
first a secretion on the outer surface of the ectoderm (PI.
II., fig. 5, ec), the cells it contains have immigrated to it
from the body ; and it has been shown by Kowalevsky
and others that many of these are mesodermal cells or
wandering amoebocytes which have passed through the

ectoderi"n. In the test they i'e[)i'oihice freely and secrete
lar^e quantities of the tuiiicine matrix.

Many of the cells in the test remain small and simple,
as the rounded, fusiform, or stellate test cells (fif^. 5, t.c).
Some become larger, much branched, elongate into fibres,
or degenerate into globular pigment cells ; others may
store up reserve products ; while others again are con-
verted into the large vacuolated "bladder-cells" which,
in the outer part of the test of Ascidia mentida, form a
well-marked vesicular layer (PI. III., fig. 9, hi.). In this the
structureless matrix contains innumerable closely packed
spherical vacuoles, each with a thin peripheral film of
protoplasm and a parietal nucleus. These bladder cells
measure from 0"10 to 0"15 mm. in diameter. Some of
them show more than one nucleus, and may be formed by
the fusion of several cells. Some of them at least appear
to be derived from the " testa-cells " of the embryo, and
are thus descendants of cells belonging to the follicle
which surrounds the ovum.

The test also becomes organised by the growth into it
of the so-called "vessels." These are out-growths of the
mesodermal body-wall, covered by ectoderm, and contain-
ing prolongations of blood channels from the connective
tissue of the body-wall. Plate II., fig. 5, shows such an
out -growth, and exhibits the general relations of test
(cuticle) ectoderm, and mesoderm. It also explains how
it is that the blood channel being pushed out as a loop
gives rise to the double or paired vessels seen branching
through the test (PL III., fig. 9). The two vessels of a
pair are one blood channel imperfectly divided by a con-
nective tissue septum. The blood courses out along one
side, round the communication in a terminal knob at the
end and back down the other side. The terminal knobs
are very numerous and form a marked feature in the outer

layer of the test (fig. 9, t.k.)', in some Ascidians they
prohably form an accessory organ of respiration. In the
adult Ascidia the vessels enter the test by a double trunk
ventrally, near the p )sterior end of the left side (see PI.
III., fig. 10).

Ectoderm and Body-wall.

Inside the test, and lying between its inner surface and
the muscular body-wall ("mantle"), is a very delicate
membrane, formed entirely of small cubical or more flat-
tened cells (PL II., fig, 5, ec), with delicate "secreting
processes " projecting into the test. This is the Ecto-
derm, the outermost complete layer of cells in the body ;
and the test lying over it has been produced as a sort of
gelatinous exudation upon the surface of the ectodermal
cells, aided by the numerous mesoderm cells which have
migrated into it, and which give it the appearance of a
connective tissue. Besides the ordinary epithelial cells, a
few gland cells and pigment cells may be found in the
ectoderm. The ectoderm is turned in for a short distance
at the branchial aperture (mouth) and atrial aperture
(cloacal), as a short stomodteum and proctodaeum, lined
in each case by a delicate prolongation of the test (PI. II.,
fig. 4, Br. and At.).

Inside the ectoderm lies a thicker layer, the so-called
" Mantle " or body-wall (paiietal mesoderm), containing a
large number of muscles, which run some along the length
of the body, and others across it, so that they form a rude
interlacing net-work (PL II., fig. 4, m.), which is much
more strongly developed on the right than on the left side
of the body. The shape of the body can be changed, to a
slight extent, by means of these muscles in the body-wall.
When the Ascidian is killed by re-agents it is generally
found that the muscles have contracted and drawn the

8

mantle away from the test, so that these layers are no
longer in continuity except on the branchial and atrial
siphons, and at the posterior end where the vessels enter
the test.

The body-wall (PI. II., fig. 5) is largely formed of
connective tissues, both homogeneous and fibrous, with
cells, blood sinuses or lacunae, nerves, and the many
muscle bundles, large and small, formed of long, fusiform,
non-striped fibres. The largest muscle bundles are found
about the centre of the right side, where they may be
0*5 mm. in thickness. This part in the living Ascidia is
often brilliantly pigmented — red, yellow, and opaque
white — the coloured cells being exactly like those found
in the blood. The connective tissue cells or corpuscles
are fusiform, stellate, or amoeboid, and may become pig-
mented or vacuolated, like the similar cells of the test.

At the anterior end the body-wall is prolonged outwards
to form the two well-marked siphons, or short wide tubes,
which lead in from the branchial and atrial apertures.
These are surrounded by strong sphincter muscles (Fl. II.,
figs. 2, 6, and 7, sph.). Inside the body- wall lies the large
cavity called the Atrium, or the peribranchial cavity, which
comnmnicates with the exterior through the atrial aperture,
and serves to convey away the water which has been used
in respiration (see PI. II., figs. 2 and 4,p.hr.).

The ectodermal lining of the atrial or peribranchial
cavity has been called by some French writers the third
tunic — tlie first being the test and the second the mantle.
The cavity of the atrium is traversed by numerous vascular
strands of mesoderm, called connectives (PI. II., fig. 2,
con.), passing from the body-wall inwards to the branchial
sac.

Figure G on PI. II. shows the relations of ectoderm (with
test over it), mesoderm, and endoderm in a section through

the antero-dorsal part of the body, where t. indicates the
test, and spli. the sphincter of the branchial aperture. The
cavity marked p. hr. is a portion of the atrial cavity lined
by ectoderm, and not to be confounded with a coelom or
body- cavity. The absence of a true coelom in the
mesoderm will be noticed in this and the other figures
{e.g. fig. 4) ; and yet the Tunicata are Coelomata — although
it is very doubtful whether an enterocoele is ever formed,
as has been described by E. van Beneden and Julin in the
development of some. The primitive coelom is, however,
largely suppressed during development, and is only repre-
sented ill the adult by the pericardium and small cavities
in the renal and reproductive organs and ducts, as will be
shown further on.

Cavities of the Body.

The following list of the cavities present in the body of
the adult Ascidia maj'' be useful at this point : —

1. The alimentary canal, including the branchial sac.
This is derived from the archenteron of the embryo, is
lined throughout by endoderm, and the system of cavities
of the intestinal gland is to be regarded merely as an out-
growth from the alimentary canal.

2. The peribranchial (atrial) cavity, derived from two
lateral ectodermal invaginations which join dorsally to
form the cloaca and open to the exterior by the atrial
aperture.

3. The original embryonic segmentation cavity (blasto-
coele) remains, where not obliterated by the development
of the mesodermal connective tissue, as the irregular
system of blood spaces, with its outgrowths in test and
branchial sac. The heart is only a special part of this
cavity which has differentiated muscular walls.

10

4. The pericardium and epicardium originate as out-
growths from the archenteron. They may therefore be
regarded as coelomic spaces. The pericardium becomes
completely closed off and separated from the alimentary
canal. The epicardium may form paired tubes of great
length, and may remain permanently connected with the
branchial sac.

5. The cavities of the renal vesicles and of the gonads
and ducts are spaces formed in the mesoblast. They have
been variously interpreted : —

(a) As of the same nature as the blood spaces (blasto-

coelic), or
(h) As formed by a splitting of the mesoblast (coelomic) .

6. The cavity of the neural gland and its duct opening
at the dorsal tubercle is derived from the primitive dorsal
neural tube of the embryo, and so may be regarded as a
part of the lumen of the cerebro-spinal nervous system.

Branchial Siphon and Tentacles.

The branchial aperture opens into a large cavity, the
Branchial Sac (PI. II., fig. 4, hr.s.), which is merely the
anterior portion of the alimentary canal, corresponding
to the pharynx or back of the mouth in man, enlarged and
greatly modified so as to act as a breathing organ, or
branchia — whence its name — in addition to performing
other important functions.

The branchial aperture itself (Br.) is thus the mouth ot
the Ascidian, and the siphon is therefore the commence-
ment of the alimentary canal. Its inner surface is lined
for a short distance by a prolongation of the test, and
where this stops, at about the line of junction of the
ectoderm of the stomodoeum with the endoderm of the
mesenteron, a circle of delicate hair-like Tentacles (PI. II.,

11

fig. 4, tn.) projects into the mouth cavity and forms a
sensitive sieve or strainer, through which all the sea-water,
and its contents, drawn into the branchial sac has to
pass.

The tentacles are simple and tapering in Ascidia, but
in many other Ascidians they are compound and may be
very elaborately branched {e.g., in Cynthia and Molgula).
They are in many cases of different sizes arranged alter-
nately or with some marked symmetry. In Ascidia
mentula there are usually from 70 to nearly 100 tentacles,
of which one-third, say 20 to 30, are much larger than the
rest. The rule is for two, occasionally three, much smaller
tentacles to be placed between each pair of larger ones
visible to the eye. Fig. 7 on PI. II. shows, what is
sometimes found, three orders of tentacles placed symmet-
rically, the middle one of each group of three smaller ones
being longer than its two neighbours. In Ascidia virginea
also the tentacles are very numerous, nearly 100, and are
of two sizes placed alternately.

Each tentacle is practically an ingrowth of the connective
tissue of the body- wall, covered by the epithelial lining of
the front of the alimentary canal. It has consequently a
connective tissue core containing muscle fibres and nerves,
and one or more blood lacunae continuous with those of
the body-wall. The delicate epithelium with which it
is covered contains some simple sensory cells. These
tentacles not only act mechanically in preventing large
objects from entering, but are also sensitive like the lobes
of the apertures, although only scattered sensory cells and
no specially differentiated sense-organs are present.

Behind the tentacles lies the plain or papillated pre-
branchial zone (PI. II., fig. l,p.hr.z.) bounded behind by
a pair of parallel and closely placed ciliated ridges with a
groove between — the peripharyngeal bands (p.p.b.) —

13

which encircle the anterior end of the branchial sac. The
anterior band forms a complete ciliated ring, but the
posterior is interrupted in the ventral and dorsal median
lines : its ends becoming continuous respectively with
the marginal folds of the endostyle (ventrally), and with
the front of the dorsal lamina, where before joining, they
bound a narrow triangular cavity lined by ciliated epi-
thelium, the epibranchial groove (see PI. II., fig. 7, above
d.l.). Behind the peribranchial bands the proper wall of
the branchial sac commences.

Branchial Sac and Atrium.

The wall of the branchial sac is penetrated by a large
number of channels, through which blood flows. Some of
these run in one direction and some in another, so as to
form complicated but perfectly definite networks, which
differ in their arrangement in different kinds of Ascidians.
Between these blood channels there are clefts (the second-
ary gill-slits or "stigmata") in the wall of the branchial
sac, by means of which the water from the interior passes
into the large external or peribranchial cavity — the atrium.

The transverse section (PI. II., fig. 2) shows how this
atrium surrounds the branchial sac on all sides except the
ventral, where the wall of the branchial sac becomes
contmuous with the body-wall. The right and left halves
of the atrium may be called the right and left peribranchial
cavities (p.br.). They unite along the dorsal edge to form
the cloaca, and there open to the exterior. The cavity
of the branchial sac communicates with the surrounding
atrium by means of the stigmata, as shown on the upper
half (left side) of fig. 2 (PI. II.). The section on the
right side is shown passing along a transverse vessel
between two of the rows of stigmata.

13

If an Ascidiaii expanded in sea-water, in a healthy
condition, be closely watched, it will soon be noticed that
there is a constant stream of water pouring in through
the branchial aperture, and another flowing out from the
atrial (as represented in fig. 1, PI. II.) ; and if some fine
and insoluble coloured powder be dropped into the water
near the branchial aperture, it will rapidly be drawn in
with the current, and after a short time some of the
powder will make its appearance in the water ejected
from the atrial aperture.

Hence, it is obvious that there is a current of sea-water
flowing through the body of the Ascidian. This current
has four distinct uses or functions : (1) it enables the animal
to breathe, by bringing in fresh supplies of oxygen ; (2) in
the first part of its course it carries the microscopic food
particles into the Ascidian's body ; (3) in the last part of
its course it carries out of the body various waste materials,
which must be got rid of; and (4) it ejects the mature ova
and spermatozoa from the body. The course which this
water-current takes is : — in through the branchial aperture
to the branchial sac, then through the clefts in the wall of
that organ into the surrounding atrium, and lastly out
through the atrial aperture to the exterior. The direction
of this current may occasionally be temporarily reversed ;
and, when the muscular body-wall contracts, a sudden
current may be ejected through both apertures simul-
taneously.

All the clefts or stigmata in the wall of the branchial
sac (shown in fig. 3, PI. II.) are bounded by cells which
bear a number of cilia (PI. IV., fig. 3) projecting across the
cleft. These cilia, so long as the animal is alive, are in
constant motion, lashing rapidly from the branchial sac
towards the atrium, so as to drive the water in the cleft
outwards ; and it is this constant action of these very

14

minute cilia which causes the regular current of water to
flow in at the branchial aperture and out at the atrial.

The branchial sac is very large, much the largest organ
of the body (it may be 15 cm. in length), and extends
almost to the posterior end of the body, while the rest
of the alimentary canal lies upon its left side imbedded in
the body-wall. The food particles, consisting of very
minute plants and animals, are carried in through the
branchial aperture by the current of water, but most of
them do not pass out into the atrium, being caught by
the ciliary action of the peripharyngeal bands, and entangled
in the viscid substance which fills the groove between
them near the anterior end of the branchial sac. This
viscid substance, or mucus, is formed in a long canal-
shaped gland called the Endostyle or hypobranchial groove
(PI. II., figs. 2 and 4, end.), which lies along the ventral
edge of the branchial sac, and terminates both anteriorly
and posteriorly in a short cul-de-sac.

The Endostyle.

On the floor of the endostylar groove is found a tract
of cells with very long cilia (PI. II. , fig. 2) . On each side are
at least two (in some Ascidians three) laterally placed
clumps of gland cells (PI. IV., fig. 5), each clump separated
from its neighbours by an area of closely packed fusiform
cells with short cilia, amongst which are found some
sensory bipolar cells. The lips of the endostyle are formed
of ciliated cubical epithelium.

This organ, on account of its thick glandular walls,
has an opaque appearance, and seems, in side view, to
run like a conspicuous solid rod in the more transparent
walls of the branchial sac — ^hence its name endostyle. It
is, however, really a gland, and corresponds to the hypo-
pharyngeal groove of Amphloxiis and the median part of

15

the thyroid gland of Veitebrata. It is interesting to
notice that the four (at least) longitudinal tracts of gland
cells are of remarkable constancy, being found not only in
all groups of Tunicata (often six tracts) including even the
pelagic, tailed, Appendicularians, but also in Amphioxus
and in the young thyroid gland of the Ammocoete.

The posterior cul-de-sac of the endostyle is quite short
in Ascidia, but in some other Tunicata it is longer and
becomes of great importance as an organ for the produc-
tion of buds. Behind the cul-de-sac the marginal folds of
the endostyle unite to form a slight ridge, the posterior
fold, which runs round the posterior end of the branchial
sac to join the end of the dorsal lamina behind the
oesophageal opening. The endostyle, in addition to its
glandular function, shares in the sensory functions of the
peripharyngeal band, the tentacles, and the dorsal tubercle,
in all of which similar sensory cells and nerve endings have
been found.

DoESAL Lamina.

The mucus formed by the glands in the lateral walls of
the endostyle is carried forwards by the lashing action of
the long ciHa placed on the floor of the organ, and so
reaches the front of the branchial sac ; here it changes
its direction of flow, and bends round to the right and left,
in the groove between the peripharyngeal bands, so as to
gain the dorsal edge of the sac, where it encounters a
projecting membranous fold, the dorsal lamina or epi-
branchial ridge, along which it is carried backwards — still
by ciliary action — to the opening of the oesophagus, the
next region of the alimentary canal after the pharynx.
The food particles become entangled in this train of mucus
chiefly in its course round the right and left sides of the
anterior end of the sac, and from this point are carried

16

back along the right side of the dorsal lamina to the
ccsophagus.

The dorsal lamina is at its widest round the oesophageal
aperture (PL 11., fig. 4, oes.). It is more or less ridged
transversely, especially on its convex left side, and may
have marginal tags or teeth (PI. IV., fig. 4), which in
some Ascidians become long processes, the langnets. In
the living animal, the lamina has its free edge curved to
the right hand side in such a manner as to constitute a
fairly perfect tube along which the train of food passes.

Wall of Branchial Sac.

Fig. 3, PI. II. shows a small part of the wall of the
branchial sac of Ascidia in which it may be seen that the
bars containing the blood channels are arranged in three
regular series : —

(1) The "transverse vessels" {tr.) which run horizon-
tally round the wall and open at their dorsal and ventral
ends into large median longitudinally running tubes, the
dorsal blood sinus behind the dorsal lamina and the ventral
blood sinus below the endostyle (see also PL III., fig. 10).

(2) The fine longitudinal or " interstigmatic vessels "
{l.v.) which run vertically between adjacent transverse
vessels and open into them, and which therefore bound
the stigmata.

(3) The "internal longitudinal bars" {i.l.) which run
vertically in a plane internal to that of the transverse and
fine longitudinal vessels. These bars (see PL II., fig. 3, B.)
communicate with the transverse vessels by short side
branches {c.d.) where they cross, and at these points are
prolonged into the cavity of the sac in the form of hollow
papilkie {p.).

In sonie Ascidians {e.g. Corella parallelog ramma and
most of the Molgulidae) the interstigmatic vessels are curved

17

so that the stigmata form more or less complete spirals.
In some species of Ascidia, and other Ascidians, the inter-
stigmatic vessels are inserted into the transverse ones in un-
dulating in place of straight lines, the result being that the
stigmatic part of the wall of the branchial sac seems to be
folded or thrown into microscopic crests and troughs.
This device for increasing the surface is known as
" minute plication," and is seen well in Ascidia mentula
(PI. IV., figs. 1 and 2). In some cases, again (Cynthiidoe),
the whole wall of the sac is pushed inwards at intervals
to form large folds visible to the eye.

The intersections of the internal longitudinal bars with
the transverse vessels divide up the inner surface of the
branchial sac into rectangular areas called "meshes" (PI.
IV., fig. 2). One such mesh, containing eight stigmata in
a row, is seen in fig. 3, PI. II. The internal longitudinal
bars bear hollow papillae at the angles of the meshes, and
occasionally in intermediate positions. There are frequently
horizontal membranes (fig. 3, B. h.vi.) attached to the
transverse vessels between the papillae, and the transverse
vessels may be of two or more sizes arranged symmetri-
cally. There are many " connectives " running from the
outer wall of the branchial sac to the body -wall, and
allowing the blood in tlie transverse vessels to communi-
cate with that in the sinuses outside (PI. II., fig. 2, con.).

In an adult medium-sized Ascidia mentula there are
about 150 transverse vessels on each side of the branchial
sac, and at least 80 internal longitudinal bars, making by
their intersection 12,000 meshes. The average number
of stigmata in a mesh is, in this species, six to eight. So
there may be as many as 96,000 stigmata present on each
side of the sac, nearly 200,000 in all. Probably these
numbers are greatly exceeded in large specimens, such as
that bliown on PI. I. These stigniata are to be regarded as

18

secondary openings due to the breaking up or subdivision
of the primary Chordate gill-clefts.

In addition to the stigmata there are generally one or
two pairs of much larger (up to 15 mm. in length^,
narrow, elongated slits placed near the posterior end of
the sac and close to the dorsal lamina, so as to be under-
neath the atrial aperture, by which water can escape
into the cloacal part of the peribranchial cavity. These
pharyngo-cloacal slits have well-marked edges which bear
much finer cilia than those of the stigmata.

The lining of the branchial sac is the pharyngeal
epithelium (endoderm) while the outer surface is covered
by the lining membrane of the peribranchial cavity in free
communication with the outer surface of the body through
the atrial aperture (see PL II., fig. 2). Both of these
epithelial surfaces are formed of squamous cells. Round
the sides of the stigmata (PL IV., fig. 3) the cells on
the longitudinal vessels become more nearly cubical in
shape and bear the cilia, while at the ends of the stigmata,
near the transverse vessels, the cells approach a columnar
form (PL IV., fig. 3). The epithelium along the internal
edge of the longitudinal bars, and on the apices of the
papillse, is also cubical or almost columnar in form.

Between the outer and the inner epithelium the wall
of the branchial sac is formed of connective tissue in
which the blood lacunae, known as "vessels," are exca-
vated. These vessels are very regular in size and
arrangement, and are so large that comparatively little
connective tissue is left, and so the blood is in close
proximity with the epithelial surface. In some branchial
sacs a few non-striped muscle fibres are found running
longitudinally in the connective tissue around the chief
vessels (see PL IV., fig. 1).

19

Alimentaky Canal.

The oesophagus is a short, narrow, curved tube which
leads ventrally to the stomach — a large, thick-walled
organ, lying on the left-hand side of the branchial sac,
imbedded in the body-wall and projecting into the atrium,
(as shown in fig. 4, PI. II.). From the other (ventral)
end of the fusiform smooth-walled stomach arises the
intestine, a long serpentine tube which ends by opening
into the dorsal or cloacal part of the atrium, from which
the undigested portions of the food are carried to the
exterior through the atrial aperture, by the water current.

The intestine is curved so as to form two loops (PL II.,
fig. 4), a first between the stomach and intestine, open
posteriorly, and in which the ovary lies ; and a second
between the intestine and rectum, open anteriorly, and in
which the renal vesicles lie. The external convex edge of
the intestine is thickened internally to form the "typhlosole,"
a large pad which runs along its entire length (PI. IL, figs.
2 and 4, tij.), reducing the lumen to a crescentic slit.

The walls of the stomach are glandular ; and, in addi-
tion, a system of delicate hyaline microscopic branched
tubules with dilated ends (the " refringent organ"),
which ramifies over the outer wall of the intestine and
communicates with the cavity of the stomach at the
pyloric end by nieans of a duct, is possibly a digestive
gland. There is in AscicUa no separate large gland to
which the name " liver " can be applied as in some other
Tunicata {e.g., Molgitla). Over these viscera, on the left
side of the body, the body- wall is thin and gelatinous, and
has usually no muscle fibres visible.

The wall of the alimentary canal, throughout its length,
consists of an epithelial lining (endoderm), a thick layer
of highly vascular connective tissue, and, over that, the
flattened epithelium lining the peribranchial cavity. The

20

only notable regional differences lie in the epithelium.
The coiuiective tissue is much the same throughout, and
is continuous with that of the body-wall in which the diges-
tive viscera are imbedded. The blood lacunae are very
numerous and communicate on the one hand with the
cardio-visceral, and on the other hand with the branchio-
visceral main blood vessels.

The connective tissue is also penetrated by (a) the caeca
of the gonads, especially the spermatic tubules, and (b)
the delicate branched clear tubules of the enigmatical
refringent organ or pyloric gland (PI. IV., fig. 7). The
tubules of this organ are lined by low cubical cells, usually
non-ciliated, containing no concretions or granules, and
having no great resemblance to gland cells. They have
been called " chylific " and absorptive, but the function is
still undetermined. There are few, if any, muscle fibres
in the connective tissue of the alimentary canal until the
rectum is reached.

The lining epithelium is for the most part ciliated,
especially in the oesophagus and intestine. The oeso-
phagus has slight longitudinal ridges and grooves, one
of which seems to continue the canal of the dorsal lamina
onwards to the stomach , while another is in relation with
the posterior fold coming from the lower end of the
endostyle. There are a few glandular (mucous) cells
scattered amongst the ciliated columnar cells of the
oesophageal wall.

The stomach has projecting folds in its interior which
unite at its pyloric end to form the intestinal typhlosole,
and, in addition to some ciliated and mucous cells, it has
large masses of highly coloured markedly glandular cells
packed with yellow granules. The intestine, again, has
ciliated and gland cells, and in the rectum the gland
cells become fewer and die out. The termination of the

21

rectum, close to the atrial aperture, is ver^^ thin- walled,
with a slight thickened edge at the anus, but no sphincter.

Heaet and Circulation.

The soluble product of the food which has been digested
passes through the wall of the alimentary canal, and
enters the numerous small blood spaces in the connective
tissue on both sides of the stomach and intestine. These
lead, by the cardio-visceral vessels, to the dorsal end of
the heart (see PI. III., fig. 10), which is merely a delicate
tube, irregularly swollen in the middle, placed behind the
stomach, and projecting into a space, the pericardium
(PI. IV., fig. 9,p.c.), which is a part of the original coelom.

The wall of the heart is continuous along one edge
(that next the stomach) with that of the pericardium, and
the heart is to be regarded as a tubular invagination of the
pericardial wall (see PI. IV., fig. 10), shutting in a portion
of the external space (the blastocoele of the embryo) and
having open ends which connnunicate with the large blood
sinuses leading to the branchial sac, to the viscera, and to
the body-wall and test. The cavity of the heart is not
divided and has no valves. Its wall is formed of a single
layer of epithelio-muscular cells, the inner (nmscular) ends
of which are cross-striated fibres running round the heart
— the only striated muscle found in the body of the
Ascidian. The larger channels through which the blood
flows are lined by a delicate endothelium, the smaller are
merely spaces in the connective tissue. All the blood
spaces and lacunoe are probably derived, like the cavity of
the heart, from the blastocoele of the embryo, and are not
(like the pericardium) a derivative of the coelom. The
wall of the pericardium is simple squamous epithelium.

From the ventral end of the heart the blood is conveyed
by the branchio-cardiac vessel and the great ventral vessel

22

lyin<^- underneath the endostyle (PI. III., fio-. 10, hr. ao.)
to the waHs of the branchial sac, where, in passing along
the transverse and interstigmatic vessels, it is purified,
and receives a supply of oxygen from the water passing
through the stigmata. It is then conveyed, by the great
dorsal vessel and the branchio- visceral vessel and its
branches, to different parts of the viscera and body-wall,
so that all the organs may receive food and oxygen, and
have their waste materials carried away. Some of these
branchio-visceral vessels from the branchial sac lead to
the walls of the stomach and intestine (see PI. III., fig. 10),
and thus bring us back to the point from which we started.

The great dorsal and ventral vessels of the branchial
sac are connected (PI. III., fig. 10), not only by the trans-
verse vessels, which run like hoops round the walls, but
also at their anterior extremities, by a circular vessel
which surrounds the front of the branchial sac, underneath
the peripharyngeal bands. A short branch runs from near
the front of the dorsal vessel to the sinuses which sur-
round the nerve ganglion.

From each end of the heart "vessels" are also given
off to supply the body- wall and test. Moreover, " connec-
tives " run from the transverse vessels of the branchial
sac directly outwards, on each side, to the body-wall and
viscera. On the left side there are three especially large
blood tubes amongst the connectives, which cross to the
alimentary canal and ovary, and branch through these
viscera.

But the course of this circulation of the blood is not
always the same ; sometimes it is exactly reversed, the
blood flowing from the branchial sac to the heart, and
from that organ to the viscera, and then back to the
branchial sac again. This curious state of affairs is caused
by the remarkable manner in which the Ascidian heart

^3

beats. If a young individual of a small transparent species,
such as Ascldla virginea, be placed alive, left-side upper-
most, in a watcb-glass of sea-water, and examined with a
low power of the microscope, the heart will be readily
seen near the posterior end of the transparent body. It
will then be noticed that the "beating" looks like succes-
sive waves of blood, which are pressed through the tubular
heart from one end to the other by the contractions of the
muscle fibres. After watching the waves passing, let us
suppose, from the dorsal end of the heart to the ventral,
for about a minute and-a-half or two minutes, it will be
seen that they gradually become slower and then stop
altogether. But now, after several seconds, a faint wave
will start from the ventral end of the heart and pass over
it to the dorsal ; and this will be followed by larger ones
for perhaps a minute or two, and then again a pause will
occur and the direction change. So that we may say, the
heart of the Ascidian beats first in one direction and then
in the other; and the reversal of the blood current takes
place every minute or two. There are generally rather
more beats in the dorso-ventral than in the opposite
direction, but there is considerable irregularity. The
numbers are usually between 60 and 80.

The cause of this remarkable reversal may possibly be
that the heart being on the ventral vessel, which is wider
than the corresponding dorsal trunk, it pumps the blood
into eitherthe lacunae of the branchial sac or those of the
viscera in greater volume than can possibly get out through
the smaller branchio-visceral vessel in the same time, the
result being that the lacunse in question will soon become
engorged, and by back pressure cause the stoppage, and
then reversal of the beat. The absence of any valves in
the heart to regulate the direction of flow obviously
facilitates this alternation of the current,

i24

When the heart is contracting ventro-dorsally it receives
oxygenated blood from the branchial sac by the branchio-
cardiac vessel (now a vein), and propels it by the cardio-
visceral trunk (now an artery) to both sides of the viscera
and body-wall. This blood, after circulating through the
system, is collected as impure blood by the branchio-
visceral vessel and conveyed to the dorsal sinus of the
branchial sac to be re-oxygenated. The heart is then a
systemic heart and contains pure blood. But after the
reversal, when the heart contracts dorso-ventrally the
veins and arteries exchange functions, the oxygenated
blood passes from the branchial sac to the viscera, the
heart receives impure blood from the system and propels
it to the ventral edge of the branchial sac, and so what
was a minute before a "systemic," is now a "respira-
tory" heart. This is a phenomenon without parallel in
the animal kingdom.

The blood of Ascidians is in the main transparent, but
contains usually certain pigmented corpuscles in addition to
many ordinary leucocytes or colourless amoeboid nucleated
cells (PI. IV., fig. 6). The pigment in the coloured cells
may be red, yellow, brown, or in some cases blue or opaque
white, and these are the result of deposition of pigment
granules in the older leucocytes. In Ascidia mentula a
large number of blood corpuscles are usually brown. The
unaltered leucocytes may be actively amoeboid, and can
proliferate. As we have seen, the blood may reach the
branchial sac either from the dorsal or from the ventral
median sinus, according to the direction in which the
heart is beating at the moment ; and it is a most interest-
hit^ and beautiful sight to watch the alternating circulation
of the variously coloured corpuscles through the transparent
vessels, and the lashing of the cilia along the edges of the

25

neighbouring stigmata, as shown in a small Ascidian under
the microscope.

Neevous System and Sense Organs.

The nervous system of the Ascidian consists of a single
elongated Ganglion or "brain," placed near the front end
of the body, in the dorsal median line, between the
branchial and atrial apertures (PI. II., fig. 4, n.g.) It
gives off several large nerves at each end, which break up
into the fibres that go to the different parts of the body,
and especially to the lobes and the muscles surrounding
the branchial and atrial apertures (PI. II., fig. (J). In
Ascidia men tula there are four chief nerves from the
anterior end of the ganglion, one to the surface of the
body and the other three to the branchial siphon ; while
three nerves run from the posterior end to the atrial
aperture and wall of cloaca. There are also one or two
smaller lateral nerves that leave the sides of the ganglion.

This jranolion is the degenerate remains of the anterior
part of the cerebro-spinal nervous system of the tailed
larval Ascidian. The posterior or spinal part has almost
entirely disappeared in most adult Tunicata ; but its
remains may be traced in a tract of degenerate nerve
tissue (the dorsal nerve cord) which runs posteriorly from
the ganglion (PI. II., fig. 6, d.n.) above the base of the
dorsal lamina towards the viscera. The ganglion has
small rounded nerve cells on its surface, while the centre
is a mass of interlacing nerve fibres. Small cells are
also found scattered along the course of the dorsal nerve
cord. The nerve cells in Ascidia vieiitula are mostly
pyriform or triangular in form, and are bi-polar or multi-
polar, and finely granular.

The Ascidian has little papillsB containing sensory cells
in its ectoderm, especially round the apertures, and has

26

isolated sense-cells and nerve-endings in various parts of
its internal epithelium ; but is very badly provided with
more definite " sense organs." It has no true eyes — for
the little brightly-coloured dots or " ocelli" placed along
the margins of the apertures (PI. I.), and formed of a
group of modified ectoderm cells, supplied by a nerve and
imbedded in a mass of red and yellow pigment (PI. IV.,
fig. 8), can scarcely be called such ; and it certainly has
no ears or otocysts. The tentacles are not very efficient
tactile organs, and the thin expanded margins of the
branchial and atrial siphons are apparently the most
sensitive parts of the body.

But there is a curiously-curled projection, the Dorsal
Tubercle (PI. II., fig. 7, d.t.), placed at the front of the
dorsal lamina, in the prebranchial zone, near the entrance
to the branchial sac, which may possibly be an organ for
testing, by smell or taste, the quality of the water drawn
in through the branchial aperture. This organ has a
narrow slit which leads by means of a ciliated funnel into
a delicate non-ciliated tube, and this can be traced back
for a distance of several centimetres to a glandular mass,
the neural gland, formed of tubules lined by small cubical
cells, lying imbedded in the connective tissue immediately
underneath the brain (see n. gl., fig. 6, PI. II.).

Hence, it has been suggested that the supposed olfactory
organ is merely the complicated opening of the duct from
the neural gland, and that this gland probably corresponds
to the hypophysis cerebri or pituitary body, which i_^s
found in all Vertebrata, from fishes up to man, attached
to the infundibulum on the lower surface of the brain. It
is probable that both views are partly right, and that
therefore the duct of the pituitary gland in the Ascidian,
opei]s into a sense organ placed on the roof of the mouth.

Sensory cells have been found amongst the ciliated

27

epithelium of the dorsal tubercle, as well as in the tentacles,
the peripharygneal bands, the endostyle, and neighbouring
parts.

As to the function of the neural gland — apart from the
dorsal tubercle — it is still somewhat mysterious. It inay
be merely to secrete viscid matter which is poured like
that from the endostyle into the peripharygneal groove, or
it may possibly be that the function is renal — for the
removal of nitrogenous waste matters in the neighbour-
hood of the nervous system.

Renal Organ.

A mass of large clear-walled vesicles, which occupies the
rectal loop and the adjacent walls of the intestine, and n^ay
extend over the whole left side, is undoubtedly a renal
organ without a duct. Each vesicle (PI. IV., fig. 11) is
apparently a little closed sack formed of modified meso-
blast cells which elnninate nitrogenous waste matters from
the blood in the neighbouring lacunae and deposit them
in the cavity, where they form one or more constantly
increasing concentrically laminated concretions of a_yel-
lowish or brown colour, sometimes coated with a chalky
deposit. These concretions (PI. IV., fig. 14) contain uric
acid, and in a large Ascidian are very numerous and of
considerable size. The nitrogenous waste products are
thus deposited and stored up throughout life in the'renal
vesicles in place of being excreted from the body.

The cells forming the walls of the renal vesicles have a
curiously wavy outline (PL IV., figs. 11, 12, and 13),
which gives them a characteristic appearance. The con-
tents of these cells seem to differ considerably (PI. IV.,
fig. 13) in different cases, probably as a result of their
functional activity.

28

■Reproductive Organs.

Ascidians are not divided into two sexes, and conse-
quently each individual is hermaphrodite, or possesses both
male and female reproductive organs (gonads), although
cross - fertilization is probably the rule as a protogynous
condition is very general — the female organs maturing
before the male. In Ascldia mentida, however, self-
fertilization does sometimes take place. The gonads in
Ascidia lie close together on the left side of the body,
alongside the stomach and intestine (PI. II., fig. 4, gon.),
and are provided with delicate ducts (gd.), which open like
the intestine into the atrium ; so that the mature ova and
spermatozoa are carried out of the Ascidian's body by the
current of water flowing from the atrial aperture. In
many Ascidians fertilisation and development take place
in the atrium, a part of which may be set aside as an
incubatory pouch. In some Ascidians (certain Molgulidae
and Cynthiidae) reproductive organs are present on both
sides of the body, and in others (Pohjcarpa) there are
many complete sets of both male and female systems
attached to the inner surface of the body-wall, on both
sides, and projecting into the peribranchial cavity.

The ovary in Ascidia mentula is a slightly ramified gland
which occupies the greater part of the intestinal loop (PI.
II., fig. 4, gon.). It contains a cavity which, along with
the cavities of the testis, is derived from an embryonic
mesodermal space which has been compared with a coelom
but may be merely blastocoelic, and the ova are formed on
its walls and fall when mature into this cavity. The
oviduct is directly continuous with the cavity of the ovary,
and leads forward alongside the rectum, and external to
the vas deferens to open into the atrium.

The testis is composed of a great number of delicate
branched white tubules, which ramify over the much

S9

coarser tubules of the ovary and the adjacent parts of the
intestinal wall. These spermatic tubules terminate in
ovate swellings, usually grouped in bunches (PI. IV.,
fig. 15). Near the commencement of the rectum the larger
tubules unite to form the vas deferens, a tube of consider-
able size which runs forward alongside the rectum, and,
like the oviduct, terminates by opening into the peri-
branchial cavity close to the anus. The lumen of the
tubules of the testis, like the cavity of the ovary, is a meso-
blastic space in the embryo, and the spermatozoa are
formed from the cells lining the wall, and are set free into
the cavity.

The mature ovum is of small size (about 0"12 mm. in
diaui.), colourless, and with little or no food yolk in the case
oi Ascidia. It is only some of the germinal cells in the
ovary that are destined to become ova. Of the rest, some
form a protecting layer, the follicle, around the young ova.
Certain of these primary follicle cells migrate inwards and
give rise, by proliferation, to a layer of cells in the superficial
part of the ovum (PI. V., fig. 1, ^.c). These are the
so-called "testa-cells" or kalymmocytes. These later on
produce a thin gelatinous layer over the surface of the
ovum and between it and the follicle, which looks like
the beginning of the test — hence the name given to the
cells, which, although so different in origin from the
mesodermal test cells of the adult, probably to some
extent give rise to " bladder cells " in the test. The rest
of the follicular " testa-cells " eventually disappear.

The follicle cells proper produce two layers, the outer of
which remains in the wall of the ovary when the ovum is
set free ; while the inner layer adheres to the surface (PI.
v., fig. l,foU.), and its cells become large and much vacuo-
lated, some of them growing out to form long papillae,
which help to sustain the floating egg in the sea-water.

30

Altogether there may he as many as seven distinct layers
around the mature egg, but they are all produced hy the
differentiation or activity of the follicle cells.

Polar bodies are formed from the maturing ovum in the
usual manner, and effect the usual reduction in the number
of chromosomes in the nucleus. In the common Styelopsis
grossularia there are two chromosomes left in the ovum,
while in some species of Ascidia there are eight.

The spermatozoa of Ascidia are of typical form (see PI.
v., fig. 1, s.).

Embryology and Life-History.

The egg (PI. V., fig. 1) after being fertilised (probably in
most cases by a spermatozoon carried by the current of
water from another Ascidian somewhere in the neighbour-
hood), proceeds to segment or divide into a number of
small pieces or young cells, thus becoming an embryo
Ascidian (PL V., fig. 2, &c.). The cells of the embryonic
body gradually come to be arranged (in a inanner the
details of which will be described below) so as to form— (1)
a skin or layer of cells— the ectoderm of the adult — covering
the outside ; (2) a tubular nervous system running along
the middle of the dorsal surface, miderneath the ectoderm;
(3) a short wide tube, placed ventrally, which is the
beginning of the branchial sac and the remainder of the
alimentary canal ; and (4) a cellular rod — the notochord —
which lies in the posterior part of the body, between the
dorsal nerve tube and the ventral alimentary canal, and is
a rudimentary or very simple back-bone, similar to that
found in the embryos of vertebrate animals (fig. 6, PI. V.).
In fact, the embryo Ascidian at this stage is comparable
with an embryo fish or frog, and is found to have the same
chief organs or parts similarly arranged ; and, moreover,

31

these parts have been formed in essentially the same
manner in both cases ; so that if similarity in structure
and development indicate relationship, it is evident that
the young Ascidian is related to the fish and the frog and
other Vertebrata, and is to be regarded as one of the
Chordata (animals which at some time of their life have a
notochord).

Turning now to the details of the development,* the
segmentation is complete, and bilateral, and nearly equal,
and results in the formation of a spherical blastula with
a small segmentation cavity (PI. V., fig. 2).

The blastula then grows larger and begins to differen-
tiate. There are slightly smaller cells which divide more
rapidly at one end of this embryo, the future ectoderm,
and slightly larger and more granular cells at the other,
which become chiefly endoderm (hypoblast). Invagination
of the larger cells then takes place (fig. 3), resulting in the
formation of a gastrnla with an archenteron. The hypo-
blast cells lining the archenteron become columnar {hy.).
The curving and more rapid growth at the anterior end of
the embryo narrow the primitively wide open blastopore,
and carry it to the posterior end of the future dorsal
surface (PI. V., fig. 4, h.p). The directions of the body
are now clear. The embryo is elongated antero-posteriorly,
the dorsal surface is flattened and the blastopore indicates
its posterior end. Around the blastopore certain of the
ectoderm cells form a medullary plate along which a
groove (the medullary groove) runs forwards, bounded
at the sides by laminae dorsales which meet behind the
blastopore. Underneath the posterior part of the medullary
groove certain of the hypoblast cells from the dorsal wall
of the archenteron in tbe median line form a band

* Tl)e early stages of C'iona, of which Castle lias given a veiy complete
account, ditfer in some [loinls from those of Asisiiliu (Jcsci'ihcd hcic.

a2

extending forwcird (iig. 4, cJi.). This baud separates off
from the liypoblast, which closes in beneath it, and thus
the notochord is formed (fig. 5, ch.). The same cells further
laterally and posteriorly become mesoblast, and separate
off as lateral plates which show no trace of metameric
segmentation (fig. 7a, m.h.). The remainder of the archen-
teron will become the branchial sac, and by further growth
bud off the rest of the alimentary canal.

The medullary groove now becomes converted into the
closed neural canal by the growing up and arching inwards
(fig. 7a, n.c.) of the laminte dorsales, which unite with one
another from behind forwards in such a way that the
blastopore now opens from the enteron into the floor of
the neural canal, forming the neurenteric passage (fig. 5,
ii.e.c). For a time the anterior end of the neural canal
remains open as a neuropore. The posterior end of the
body is now elongating to form a tail, and the embryo is
rapidly acquiring the tadpole shape (fig. 6) characteristic
of the free larva.

The tail grows rapidly, curves round the body, and also
undergoes torsion so that its dorsal surface comes to lie
on the left side. It contains ectoderm cells on its surface,
notochordal cells (in single file) up the centre (see fig. 7,
n.ch.), a neural canal dorsally, and a row of endo-
derni cells (/«?/.) representing the enteron ventrally to the
notochord. Later on the mesoblast also is prolonged into
the tail where it forms a band of striated muscle-cells at
each side of the notochord. When the ectoderm cells
begin to secrete the cuticular test it forms two delicate
transparent longitudinal (dorsal and ventral) fins in the
tail (fig. 7), and especially at its extremity where radial
thickenings form striae resembling fin rays. The ecto-
derm on the anterior end of the body grows out into three
adhering papillae (figs. 8 and 9).

33

The neural canal now differentiates into a tubular dorsal
nervous system. The anterior end dilates to form the
thin-walled cerebral vesicle (fi<::js. 8 and 9, n.v.), containing
later the intra-cerebral, dorsal, pigmented eye (oc.) and the
ventral otolith {an.) of the larva. The next part of the canal
thickens to form the trunk ganglion, and behind that is
the more slender spinal cord {n.c), which runs to the extre-
mity of the tail. A ciliated diverticulum of the anterior
end of the enteric cavity (future pharynx), which enters
into close relations with the front of the cerebral vesicle,
and later opens into the ectodermic invagination which
forms the mouth at that spot, is evidently the rudiment
of the neural or hypophysial gland and canal.

The future branchial sac (fig. 9, mes.), with a ventral
median thickening which will be the endostyle, is by this
time clearly distinguishable, by its large size, from the
much narrower posterior part of the enteron which grows
out to become the oesophagus, stomach, and intestine.

The notochord does not extend forward into the
branchial region, but is confined to the posterior or
caudal part of the embryo. It now shows lenticular
pieces of a gelatinous intercellular substance secreted by
the cells and lying between them (fig. 8). The mouth
forms as a stomodseum or ectodermal invagination antero-
dorsally, in the region where the neuropore had closed
up ; and, about the same time, two lateral ectodermal
involutions appear (fig. 9, at.) which become the atrial or
peribranchial pouches, at first distinct, afterwards united
in the mid-dorsal line to form the adult cloaca and atrial
aperture.

In-growths from the atrial pouches and out-growths from
the wall of the pharynx coalesce to form the proto-
stigmata (primary gill slits) by which the cavity of the
branchial sac is first placed in communication with the

34

exterior tbrougb the atrial apertures. Opinions differ as to
whether only one or a few pairs of true gill clefts are
represented in the young Ascidian, and the actual details
of their formation and sub-division differ greatly in differ-
ent forms. To what precise extent the walls of the atrial
or peribranchial cavities are formed of ectoderm or endo-
derm, is also still doubtful.

The embryo is hatched, about two or three days after
fertilisation, as a larva or Ascidian tadpole (fig. 9) which
leads a free-swimming existence for a short time during
which it develops its nervous system and cerebral sense
organs, and the powerful mesoblastic muscle bands lying
at the sides of the notochord (now a cylindrical rod of
gelatinous nature surrounded by the remains of the
original cells) in the tail, and forming the locomotory
apparatus. Figure 9 shows this stage, the highest in its
chordate organisation, when the larva swims actively
through the sea by vibrating its long tail provided with
dorsal and ventral fins.

In addition to the structures already mentioned, the
mesoderm has formed the beginning of the muscular body-
wall and the connective tissue around the organs, and has
given rise to the blood, the endostyle has developed as a
thick-walled groove along the ventral edge of the pharynx,
which now becomes the branchial sac, and the pericardial
sac and its invagination the heart have formed in the
mesoblast between the endostyle and stomach. The
unpaired optic organ in the cerebral vesicle, when fully
formed, has a retina, pigment layer, lens and cornea ; while
the ventral median sense-organ is a large spherical, partially
pigmented otolith, supported by delicate hair-like processes
on the summit of a hollow "crista acustica " (fig. 9).
Both the otolith and the retina and lens of the eye are
formed oiigmally by the differentiation of a group of cells

35

in the epitbeliuin lining tiie cerebral vesicle — they are
my clonic sense-organs.

After a few hours, or at most a day or so, the larva
attaches itself by one or more of the three anterior ecto-
dermal glandular papillae (one dorsal and two lateral)
to some foreign body, and commences the retrogressive
metamorphosis which leads to the adult state. The
adhering papillae having performed their function begin to
atroph}^ and their place is taken by the rapidly increasing
test. The tail, which at first vibrates rapidly, is partly
withdrawn from the test and absorbed and partly cast off
in shreds (figs. 10 to 12). The notochord, nerve tube,
muscles, &c., are withdrawn into the body, where they
break down and are absorbed by phagocytes, or dissolved
in the fluid of the body-cavity. The posterior part of the
nerve tube and its anterior vesicle with the large sense
organs disappear, and the middle part undergoes prolifera-
tion dorsally to form the relatively small ganglion of the
adult, underneath which the neural tube gives rise to the
hypophysial gland. While the locomotory, nervous, and
sensory organs are thus disappearing or being reduced,
the alimentary canal and reproductive viscera are growing
larger. The branchial sac enlarges, its walls become
penetrated by blood channels and grow out to form bars
and papillae, and the number of openings greatly increases
by the primary gill slits becoming broken up into the
transverse rows of stigmata.

The stomach and intestine, which developed as an out-
growth from the back of the branchial sac at the dorsal
edge, become longer and curve so that the end of the
intestine acquires an opening into at first the left-hand
side, and eventually the cloacal or median part of the atrial
cavity. The adhering papillae have now disappeared, and
are replaced functionally by a growth of the test over

36

neighbouring objects ; and, at the same time, the region
of the body between the point of fixation and the mouth
(branchial aperture) increases rapidly in extent so as to
cause the body of the Ascidian to rotate through about
180°, and so carry the branchial siphon to the opposite
end from the area of attachment (see figs. 10, 11, 12, and
13 on PI. v.).

Finally, the gonads and their ducts form in the meso-
derm between stomach and intestine, and so bring us to
the sedentary degenerate fixed adult Ascidian with little
or no trace of the Chordate characteristics so marked in
the earlier larval stage (compare figs. 13 and 9). The free-
swimming tailed larva shows the Ascidian at the highest
level of its organisation, and is the stage that indicates the
genetic relationship of the Tunicata with the Vertebrata.
In some Ascidians with more food-yolk in the egg, or in
which the development takes place within the body of the
parent, the life-history as given above is more or less
modified and abbreviated, and in some few forms the
tailed larval stage is missing.

The remarkable life-history of the typical Ascidian, of
which the outlines are given above, is of importance from
two points of view : —

1st. It is an excellent example of degeneration. I^he
free-swimming larva is a more highly developed animal
than the adult Ascidian. The larva is, as we have seen,
comparable with a larval fish or a young tadpole, and so is
a chordate animal showing evident relationship to the
Vertebrata ; while the adult is in its structure non-chordate,
and may be regarded as being on a level with some of the
worms, or with the lower Mollusca, in its organisation—
although of an entirely different type.

37

2nd. It shows us the true position of the Ascidians
(Tunicata) in the animal series. If we knew only the
adult forms we might regard them as being an aberrant
group of the Vermes, or possibly as occupying a position
between worms and the lower Mollusca, or we might place
them as an independent group ; but we should certainly
have to class them as Invertebrate animals. But when
we know the whole life-history, and consider it in the
light of "recapitulation" and "evolutionary" views, we
recognise that the Ascidians are evidently related to the
Vertebrata, and were at one time free-swimming Chordata
occupying a position somewhere below the lowest Fishes.

N.B. — In an account of this nature, in which I have
obviously made the fullest use of the published works
of my predecessors, I have not considered it necessary
to burden the text with frequent references to original
memoirs. While accepting, then, full responsibility
for my statements — nearly all of which I have taken
occasion to verify by personal observation— I do not,
of course, claim any originality in regard to them.
Several excellent bibliographies of the Tunicata have
already been published : another seems superfluous.

APPENDIX.

It may be useful to add here a brief statement of the
classification and characters of the Tunicata, in order to
indicate the position of AscicUa as a type of the group,
and its relations to the other British Ascidians.*

TUNICATA.

The Tunicata (or Urochorda) are hermaphrodite marine
chordate animals, which show in their development the
essential vertebrate characters, but in which the notochord
is restricted to the posterior part of the body, and is in
most cases present only during the free-swimming larval
stage. The adult animals are usually sessile and degener-
ate, and may be either solitary or colonies, fixed or free.
The nervous system is in the larva of the elongated,
tubular, dorsal vertebrate type, but in most cases degener-
ates in the adult to form a small ganglion placed above
the pharynx. The body is completely covered with a thick
cuticular test ("tunic "), which contains a substance similar
to cellulose. The alimentary canal has a greatly enlarged
respiratory pharynx (the branchial sac), which is perforated
by two, or many, more or less modified gill slits, opening
into a peribranchial or atrial cavity, which communicates
with the exterior by a single dorsal exhalent aperture,
rarely two ventral apertures. The ventral heart is simple
and tubular, and periodically reverses the direction of the
blood current.

* For a more detailed classification, with definitions of all the groups and
analytical keys to the species, see Herdnian's Revised Classification of the
Tunicata. .lourn. Linn. Soc. -Zool., vol. XXIII , p. 558, 1S91.

39

The leading vertebrate characteristics of the Timicata
are the uotochord, the dorsal nervous system, the ventral
heart, and the respiratory pharynx with gill clefts, but
these all disappear or undergo modification to such an
extent, during the metamorphosis, that the degenerate
adults would not, in most cases, be recognised as belong-
ing to the chordata were it not for our knowledge of the
life-history.

The class Tunicata may be divided into three orders : —

Order I. LARVACEA.

This comprises the free-swimming, permanently-tailed,
larva like, mostly minute Appendicularians. A relatively
large test or "House" is formed with great rapidity as a
secretion from the surface of a special part of the ecto-
derm ; it is, however, merely a temporary structure, which
may be cast off and afterwards replaced by another
*' House." The branchial sac is simply an enlarged pharynx,
with two ventral ciliated openings (stigmata) leading to
the exterior. These open independently on the ventral
surface, and there is no separate peribranchial cavity.
The tail is a large locomotor appendage, in which there
is a skeletal axis, the urochord, comparable with the noto-
chord of Vertebrata. The nervous system consists of a
large anterior and dorsally-placed ganglion, and a long
nerve cord with smaller ganglia stretching backwards
from it over the alimentary canal to reach the tail, along
which it runs on the left side of the urochord. The
alimentary canal lies behind the branchial sac, and the
anus opens ventrally on the surface of the body in front
of the stigmata (or atriopores). The gonads are placed at
the posterior end of the body. Gemmation does not take
place, and alternation of generations and metamorphosis
do not occur in the life-history.

40

This group contains a single family, the Appendi-
culariidae, all minute (about 5 mm. long), tailed, free-
swimming forms which have undergone comparatively
little degeneration, and, consequently, correspond more
nearly to the tailed larval condition than to the adult
forms of the other groups of Tunicata. There are nearly
a dozen genera known, of which at least two, probably
more, inhabit British seas. In the genus Oihopleura, to
which our commonest Appendicularians belong, the body
is short and ovoid, and no anterior fold or "hood" is
present. The tail is three or four times the length of the
body, and four to six times as long as it is broad.
In Fritlllaria, on the other hand, the body is elongated,
and somewhat constricted in the middle where the tail is
attached. A fold of integument on the front of the body
forms a "hood." The tail is short and wide, not twice
as long as the body.

The British species of Larvacea are still insufiticiently
known.

Order II. THALIACEA.

Free-swimming pelagic forms of moderate size, which
may be either Simple or Compound, and in which the
adult is never provided with a tail or notochord. Conse-
quently the whole body here corresponds to the tnmlc
only of the Appendicularian, without the tail. The test
is permanent, and may be either well-developed or very
slight. In all cases it is clear and transparent. The
musculature of the body-wall is in the form of more or
less complete circular bands, by the contraction of which
water is ejected from the body, and so locomotion is
effected. The branchial sac has either two large or many
small apertures (stigmata), leading to a single peribranchial
cavity, into which the anus also opens. Alternation pf

41

generation!? occurs in the life-history, and may be compli-
cated by polymorphism. The order Thaliacea comprises
two groups, Cyclomyaria and Hbmimyaria.

Sub-order I. Cyclomyaria.

Free-swimming pelagic forms, which exhibit alternation
of generations in their life-history, but never form perman-
ent colonies. The body is cask-shaped, with the branchial
and atrial apertures at the opposite ends. The test is
moderately well developed. The musculature is mostly in
the form of complete circular bands surrounding the body.
The branchial sac is fairly large, occupying the anterior
half or more of the body. Stigmata are usually present
in its posterior part only. The peribranchial cavity is
mainly posterior to the branchial sac. The alimentary
canal is placed ventral ly close to the posterior end of the
branchial sac. Hermaphrodite reproductive organs are
placed ventral ly near the intestine.

This group contains only one British genus, Duliolum,
and even that is probably only an occasional visitant. It
has a cask-shaped body, from 1 to 2 cm. in length, with
lobed terminal branchial and atrial apertures. The body-
wall contains eight or nine circular muscle bands, by the
contraction of which tlie animal swims.

The best-known British form, Doliolum tritonis, has
been captured on occasions in thousands off our N.W.
coast, between the Hebrides and the Faroes. This species,
and the closely allied D. nationalis, have also been found
in the English Channel and off the S.W. coast of Ireland.

Sub-order II. Hemimyaria.

Free-swimming pelagic forms, which exhibit alternation
of generations in their life-history, and in the sexual con-
dition form colonies. The body is more or less fusiform.

42

with the long axis antero-posterior, and the branchial and
atrial apertures nearly terminal . The test is well developed.
The musculature of the body-wall is in the form of a series
of transversely-running bands, which do not form complete
independent rings, as in the Cyclomyaria. The branchial
and peribranchial cavities form a continuous space in the
interior of the body, opening externally by the branchial
and atrial apertures, and traversed obliquely from the
dorsal and anterior end to the ventral and posterior by a
long, narrow, vascular band, which represents the dorsal
lamina, the dorsal blood sinus, and the neighbouring part
of the dorsal edge of the branchial sac of an ordinary
Ascidian. The alimentary canal is placed ventrally. The
embryonic development is direct, no tailed larva being
formed.

The Salpidfe, the chief family in this sub-order, includes
the single genus Salpa, which, however, may be divided
into two well-marked groups of species — (1) those, such
as S. plnnata, in which the alimentary canal is stretched
out along the ventral surface of the body, and (2) those,
such as S.fiisiformis, in which the alimentary canal forms
a compact globular mass, the "nucleus," near the posterior
end of the body. About fifteen species altogether are
known; they are all pelagic forms, and are found in many
seas. Each species occurs in two forms — the solitary
asexual (proles solitaria) and the aggregated sexual (p)-oles
gregaria) — which are usually quite unlike one another.
The so'itar}^ form gives rise, by internal gemmation, to a
complex tubular stolon, which contains processes from
all the more important organs of the parent body, and
which becomes segmented into a series of buds. As the
stolon elongates, the buds near the free end, which have
become advanced in their development, are set free in
groups, the members in which remain attached together

, 43

by processes of the test, each enclosing a diverticulum
from the body -wall, so as to form " chains." Each
member of the chain is a Salpa of the sexual or aggre-
gated form, and when mature may — either still attached
to its neighbours or separated from them — produce one
or several embryos, which develop into the solitary Salpa.
Thus the two forms alternate regularly.

Salpa, like Doliolum, is probably only an occasional
visitor in our seas, but several species of the genus —
Salpa democratica-mucronata, S. ruucuiata-fusiformis,
and S. zo)iaria^\SbYe been found on occasions in the seas
of the Hebrides, or cast ashore on our southern and
western coasts.

Order III. ASCIDIACEA.

Fixed or free-swimming Simple or Compound Ascidians,
which, in the adult, are never provided with a tail, and
have no trace of a notochord. The free-swimming forms
are colonies, the Simple Ascidians being always fixed.
The test is permanent and well developed ; as a rule, it
increases with the age of the individual. The branchial
sac is large and well developed. Its walls are perforated by
numerous slits (stigmata) opening into the peribranchial
cavity, which communicates with the exterior by the atrial
aperture. Many of the forms reproduce by gemmation,
and in most of them the sexually produced embryo develops
into a tailed larva.

The Ascidiacea includes three groups — the Simple
Ascidians, the Compound Ascidians, and the free-swim-
ming colonial Pijrosoma.

Sub-order I. Ascidi^ Simplices.

Fixed Ascidians which are solitary and very rarely
reproduce by gemmation ; if colonies are formed, the

44

members are not buried in a common investing mass,
but each has a distinct test of its own. No strict line of
demarcation can be drawn between the Simple and Com-
pomid Ascidians, and one of the families of the former
group, the Clavelinidse (the Social Ascidians), forms a
transition from the typical Simple forms, which never
reproduce bj' gemmation to the Compound forms, which
always do.

The Ascidise Simplices may be divided into the follow-
ing families : —

Family I. Clavelinid^ : — Simple Ascidians which
reproduce by gemmation to form small colonies, in which
each Ascidiozooid has a distinct test, but all are connected
by a common blood-system. Buds formed on stolons,
which are vascular out-growths from the posterior end of
the body containing prolongations from the ectoderm,
mesoderm, and endoderm of the Ascidiozooid. Branchial
sac not folded ; internal longitudinal bars usually absent ;
stigmata straight ; tentacles simple.

This family contains three chief genera — Ecteinascidia,
with internal longitudinal bars in branchial sac ; Glavelina,
with intestine extending behind branchial sac ; and Pero-
phora, with intestine alongside branchial sac. Glavelina
lepadiformis and Perophora listeri are common British
species found at a few fathoms depth off various parts
of our coast. Both occur round the south end of the Isle
of Man,

Family II. Ascidiid^e : — Solitary fixed Ascidians with
gelatinous test ; branchial aperture usually 8-lobed, atrial
aperture usually 6-lobed. Branchial sac not folded ; in-
ternal longitudinal bars usually present ; stigmata straight
or curved ; tentacles simple.

This family contains, along with several other genera,

45

our typical fonn AsckUa, of which there are many species
ill British seas, widely distributed round our coasts. Two
other common British forms, belonging to this family, are
Ciona intestinalis, with a very soft, pale-green test and
languets in place of a dorsal lamina, and Corella parallelo-
gramma, in which the stigmata of the branchial sac are
curved to form beautiful spirals.

Family III. Cynthiid^ : — Solitary fixed Ascidians,
usually with leathery test ; branchial and atrial apertures
both 4-lobed Branchial sac longitudinally folded; stig-
mata straight ; tentacles simple or compound.

This is the largest family of Simple Ascidians, and
contains a number of genera, about six of which are
British. Stijela has simple tentacles, and not more than
4-folds on each side of the branchial sac. A very common
species all round our coasts, between tide marks, is the
little red Stijela (or Styelopsis) grossularia. In Cynthia
the tentacles are compound, and there are more than
4 folds, usually 7 or 8, on each side of the branchial sac.
The curious little Forbesella tessellata, from deep water
in the Irish Sea, is in some respects intermediate between
Styela and Cynthia.

Family IV. Molgulid^ : — Solitary Ascidians, often
not fixed; branchial aperture 6-lobed, atrial 4-lobed. Test
usuall}^ encrusted with sand. Branchial sac longitudinally
folded ; stigmata more or less curved, usually arranged in
spirals; tentacles compound.

Several species of Molgula, all looking when dredged
like little sandy balls, and one of Eugyra (E. glutinans) ,
are common at a few fathoms depth round most parts of
our coasts.

46

Sub-order II. AsciDi^ Cuaipositji:.

Fixed Ascidians which reproduce by gemmation, so as
to form colonies, in which the Ascidiozooids are buried in
a common investing mass, and have no separate tests.
This is probabl}' a somewhat artificial assemblage, formed
of two or three groups of Ascidians which produce colonies
in which the Ascidiozooids are so intimately united that
they possess a common test or investing mass. This is
the only character which distinguishes them from the
Clavelinidfe., but the property of reproducing by gemma-
tion separates them from the rest of the Ascidia3 Simplices.
The Ascidiae Compositae may be divided into the following
families : —

Family I. Distomid^ : — Ascidiozooids divided into two
regions, thorax and abdomen; testes numerous ; vas deferens
not spirally coiled. The chief genera are — Distoma, Dis-
taplia, Colella, the last forming a pedunculated colony,
in which the Ascidiozooids develop incubatory pouches,
opening from the peribranchial cavity, in which the
embryos undergo their development.

Family II. Ccelocormid^ : — Colony not fixed, having
a large axial cavity with a terminal aperture. Branchial
apertures 5-lobed. This includes one species, Ca'locornms
huxleyi, which is a transition form between the ordinary
Compound Ascidians (e.g., DistomidseJ and the Ascidioe
Luciae (Pyrosoma).

Family III. Didemnid^ : — Colony usually thin and
incrusting. Test containing stellate calcareous spicules.
Testis single, large ; vas deferens spirally coiled. The
chief genera are — Didemnum, in which the colony is thick
and fleshy, and there are only three rows of stigmata on
each side of the branchial sac ; and Leptoclimnn, in which
the colony is thin and incrusting, and there are four rows

47

of stigmata on each side of the branchial sac. Colonies of
Leptoclinum, forming thin, white, grey, or yellow crusts,
under stones at low water, are amongst the connnonest of
British Compound Ascidians.

Family IV. Diplosomid^ : — Test reduced in amount,
rarely containing spicules. Vas deferens not spirally
coiled. In Diplosoma , the most important genus, the
larva is gemmiparous. This is a common British form,
especially on Zosfera beds, and amongst sea-weeds.

Family V. Polyclinid^ : — Ascidiozooids divided into
three regions, thorax, abdomen, and post-abdomen. Testes
numerous ; vas deferens not spirally coiled. The chief
genera are — Pharyngodictyon, with stigmata absent or
modified, one species : the only Compound Ascidian
known from a depth of 1000 fathoms ; Polyclinum, with
a smooth-walled stomach ; Aplidium, with the stomach-
wall longitudinally folded; and Amaroucium,m. which the
Ascidiozooid has a long post-abdomen and a large atrial
languet. The last three genera contain many common
British species.

Family VI. Botryllid.^ : — Ascidiozooids having the
intestine and reproductive organs alongside the branchial
sac. Dorsal lamina present ; internal longitudinal bars
present in branchial sac. The chief genera are — Botryllus,
with simple stellate systems, and Botrylloides* with
elongated or ramified systems. There are many species
of both these genera, which form brilliantly coloured fleshy
crusts under stones and on sea-weed at low tide. They are
amongst the commonest and the most beautiful of British
Ascidians.

* It is intended that a future L.M.B.C. Memoir will deal with But rijllu ides
as a type of the Compound Ascidians.

48

Family VII. Polystyelid^ : — Ascidiozooids not
grouped in systems. Branchial and atrial apertures 4-
lobed. Branchial sac may he folded; internal longitudinal
bars present. The chief genera are — Thylacium, with
ascidiozooids projecting above general surface of colony ;
Goodsiria, with ascidiozooids completely imbedded in
investing mass ; and Clwrizocormus, with ascidiozooids
united in little groups, which are connected by stolons.
The last genus contains one species, Cli. reticulatus, a
transition form between the other Polystyelidas and the
Cynthiidte among Simple Ascidians. Tlujlacium is the
only British form.

Sub-order III. AsciDi^ Luci^.

Free-swimming pelagic colonies having the form of a
hc^Uow cylinder closed at one end. The ascidiozooids
forming the colony are injbedded in the common test in
such a manner that the branchial apertures open on the
outer surface, and the atrial apertures on the inner surface
next to the central cavity of the colony. The ascidiozooids
are produced by gemmation from a rudimentary larva (the
cyathozooid) developed sexually.

This sub-order includes a single family, the Pyrosomid^,
containing one well-marked genus, Pyrosoma, with several
species. They are found swimming near the surface of
the sea, chiefly in tropical latitudes, and are brilliantly
phosphorescent. A fully developed colony may be from
an inch or two to upwards of four feet in length. Pyro-
soma does not occur in British seas.

49

Explanation of Plates.
Plate I.

Large specimen of Ascidia mentula, from the right side,
natural size, from Hfe.

Length 11"5 cm. Breadth 6 cm. Thickness 3 cm.

A, shows the sensory edge of the branchial aperture
with lobes and coloured lines (the " ocelli "), enlarged.

Plate II.
Reference Letters in Plates II., III., and IV.

g.d. genital ducts.

a. anus.

At. atrial aperture.

at. I. atrial lobe.

hl.c. bladder cell.

hl.s. blood sinus.

Br. branchial aperture.

hr.car. brancliio-cardiac vessel.

hr.ao. branchial (veutral) vessel

Br.s. branchial sac.

br.visc. branchio- visceral vessel

card. vise, cardie- visceral vessel

c.c. common cloaca.

c.d. connecting duct.

cl. cloaca.

C071. connective.

c.t.c. connective tissue cell.

d.ao. dorsal vessel.

d.bl.s. dorsal blood sinus.

d.l. dorsal lamina.

d.n. dorsal nerve cord.

d. t. dorsal tubercle.

ec. ectoderm.

eii. endoderm.

end. endostyle.

ej).c. epicardial tube.

gl. pyloric gland.

gl.d. duct of neural gland.

goii. gonads.

h. heart.

hm. horizontal membrane.

hijp. hypophysial (neural) gland

liyp.d. hypophysial duct.

i.i' . intestine.

i.l. internal longitudinal bars.

l.v. interstigmatic vessels.

m. body-wall, or mantle.

m.h. muscle bands.

mes.c. mesoblast cell.

VI. f. muscle fibres.

n. nerves.

ii.g. nerve ganglion.

n.gl. neural gland.

(ss. oesophagus.

ov. ovary.

jy.br. peribranchial cavity.

2). br.s. prebranchial zone.

2).c. pericardium.

p.p'. papilliE.

2)2).b. peripharyngeal bands.

50

r. rectum. tn. tentacle.

ren. renal vesicles. t.k. terminal knobs on vessels.

sg. stigmata. tr. transverse vessels.

sp. testis. t.v. test vessel.

sph. sphincter. ty. typhlosole.

St. stomach. v. vessel.

t. test. v.app. vascular appendage.

t.c. test cells. v.bl.s. ventral blood sinus.

Fig. 1. Diagram of the outside oiAscidia, from right side.

Fig. 2. Transverse section through the atrial aperture
to show arrangement of internal cavities and
organs. (To avoid complication the internal
longitudinal bars are not represented in the
branchial sac.)

Fig. 3. A mesh of the branchial sac, diagrammatic.

A. From the inner surface. B. In section.

Fig. 4. Diagrammatic dissection to show the structure
oi Ascidia (compare with fig. 2).

Fig. 5. Section through test and mantle (body-wall) to
show the relations of the ectoderm and meso-
derm to the test.

Fig. 6. Sagittal section through antero-doi'sal part of
body to show relations of nerve ganglion,
neural gland, &c.

Fig. 7. Dorsal front of pharynx from inside, to show
dorsal tubercle, tentacles, and neighbouring
parts. X 50.

Plate III.

Figs. 1 to 4 show typical examples of four important

families of Compound Ascidians, natural size.
Fig. 1 is a Distomid colony {Colella).
Fig. 2 is a Didemnid colony {LeptocUmun).
Fig. 3 is a Polyclinid colony {Pharyngodictyon) .

51

Fig. 4 is a Botryliid colony {Botryllus) .

Fig. 5 shows the union of two Ascidiozooids, in a Com-
pound Ascidian colony, to form a common
cloaca (c.c.) in the common test.

Figs. 6 to 8 show the three forms of body found amongst
the Ascidiozooids of Compound Ascidians.

Fig. 6 is from a Botryliid colony.

Fig. 7 is from a Distomid colony.

Fig. 8 is from a Polyclinid colony (all from right side,
magnified).

Fig. 9. Section through outer part of test of Asoidia to
show " vessels " and " bladder " cells. X 50.

Fig, 10. Diagram of Ascidia to show the arrangement
of the blood system, from left side, in correct
morphological position for comparison with
Vertebrates.

Plate IV.

Fig. 1. Part of branchial sac of Ascidia vienttda, from

the outside. X 50.
Fig. 2. Part of branchial sac of Ascidia mentiila from

the inside. X 50.
Small part of last. X 300.
Part of the dorsal lamina. X 40.
Transverse section of the endostyle, X 200.
Blood corpuscles oi Ascidia.
Part of section of wall of intestine showing

refringent organ, spermatic tubules, &c.
Ocellus from the branchial aperture, in section.
The heart and pericardium, in side view.
The heart and pericardium, in section.
Group of renal vesicles, X 50.
Part of a renal vesicle, more highly magnified.
Two cells from wall of renal vesicle,

Fig.

3.

Fig.

4.

Fig.

5.

Fig.

6.

Fig.

7.

Fig.

8.

Fig.

9.

Fig.

10.

Fig.

11.

Fig.

12.

Fig.

13.

52

Fig. 14. Concretions from renal vesicles. X 200.
Fig. 15. Bunch of spermatic tubules from testis. X 50.

Plate V.
Reference letters. — ar. arcbenteron, at. atrial in-
volution, . ail. otolitb, be. blastocoele, bp. blastopore,
ch. notochord, ep. epiblast, /. tail fin, foil, follicle cell,
hy. hypoblast, i. intestine, ??i. mouth, 7nes. mesoblast,
muse, muscle cells, mb. mesoblast layer, mc. mesoblast
cell, n. nucleus of ovum, nc. neural canal, n.ex. neuren-
teric canal, 7i. ch. notochord, ?i.v. cerebral vesicle, oc.
cerebral eye, p. protoplasm of ovum, s. spermatozoon, t.c.
" testa " cells of ovum.

Fig. 1. Mature ovum and spermatozoon of Ascidia,
Fig. 2. A segmentation stage, in section, to show blas-

tula.
Fig. 3. Early gastrula stage.
Fig. 4. Later gastrula stage.

In these and some of the other figures, the
clear cells indicate the notochord, the hypo-
blast cells have small circles, and the nerve
cells are marked with obhque lines, while the
mesoblast cells are cross hatched.
Fig. 5. A later embryo, showing rudiments of noto-
chord and nervous system.
Embryo showing body and tail, &c.
Transverse section of the tail of larva.
7a. Transverse section of the body of embryo.
Embryo ready to be hatched.
Free-swinnning tailed larva.
The metamorphosis — larva attached to a stone.
Tail and nervous system of larva degenerating.
Further degeneration and metamorphosis of
larva into
Fig. 13. tlie young fixed Ascidian.

Fig.

6.

Fig.

7.

Fig.

7a

Fig.

8.

Fig.

9.

Fig.

10.

Fig.

11.

Fig.

12.

0.

.-^

■^

5

u

CQ

'' vSps**-'^'

L.M.B.C. Memoir I.

Plate II.

sg /-j?/Z. l'. ov. J

W.A.H. del.

dl-

ASCIDIA.

L.M.B.C. Memoir I

Plate III.

en-d.

-cytrcL. vise
PosteJ~ior

Ve^li7~(zZ

W.A.H. del.

Fig. 10.

ASCIDIA.

L.M.B.C. Memoir I.

Plate TV.

W.A.H.del

ASCIDIA.

S.B.Iifh.

L.M.B.C. Memoir I.

Plate V.

W.A.H. del.

Fiy. 10.

Fig. 11. Fig. 12.

ASCIDIA.

Fig. 13.

3 2044 093 366 359

s: