BRITISH MUSEUM (NATURAL HISTORY)

BRITISH ANTARCTIC (“TERRA NOYA”) EXPEDITION, 1910.

NATURAL HISTORY REPORTS.

ZOOLOGY. VOL. V.

COELENTERATA.

LONDON :

PRINTED BY ORDER OF THE TRUSTEES OF THE BRITISH MUSEUM.

Sold by B. Quaritch, Ltd., 11, Grafton Street, New Bond Street, London, W.l ;

Dulau & Co., Ltd., 2, Stafford Street, London, W.l ;

Oxford University Press, Warwick Square, London, E.C.4 ;
and at the

British Museum (Natural History), Cromwell Road, London, S.W.7.

1918-1930.

[All rights reserved.]

.

■

A

'

*

*

.

ZOOLOGY. VOL. V.

CONTENTS.

1. Stephenson, T. A. Coelenterata. Parti. — Actiniaria. [Issued April 27, 1918.]

Pp. 1-68, 6 pis.

2. Matthai, GeoPvGE. Coelenterata. Part II. — Madreporaria.

(a) On Favia confer ta, Yerrill, with notes on other Atlantic Species of Favia.

[Issued November 22, 1919.]
Pp. 69-96, 4 pis.

3. Totton, A. Knyvett. Coelenterata. Part III. — Antipatharia (and their Cirripede

Commensals).
[Issued October 27, 1923.]
Pp. 97-120, 2 pis.

4. Gardiner, J. Stanley. Coelenterata. Part IV.-
(6) Turbinolidae and Eupsammidae.

-Madreporaria.

[Issued August 17, 1929.]
Pp. 121-130, 1 pi.

5. Totton, A. Knyvett. Coelenterata. Part V. — Hydroida.

[Issued February 22, 1930.]
Pp. 131-252, 3 pis.

1

COELENTERATA.

PART I— ACTINIARIA.

BY T. A. STEPHENSON,

Demonstrator of Zoology, University College of Wales, Aberystwyth.
WITH SIX PLATES.

PAGE

I. — Introduction .......... 1

II. — General Notes .......... 2

III. — Definitions of Terms ......... 6

IV. — Descriptions of Species ........ 8

V. — Literature .......... 62

Index ........... 65

I.— INTRODUCTION.

Throucih the kindness of Dr. S. F. Harmer, Keeper of Zoology, British Museum
(Natural History), I have had the pleasure of studying the Actiniaria collected by
the “ Terra Nova ” Expedition. The collection, though not a large one, contains
representatives of eight families, so that the fifteen species which it includes show, a
fairly wide range of variation. All the species belong to the Actiniina, the
Stichodactylina being quite unrepresented, as are also the orders Edwardsiaria
and Zoanthinaria.

The most interesting part of the collection was made in the Antarctic- itself. In
the vicinity of McMurdo Sound seven species were obtained — Ilalcampoides aspera,
Dactylanthe antarctica, Cymhactis selaginella, Lilliella lacunifera , Leptoteichus insignis,
Ilormosoma scotti and Artemidcictis victrix. Two of these species — namely, Cymhactis
selaginella and Artemidactis victrix — were also obtained in Ross Sea. All were dredged
from deep water, and most of them are animals with a thick, stiff body-wall. In
addition to these, specimens were collected in the following localities : —

1. New Zealand, on shore. Aipjtasia minima.

Sagartia, sp.

2. Off New Zealand in 70 fms. Halcurias endocoelactis.

Epiactis novo-zealandica.

B

VOL. V.

2

TERRA NOVA

EXPEDITION.

3. Off Falkland Islands.

4. Off Eio de Janeiro.

5. Island of S. Trinidad, on shore.

Bolocera longicornis.
Calliactis reticulata.
Phymactis clematis.
Aiptasioides prima.

The majority of the specimens are excellently preserved in a state of expansion
in formalin, and the large species make very fine specimens. On the other hand, two
or three of the small ones are in such poor condition that it is impossible to make the

descriptions of them complete. In preparing sections of the material I have used

several methods of staining, but have found double-staining with haematoxylin and
eosin by far the most useful for all purposes. The sections were immersed in a 2 per

cent, solution of ammonia iron alum before staining with haematoxylin : they were

allowed to remain in a ’5 per cent, solution (or weaker) of haematoxylin for ten minutes
or upwards, till they were a good deal overstained, and were then destained to the
required shade with 2 per cent, ammonia iron alum. After clearing with xylol they
were stained for about fifteen minutes with a saturated solution of eosin in clove oil,
and mounted. This method brings out nuclei, nematocysts, gland-cells and muscula-
ture extremely well. Its general result may be seen from Plate II.

I have found it necessary to introduce five new genera and eleven new species. I
have considered very carefully the propriety of instituting each one, and to the best of
my belief they are valid.

With one exception (PI. VI, fig. 16), the figures are all drawn from “Terra
Nova” material.

I should like to offer my most sincere and hearty thanks to Dr. H. J. Fleure and
Mr. C. L. Walton of this College, and also to Prof. J. Playfair McMurrich and Prof.
G. C. Bourne, for the kindness and help which they have given me in various ways
during the course of my work. Dr. Fleure in particular has been tireless in obtaining
for me necessary notes from literature which I was unable to see myself : and Prof.
McMurrich and Prof. Bourne have given me freely both information and their opinions
on several points.

I have included a coloured figure of a preserved specimen of each species described
(showing, of course, only what colour it retained in the preserved state) in order to
facilitate identification.

II -GENERAL NOTES.

A.

In a paper (54a)* on Alcyonium digitatum Miss Edith M. Prattf has clearly shown
that in that species (and probably in other Alcyonaria) the function of the curious stellate

* Numbers in parentheses refer to numbers of papers in the bibliography, to which reference should be

made for details.

f Now Mrs. Musgrave.

ACTINIAE! A — STEPHEN SON.

3

cells forming the mesogloeal cell-plexus is primarily nutritive, and probably secondarily
nervous ; and that they are amoeboid cells. She has also shown that the granular
gland-cells of the actinopharynx and mesenterial filaments secrete a digestive fluid
which acts on the food before it is ingested by the amoeboid endoderm cells. There
seems no reason to doubt that the cells of the mesogloea of the Actiniaria (such as I
have described in many of the species dealt with in this paper, and such as have been
from time to time described by various authors) are similarly nutritive amoeboid cells,
and their appearance often suggests that this is so : some contain what are very
probably food-vacuoles and particles. In this case they are not, as has been supposed,
either connective tissue cells or nerve-cells, though they have a secondary nervous
function. In some cases I have found similar cells in the endoderm or ectoderm, but
they do not seem to be conspicuous there.

B.

In three species I have noted a curious phenomenon. In Leptoteichus insignis and
Bolocera longicornis there are numerous zooxanthellae in the gonads, lying round the
testes. In Cymbactis selaginella they are even mixed up with the sperms, but lie
outside the ova. It is interesting to note in this connection that in many animals the
gonads tend to be placed in such positions that they obtain a good supply of oxygen.
To take a few instances : in Amphioxus the gonads lie in the wall of the atrial cavity,
where the stream of water passing out through the gill-slits reaches them ; in
Balanoglossus the arrangement is rather similar, the gonads being placed in genital
wings on each side of the branchial region. In Chiton, they lie immediately against
the aorta ; while in many Decapod Crustacea they are immediately below the floor of
the pericardial sinus. It seems fair to deduce, therefore, that the presence of
zooxanthellae in the gonads of some Actiniaria may be useful to the gonad in a similar
way, since the zooxanthellae give off oxygen. It is, of course, possible that in the
female the ova may finally absorb some of these algal cells, as is the case in Millepora
(36a), but of this there is at present no evidence.

C.

In the descriptions which follow I have omitted details of the basilar (not parieto-
basilar) muscles in those species which possess them, and also exact measurements of
cnidae and spirocysts, as those seem to me to be non-essential points. In the Plates,
I have treated the musculature severely in black and white, not showing individual
fibres as a rule, but only whole processes. As regards mesogloeal sphincters, in each
case I have given two figures — one to show the outline of the whole, another to
show exactly the detail of a small part much enlarged. I hope this method will prove
satisfactory, as it is difficult to give accurately the general appearance of a mesogloeal
sphincter when the fibres are so numerous that they cannot be drawn individually.

b 2

4

“TERRA NOVA” EXPEDITION.

I).

It is often very difficult to know what characters are of most value as distinctions
of families, genera, and species in dealing with Actiniaria — or indeed what characters
are of any classificatory value at all. I have given some attention to discriminating
between the relative values of different characters in this group, and one or two points
have occurred to me as worth noting. It must of course be understood that here as
elsewhere it is impossible to lay down hard and fast rules. The following remarks
do not pretend to be anything like a thorough treatment of the subject, but I hope they
may be of some use. There is no single character that is used in classifications of this
group to which at least one exception cannot be found, but some are more constant
than others. I hope to enter much more fully into this subject at an early date, giving
the reasons for my conclusions in detail.

(i) External characters. — The possession by an anemone of acrorhagi, vesicles, a
collar, or similar well-marked structures of course separates it from those sorts which
either possess none of these structures, or possess different ones. And species, and
often genera, may be easily distinguished by the arrangement and form of such
structures when present. Then, again, sometimes the oral disc may be lobed, or it may
widely exceed the column, as in Artemiclactis and Cereus, and these features form
sound characters of distinction. Capitular ridges seem variable, and of doubtful value.
But there is one character which has sometimes been quoted in generic definitions
which is emphatically of no value whatever, and that is the height of the body in
relation to its width. If I may take one example, I will mention the case of
Parantheoides Carlgren. In 1898, Carlgren (15) noted in generic definitions that
Parantlieoides has the body at most twice as high as broad, while in Paranthus it is
several times as high as wide, with a weak base.

Now for some years I have studied living Actiniaria in their natural habitats, and
have successfully kept numbers of them for periods exceeding a year in captivity.
They impress one strongly with the fact that but for special features, such as those
mentioned above, they have no set form. For instance, I have seen a specimen of
Sagartia miniata assume almost every conceivable form within a day : it would close
to a button, stand erect as a short thick pillar, become urn-shaped, vase-shaped, or
constricted in the middle at one or even several points, like an hourglass. It could
elongate itself till it became as narrow as a worm, and even assumed the shape
characteristic of the genus Kodioides — that is to say, it developed a narrow stem not a
quarter of an inch in thickness, with a great globular head at the end of it. If this
specimen had been preserved it would naturally have become fixed in any position
which it had happened to assume at the moment of death, or which any anaesthetic
used had caused it to take up. On one occasion I preserved another specimen of the
same species. While alive, it always remained flat and dish-like, and even after feeding
(which is the time when anemones assume their finest shapes) it did not rise above a

ACTINIAE! A — STEPHENSON.

5

very low pillar. But in this case, when preserved, it stretched itself out till it was as long
and narrow as a thick worm. Similarly, Antliopleura alfordi, when alive, will change
its shape within ten minutes from that of a flat crateriform dish, not half an inch thick
though fully expanded, to that of a tall graceful pillar expanded at the top. Sagartia
viduata may change from being five inches high, and shrink in one minute till it is only
two or three millimetres high. Even the thick- walled Urticina crass icornis may extend
its body into a tall pillar several times as high as wide, though it is not a frequent
occurrence. My point is, that when a specimen is once preserved it may be fixed at
any stage of extension, and it is not possible to tell how long it could be if it wished.
So that although Parantheoicles is doubtless distinct from Paranthus, the fact that when
preserved one is twice, the other more than twice, as long as wide, seems an unsuitable
character for generic distinction.

(ii) Gonads. — I have intentionally omitted any mention of these in my new
generic definitions, but hope to enter into this question at a later date.

(iii) Perfect mesenteries. — Whether the anemone has only six pairs of perfect
mesenteries or numerous perfect pairs seems on the whole a good and constant generic
character. Even to this, Dactylanthe antarctica is an exception. The number of perfect
mesenteries is erratic in the genus Sagartia, but that is a genus which has to be
considered by itself as a special study, and it may be in a fluid state at the present time.

(iv) Sphincter -muscles. — Families are of course easily separated into two sets —
those with mesogloeal and those with endodermal sphincters. There are two exceptions
to this (Sagartiidae and llyanthidae), but the character of the sphincter is an invariably
good generic distinction. The sphincter may vary a good deal as to detail in one genus
( e.g ., Epiactis and Actinostola), but as far as I am aware it is always the same (within
narrow limits) for the various individuals of one species. So that if the sphincter
of any given Paractid, say, is alveolar and vertically layered (as in Cymbactis selaginella
and Actinostola callosa) that is a good specific character.

(v) Tentacles. — Speaking generally, forms with the longitudinal muscles of the
tentacles ectodermal fall into different genera from those which have the muscles
embedded in the mesogloea. There are one or two cases where some species have them
ectodermal, some partly ectodermal and partly mesogloeal. Paradis, as at present
limited, seems to have both types.

(vi) Mesenterial musculature. — The shape of the retractor muscle may vary a
good deal in one species, but generally two species may be distinguished by their
retractors. Two very distinct types of retractor are found : those which are
circumscribed, i.e., have a reniform outline in transverse section ; and those which are
diffuse. These constitute stable and important characters for generic differentiation.
The nature of the individual muscle-processes of a retractor muscle is a useful one for
distinguishing species, though it can hardly be used as a generic character. Reference
to the descriptions and figures of Artemidactis, Leptoteichus , Halcampoides aspera and
Cymbactis selaginella will make clear some of the main types of muscle-processes.

6

TERRA NOVA” EXPEDITION.

(vii) As I have mentioned in connection with Halcurias endocoelactis, it does not
seem to me that the possession by the body-wall ectoderm of spirocysts and
musculature can be considered of more than specific or sometimes generic value.
Halcurias endocoelactis has no ectoderm spirocysts in the body-wall, but they are found
in H. pilatus and II. carlgreni. Bolocera brevicornis has muscles in the body-wall
ectoderm, other species of Bolocera have not. Boloceroides has this musculature also,
but seems to me as nearly related to Bolocera as to Protanthea , so that there it
is only of generic value, if any.

(viii) Summary. — It would seem then that the most generally reliable generic
characters were to be found in external features of the body other than its length ; in
the number of perfect mesenteries ; and in the musculature. That the position of the
gonads, and the presence or absence of spirocysts and ectodermal musculature (though
no doubt the latter represent primitiveness or the survival in some cases of primitive
traits) in the body-wall, were less useful and of secondary value. And that, finally,
the length of the body in proportion to its width is quite useless, unless the animal
is studied alive, and it is proved that it cannot change its proportions. Of course
in special families other characters come in, such as the distribution of cinclides in the
Sagartiidae ; but these hardly need enumeration. I do not make the above statements
dogmatically, but have endeavoured to draw fair conclusions from the evidence at my
disposal.

Ill— DEFINITIONS OF TERMS.

Acontia. — Slender white or coloured threads attached to the borders of the
mesenteries in some families of Actiniaria, just below the mesenterial filaments.
They are loaded with nematocysts, and can be protruded through the mouth, and in
some cases through special pores (cinclides) in the body-wall, for purposes of defence or
to paralyse prey. In histological structure they differ completely from the mesenterial
filaments.

Acrorhagi. — Marginal outgrowths of the body-wall found in some genera of
Actiniaria, containing a strong concentration of nematocysts. They may be simple
and spherical, slightly branched, or even frondose.

Actinopharynx. — The “gullet,” “oesophagus,” “pharynx” or “ stomodaeum ” of
the Actiniaria.

Actinostome. — The mouth, or upper aperture of the actinopharynx, in the centre of
the oral disc.

Alveolar. — In transverse section the mesogloeal sphincters of many anemones
are composed of numerous small more or less circular groups of muscle-fibres, these
small circles having the appearance of alveoli. In a sphincter in wrhich the “ alveoli ”
are fairly scattered, the structure may be termed “alveolar.” When the “alveoli ” are

ACTINIARIA — STEPHENSON.

7

so close together that the strands of mesogloea between them are very reduced, and
form a network, the structure may be termed “ reticular.”

Capitulum. — The bodies of some Actiniaria show a distinction into three regions :
the cylindrical central part of the body, which is often provided with a cuticle, is
termed the “ scapus.” The distal extremity, which bears the tentacles, is termed the
“ capitulum ” ; it is not furnished with a cuticle, is usually rather different in structure
from the scapus, and can be completely introverted, as a rule, into the latter. The
aboral end of the body if rounded and swollen out iuto a bladder-like structure is called a
physa. Many species possess scapus and capitulum, but an ordinary pedal disc instead
of a physa. In some cases a pedal disc can be temporarily inflated to form a physa.

Ciliated lobes of mesenterial filament.— In transverse section a simple filament
presents a trefoil-like form : the two lateral lobes consist of plain ciliated cells, the
central lobe consists of ciliated cells, but among them occur numerous gland-cells and
often nematocysts. The central lobe is referred to in this report as the “ glandular
lobe,” and the lateral lobes as the “ ciliated lobes.”

Ci riel ides. — Small apertures or pores in the body-wall of some Actiniaria, through
which the acontia, if present, may be protruded.

< 'nidae are the same as nematocysts.

Coelenteron. — The general cavity of the body of a Coelenterate.

Endocoel. — The space between two mesenteries of a pair.

Enterostome. — The lower or internal orifice of the actinopharynx, by which the
cavity of the latter is put into communication with the coelenteron.

Exocoel. — The space between two pairs of mesenteries.

Fosse. — Some anemones have the margin of the body raised into a distinct rim, or
parapet, outside the liases of the tentacles ; the circular groove between this parapet
and the tentacle-bases is known as a “ fosse.”

Glandular lobe of mesenterial filament. See “ Ciliated lobes.”

Limbus. — The margin of the pedal disc.

Marginal stomata.— Large or small apertures in large mesenteries (one aperture
per mesentery), near the margin of the body, by means of which water may pass through
the mesenteries.

Nematocysts. — Stinging capsules found in various parts of the bodies of most
coelenterates, and contained in special cells known as nematoblasts.

Oral stomata. — Apertures in perfect mesenteries (one aperture per mesentery), in
the angle formed by the oral disc and actinopharynx, by means of which water may
pass through the mesenteries.

Pennon. — The retractor muscle of a mesentery in transverse section often has the
form of a flag or a pennon, and is therefore referred to as the “ muscle-pennon.”

Perfect mesentery. — A mesentery whose distal edge (the edge furthest from the
body-wall) is at least in part attached to the actinopharynx-wall.

Physa. See “ Capitulum.”

8

“TERRA NOVA” EXPEDITION.

Pseudo-acrorhagi. — Marginal outgrowths of the body-wall of some Actiniaria,
which have the appearance of acrorhagi (see above), but contain no nematocysts. They
may be perforated.

Reticular. See “ Alveolar”

Scapus. — See “ Capitulum."

Spirocysts. — Thin-walled nematocysts containing a spirally arranged thread which
is clearly visible through the wall of the capsule.

Verrucae. — Thick-walled, slightly hollow, muscular outgrowths of the body- wall,
which often have the power of attaching themselves to foreign bodies.

Vesicles. — Thin-walled, bladder-like, often very hollow outgrowths of the
body-wall. Not muscular (always ?).

IV.— DESCRIPTIONS OF SPECIES.

SUB-ORDER ACTINIARIA.

I. ACTINIINA, M. Edw.

Family ILYANTHIDAE, Gosse.

Actiniina with the aboral extremity of the body rounded, so that there is no
pedal disc. No acontia.

I have ventured to enter into a certain amount of discussion of this family and
some of the genera which it contains, and to go over again ground which has been
partly covered by various authors, in the hope of simplifying matters a little.

I have retained the original family Ilyanthidae of Gosse (excepting of course
Edwardsia and Cerianthus), in preference to the three families Halcampidae, Halcampo-
morphidae, and Ilyanthidae, as it seems to me unwise, considering the very close
general relationships of the genera Halcampa, Halcampoides , and Ilyanthus, to place
them in three different families on account of the differences in their sphincter-
muscles. For although the character of the sphincter is as a rule a very reliable one,
exceptions must be allowed. In the family Sagartiidae, although its members have
almost universally a mesogloeal sphincter, the genus Aiptasia includes forms with no
sphincter, or a weak endodermal one. Again, most of the Boloceridae have a diffuse
endodermal sphincter, but in Bolocera pollens it is strongly circumscribed. So that
although Halcampa has a feeble mesogloeal sphincter, Halcampoides a diffuse, and
Ilyanthus a feebly circumscribed endodermal one, it seems to me advisable, on account
of their close general affinities, to include them all in one family, retaining, however,
Carlgren’s sub-families Halcampinae and Halcampomorphinae.

A considerable amount of discussion has taken place from time to time in
connection with the genera Halcampa and Halcampoides, but I think the question may
now be satisfactorily settled. In 1893 Carlgren (8) described two species of

ACTINIARIA— STEPHENSON.

9

Halcampa, H. duodecimcirrata and H. arctica, and showed that they possessed
mesogloeal sphincters. He expressed the opinion that II. chrysant helium, the type of
the genus, would also be found to possess such a sphincter, and that therefore the
genus Halcampa must l^e restricted to forms with a mesogloeal circular muscle. For
forms with an endodermal sphincter he proposed the name Ilalcamptomorphe, with
H. clavus Hertw. as its type. He further divided the family Ilyanthidae into two
sub-families, according to the character of the sphincter-muscle — Halcampinae and
Haleampomorphinae (8, pp. 36-38).

In 1896 Kwietniewski pointed out that according to the observations of Haddon
and Faurot, II. chrysanthellum, the type, had an endodermal sphincter : that therefore
Carlgren’s Halcamptomorphe was a synonym of Ilalcampa Gosse : and that a new
name was needed for forms with a mesogloeal sphincter, for which he proposed
Halianthus, and for the sub-family Halianthinae. Haddon later (28, pp. 412, 413)
confirmed his observation regarding II. chrysanthellum.

In 1900, however, Carlgren (ll) summarised this, and stated that he had been
able to examine sections of H. chrysanthellum, and had found that his original surmise
was correct, and that its sphincter was indeed mesogloeal. It had been overlooked, as
it was rather small.

I have made preparations of seven undoubted specimens of II. chrysanthellum
from Plymouth, and can confirm Carlgren’s remarks ; there is certainly, in my opinion,
a very distinct mesogloeal muscle. (See PL VI, fig. 16, which shows its position.)

In 1896 Appellof (4) showed that the type of the genus Halcampomorphe
(II. clavus Hertw.) was really a Halcampoides Dan. (a fact which it was quite impossible
for Carlgren to have known when he instituted Halcampomorphe), and that therefore
Halcampomorphe must lapse ; and he re-named the sub-family Halcampoidinae. But
in 1900 (11) Carlgren pointed out that even if the two genera were synonymous
there was no need to re-name the sub-family, since the name Haleampomorphinae was
not misleading, and was as good as Halcampoidinae, to which it had priority.

Kwietniewski proposed to separate forms with twelve tentacles from forms
with more than twelve, and for forms with a mesogloeal sphincter and more than
twelve tentacles he proposed the genus Halianthella. But, as McMurrich remarked
in 1905 (40), the number of tentacles alone is hardly, a valid generic character; and
it seems to me that there is absolutely no need to separate Haliahthella from
Halianthus. Acting on this principle, I have included my new Halcampoides aspera
in the genus Halcampoides , although the forms hitherto referred to that genus have
only twelve tentacles. There are no other differences to support this difference in the
number of tentacles as a generic character.

There remains the question of Halcampella Andres. Hertwig’s II. maxima is, in
my opinion, a Halcampoides. The anatomy of the type, II. endromitata, is not yet
known, but whether its sphincter proves to be mesogloeal or endodermal, it must
inevitably be included in either Ilalcampa or Halcampoides, since the fact that it has

c

VOL. V.

10

“TERRA NOVA” EXPEDITION.

more than twelve tentacles is not enough to separate it — that is to say, unless it
should prove to possess other unsuspected characters which will quite remove it from
the forms under discussion.

To summarise this, we have now reduced the number of genera to two, Halcainpa
and Halcampoides, and their synonymy is as follows : —

1. Sub-family Halcampnnae , Carlgren.

=3 Halianthinae, Kwietniewski.

Ilyanthidae with a mesogloeal sphincter-muscle.

Genus Halcainpa, Gosse.

3= Halianthus, Kwietniewski.

= Halianthella, Kwietniewski.

Halcampinae with eight, twelve, or more tentacles; not more than twelve perfect
mesenteries ; secondary imperfect mesenteries may be present. No conchula.

2. Sub-family Ilalcampomorphinae, Carlgren.

= Halcampoidinae, Appellof.

= Halcampinae, Kwietniewski.

Ilyanthidae with the sphincter either absent, or if present endodermal.

Genus Halcampoides , Danielssen.

=3 Iialcampomorphe , Carlgren, 1893.

=3 Halcainpa, Kwietniewski.

—— Iialcampella, Hertwig.

Halcampomorphinae with twelve or more tentacles ; twelve perfect mesenteries ;
secondary imperfect mesenteries may or may not be present. No conchula.
Sphincter absent, or if present diffuse, not circumscribed.

3. Genus Halcampella, Andres.

Probably a synonym of Halcainpa or Halcampoides.

HALCAMPOIDES, Danielssen..

1. Halcampoides aspera, n. sp. (PI. I, fig. 1 ; PI. Ill, figs. 1, 2, 3, 4 ; PL VI, fig. 13.)

One specimen. Station 314, 5 miles N. of Inaccessible Island, McMurdo Sound,
222-241 fathoms, Jan. 23, 1911. Agassiz trawl. Bottom— mud.

Measurements. — Total length (contracted), 2 • 5 cm. Diameter of crown, * 9 cm. ;
of column, 1 * 2 cm. ; of physa, 1 • 4 cm.

The single specimen of this species was strongly contracted, and its internal
preservation extremely poor. It was thus too brittle to dissect, and I can describe its
anatomy only from sections. The physa was damaged and imperfect.

External characters. — The body is longer than wide, and is broader at the aboral
end than at the other. The aboral end is rounded, forming a physa which is not sharply
marked off from the main body or scapus. I can give no details about it, as it is much

ACTIETARIA — STEPHENSON.

11

torn. The upper part of the column, or capitulum, is completely introverted, so that
it is not visible till the animal is opened : in structure its wall is not different from that
of the scapus. The capitulum is traversed by eight distinct but rather irregular
longitudinal ridges, each of which divides into two at the margin of the scapus. The
sixteen ridges run down the scapus, becoming less distinct at its aboral end, but they
are extremely irregular and much interrupted by deep transverse wrinkles of the body-
wall, due to contraction : they die out by the time the physa is reached. All are not
equally distinctly developed. The body-surface has no verrucae, and there are no
cinclides. The wall is hard and stiff, quite thin in the deepest furrows, but thick on the
ridges. The ectoderm has been removed from a large portion of the surface, but where
it is present it is rough, brown, and corrugated, and is mixed up to some extent with
foreign bodies : a little cuticle is present in small patches. The surface of the capitulum
does not differ from that of the scapus. For the general appearance, see PI. I, fig. 1.

The oral disc and tentacles are completely hidden by the contracted state of the
capitulum, and so far is this latter introverted that they lie about half-way down the
length of the body. The disc is narrow and slightly developed. The tentacles are
small, short, stiff, thick- walled, fairly slender, and bluntly pointed. It is impossible to
enumerate them exactly without damaging the specimen, but there appear to be about
thirty. From a study of sections, and from the number of mesenteries present, there
are probably thirty-two.

Structure, (i) Mesenteries. — The total number of mesenteries is thirty-two. Of these,
thirteen are large and perfect, and bear muscle-pennons : twelve of these form the usual first
cycle of six pairs, including two pairs of directives. The thirteenth occurs in the left
ventro-lateral exocoel, and its pennon is ventrally directed. It seems to me that since
it is an odd mesentery, occurring on one side of the animal only (so that the animal is
not bilaterally symmetrical), it is probably accidental, and not an important development.
Gonads and mesenterial filaments are confined to these perfect mesenteries. The
specimen is a female. The remaining nineteen mesenteries are small, and form the second
cycle. They have neither muscle-pennons, gonads, nor filaments. Their exact arrange-
ment between the pairs of perfect mesenteries I cannot give, on account of the bad state
of preservation. The primary mesenteries join the oesophagus at various different
levels. The muscle-pennons end at different levels in different mesenteries, some of
them being very short indeed. In the highest section I have, only one is present : at a
rather lower level, thirteen are visible : still lower, five : then three, and finally none.
The mesenteries themselves are thin, the perfect ones bearing well-developed parietal
muscles, as well as the large circumscribed pennons. These pennons vary as regards
their exact shape and size in different mesenteries : they are sometimes round, some-
times a long irregular oblong, but always circumscribed : and in the case of one pair of
directives (the ventral) the inequality in their size is very marked, one pennon being
considerably larger than the other. The structure of a pennon is diagrainmatically
represented in PI. Ill, fig. 3. In this particular mesentery it will be seen that there

c 2

12

“TERRA NOVA” EXPEDITION.

are four stout main lamellae radiating from the centre, which have secondary and tertiary
branches, giving them an arborescent appearance ; and a few* smaller processes between
the bases of these primary lamellae, The same general plan, with small variations m
detail, is universal. The parietal muscle is strongly developed (PI. Ill, fig. 2), and
bears numerous fine processes on either side. These are simple or slightly branched
where the mesentery leaves the body-wall, but many of them become higher and more
branched at the outer edge of the muscle. In number they vary a good deal in
individual mesenteries. Their exact number in the case figured may be seen from the
figure. It will also be noticed that the processes of the parietal muscle are much finer
than those of the pennon, even though fig. 2 was drawn under higher magnification
than fig. 3. The imperfect mesenteries bear parietal muscles (which resemble those of
the perfect mesenteries, but are larger) and nothing else. The mesenterial filaments
bear well-developed cihated lobes : where a filament-trefoil is accurately cut across, it
presents a characteristic shape, the glandular lobe having a very stout, pillar-like stem,
nearly as wide as its own diameter, the ciliated lobes being pushed rather far back,
and separated by this wide stem. The nuclei of the ciliated lobes are mainly concen-
trated in a band in their inner parts. The glandular lobe contains fairly large, irregular,
clearly granular gland-cells, which become, in the larger filaments, oval and compact
like those of the actinopharynx-ectoderm. Rather short, narrow, but not quite needle-
like thick walled cnidae, blunt ended, and with an unusual vacuolated appearance, also
occur in the glandular parts of the filaments.

(ii) Sphincter. — This is endodermal, weak and attenuated. Its processes (PI. Ill,
fig. l) are hardly better developed than those of the general body-wall circular muscle.

(iii) Endoderm. — The endoderm of the coelcnteron is not particularly interesting :
it is of a loose, spongy nature with a thicker crust.

(iv) Mesoyloea. — This is almost homogeneous as regards the matrix, but more or
less fibrous in places. It contains numerous cells showing wide variation in shape and
size. Some are clear, some finely mealy or almost granular ; some are elongate and
bipolar, some unipolar, some rounded and without the processes of the bipolar and
unipolar ones. In some places there are clear spaces containing nuclei, and there are
also large granular patches of indefinite shape and variable size : these latter are most
numerous in the body-wall, whereas there are more of the cells in the mesenteries,
actinopharynx, and ectoderm.

(v) Ectoderm of body-wcdl. — High. Contains many large, loose, irregular granular
gland-cells, with large scattered granules, and medium sized thick walled, blunt-ended
cnidae, not needle-like, but fairly narrow. The cnidae are few' in number except in the
upper part of the capitulum, where they are numerous. Many of the supporting cells
contain brown pigment-granules. Foreign bodies, such as sand-grains and diatoms, are
present, more or less mixed up with the ectoderm, and here and there are patches of
cuticle.

(vi) Actinopharynx. — Much folded. The ectoderm with very numerous large, oval,

ACTTNTABIA — STEPHENSON.

13

compact gland-cells with large granules, which are unlike those of any other part of
the body, except the larger mesenterial filaments. Unfortunately, the preservation of
the material allowed no dissection with regard to siphonoglyphes, so I can give no
data : probably none.

(vii) Tentacles. — The proportions of the various layers of tissue may be seen from
PI. Ill, fig. 4. The longitudinal muscles are ectodermal, supported on very short,
rather club-like, simple or slightly branched processes. Endoderm forming a fine and
rather regular network. Nerve-layer of ectoderm broad and very well defined.
Ectoderm with a dense battery of long, large, almost cylindrical spirocysts in its outer
part, and a fairly well defined zone of irregularly shaped, finely granular gland-cells in
its inner part ; these gland-cells are not like those of the actinopharynx, etc., being
smaller and finer. Between the gland-cell and spirocyst zones the nuclei of the
supporting cells are more numerous than elsewhere.

A point of interest in connection with this species is the presence of a parasite
embedded in the mesenteries. The parasite is probably a Crustacean, but more than
this I cannot certainly state. It seems to be surrounded by a sort of pocket, formed of
the tissue of the anemone, probably a defensive measure on the part of the latter.
Carlgren (15) has recorded (with more detail) a similar instance in the case of
Actinostola intermedia.

I have named this species aspera on account of its rough external appearance and
texture. It differs from II. claims and II. kerguelensis in the possession of more than
twelve tentacles and mesenteries. From II. maxima it is to be distinguished by its
general build, and by the difference in the branching of the muscle-processes of the
pennon. II. minuta, Wass., as Carlgren has pointed out (12), is a llaloclava.
Ilalcampella ostrooumowi Wyragevitsch is a minute species living on mussel-shells, and
though I have not seen an account of its anatomy, it is certainly distinct, by external
characters and habitat, from II. aspera.

In PI. VI, fig. 13, I have shown, very greatly enlarged, the tip of one of the
processes of the mesenterial pennon. This shows clearly the typical arrangement of
muscle-fibres (here seen in transverse section, of course) in the endoderm, supported
by a finger of mesogloea. It also serves to emphasise the difference in appearance
between an ordinary process and a process of Cymbactis selaginella, which is shown
in fig. 14.

Family ENDOCOELACTIDAE, Carlgren.

Actiniina with thin or thick, sometimes cartilaginous body-wall, without sphincter
or fosse, and usually with spirocysts in the ectoderm of the body-wall and actino-
pharynx. Arrangement of mesenteries quite different from the normal Actiniarian
type, owing to the development of the second and third orders of mesenteries in the
endocoels. In consequence of this the arrangement of tentacles is very different from
the normal type. (Among others, ten tentacles of the first cycle border immediately

14

“TERRA NOVA” EXPEDITION.

on those of the second.) Sex-organs present on all the stronger mesenteries, including
the directive mesenteries. True basilar muscles absent, base present.

The above definition is that given by Carlgren (12), with a few modifications. I
have written “ Actiniina ” instead of “ Athenaria,” since I agree with McMurrich in
keeping to the two old classes Actiniina and Stichodactylina rather than Carlgren’s
Protantheae and Nynantheae ; for a discussion on this subject, see McMurrich (40). The
other alterations are those necessitated by the facts that the new Ilalcurias, which is
described below, possesses a thin body-wall, which has no spirocysts in its ectoderm.

HALCURIAS, McMurrich.

Endocoelactis, Carlgren.

Endocoelactidae with twenty large perfect mesenteries which extend from base to
oral disc, and bear strong muscle-pennons. Mesenteries in four cycles, those of the
third and fourth cycles being small and confined to the uppermost part of the body.
Individual mesenteries of one pair equally developed. Tentacles not thickened outside.
Spirocysts may or may not be present in the ectoderm of the body-wall and
actinopharynx.

2. Ilalcurias endocoelactis, n. sp. (PI. I, fig. 4; PI. II, figs. 8, 9, 10, 11, 16 ;
PI. Ill, figs. 22, 23, 25, 26.)

One specimen. Station 96, seven miles E. of North Cape, New Zealand, 70 fathoms,
Aug. 3, 1911. Agassiz trawl. Bottom — sand and rock.

Measurements. — Height of column, 4 • 8 cm. Least diameter of column, ‘7 cm.
Greatest diameter of column, 1 ■ 5 cm. Diameter of pedal disc, 1 cm. ; of tentacular
crown, 1 • 5 cm.

External characters. — The general form may be seen from PI. I, fig. 4. Pedal
disc small, slightly exceeding the lowest part of the column, nearly circular,
apparently incapable of strong adhesion ; rather thin, wrinkled, much exceeded by the
uppermost part of the column in width. Column several times higher than wide,
rather worm-like, of different diameter at different heights. It is narrowest just above
the limbus, and increases very slightly in width from there upwards, till at about two-
thirds of the whole height it swells out rather suddenly into a bulbous portion, below
the margin. This upper portion has a very thin collapsible wall, which is semi-
transparent : below, the wall is rather thicker. Margin tentaculate. There is a strong-
fold just outside the bases of the tentacles, which in the figure has rather the
appearance of a narrow collar ; but it is only an accidental contraction-fold, and
does not run round to the other side of the anemone. Surface soft and smooth,
without verrucae, etc., but divided by many fine transverse and longitudinal furrows
into small inconspicuous papilla-like areas, more prominent in some places than in
others, and probably due to contraction. The oral disc is thin and smooth, and does

ACTINIARIA— STEPHENSON.

15

not exceed the column : radial grooves are present running from the bases of the inner
tentacles towards the mouth, but they do not reach it. Mouth raised on a cone, with a
prominent lip. The tentacles are short, stout, blunt and conical, rather thin walled,
but erect and self-supporting, not flaccid. Their surface is transversely striated. In
size they are irregular, the inner being on the whole the largest. Unfortunately, I can
give only rather imperfect data as to their exact arrangement and relations to the
mesenteries, as I had sectionised more than half of the upper part of the animal before
I had identified it, and consequently was unaware of the interest involved. There
are sixty tentacles, and eighteen of these are more internal than the others ; of the
remainder, thirtv-nine communicate with the margin, and three do not. But since the
arrangement of mesenteries is so irregular, that of all but the first cycle of eighteen
must necessarily be very confused.

Structure, (i) Mesenteries. — There are twenty large mesenteries, which extend
from base to oral disc, and which all bear strong muscle-pennons and are all fertile.
These form the first two cycles, 6 p. + 4 p., comprising eight pairs and two pairs of
normal directives, and in this respect the animal agrees with the two species of
Halcurias hitherto described. But the arrangement of the third and fourth cycles is
quite extraordinary. The mesenteries composing these cycles are very small and thin,
and are attached to the oral disc on its underside. Those of the third cycle run a little
way down the actinopharynx and body- wall, those of the fourth cycle do not reach the
actinopharynx at all. Their arrangement is exceedingly irregular, not being the same
in any two endocoels. But two characters they possess in common — they are all
developed in the endocoels of the larger mesenteries : and in every case they have the
character of directives. (They bear neither gonads nor filaments, and the longitudinal
musculature is reduced to a low fringe of processes on one face of each mesentery.)
That is to say, the longitudinal muscles of the two mesenteries of each pair face away
from each other. In all members of the family previously studied these small
mesenteries form normal pairs, with their muscles facing each other, and not directives
as in this case. The exact arrangement of the mesenteries in one half of the animal is
shown in PI. Ill, fig. 26. On the other side, two endocoels were lost in longitudinal
sections of the margin before I knew the animal’s genus, but in the other two endocoels
the arrangement was as follows : —

(1) One errdocoel contained a single pair of tertiary and no quarternary
mesenteries.

(2) The other contained one pair of tertiary besides one pair of quarternary
mesenteries.

It will be further noted from PL III, fig. 26, that in one case a pair of small
fourth-cycle mesenteries (without musculature of any sort) occurs between two
mesenteries of a third-cycle pair : and that in another endocoel the small fourth-cycle
pair is attached to the actinopharynx, but does not reach the disc or body-wall.

The state of affairs here described is extremely interesting, since it confirms and

16

TERRA NOVA” EXPEDITION.

adds to some of the ideas expressed by Carlgren (12) on the subject. Carlgren
supposes that probably the first six pairs of mesenteries are formed in the usual way
in this genus : that the second cycle to appear consists of four pairs having the
character of directives, which form with the six original pairs ten apparently normal
pairs, in whose endocoels the subsequent mesenteries grow. This is confirmed by the
fact that in II. endocoelactis, the species under discussion, the tertiary and quarternary
pairs of mesenteries, as icell as the secondary ones, are formed as directives.

The specimen is a female. The large mesenteries themselves are very thin, mere
films, but they bear very powerful muscle-pennons, and have enormous oral stomata in
an unusal position, well below the mouth. Where each large mesentery leaves the
body-wall it bears on each side an about equal development of rather short, fine,
simple or slightly branched muscle-processes, so that it has the appearance of a parietal
muscle (see PL III, fig. 23). These processes are very short where the mesentery joins
the body-wall, becoming higher about the middle of the muscle, and then dying off
again. Further out, each large mesentery bears a large pennon ; this is circumscribed,
but varies in shape according to the level at which sections are cut. In the region
below the level of the actinopharynx, where the gonads and filaments are present, it
has the form shown in PI. Ill, fig. 22. The individual pennons vary of course to some
extent in shape and size. The muscle-processes are fine and numerous (for exact
number see plate), and some are simple, some a little branched, some much branched.
The simple ones are as long as the others, and the compound ones often branch very
near their origin, the various branches running nearly parallel with each other, so that
there seem to be more simple processes than there really are. Above the level of the
enterostome, on a level with the upper part of the actinopharynx, the pennons are
stretched out straight, and cover most of the surface of the mesentery, starting
immediately outside the actinopharynx. Their form here is elongate oblong, almost
diffuse, and their processes are shorter than lower down. Their form changes gradually
to the typical circumscribed outline as they reach lower levels. The mesenterial
filaments are well developed. Their glandular portions contain gland-cells of two kinds
— large granular ones of various shapes and sizes, and numerous small ones such as are
characteristic of other parts of the animal. The endoderm of the body-wall contains
these small gland-cells also, but they are absent from that of the greater part of the
mesentery, which contains, however, numerous large dimly visible non-granular gland-
cells, such as are found only here and there in the body-wall endoderm.

(ii) Sphincter. — Absent.

(iii) Mesogloea. — In longitudinal sections passing through the margin this has the
appearance of a rough network with large meshes, being very fibrous. In the clearer
parts lie numerous cells with processes, and large oval or rounded nuclei which (varying
in size in different cases) are usually deeply stained with haematoxylin. A typical
bipolar cell from this region is shown in PI. II, fig. 16. In transverse sections lower
down the fibrous nature of the matrix is less conspicuous, but the cells are extremely

ACTINTARIA— STEPHENSON.

17

interesting. Near the enclodermal side of the mesogloea, which is more deeply stained
with eosin than the ectodermal half, are very many small cells possessing apparently a
large nucleus and very little protoplasm (PL II, fig. 11). These often lie in a clear
cavity or capsule, distinctly shown in the figure, which reminds one of what Carlgren
has described in Bolocera longicornis (for more detail see description of that species
below), and which is probably a post-mortem development. In the central part of the
mesogloea, and the part near the ectoderm, cells such as that shown in PI. II, fig. 8, are
very frequent. They consist of a large, clear, often mealy-looking nucleus, surrounded
by a delicate vacuolated network of protoplasm, of different size and form in different
cases. Thirdly, immediately below the ectoderm we find cells (PI. II, figs. 9, 10) of a
rounded form and blaclder-like appearance, with indications of network within. It
seems likely that the last two sorts of cell described represent different phases of the
same thing, particularly if, as is probable, they are amoeboid nutritive cells (see p. 2).
Possibly this cell-plexus of the mesogloea of Antinians represents an early stage in the
evolution of what has become connective tissue in higher animals, though it cannot be
called connective tissue as it is.

(iv) Ectoderm of body -wall. — This is high and regular, and full of two sorts of
gland-cells — the small sort found in the tentacles and elsewhere, and also large irregular
ones with dimly defined, soft granules which do not absorb haematoxylin. I believe I
can state with certainty that no spirocysts are present. I have taken the greatest
care to confirm this statement by making four sets of preparations of different parts
of the body, and by using four different methods of staining. I have stained with
haematoxylin alone, with eosin alone, with the two together, and with Picrosaurefuchsin.
The sections seem to be good ones, and I have searched them through and through for
spirocysts, but there is no trace of one. Further, in sections which pass through the
tentacles and the body-margin, the spirocysts are clearly seen crowding the tentacle-
ectoderm, but die out at the edge of the body. So that I believe it to be a fair
statement that spirocysts are absent from the ectoderm of the body-wall. (This applies
also to that of the actinopharynx. ) I have taken particular care in this case, because
spirocysts have been demonstrated in the body- ectoderm of all members of the family
hitherto described, and Carlgren has used this in characterising the family
Endocoelactidae. It seems to me that this case shows that although spirocysts when
present indicate primitiveness, their presence cannot be used as a classificatory
character, since one member of a family (and genus) may retain them while another
has lost them. With regard to the position of my species as a Halcurias there can be
no doubt. It shows all the usual Halcurias characters except this one. It has ten pairs
of large mesenteries, including two pairs of directives, two cycles of small ones developed
in the endocoels and restricted to the upper part of the body : it has no sphincter and
no basilar muscles, and has all the facies of a Halcurias. Its body-form is elongate,
but I consider this of no importance in classification. I have elaborated this remark in
a general note earlier in the paper (see p. 4).

VOL, V.

D

18

TERRA NOVA” EXPEDITION.

As far as I can tell, the bases of the ectoderm cells are thickened as Carlgren (12)
describes for Porponia, etc. At any rate, there is no indication of true ectodermal
muscles in the body-wall, though in the pedal disc there are clearly visiblt fine short
processes running from the mesogloea into the ectoderm, which appear to be muscular.
It seems to me that the presence of ectodermal longitudinal muscle in the body -wall,
though a primitive trait, cannot be used for classification any more than the presence
of spirocysts, since McMurrich (38) has shown, and later (40) confirmed, that Bolocera
brevicornis possesses this musculature, whereas other Bolocera do not. So that probably
these two characters are not of more than specific value.

(v) Actinopharynx. — Well developed and much folded, running well down the
coelenteron. One siphonoglyphe. The ectoderm is very regular and well developed,
and contains a very large number of closely packed gland-cells, so that it is nearly black.
There are large very granular ones near the surface, and small ones (many times
smaller than the granular ones) concentrated chiefly in the central zone of the layer.
(See Tentacles.)

(vi) Tentacles. — Endoderm thin, mesogloea thicker, ectoderm thickest. The
longitudinal musculature is ectodermal, and is supported on short, rather stout, simple
and slightly branched processes (PI. Ill, fig. 25), which vary in height in different
parts of the same tentacle, but are usually less than one-third as high as the ectoderm.
Nerve-layer distinct, broad, fine. The outer part of the ectoderm contains a battery
of large rather stout spirocysts. The central part is densely crowded with the nuclei of
the supporting cells, and with numerous small elongate regular cells which stain with
haematoxylin less deeply than nuclei, and often seem faintly granular and mealy, but very
finely so. They seem to be small gland-cells, but I will not emphasise that ; I have
.referred to them in other parts of the body as “ small gland-cells such as are found in
the tentacles,” but their nature remains rather uncertain. Cells very like them occur
in the mesogloea of the tentacles.

This species is clearly distinguished from IT. pilatus and II. carlgreni anatomically
by the arrangement of the third and fourth cycle mesenteries and by the absence of
spirocysts in the body-wall ectoderm. Its elongate form and short tentacles may be
of specific value, but unless the animal is studied alive this is uncertain (see p. 4).

Family PTYCHODACTIDAE, Appellof.

Actiniina with weakly-developed longitudinal muscle- and nerve-layer in body-
wall. No ciliated lobes to mesenterial filaments. No sphincter or a weak one. Sex-
organs on proximal parts of mesenteries, glandular mesenterial filaments on distal
parts. Actinopharynx variable, arrangements for circulation always strong. Character
of incomplete mesenteries peculiar ; they carry, beyond the glandular parts of the
mesenterial filaments, curious structures giving the appearance of a half-funnel. Base
present, sometimes feebly developed, basilar muscles absent. The above are the

ACTINIAR I A— STEPHENSON.

19

characters given by Carlgren (14) for this family. I have written “ Actiniina” instead
of “ Protactinina,” and added “ basilar muscles absent.”

DACTYLANTHUS, Carlgren.

Cystiactis, pars, Clubb.

Ptychodactidae with the body provided writh many conical or cylindrical out-
growths, one row to correspond to each mesenterial chamber. Outgrowths of body-wall
are like the tentacles, and have spirocysts, which are absent from the rest of the body-
wall. Xo sharp distinction between foot-disc and body-wall, or between body -wall and
oral disc. No sphincter or a weak diffuse one. Actinopharynx strong, broad, two well-
developed sipkonoglyphes with long ends ; with pocket-like outgrowths at insertions
of perfect mesenteries. Few mesenteries, two cycles, all fertile. All or only the
primaries perfect. Mesenterial muscles w'eak, filaments meandering. Imperfect
mesenteries, with peculiar half-funnel -like arrangements along free edges beyond
filaments. Proximal halves of the mesenteries mostly fused at their edges (constant
for genus ?).

The above list of characters is given for the genus by Carlgren (14).

3. Dactylanthus antarcticus, Clubb. (PL I, fig. 10.)

Cystiactis antarcticus, J. A. Clubb. “Actiniae.” Reports, Nat. Antarctic Expedition, 1901-4
(“ Discovery ”) ; Nat. Hist., vol. IV. 1908.

One specimen. Station 338, 77° 13' S. ; 164° 18' E., entrance to McMurdo
Sound, 207 fathoms, Jan. 23, 1912. Agassiz trawl. Bottom — mud.

Measurements. — Diameter of tentacular crown, 1 ’ 9 cm. ; of pedal disc, 1 • 9 cm. ;
of column, 2 • 8 cm. Height of column, 3 • 2 cm. Length of one tentacle, about * 2 cm.

This species has been completely described by Carlgren (14) and Clubb (18), so
I will merely say that my specimen agrees with Carlgren’s description in all respects.
Tentacles 12 + 12 = 24. Twenty-four rows of body-tubercles, twelve complete, twelve
running from base only part way up column. Mesenteries 6 p. +6 p. = 12 p., first cycle
perfect, all fertile. Fusion of mesenteries at base, pouches of actinopharynx, siphono-
glyphes, curious free edges of imperfect mesenteries, etc., all clearly visible. Appearance
shown on PL I, fig. 10.

Family BOLOCERIDAE, McMurrich.

Actiniina with well -developed base and basilar muscles. Sphincter endodermal,
variously developed, or absent ; if present, may be diffuse or circumscribed. Usually
numerous mesenteries perfect. Tentacles attached to the oral disc by a very short
tkin-walled neck, and usually provided each with a sphincter muscle above the neck,
by means of which the neck may be torn, in which case the tentacle is lost.

The genera related to Bolocera have been referred to various families at different
times. For instance, Carlgren included Bolocera in the Antheadae in 1891 and 1893,

d 2

20

“TERRA NOVA” EXPEDITION.

and Gosse originally placed it in the Bunodidae. There can be no doubt, however,
that a family Boloceridae is advisable. I consider that it should contain the genera
Bolocera , Boloceropsis, and probably Polystomidium and Polyopis. I cannot agree
with Carlgren in placing Boloceroides in the Gonactiniidae, because it seems to me
related as much to Bolocera as to Gonactinia and Protantliea, and I do not regard the
fact that it possesses longitudinal ectodermal body-wall musculature as of great
classificatory importance. But it is doubtful whether it can be included in the
Boloceridae, and its position needs reconsideration. Boloceropsis has no tentacular
sphincter, but the tentacles are typically Boloceroid in form, and have a short thin-
walled neck attaching them to the oral disc.

BOLOCERA, Gosse.

Boloceridae with well-developed endodermal sphincter-muscle, which may be
diffuse or circumscribed. Tentacles with well-developed sphincters.

4. Bolocera longicornis, Carlgren. (PI. I, fig. 5 ; PI. Ill, figs. 24, 27.)

Bolocera longicornis. (“ Beitrage zur Kenntniss der Actinien-Gattung Bolocera, Gosse.” Ofv.

K. Vet.-Akad. Forh. No. 4, 1891.)

Five specimens. Station 38, off the Falkland Islands, 52° 23' S. ; 63° 50' W.,
125 fathoms, April 13, 1913. Agassiz trawl.

Measurements (of the best specimen, a small one). — Diameter of oral disc and ten-
tacles, 5 • 5 cm. ; of column, 2 cm. ; of pedal disc, 2 cm. Height of column, 2 cm.
Length of a large tentacle, 2 cm.

In 1893 Carlgren (8, pp. 50-57) gave an excellent description of this species. I
may thus merely add a few points as to its variation.

Firstly, I have been able to examine numerous specimens of this species (apart
from the present collection) dredged off the West of Ireland. These agree perfectly with
Carlgren’s description on the whole ; I have not yet made a complete study of them,
but I have examined two large specimens in detail. In one of these the last cycle of
mesenteries is imperfect, in the other perfect ; the second being considerably larger
than the first. One of these specimens was unusually well preserved, and bore gonads
on the mesenteries of all cycles, those of the primary mesenteries being, however, small,
fewr, and scattered. In the other, the last cycle wras sterile, and one mesentery at least
of the first cycle fertile. I sectionised the sphincter of one of these. It is perfectly
regular and diffuse as shown in Carlgren’s figure (Taf. VII, fig. 6), but the processes are
coarser and more branched. Wassilieff (66) has noted this same thing with regard to
Japanese specimens of B. longicornis, and the sphincter of my example is more like his
figure than Carlgren’s, which must have been drawn from a rather unusually delicate
individual. Another interesting point is that in my specimen the gonads contain
numerous small zooxanthellae, which lie round the testes. I have observed this pheno-
menon in several other species, and have referred to it in more detail earlier (see p. 3).

ACTINIAE I A— STEPHENSON.

21

Secondly, as regards the “ Terra Nova ” specimens of B. longicorni.s : — These are
from a widely different locality from that of any specimens of the species hitherto
recorded, and differ from Northern specimens in a few minor directions, hut I do not think
they can he regarded as a separate species. The general form, the arrangement of tentacles
and mesenteries, the number of perfect mesenteries, and the distribution of gonads are
the same in both. In the “ Terra Nova ” specimens, of which all but one are much
damaged and battered, and have lost most of their tentacles, the tentacles differ from
those of Northern examples by being flaccid and thin-walled, and not fluted. But since
most of their ectoderm has been scraped off’ by rough treatment, this is perhaps only to
be expected. I have drawn the best specimen as it appeared in spirit (PL I, fig. 5), but
although the liquid supports the tentacles, their ragged appearance is obvious. The
figure shows a curious abnormality of this particular individual — a small bulge 'of the
body- wall on the right, just below the margin. This is in reality a pouch pushed out
from the coelenteron, and possibly the beginning of a bud. The sphincter of one of
these specimens is shown in detail in PL III, fig. 27, which also shows half the
tentacular sphincter. This figure agrees with the sphincter of my Northern specimen of
B. longicornis and with Wassilieff s figure better than with Carlgren’s, and I give it to
show the possible variation from the originally described specimen. The sphincter
reminds one strongly of that of B. brevicornis, as figured by McMurrich (38, PL XXIII,
fig. 31).

Carlgren gave no figure of the muscle-pennon of the mesentery of B. longicornis,
so I have drawn one (PL III, fig. 24), and will add a few details. In my opinion the form
of the muscle-pennon, sphincter, and longitudinal muscle of the tentacles of any
anemone are the most constant specific characters, in many cases the only ones, and
should therefore always be figured when a new species is described. The whole muscular
surface of the mesentery is occupied by longitudinal muscle on one face : this is
supported on very short processes where it leaves the body-wall, and these increase very
gradually in size, forming a low diffuse pennon. The manner in which this pennon
ends at the edge furthest from the body-wall varies in individual mesenteries.
Typically it ends quite abruptly, with little or no tapering, but in some cases it
tapers off before ending, though not nearly so gradually as it does toward the body-
wall. Sometimes the individual mesenteries of a pair are slightly unequal in length,
and on one the pennon may end abruptly, while on the other it is tapering. The
muscle-processes are rather similar to those of the sphincter, and are mostly a good deal
branched. The parieto-basilar muscle is fairly developed, though feeble, and it varies
greatly in different mesenteries. Its processes are short and simple or only very slightly
1 (ranched ; in some cases it has no clearly marked edge ; in others the edge forms a
slight prominence (as in Pl. III, fig. 24) ; again, it may have a slight but distinct free
fold, or, lastly, often the free fold has joined up in process of growth with the body of
the mesentery, and so encloses a cavity, rather as in B. pannosa.

In these specimens, as well as in the Northern one which I have mentioned above,

22

“TERRA NOVA” EXPEDITION.

the gonads are crowded with zooxanthellae, which lie round the testes — the specimen
investigated being a male.

Lastly, a few points about the cells of the mesogloea. Wassilieff (66) stated with
regard to Japanese specimens of B. longicomis that the cells in the tentacle-mesogloea
were less numerous than in the specimen figured by Carlgren (1893, Taf. VI, figs. 4, 5).
In my preparations of both Northern and Southern forms of this species numerous
cells are clearly visible in the mesogloea, but they are as plentiful in the other parts of
the body as in the tentacles, and are similar in character to the mesogloeal cells of
many other species. In my Northern specimen they are less numerous than in the
Southern one, but in neither are they as crowded as in Carlgren’s figure ; which shows
that their number varies a good deal. In my specimens these cells are of almost every
conceivable shape, and their size varies greatly. Some are rounded, some rounded
with one or two short projections, some pyriform, some spindle-shaped with a process at
each end, some of irregular form with several processes ; often they are strung together
in chains with the tips of their processes touching each other, and, indeed, all seem more
or less connected by fine strands of protoplasm. A few examples are shown on PI. Ill,
figs. 11, 12, 13, 14, 15 and 16. All those figured are from the mesogloea of a tentacle,
but are quite similar to those of the body-wall. Carlgren describes his cells as lying
in capsules (Kapseln). In my specimens many of the cells lie in clear spaces, which
suggest the capsules shown in his figures. I have indicated the outlines of these spaces
by dotted lines in the figures on PI. Ill, but their actual appearance may be better
understood by reference to figures of cells of other species (PI. II, figs. 11, 13, 14). If
these are indeed, as I suppose, the “capsules,” then I think they are not capsules at all,
but simply post-mortem splittings-away of the mesogloeal matrix from the cell. They
seem to have no definite limiting membrane, and do not occur round more than about
half the cells: also, they are often only on one side of the cell, as in PI. Ill, fig. 16,
which suggests strongly a slight drawing back or contraction of the matrix ; as does
fig. 11. I consider these cells to be probably amoeboid cells carrying or transmitting
nutriment, as Miss Pratt (54a) has proved to be the case in Alcyonium, and not
connective-tissue cells (see p. 2). In my material they are rather more numerous
near the ectoderm and endoderm in one case, evenly distributed in the other. This
may be due to a difference in the time which has elapsed since the last meal. Note
the strong suggestion of a food-carrying cell given by PI. Ill, fig. 15.

Although it seems that the sphincter may be more branched than Carlgren at first
thought, Bolocera longicomis may still be easily distinguished from other species of
Bolocera. From B. multicornis, B. pannosa, B. brevicornis, and B. africana it
differs in general form, and has far longer tentacles than any of them. From B.
g)ollens it is marked off, since that has a circumscribed sphincter. In B. kerguelensis
and B. multipora the sphincter though diffuse has a strong apical mesogloeal lamella,
much larger and more branched than any of the others, in sections cut behind the
insertion of a tentacle. (This is also the case in B. pannosa, and sometimes at any rate

ACTINIARIA— STEPHENSON.

23

in B. multicornis.) B. norwegica has only six, and B. occidua only twelve pairs of
perfect mesenteries, whereas in B. longicornis all are perfect. The anatomy of B.
tuediae is not yet fully known, but I hope to contribute a few notes on it before long.

B. longicornis, then, may be defined as follows : — Tentacles in six cycles (in well-
grown specimens), the inner ones about as long as the height of the body. Sphincter
diffuse, without any strong apical lamella, but varying in the amount of branching of
its processes. Mesenteries all or nearly all perfect.

Family ALICIIDAE, Duerden.

“ Actiniina with an adherent base. Column- wall with simple or compound hollow
tubercles, or with vesicular outgrowths. Sphincter endodermal, diffuse. Margin with
or without acrorhagi. Tentacles simple. Mesenteries arranged in several cycles, of
which usually more than one is perfect ; longitudinal muscles diffuse ; parieto-basilars
and basilars unequally developed. No acontia.” (McMurrich, 40.)

PHYMACTIS, Milne-Edwards.

Aliciidae with the column-wall entirely covered with simple and compound
vesicles, without definite arrangement, Acrorhagi present. Numerous perfect
mesenteries.

5. P. clematis, Drayton. (PI. I, fig. 6 ; PI. VI, fig. 15.)

P. clematis, Drayton. Drayton in Dana. “ Zoophytes.” U.S. Explor. Expect, Philadelphia,
1846.

Three specimens. Shore between tide-marks, South Trinidad Island, July 26, 1910.

Measurements. — (i) Largest specimen. — Greatest diameter, 4*2 cm. Diameter of
oral opening, 2 • 3 cm. ; of pedal disc, 2 ' 1 cm. Length of a tentacle, * 5 cm. Greatest
height, 2 * 3 cm.

(ii) Smallest specimen. —Greatest diameter, 2 • 8 cm.

There is no need to describe this species, as that has been done fully by McMurrich
(40) and Carlgren (15). I will merely add a point of interest which struck me. My
specimens agree exactly with McMurrich’s description, but that the mesenteries are
thicker and more compact (perhaps due to contraction), and the vesicles of the column
are inseparably fused together in a mass, some being more completely attached to each
other than others. Prof. McMurrich assures me that this was the case in some of the
specimens which he examined.

The mesenterial filaments present a point of unusual interest. The greater part
of their mesogloea is crowded with small zooxanthellae (PI. VI, fig. 15). I have not
hitherto seen any record of zooxanthellae in the mesogloea of an anemone. It will be noted
that there is a narrow area free from them at the back of the filament. Furthermore,
the distribution of the nuclei of the endoderm-cells of the filament is rather striking,
and this is also shown in the figure. In the ciliated lateral lobes some are scattered

24

“TERRA NOVA” EXPEDITION.

throughout the height of the layer, but by far the greater number of them are
concentrated in a dense zone in the inner part. In the glandular lobe of the filament,
on the other hand, they are found in an irregular belt in the middle of the layer.

The general appearance of the largest specimen is shown in PI. I, fig. 6.

Family CRIBRINIDAE, McMurrich.

“ Actiniina with an adherent base. The column usually more or less verrucose
and frequently with acrorhagi at the margin, these, however, never being ramose or
frondose. Sphincter endodermal, circumscribed. Usually more than the first cycle of
mesenteries perfect. No acontia.” (McMurrich, 40.)

EPIACTIS, Verrill.

Leiotcalia, Hertwig.

Leiotealia, Hertwig. “ Report on the Actiniaria dredged by H.M.S. * Challenger.’ ”
“ Challenger ” reports, Zoology, Vol. VI, 1882.

Cribrinidae with smooth body-wall, without verrucae or acrorhagi. Tentacles
simple, in several cycles.

6. E. novo-zealandica, n. sp. (PI. I, fig. 8 ; PL III, fig. 28 ; PI. VI, figs. 3, 4.)

One specimen. Station 96, seven miles E. of North Cape, New Zealand, 70 fathoms,
Aug. 3, 1911. Agassiz trawl. Bottom — sand and rock.

Measurements.— Diameter of crown, 2 * 1 cm. ; of column, 2 • 3 cm. ; of base,
2*3 cm. Height of column, 2'7 cm. Length of a large tentacle, ’9 cm.

External characters. — Pedal disc irregular in outline, slightly exceeding the lower
part of the column, without cuticle. It has been torn in detaching the animal, but
where any ectoderm remains it is thick and firm, a little corrugated. The column is
barrel -shaped, being slightly wider in the middle than above and below (PI. I, fig. 8),
a little higher than wide. Substance firm, tough, opaque and self-supporting. Colour
of whole animal yellowish. Body-wall fairly thick. Margin a distinct parapet, within
which is a deep fosse, the tentacles arising from the disc on the inner side of it. There
are no verrucae or acrorhagi or pseudo-acrorhagi, but the whole surface of the column is
so wrinkled in all directions that it has a corrugated appearance, as if covered with
irregular papillae. Some of the wrinkles are so deep that in transverse section their
ectoderm gets isolated in places, and little islands of it appear in the mesogloea, as
Stuckey has recorded for E. tliompsoni. Tentacles covering most of the disc, the inner
a good deal larger than the outer on the whole, but the size rather irregular. They are
stiff and self-supporting, fairly thick walled, fairly long, graceful, slender, tapering to a
point. They are longitudinally and transversely striated, the longitudinal striation
practically amounting to fluting. They are in four cycles — 12, 12, 24, 48 = 96. Oral
disc narrow, radially grooved as usual. Mouth wide open, with a corrugated lip.

ACTINIARIA — STEPHENSON.

25

Structure, (i) Mesenteries. — The specimen is not very well preserved, and is very
brittle, so that the mesenteries are hard to count. There are at least forty pairs, in four
cycles, and all are perfect. Probably the formula is 6 + 6 + 12 + 24 = 48 (pairs).
The mesenteries of the last cycle run about half-way down the actinopharynx. The
specimen is infertile, and I can find no mesenterial stomata. The musculature of the
mesenteries is strongly developed, and resembles that of E. thompsoni.

(a) A large mesentery (PI. VI, fig. 4). There is a large, long, broad peunon,
occupying half or more than half the muscular surface of the mesentery, and ending at
both ends quite abruptly in a strong projection. There are cracks in its surface due to
infolding of the endoderm. Its processes are very numerous, long and fine, and highly
branched. The part of this side of the mesentery between the pennon and the body-
wall is fringed with short simple or slightly branched processes of the longitudinal
musculature, which are highest at a point about half-way between the pennon and the
body-wall. The parieto-basilar muscle is very well developed, and has a strong free fold,
quite a flap, often enclosing elongate cavities in the mesogloea of the mesentery. Its
processes are low and simple or slightly branched : they cease at the apex of the fold.
The mesentery is strong and firm, with thick mesogloea. The above remarks apply to
all the larger mesenteries.

(b) The smaller mesenteries. In each succeeding cycle the pennon of course
becomes shorter, as does the mesentery, but it does not become narrower, so that on
the mesenteries of the last cycle the pennon has a rounded circumscribed form, without
a large surface of attachment to the mesentery.

(ii) Sphincter. — This is very interesting. It not only forms a link between the
sphincters of E. prolifera and E. ritteri, but seems to show an intermediate step
between two stages of sphincter-evolution. It is strongly circumscribed (PL VI, fig. 3),
but is attached to the wall by a sort of “ tail,” which tapers off gradually. It is sub-
pinnate, the main lamella from which the processes arise being slender, and near one
side of the muscle. The processes arise mainly from its end in a radiating manner, and
are exceedingly fine and numerous : they are much branched, and a good deal of
anastomosis takes place between them. The figure gives accurately the general
appearance and approximate number of them, since it would be impossible to draw each
process individually as in Bolocera, for instance. This sphincter differs from that of E.
ritteri in having even finer and more numerous processes, and by being fastened to the
wall by a “ tail,” thus having a much wider area of attachment. It rather resembles the
sphincter of E. thompsoni, but again its processes are far more numerous and more
anastomosing, and in E. thompsoni the main lamella is central and there is no tail. It
does not resemble the sphincter of E. prolifera at all, and E. nymphaea, E. badia, E.
dubia, E. spetzbergensis have quite a distinct type of sphincter with far fewer processes
than the three above-mentioned species. E. marsupialis is insufficiently described.
It seems to me that the sphincter-muscle of E. novo-zealandica represents a stage
between, for instance, the two types of sphincter found in Bolocera Jcerguelensis and

E

VOL. V.

26

“TERRA NOVA” EXPEDITION.

Epiactis badia. In B. kerguelensis the sphincter is diffuse, hut it has diverged from the
truly and typically diffuse state, such as in B. longicornis, towards circumscription, by
producing one large mesogloeal lamella at the upper end, which far exceeds and is much
more branched than the others. Then in E. novo-zealandica it seems as if this strong
upper lamella had grown out enormously, and the lower diffuse part of the sphincter
had shrivelled up like the tail of a tadpole ; but still a small tail is left. In E. badia, on
the other hand, even this little tail has gone, and the sphincter is attached to the wall
by a short peduncle, and the lamella has become central with the processes truly
pinnately arranged (McMurrich, 1893, PI. XXXII, fig. 104).

(iii) Endoderm. — Endodermal circular muscle of body-wall well developed, on
short stout processes almost approaching those of Artemidactis victrix , etc. Endoderm
of body-wall low, with numerous small finely granular gland-cells in its outer part,
varying in shapes and sizes as to detail. Endoderm of mesenteries rather better
developed, and crowded with similar small gland-cells.

(iv) Mesogloea. — The matrix is finely fibrous. There are numerous cells, and in
the body-wall and tentacles they are quite crowded — fe'wer in actinopharynx and
mesenteries. Many lie in clear spaces, 'which I attribute to post-mortem contraction.
Some are large and dim with clear nuclei, bipolar or stellate. Many are irregular and
granular. The majority, however, are rather small and very clear, oval, rounded,
pyriform or bipolar, deeply stained with haematoxylin, and often finely granular. It is
possible that these are only large nuclei of invisible cells, but they are just like
numerous small gland-cells of unusual sort occurring in the endoderm and ectoderm of
the tentacles. There is often a darker spot which may be the nucleus, but on the other
hand there sometimes seems to be some very dim protoplasm round them.

(v) Ectoderm of body-icall. — High ; higher than the endoderm. With numerous
small gland-cells and nuclei in the inner part, some of the former finely granular, but
some not visibly so.

(vi) Actinopharynx. — Descending more than half-way down the coelenteron : finely
ribbed and corrugated, but without distinct ridges. Two well-marked siphonoglyphes.
Its ectoderm is densely crowded with small deeply staining finely granular gland-cells.

(vii) Tentacles. — Endoderm thin : mesogloea and ectoderm regularly folded by the
longitudinal flutings (PL III, fig. 28). Mesogloea thin in the furrows, thick on the
ridges, and where thick about equal to the ectoderm. Longitudinal muscles
ectodermal, best developed on the ridges and wreak in the furrows. The processes are
rather scattered, fairly high, slender, simple or a little branched. In E. thompsoni they
are higher, and the mesogloea is thinner. Endodermal circular muscle w-ell developed.
Endoderm with a thick belt in its central part of small regular cells, which I take to be
unusual gland-cells, though they rather resemble large nuclei of various shapes. They
are rounded or oval as a rule, or elongate, some finely granular, some not visibly so.
The ectoderm has a striking appearance, being clearly divided into three zones of about
equal width : the inner is the well-marked nerve-layer (broader than in E. thompsoni) :

ACTIN I ARIA— STEPHENSON.

27

the next is composed of a crowd of the small gland-cells found in the endoderm, but
their shapes are more various : the outer zone is a densely-packed battery of spirocysts.
These latter are about as usual, large, long, subcylindric. Here and there occurs a
thick walled nematocyst, rather narrower and shorter than the spirocysts. The outer
half of the ectoderm is occupied by the nuclei of the supporting cells.

I have followed Torrey (60) in keeping to the genus Epiactis, which name has
priority over Hertwig’s Leiotealia. Epiactis prolifera, the type species of the genus,
has certainly a rather different sphincter from all the other species, but some of the
others have sphincters intermediate in type between that and E. nympliaea, which
represents the opposite extreme. Furrows of the body-wall have no value to separate
E. nympliaea generically, as Torrey pointed out, and the lower sphincter of its body is a
very feeble and indefinite one. Moreover, in general, all the species described at
various times under Leiotealia and Epiactis resemble each other considerably, and
form one good genus. Therefore Epiactis contains the following species

E. pjrolifera, Verr.

E. ritteri, Torrey.

E. thompsoni, Stuckey.

E. novo-zealandica, n. sp.

E. fecunda, Verr.

(= E. regular is, Yerr.)
E. dubia, Wassilieff.

E. spetzbergensis, Kwietn.
E. marsupialis, Carlgr.

E. badia, McMurrich.

E. nympliaea, Hertwig.

It is possible, as Carlgren (15) has pointed out, that E. badia is identical with
Isotealia antarctica, but as yet that is not proved.

E. novo-zealandica is at once divided from most of the above species by its sphincter.
From E. thompsoni and E. ritteri its sphincter also divides it ; E. ritteri differs from it,
so far as one can tell from Torrey ’s description, in the pennons of the mesenteries, which
in that species are flat ; and E. thompsoni has also very stout instead of slender
tentacles, which are arranged in a curious and quite unusual manner.

Family PARACTIDAE, Hertwig.

Actiniina with well-developed base and basilar muscles. Sphincter mesogloeal.
Acontia absent.

Since I am introducing three new species (including two new genera) into this
family, it is desirable to point out how they may be distinguished, by means of a key,
from other members of the Paractidae, as that family is at present understood. I hope,
at an early date, to completely revise the classification of Paractidae and Sagartiidae.

The family falls naturally into two divisions, the Actinostolinae and Paractinae :
with the former we are not concerned.

In the Paractinae, two groups of genera at once distinguish themselves firstly,

E 2

28

“TERRA NOVA” EXPEDITION.

those which have a smooth body-wall, and secondly, those which are provided with
papillae or verrucae.

I. The smooth-walled genera.

A. Here two genera are marked off by the possession of a cuticular covering —
Phelliopsis Verrill and Marsupifer Carlgren, of which the former has a simple sphincter,
while the latter has the sphincter doubled, and its female develops extraordinary brood-
pouches.

B. Among the forms with no cuticle are many which have a perfectly smooth
wall with no specialised margin and with normal tentacles.

a. But in Sycyonis R. Hertwig, the terminal openings of the tentacles are so large
that the tentacles are almost reduced to collars. In Pycnanthus McMurrich
there are capitular ridges. Here my first genus may be placed — Hormosoma,
which has a well-marked collar at the margin of the body, lapping back
over the column in expansion, as well as other striking characters (see below).

h. If we now consider the forms which have no specialised external features, the
first to be separated off is Antholoba Hertwig, which has the disc lobed and
the tentacles very small and numerous. Of the remainder with a non-lobed
disc, Alloactis Verrill has specialised tentacles, while in the following genera
they are quite normal. First, we have the genera with a thick body- wall —
Cymbactis McMurrich, with numerous perfect mesenteries, longitudinal
muscles of the tentacles mesogloeal, and mesenterial musculature of the usual
type. Secondly, we have the thin walled genera : Paranthus, which has only
six pairs of perfect mesenteries, and the longitudinal muscles of the tentacles
ectodermal (see Maguire, 46) ; and Paradis M.-Edwards, which should be
confined to forms with numerous perfect mesenteries, longitudinal muscles of
the tentacles ectodermal or mesogloeal.

II. The genera with papillae or verrucae.

A. Genera with more or less cuticular covering. Here we have Phelliomorpha
Carlgren with a true cuticular sheath, elongate body, papillae as in Ilalcampa, and a
small sphincter lying directly at the bases of the tentacles ; and my genus Lilliella,
with a feebly developed and patchy encrustation, a large and well-developed sphincter,
and in the mesogloea of the body -wall very numerous lacunae, some of which are so
large and so near the surface that they form externally visible hollow papillae.

B. With the genera which have no cuticular covering we are not specially

concerned here. There are several of them, none very well known, but they are easily
distinguished from one another : Cyathadis Danielssen ; Cadosadis Danielssen ;

Tealidium Hertwig ; Raphadis Verrill ; Ammophiladis Verrill.

III. There are several genera which I have not included in the above key, as the
present knowledge of them is not complete enough to allow of assigning to them any
definite positions — viz. , Parantheoides Carlgren ; Par actinia Andres ; Archadis
Verrill ; and Synanthus Verrill.

ACTINIARI A— STEPHENSON.

29

Sub-family PARACTINAE (Carlgren).

Paractidae in which the individual mesenteries of each pair are typically equally
developed ; if any inequality occurs it is slight, and follows no definite rule as in the
Actinostolinae, where the inequality is striking. I do not agree that the Paractinae and
Actinostolinae should have higher rank than that of sub-families.

HORMOSOMA,* n. gen.

Paractinae with thick, smooth body-wall, without verrucae or capitular ridges, but
with a prominent and well-developed marginal collar, which laps back over the column
in expansion. Tentacles normal, their longitudinal muscles mesogloeal. More than six
pairs of mesenteries perfect. No cuticle. Sphincter simple. Margin not lobed.
Mesenterial musculature extraordinary, supported on high ridges (see below), the
retractors more or less circumscribed. Genotype, Id. scotti, n. sp.

This genus differs from Cymbactis by the possession of a collar, from Pycnanthus
by the absence of capitular ridges, and from Paradis and Paranthus by its thick wall
and collar. It also differs from all anemones I have seen described in the character
of the mesenteries : it is difficult to put this last character into two or three words for
the generic definition, but reference to the figure and the description below will make
it clear. I have honoured the species by calling it Scotti, after the heroic leader of the
Expedition, the late Captain R. F. Scott, C.Y.O.

7. Hormosoma scotti, n. sp. (PI. II, figs. 2, 17,18 ; PL III, figs. 17, 19, 21;
PI. IV, figs. 4, 7, 8, 9 ; PI. VI, fig. 10.)

Three specimens. Station 338, 77° 13' S. ; 164° 18' E., 207 fathoms, entrance
to McMurdo Sound, Jan. 23, 1912. Agassiz trawl. Bottom — mud.

Measurements, (i) Largest specimen {expanded). — Diameter of oral disc and
tentacles, 10 '2 c.m. ; of pedal disc, 6’2 cm. ; of column, 4‘8 cm. Height of column,
2 • 8 cm. Length of a large tentacle, 1 • 3 cm.

(ii) Smaller expanded specimen. — Oral disc and tentacles ; diameter, 6 ' 5 cm.

External characters. — Base well marked, broad, thick, its edges incurved in two
specimens and in part of the third, expanded and exceeding the column in the other
part of the third. Its outline undulate. Its ectoderm intact, resembling that of the
column, much corrugated and radially furrowed, more distinctly in two than in the
third. The column varies in shape according to the state of contraction. In the
contracted specimen it is about as high as wide, in the expanded ones much wider than
high, especially in the largest (PI. II, fig. 2). The smallest specimen is not much
smaller than the one figured. In the fully open specimen it is rather larger above than
below, the disc being enormous. The wall is very thick. In the two smaller specimens
most of the ectoderm has been rubbed off, but in the largest it is intact : it is rather

*op/uos, collar, and aw/xa, body : “ with a collared body.”

30

TERRA NOVA” EXPEDITION.

thick and soft, reddish brown, and much wrinkled and irregularly corrugated : it
becomes thinner and smoother at the margin on the underside of the collar. The body
has transverse contraction-folds, and many of the wrinkles take a longitudinal direction.
The upper margin is developed into a very distinct and well-marked collar, quite as
distinct as in Metridium, perhaps even more so. It has a thin edge, and its upper
surface is lined by delicate smooth transparent capitular ectoderm, such as covers the
disc and lines the broad fosse separating the collar from the outer tentacles. In the
contracted specimen it is drawn round so as partly to cover and conceal the tentacles,
and in the expanded one it laps back over the column. In the third specimen it is
imperfect on one side of the animal, but the creature has clearly been damaged, so that
an abnormal growth has taken place. Unfortunately the figure hardly shows a trace
of the collar, as it is hidden by the downwardly pointing tentacles. The oral disc when
fully expanded is very broad, and projects on all sides above the collar beyond the
column, but it is perfectly circular in outline, not at all lobed or undulate. In the fully
expanded specimen it is very thin, flabby and semi-transparent, in the contracted state
of course it is rather thicker. It is convex, and forms a dome above the tentacles. It
has shallow radial ridges running towards the mouth from the bases of all the tentacles,
those running from the primary tentacles being of course the largest. The mouth is
raised on a cone, is roughly circular, and has a prominent ridged lip. The tentacles are
short, stout, and very blunt. Most of them are rather club-shaped (though a few are
more conical), being rather narrower at the base than in the middle or near the tip.
They are strongly longitudinally fluted, and also transversely striated, and are thick-
walled, but not stiff enough to support themselves out of water if well expanded. The
inner are a good deal longer than the outer. Their colour, like that of the disc, is a
.sort of pinkish buff. They are very contractile, and in the contracted state become
mere knobs. The disc can be completely infolded, and the collar pulled in over the
edges of the tentacles. The tentacles are large when well expanded. Sometimes the
blunt tip is convex, sometimes there is a depression in it, but I cannot demonstrate any
terminal pore. Tentacles regularly arranged in four cycles, 12 + 12 4- 24 + 48 = 96.

Structure, (i) Mesenteries. — These are in four cycles, 6 + 6 + 12-1-24 = 48 (pairs).
The first three cycles perfect. All are fertile. The mesenteries are thick below, thinner
above, and they are strongly ridged. The ridges are clearly visible to the naked eye,
and run longitudinally : in transverse section they form extraordinary lobes at each side
of the mesentery, like those of a deeply cut oak-leaf (see PI. IY, fig. 9). They are
more prominent below than above. There is no trace of mesenterial stomata. The
retractor-muscle forms a strong, rounded, prominent ridge on one side of the mesentery
just behind the gonad, and the parieto-basilar muscle has a large free flap on the other
surface, which can be pulled out with a pair of forceps : this muscle covers most of the
surface of the mesentery and joins the body-wall just below the oral disc. As to the
musculature, it would perhaps make it most clear if I first describe a single mesentery
of one of the higher cycles, such as is shown in PI. IV, fig. 9. The mesogloea is very

ACTINIARIA— STEPHENSON.

31

thin where the mesentery leaves the body-wall, but throughout most of the muscular
part is extraordinarily thick. This section is taken about half-way up the mesentery.
The longitudinal musculature covers the whole of one side of the mesentery behind the
gonad, the parieto-basilar the whole of the other. The mesogloea of the mesentery
sends out on each side huge lobe-like lamellae, which have thick lobe-like branches,
and give the mesentery its extraordinary oak-leaf-like appearance. The longitudinal
muscle-fibres, on the greater part of the surface of the mesentery, are borne on small,
short, tertiary, simple or slightly branched processes, arising like a fringe round the
lobes of mesogloea : but just behind the gonad the processes become high and ordinary
and branched, and form a small, strong, circumscribed pennon. The opposite face of the
mesentery is fringed with parieto-basilar muscle similar to the longitudinal, but opposite
the pennon this muscle ends in a large free flap with higher processes fringing the
finger of mesogloea which runs out into it. It can be seen from the figure that the
outlines of the mesogloeal lobes have a more spiky form on the parieto-basilar side of
the mesentery than on the other. The narrow neck of mesogloea uniting the mesentery
and the body-wall bears a few little processes on each side. In the small imperfect
mesenteries the lobing in transverse section is not distinct, and there is no pennon : but
the flap of the parieto-basilar muscle is well defined. Larger mesenteries than the one
described present essentially the same characters. In some cases one mesentery of a
pair is a little longer than its neighbour, but there is no consistent irregularity as in
Actinostola, and when irregularity occurs it is not according to rule. I noticed the
same thing in Bolocerci longicornis, and it is perhaps faintly visible in Cymbactis
selnginella.

(ii) Sphincter. — This has a very definite and characteristic outline in transverse
section, rather like a bird’s head (PI. IV, fig. 7). It is rather similar in some respects
to that of Actinostola spetzbergensis (Carlgren 1893, Taf. IX, fig. 1), but besides other
differences its structure is even finer. Throughout its length one side of the sphincter
lies up against the endoderm. Above, it fills the whole thickness of the mesogloea,
running out into the collar and completely filling it. Below, it rapidly tapers off to a
sharp point, and so does not run far down the wall. A very small portion of it is
given in detail in PL IV, fig. 4, to show its minute structure. It will be seen that the
mesogloea, which is dotted, is divided by numerous clear spaces into fine strands, which
form a network, fringed by muscle-fibres. I propose to call this type of sphincter-
structure “reticular” for convenience of reference, as opposed to the “alveolar”
type shown in PI. IV, fig. 15, for another species in which the spaces lie further
apart and look rather like alveoli. A third type of sphincter-structure is shown
for Leptoteichus in PI. IV, fig. 2, in which the muscle-fibres are merely single or in
small clumps.

In the case under discussion the whole sphincter is a network of exceedingly fine
and numerous strands. The meshes are rather larger on the endodermal than on the
ectodermal side. The general tendency of the stronger fibres is to radiate from the

32

“TERRA NOVA” EXPEDITION.

basal point of the sphincter, and to sweep outwards into the collar, and this gives the
sphincter a characteristic appearance, wavy and fibrous.

(iii) Endoderm. — That of the body-wall is higher than that of the muscular part
of the mesenteries, and contains large granular gland-cells which are on the whole oval
in shape, though varying in detail. These are not numerous in the endoderm of the
mesenteries and body-wall, but are plentiful in the mesenterial filaments. The glandular
lobes of the filaments are clearly marked off and have a rounded kidney-like outline
with a very short stem. They contain two sorts of gland-cells : a fringe of the large
coarse sort above mentioned, and also delicate, narrow, elongate ones such as are not
found elsewhere in the endoderm. They have much smaller granules than the others,
and though much longer are far narrower. Numerous spaces which have presumably
contained gland-cells occur in the endoderm, and are perhaps empty. The endoderm
everywhere has a tendency to be folded. The endodermal circular muscle of the wall
is well developed.

(iv) Mesogloea. — The matrix is almost homogeneous, a little more fibrous in the
mesenteries. Amoeboid cells of all shapes and sizes are found in it. Examples are
shown on PI. Ill, figs. 17, 19, 21, and PI. II, figs. 17, 18. The two kinds shown in
PI. II, fig. 18, and PL III, fig. 17, are particularly numerous in the mesenteries. All the
cells seem connected together with very fine processes. Some are rounded, some bipolar,
some stellate, often they are strung together in chains and groups. They possess, in
this species, no haematoxylin-staining nuclei, but it seems as if their nuclei stained red
with eosin only (see PI. II, fig. 17).

(v) Ectoderm of body-wall. — This contains very numerous large gland-cells of
various shapes, and often a deeply embedded cell has a clearly visible “ duct.” Many
are superficial, many deeply embedded in the very thick ectoderm. The ectoderm of
the fosse is more delicate than that of the rest of the body.

(vi) Actinopharynx. — With strong longitudinal ridges and slight transverse folds.
Two large very clearly marked siphonoglyphes.

(vii) Tentacles.— Ectoderm rather higher than endoderm. Mesogloea in its thicker
parts much thicker than either. The fluting of the tentacles causes deep folding of the
ectoderm, so that alternating high places and valleys occur in section (PI. IV, fig. 8).
The longitudinal musculature is embedded in the mesogloea, and fills its outer half,
lying against the inner side of the ectoderm, and separated from the endoderm by a
broad belt of mesogloea. It is rather better developed in some parts of the tentacle
than in others, and is poor in the furrows and strong on the ridges. The muscle-fibres
are arranged in complete or incomplete rings which are compressed into various shapes.
Towards the edges of the layer they become scattered in small feeble clumps (PI. IV,
fig. 8). The nerve-layer of the ectoderm is well defined, forming a faint open network.
There are spirocysts in a dense battery in the outer part of the ectoderm, mostly ; they
are long and nearly cylindrical. Numerous coarse granular gland-cells are present,
some among the cnidae of the battery, some below it. The endoderm contains a few

ACTINIARIA— STEPHENSON.

33

gland-cells, and certain cells with processes which resemble more or less the amoeboid
cells of the mesogloea (PI. VI, fig. 10).

LILLIELLA,* n. gen.

Paractinae with thick body-wall which is feebly divided into scapus and
capitulum. Scapus with a poorly developed patchy incrustation, not a cuticular
sheath. The mesogloea of the scapus contains very numerous lacunae of all shapes
and sizes, some of which are so large and so near the surface that they form externally
visible papillae. Only six pairs of perfect mesenteries. Mesenterial musculature
normal. Sphincter simple, large, and well developed. Longitudinal muscles of
tentacles ectodermal ? Genotype, L. lacunifera, n. sp.

I believe I am justified in establishing a new genus for the single badly preserved
specimen which is described below, though it is possible that more and better preserved
material would present differences. The lacunae which are its most striking character
seem to be genuine normal structures with a cellular lining, and not due to bad
preservation. Apart from its extraordinary lacunae, the genus differs from
Phellioniorpha by the character of its sphincter and the absence of a true cuticle-sheath
and delicate capitulum. Cyathactis, etc., have no incrustation at all, and possess
numerous perfect mesenteries.

8. Lilliella lacunifera , n. sp. (PI. I, fig. 13 ; PL II, figs. 12, 13, 14 ; PI. Ill, fig. 5 ;
PI. IV, figs. 10, 11, 12, 13.)

The significance of the specific name is obvious, on account of the lacunae
characteristic of the animal.

One specimen. Station 348, off Barne Glacier, McMurdo Sound, 200 fathoms,
Feb. 13, 1912. Agassiz trawl. Bottom — mud.

Measurements. — Height of column, 7 * 3 cm. Diameter of column, 3 • 6 cm. ; of
pedal disc, 2 * 8 cm. ; of capitulum, 2 * 8 cm. ; of oral opening, 1 ' 5 cm. Length of a
tentacle, 1 • 5 cm.

External characters. — Unfortunately all but the body-wall and the frames of the
mesenteries of the only specimen are absolutely decomposed into a homogeneous mass,
so I cannot make a complete description.

Pedal disc roughly oval, the edges of the column a little inrolled over it. Traces
of cuticle or mucus near its edge. It is smooth, whitish, thin and membranous.

Column much higher than wide. It has no definite shape, as it has been crushed
(PL I, fig. 13). The base would slightly exceed it if the edges were unrolled. The
column narrows slightly upwards, the summit being contracted, and only a few
tentacles peeping out. The wall is thick, firm, and self-supporting. The uppermost
portion, the capitulum, is smooth and devoid of ectoderm : it is not thinner or more

* Named in honour of Mr. D. G. Lillie, Biologist to the Expedition, to whose care and skill the
formation of the collection and the excellent preservation of the specimens are due.

von. v.

F

34

“TERRA NOVA” EXPEDITION.

delicate than the scapus, if anything thicker : its margin is tentaculate. The scapus
has a rough appearance from various causes. A good deal of the ectoderm has gone,
but where it remains in patches or in cracks it is rough and brown and gritty, and
when sectionised can be seen to be a good deal mixed up with sand-grains, mucus, and
foreign bodies such as diatoms and spicules : so that although there is no true cuticle,
the animal when whole must have had a good deal of incrustation.

The surface has more or less regular longitudinal and transverse furrows, dividing
it up into numerous small rectangular or irregular areas. In addition to this, here and
there, irregularly distributed but fairly numerous, are little irregularly shaped hollow
vesicles or papillae : they are inconspicuous and do not project beyond the surface, and
are by no means true verrucae or vesicles : if one is cut open, it is seen to be caused

simply by a large lacuna lying in the mesogloea quite near the surface. What is
visible of the oral disc is thin and radially grooved. Mouth and lip degenerated.
The tentacles have lost all their ectoderm, and are brittle, flaccid, semi-transparent,
long, tapering, and pointed. They are too disintegrated to count accurately, but there
are over eighty in three or four cycles. Their arrangement may have been 12, 12, 24,
48 = 96.

Structure, (i) Mesenteries. — In four cycles. 6 + 6 + 12 + 24 = 48 pairs. Those of
the first cycle alone are perfect, and are connected with the actinopharynx throughout
its length. Those of the last cycle are very small and hardly project beyond the
encloderm. The filaments and gonads form a solid, brittle, structureless mass, so I can
give no details of fertility. In sections, remains of ova, granular gland-cells, and
filaments are distinguishable, but no trace of large nematocysts or acontia : from this I
conclude the animal is a Paractid, but if acontia should turn up in better material the
genus can easily be transferred to the Sagartiidae. I can give no description of
musculature, as all has vanished, but what remains shows that it was not of the
Leptoteiclius or Hormosomci type.

(ii) Sphincter. — This is pretty well preserved. It is long, running through the
whole thick wall of the capitulum and just into the top of the scapus. It does not at
any point fill the whole mesogloea, but lies right up against the ectoderm, and is
separated throughout by a broad band from the endoderm. Its shape is outlined in
PI. IV, fig. 13. The outer side is irregular and follows the depressions or elevations of
the outer side of the mesogloea : about half-way down it begins to separate from the
ectodermal side (the ectoderm itself has vanished), and gradually tapers off to a point
lying in the middle of the mesogloea. Its structure is moderately clear, though the
fibres are somewhat degenerated. A small portion is shown in PI. IV, fig. 10. It is
alveolar, but many of the alveoli have collapsed into clumps, particularly the larger
ones — some more and some less. The development is best and densest a little below
the margin, and the alveoli tend to an indistinct arrangement in transverse rows in the
upper part. The general appearance is irregular and rough, the alveoli and clumps
being often small and scattered, often massed together.

ACTINIARIA— STEPHENSON.

35

(iii) Mesogloea. — This is well preserved in the body-wall, although most of the
animal is badly preserved. Its matrix has a curious uneven appearance which is
enhanced by the various structures which it contains. Firstly, there are numerous
lacunae of all sizes and many shapes, many of them visible to the naked eye. A small
portion of mesogloea is shown in PI. IV, fig. 12 : this shows the lacunae, and the dotted
portions inside them represent large masses of their lining. I do not think these lacunae
can be merely accidental, or due to bad preservation, because they are so regularly and
uniformly developed, and are lined or half full of some sort of epithelial lining : this
consists of a fine protoplasmic network which stains more or less with haematoxylin,
and contains numerous nuclei. This network may merely lie in a narrow band round
the edge of the lacuna, or may partly or wholly fill it. A small portion of it is shown
in PI. IV, fig. 11. Although fig. 12 does not actually show one, it will be easily under-
stood that when a large lacuna (many are considerably larger than those figured)
occurs just below the surface, it produces one of the little blisters described above as
being externally visible on the scapus. The lacunae occur less frequently in the
capitulum amongst the sphincter than elsewhere, and do not there form blisters on the
surface — nor do these latter occur in every part of the scapus equally. Secondly, here
and there a little crack or pit in the body-surface becomes isolated in a section, so that
it has the appearance of a small island of ectoderm near the surface of the mesogloea,
but is of course not like a lacuna. Thirdly, the mesogloea between the lacunae contains
many cells and curious blackish (probably haematoxylin-stained) networks ; the cells
may be rounded or oval or amoeboid (see PI. II, figs. 12, 13, 14). Often they are
surrounded by a little of the curious black reticulum (PI. II, figs. 13, 14), and often
they lie in clear spaces, probably due to post-mortem contraction of the cell or the
matrix. Very frequently the black networks occur free, and unconnected with a cell, as
in PI. Ill, fig. 5, which is a small example. Some of the cells are granular. Their
nuclei are usually clear, and stain with haematoxylin. Lastly, the matrix does not
stain equally deeply everywhere with eosin, but narrow bands lying just beneath the
ectoderm and endoderm stain more darkly than the central part. The mesogloea of
the mesenteries does not seem to contain any lacunae, but there are curious patches here
and there of a coarse network not quite like that contained in the lacunae or surround-
ing cells : it occurs here and there in the body-wall also, and the patches look like
“ soft ” spots, perhaps degenerated a little.

Tentacles. — There are remains of processes of the longitudinal muscle in a fringe
on the outer side of the mesogloea, so that presumably it was ectodermal.

CYMBACTIS, McMurrich.

Paractinae with thick smooth body-wall, without verrucae, collar, or capitular
ridges : margin tentaculate, not lobed. Tentacles normal, their longitudinal muscles
mesogloeal. No cuticle. Sphincter simple. More than six pairs of mesenteries perfect.
Musculature of the mesenteries ordinary.

f 2

36

“TERRA NOVA” EXPEDITION.

I have slightly widened the definition of this genus to include forms discovered
since the species originally described by McMurrich in 1893. The genus at present
contains the following species : —

C. faeculenta, McMurrich.

C. actinostoloides, Wassilieff.

C. maxima, Wassilieff.

C. selaginella, n. sp.

9. Cymbactis selaginella, n. sp. (PI. I, figs. 7, 11 ; PI. IV, figs. 14, 15 ; PI. \,
figs. 1, 2, 4 ; PI. VI, fig. 14.)

I have given this species the name selaginella because of the characteristic
appearance of the processes of its retractor-muscles (PI. VI, fig. 14), which strongly
remind one of a sprig of Selaginella or club-moss. Of course they are not necessarily
confined to this species, but provide a convenient character to label it by.

In all there are seven specimens.

A. Five specimens. Station 294, Ross Sea, 74° 25' S. ; 1 7 9° 3' E., 158 fathoms,
Jan. 15, 1913. Agassiz trawl.

B. One specimen. Station 338, entrance to McMurdo Sound, 77c 13' S. ; 164" 18'
E., 207 fathoms, Jan. 23, 1912. Agassiz trawl. Bottom — mud.

C. One specimen. Station 339, entrance to McMurdo Sound, 77° 5' S. ; I64c 17' E.,
140 fathoms, Jan. 24, 1912. Agassiz trawl. Bottom — mud.

Measurements, (i) Largest specimen. — Diameter of oral disc and tentacles,
4’ 5 cm. ; of column, 3 • 7 cm. Height of column, 4 cm. Length of a large tentacle,
1 • 2 cm. Diameter of pedal disc, 3 1 9 cm.

(ii) Smallest specimen.— Height of column, 1 * 7 cm. Diameter of column, 1 ’6 cm.

External characters. — Six of the seven specimens are similar in general aspect, but
the seventh is more contracted, and in general appearance is so different that at first I
thought it was another species. This difference will be apparent from a glance at PI. I,
figs. 7, 11, the larger of which is a typical specimen, the smaller the one least like it ;
but in arrangement of tentacles, structure, and histology the two are identical, hardly
the slightest difference occurring between them histologically, and certainly they are
the same species. Pedal disc concave, rather drawn in at the edges ; if fully expanded
would slightly exceed the column. It is thick and firm, though not as thick as the
column, and is slightly wrinkled and furrowed, radially and otherwise. In five of
the seven specimens it has a curious little mound arising from the concave
centre, but the mound is imperforate. Only one specimen is at all contracted,
the others being well open. In these six the column is a little broader than high,
in the contracted one about as broad as high. When expanded, it is a little wider
at the margin than just above the base. The largest specimen is flattened by pressure.
Limbus thick-edged and well marked off. The wall is very thick and firm, tough and
pliable ; in the smallest specimen its thickness in proportion to the size of the animal

ACTIN'! ARIA— STEPHENSON-.

37

is enormous. The margin has a very slightly thickened rim, but is truly tentaculate,
without any parapet. There are no verrucae, capitular ridges, etc., but the surface
is broken up by longitudinal and transverse furrows into low', irregular areas, sometimes
like tubercles. In one specimen there is a perfectly circular deep constriction about
half-way up the body, but this is only a “ waist ” such as a living anemone often
produces for itself temporarily, and which soon passes off upwards. The mouth gapes,
so that the oral disc is rather narrow'. It is fairly thin, and has prominent radial
ridges running from the bases of the tentacles inwards towards the mouth ; these, of
course, are larger for the tentacles of the first two cycles than for those of the third.
The radii have fine transverse striations. The tentacles vary greatly in form in different
specimens, and in different parts of the same specimen, according to the state of
expansion. In PI. I, fig. 7, they happen to be all a good deal contracted, and are then
short and stout and conical, either blunt or w'ith a tapering point ; but in PI. I, fig. 11,
they are fully extended (though mostly hidden) and are quite long and slender, and
rather flaccid. They are thick walled. In many cases they are half expanded, and
then they have a stout base, but are on the w'hole moderately slender, and taper
gracefully and regularly to a blunt point. In PI. I, fig. 11, they are a little damaged.
They are rather soft, and have a curious mealy appearance due to irregular longitudinal
and transverse furrows, but are not distinctly fluted. I can demonstrate no terminal
pores. They are arranged in three cycles close together at the edge of the disc ; the
typical formula is 16 + 16 + 32 = 64, but in the smallest animal a tentacle of a
fourth cycle has appeared here and there, and in one other case slight divergences from
the regular arrangement are met with ; but the arrangement is never hexamerous or

o o J o

decamerous. Three specimens have the body, disc, and tentacles pale flesh-buff, the
tentacles a darker shade of the same. Four are pure cream- white, except that in one of
them the actinopharynx is striped with browmish violet.

Structure, (i) Mesenteries. — These are thick and excessively flabby, so that it is
hard to cut accurate sections of them. I can demonstrate no stomata. In the large
specimen examined there are four cycles of mesenteries : each of the first two cycles
contains eight pairs, and these are all perfect sterile. The third cycle contains seventeen
imperfect fertile pairs, one extra pair being accidentally present in one exocoel. The
fourth cycle consists of small single mesenteries, not pairs, alternating regularly with
the other cycles. They are entirely confined to the proximal part of the body, and
are fertile. From transverse sections it is evident that they are truly only single
mesenteries, and there is no trace of a young partner not yet grown beyond the
endoderm, as in the last cycle of mesenteries in some species of Actinostola. In
the smallest specimen there are the first three cycles as before, — 8 p. + 8 p. perfect
and sterile, 16 p. imperfect and fertile, but the fourth cycle only appears here and
there, as the anemone is very young. I will describe the musculature from sections
of the smallest specimen, as in the larger ones the mesenteries are so tw'isted and
flabby that good preparations cannot be made from them. The only difference is

38

“TERRA NOVA” EXPEDITION.

that in the larger specimen the pennons are larger and broader. They are the
same shape and have the same curious Selaginella- like processes. The longitudinal
musculature occupies the whole of one surface of the muscular part of the mesentery,
and forms a rather distinctly shaped diffuse pennon (PI. V, fig. 4). It is narrow
where it leaves the body-wall, but rapidly increases in breadth, so that at a very
short distance from the wall it is quite wide, and this is the highest point. From
here it tapers off very gradually to the other edge, and here and there its sequence
is broken by depressions lined by endoderm, and caused by the irregular surface
of the mesentery. The processes are rather fine, for the most part branched,
and have the curious appearance of a club-moss or a Selaginella, the sharp fibres
fringing them like pointed leaves. A small one is shown in PI. VI, fig. 14, much
enlarged. The parieto-basilar muscle is fairly developed, but where its processes
are cut across squarely they are very low and simple or only a little branched.
The muscle, however, encloses in the mesentery a series of cavities (clearly shown
in the figure), presumably by developing a free edge and repeatedly joining up
again with the body of the mesentery in course of growth. In the upper part of the
mesentery, where this muscle covers only a part of the surface, these cavities never occur
in the mesogloea beyond the point where it ends in a slight protuberance or faint free fold.

(ii) Sphincter. — I have figured the form of this in a large and small specimen, but
it will be obvious that in PI. V, fig. 2, the sphincter would be the same shape as in
fig. 1 if it were straightened out — i.c., if the animal were not wrinkled by contraction.
The inner edge lies up against the endoderm throughout, the lower part passing
insensibly into the ordinary endodermal circular muscle of the wall. It is wedge-
shaped, almost filling the mesogloea at the broad top, and dying out very gradually
and regularly below, and of course widely separated from the ectoderm throughout most
of its length. In structure it is “ alveolar” (in the small specimen ; see PI. IV, fig. 15,
for exact detail of a small portion) to “ alveolo-reticular ” (in the larger specimen). The
alveoli vary greatly in shape and size as to detail, but the largest ones occur on the
upper and inner margins ; they are very distinctly arranged in vertical rows, in the
upper part, with clear bands free from muscle-fibres between the rows. This is shown
in the small piece of detail in PI. IV, fig. 15. That is to say, the sphincter is layered
vertically. The majority of the alveoli are transversely elongate with a secondary
tendency to be arranged in irregular transverse rows : this elongation being less distinct
in the larger specimen than in the smaller.

(iii) Endoderm. — Endodermal circular muscle well developed. Body-w'all endoderm
high and deeply folded and devoid of granular gland-cells, though in one specimen
there are certain mealy, indefinite bodies which may be mucus-cells. When a mesenterial
filament has the simple trefoil-form it is fairly distinct, the three lobes being rather
angular : the glandular lobe has a stout stem, and the ciliated lobes are well developed,
with their nuclei concentrated in a dense zone in the inner part. The glandular
lobe has numerous granular gland-cells, mainly superficial, and its nuclei are mostly

ACTINIARIA— STEPHENSON.

39

concentrated in two masses at the back of the lobe, one on each side of the stem.
Gland-cells are present in the endoderm of the mesenteries, but whether they are
granular or not I cannot be sure.

(iv) Mesogloea. — The matrix varies a great deal in different parts of the same
section, often seeming homogeneous, often visibly and finely fibrous, sometimes strongly
so, sometimes faintly. It stains with eosin in an interesting manner : in the sphincter-
region, in one specimen, six distinct zones each stained a different depth can be made
out : the deepest layers are those just outside the endoderm and inside the ectoderm.
The cells are not particularly interesting : they are rather scattered, rather small, of all
shapes and sizes as to detail. They may be bipolar or multipolar, or may form chains
or lie in more or less definite clear spaces. Often they are surrounded by a complex
network of very fine processes.

(v) Ectoderm of body-wall. — This is higher in the old specimen than in the young
one : in this latter, where any ectoderm remains, it has the nuclei confined to the inner
part, and forming a well-marked zone. The outer part contains numerous ill-defined
gland-cells of all shapes, but typically large and oval : these are mostly stained with
eosin and not haematoxylin, and though not definitely and clearly granular, look as if
they might be so if one could see them more clearly. In the older specimen the ectoderm
is very regular, the nuclei are well scattered, most numerous in the middle part, and
there is a superficial band of gland-cells. These are fairly numerous, smaller than those
of the filaments, in the lower part of the wall. But higher up they are larger, and the
cells less regular. They are elongate and granular below, more various in shape above,
and granular or not visibly so in different cases.

(vi) Actinopharynx. — Furrowed longitudinally and wrinkled transversely as usual,
generally with a prominent lip, soft and rather thill. In all cases there are two well-
marked siphonoglyphes.

(vii) Tentacles. — Ectoderm higher than endoderm : mesogloea in its thicker parts
higher than either. All the layers are a good deal folded irregularly, by the wrinkling
of the tentacle-surface, and so vary in thickness at alternating points. There are two
quite distinct sets of longitudinal muscle : the main part of it is completely embedded
in the mesogloea, filling up about half that layer, and varying in position, even in the
same tentacle. Sometimes it lies in the outer half, only separated from the ectoderm
by a narrow strip of clear mesogloea, sometimes it lies right in the middle of the
mesogloea, separated about equally from ectoderm and endoderm (PI. IV, fig. 14).
It is formed of large oval or rounded alveoli of different sizes (lined by the muscle-
fibres), which are so close together that they rather press each other out of shape and
make the structure “ reticular.” But quite apart from this mesogloeal musculature, and
not connected with it, is a secondary fringe of tiny muscle-processes projecting into the
ectoderm. These are shown in the figure : they are very small, and simple for the most
part. The nerve-layer of the ectoderm is narrow but clear. The outer part of the
ectoderm has a dense battery of spirocysts, which are larger in the old specimen than

40

TERRA NOVA” EXPEDITION.

in the young one, and in the former are long, large, subcylindrical, and seem rather
more narrow, tapering and pointed at one end than at the other. Nuclei of supporting
cells most numerous in a belt inside the spirocyst-battery. In one specimen granular
gland-cells can be made out superficially, but not in the other, except in the disc. But
in both, there are curious homogeneous gland-cells just outside the nerve-layer ; they
are of various shapes and rather mealy-looking, and either remain grey or absorb eosin.
In the larger specimen the endoderm is a fine open network, with a dense outer crust
containing a few granular gland-cells and a large number of round oil-globules, which
are either large and separate, or smaller and clustered.

(viii) Two specimens examined were female, a third male.

This species is easily to be distinguished from the other three species in the genus,
as will be clear from reference to the plates and descriptions of them given by McMurrich
(1893) and Wassilieff (1908). C. selaginella is characterised by the possession of few
tentacles, which are octamerously arranged, as are also the mesenteries ; by a well-
developed, alveolar to alveolo-reticular, vertically layered sphincter ; and by the distinct
form of the retractor-muscle, and the curious club-moss-like appearance of its processes.

Family SAGARTIIDAE, Gosse.

Actiniina with well-developed base and basilar muscles. Sphincter mesogloeal in
all genera but one, Acontia present. Cinclides present or absent.

Sub-family SAGARTIINAE, Verrill.

Sagartiidae with soft body-wall, possessing cinclides but no cuticle ; sphincter
present, mesogloeal ; more than six pairs of perfect mesenteries, of which the six pairs
of primary mesenteries are fertile, except for the directives in some cases. Suckers
may or may not be present.

ARTEMIDACTIS, n. gen.

Sagartiinae with thin body-wall (in proportion to the size of the animal), without
verrucae or collar. Body cylindrical below, but widely expanded above, so that the
disc greatly exceeds the column, although it is not lobed : margin tentaculate.
Longitudinal muscles of the tentacles ectodermal. Musculature of mesenteries very
weak, forming no distinct pennons ; but borne on such stout lobe-like processes of the
mesogloea that the musculature appears as if reversed, and apparently projects from
endoderm to mesogloea. The mesenterial filament is confined to a certain definite area
on each mesentery, the edge of the mesentery above and below being quite free.
Tentacles very numerous. Cinclides irregularly distributed. Genotype, A. victrix, n. sp.

The only genus near this is Cereus. From Sagartia, etc., the widely expanded
upper part of the body quite separates it. Cereus possesses this expanded part, but
does not resemble Artemidactis in structure. Cereus pedunculatus, the type, is a

ACTINI ARIA— STEPHENSON.

41

strongly muscular perfectly retractile animal, with strong muscle-pennons with processes
of exactly the normal type. Artemidactis seems unable to shorten its body at all, from
the absence of retractors, although the sphincter is strong enough to conceal the
tentacles. Moreover, the restriction of the mesenterial filaments is curious (see below).

Even though the external appearance of this creature when dead may not be
particularly fine, it is enough to give to anyone who has known anemones alive and
healthy a key to the beauty which it would display in a free and living condition : and,
even now, its histology is wonderful almost beyond imagination. The species is named
after Diana, since we have to deal with a large anemone, which is evidently numerous in
its own locality, and therefore presumably successful : and is not a large anemone a
huntress ?

10. Artemidactis victrix, n. sp. (PI. II, figs. 1, 3, 7, 15 ; PI. Ill, figs. 6, 7, 8, 9, 10 ;

PI. V, figs. 3, 5, 6, 7, 8 ; PI. VI, figs. 5, 6, 7, 8, 9, 12.)

A. Eight specimens. Station 294, Ross Sea, 74° 25' S. ; 179° 3'E., 158 fathoms,
Jan. 15, 1913. Agassiz trawl.

B. One specimen. Station 338, entrance to McMurdo Sound, 77° 13' S. ;

164° 18' E., 207 fathoms, Jan. 23, 1912. Agassiz trawl. Bottom — mud.

Measurements, (i) Largest specimen. — Diameter of tentacular crown, 8 1 4 cm.
Length of a large tentacle, 2*1 cm. Diameter of extruded actinopharynx, 5*7 cm.
Height of column, 8 * 8 cm. Diameter of lower part of column, 4‘ 5 cm. ; of pedal disc,
6 • 1 cm.

(ii) Smallest specimen. — Diameter of tentacular crown, 2’ 7 cm. — Total length,
5 • 2 cm.

External characters. — Most of the specimens are very large, several medium, and
one small. A medium-sized one is shown in PI. II, fig. 1. The pedal disc is thin, and
secretes a sort of soft cuticle, formed by accumulated layers of mucus. It is often
contracted and crumpled, with an irregular edge, but when expanded it well exceeds the
column. In one case it was very firmly adherent to a large Tunicate. The column is
more or less cylindrical throughout the lower and greater part, but above it suddenly
dilates into a wide expanded umbrella-like portion, which is generally folded backwards
so that it overlaps and hides the upper part of the column : the disc and puffed out
actinopharynx in turn hide it, taking the form of a dome, so that the tentacles all point
downwards, in a fringe at the edge of the disc. This is rather as in Cereus pedunculatus.
The body is a good deal higher than wide, very much so if the disc be turned right way
out. It does not seem, as far as one can tell from preserved material, as if the animal
had strong enough retractor-muscles to shorten the long body : even in the one specimen
in which the tentacles are retracted, and covered by the strong sphincter, the body is
as long as ever ; and all the other specimens are widely expanded. The wall is very
thin and baggy for such a large animal, and cannot support its own weight out of spirit,
but it is tough. It is smooth, without verrucae. The ectoderm is pale brown, but is

G

VOL. V,

42

“TERRA NOVA” EXPEDITION.

often gone from the surface, which is then pinkish buff in some specimens, cream-
coloured in others. The margin varies in different specimens according to the state of
its expansion. Often there is a very distinct parapet : in one case the margin is a little
inflated and rounded : in some there is a parapet on one side and no parapet on the other.
The part just below the margin has sometimes very inconspicuous low knobs or
tubercles and very indistinct solid or more or less hollow short ridges, but they are of
no importance. The cinclides are fairly numerous, scattered irregularly over the
column, and possibly rather more numerous above than below. Often acontia are
protruded through them. The oral disc greatly exceeds the column in expansion, but
the actinopharynx is so much puffed out in all cases that the free surface of the disc is
quite narrow. It slopes upwards from the tentacles to the edge of the actinopharynx.
If at all inflated it is very thin, and has slight radial furrows running inwards from each
side of the tentacles of the first three cycles. The extruded part of the actinopharynx
is flabby, and forms two great rounded inflated lips, which are strongly and regularly
ridged and furrowed. There are three alternating sets of these ridges, large, medium,
and small — one ridge for every pair of mesenteries in the body : the siphonogiyphes
interfere a little with the exact number. The tentacles have a graceful form, tapering
from a broad base to a blunt or pointed end. They vary in size and in the thickness
and stiffness of their wall, according to the amount of inflation. They are fairly large
actually, but compared with the size of the body quite small. They are finely striated
longitudinally and transversely. As for as I can tell, their ends are imperforate. They
are arranged in four cycles in the smallest specimen— 24 4- 24 + 48 + nearly 96 =
nearly 192. In the larger specimens there are five cycles, in all cases twenty-four
primary ones. In one specimen of which I made an exact count the arrangement was
24 + 24 + 48 + 96 4- 100 = 292. The first four cycles are perfectly regular, but
the fifth is very unevenly distributed ; it is evidently incomplete, and will grow to 192
when perfect. Two quarters of the animal had each twenty-four tentacles of this cycle,
one had thirty, and the other twenty-two.

Structure, (i) Mesenteries — In the large specimen examined, as follows : — Six
cycles — 6 4- 6 + 12 4 24 4-48 + 96 pairs = 192 pairs. The first three cycles are
perfect : the mesenteries of the sixth cycle are small, and are entirely confined to the
lowest part of the body. All cycles bear gonads, filaments, and acontia. Small
internal stomata are present. The first two cycles are attached to the actinopharynx
throughout its length, the third does not run so far down, the fourth does not reach it
at all. The filaments and gonads are strictly confined to a certain small part of each
mesentery, and the edge of the mesentery above and below that part is absolutely
unbound by even the simplest strand of filament. The filaments of the perfect
mesenteries form a large knot on the edge of the mesentery just after it leaves the
actinopharynx, the short space between the knot and the edge of the gullet being
bound by a simple bit of filament, which in the first and second cycles is continuous
with a ridge of pharynx. Below this single knot there is no gonad or filament. In the

ACTINIARIA — STEPHENSON.

43

imperfect mesentery there is a short knotted filament and a row of gonads, above and
below which there is nothing. But it is a curious fact that the mesenteries of each
cycle have their knots of filament at a different level (the same for each member of the
cycle), so that the various knots do not interfere with each other. Those of the sixth
cycle are near the base, of the fifth half-way between base and lower end of actino-
pharynx, of the fourth partly above, partly below, the level of the bottom of the
ac-tinopharynx, of the perfect mesenteries between the zones of the fourth and fifth.
This is diagrammatically represented in PI. VI, fig. 12, which shows a small piece of
wall, with some mesenteries of each cycle.* The thickened portion of each mesentery
represents the filament and gonad, the line (E the base of the actinopharynx, and
everything which in reality would be hidden by the latter is given in broken lines.
In a smaller specimen examined the various notes given above hold good, but the zones
of filament are rather more diffuse, and the zones of the fourth and fifth cycles more or
less coincident, though that of the sixth is quite distinct. Now as to the musculature.
A whole mesentery of the third cycle is shown in PI. II, fig. 3. It will be seen that
the greater part of the muscular portion of the mesentery (/.<?., the part between gonad
and body- wall) is very thin — indeed, it is a mere flimsy, transparent membrane— but
that where it joins the body-wall it is rather thicker. This broader basal part bears
on each side an approximately equal development of very stout, lobe-like processes of
the mesogloea, which have short stout primary and secondary branches, and are fringed
by the muscle-fibres. They are generally rather larger on the longitudinal than on the
parieto-basilar face of the mesentery. The main thin part between this and the gonad
has practically no muscle, but just behind the gonad a few very low and inconspicuous
rounded processes arise on the longitudinal side, but make no approach to a pennon.
In sections taken through the part of a mesentery which has no filament or gonad, the
thin portion ends in a slight spoon-shaped swelling with short stout processes all round
the edge of it. In the mesenteries of all cycles the general arrangement is as in
the case figured, the basal part near the body-wall, bearing the musculature, varying
in extent according to cycle, but the processes are always very stout. The processes
from one side of this part of a directive mesentery are shown in PI. V, fig. 8, and it
will be seen that they are similar to those of Leptoteichus, so much mesogloea being
present that the processes seem reversed, and appear to project from endoderm to
mesogloea. The letter m is in the mesogloea. In sections at a low level, the fibres of
the parieto-basilar muscle run transversely of course. The parieto-basilar muscle is
nowhere better developed than the basal part of the longitudinal muscle. In the
largest mesenteries of all there is rather more attempt at a pennon than in PI. II,
fig. 3, the processes behind the gonad being a little larger and more numerous, but
never do they reach the size of those on each side of the mesentery where it leaves the
wall, and so no real pennon exists. The specimen sectionised is a female.

* It must be understood that this diagram shows the various details with considerably more
regularity than does the actual specimen.

G 2

44

“TERRA NOVA” EXPEDITION.

(ii) Endoderm. — The mesenterial filaments deserve special description in this
species, as they are very beautiful (see PI. II, fig. 3). If a simple filament be sectionised
it represents, of course, the form of a single trefoil. The glandular lobe is transversely
oval, with a very narrow stem connecting it to the other parts, and on each side of this
stem is a kidney -shaped, curved, ciliated lobe, the two very near together. Proximal
to each ciliated lobe is a delicate patch of network, which projects a little behind the
trefoil on each side. The mesogloea of the mesentery sends a narrow strand into the
back of the filament, which has a narrow branch to each lobe, and in the “pelvis”
of each “ kidney ” this forms a small expansion. This, of course, is mostly only the
typical structure of many Actinian filaments, but here the form, regular development,
and colour when stained give it greater clearness than usual. The nuclei of the
ciliated lobes are concentrated in their inner parts. The glandular lobe is evenly
crowded with fairly large elongate granular gland-cells. When a compound filament
is cut across, in many cases the individual trefoils remain distinct, and form beautiful
bunches, such as are shown in fig. 3. In most species they lose all form, and the
compound filament becomes a shapeless mass. In this species, when the trefoils do
lose their form, a splendid folding of the ciliated parts becomes visible ; this is not
confined to this species, but is here better and more regularly developed than usual.
A little of it is shown in PI. II, fig. 7. This figure demonstrates the regular alternation
of long lozenges of clear tissue, bounded by pillars of cells stained pale pink or lilac,
and the pattern which the stain forms (in all cases the same) in the more deeply
stained outer part. The endoderm of the body- wall is uninteresting : it is high, and
contains very few gland-cells. That of the thin part of the mesenteries is low, with a few
granular gland-cells. But that of the gonads and the part of the mesentery proximal
to the filament-trefoils is high and beautifully developed. It is arranged in alternating
low hills and valleys, and in many places is folded. It has throughout the character
of a fine and rather regular close network, only clearly seen under high powers, and
quite different from the network-lobes of the filaments. There are very numerous
granular gland-cells which are large, and though various in form are mainly elongate
and club-shaped (see PL II, fig. 3), and have a tendency to be radially arranged in the
ridges and hillocks. Often ducts are traceable in such positions. The endoderm contains,
here and there, curious amoeboid cells : these are only slightly stained and mostly
dimly visible against the reticulate endoderm background, but I have managed to see
two or three clearly enough to figure them (PI. VI, figs. 7, 8, 9). In colour they are
grey, often with vacuoles and a mealy appearance or a few granules. They are really
almost exactly like the cells of the mesogloea, and it seems quite possible that they are
food-transmitting cells such as occur in the endoderm of Alcyonium. A few odd scraps
of gland-cells from this endoderm are figured on PL VI, figs. 5, G.

(iii) Mesogloea. — The cell-plexus here is particularly well developed, more so in
the mesenteries than elsewhere. The matrix is almost homogeneous in the mesenteries,
more fibrous in the body -wall. The cells are unusually large and numerous : there are

ACTINIARIA— STEPHENSON.

45

two sorts — small transparent bipolar ones, and much larger, irregular, mealy grey ones.
Many have very dim outlines. Some are rounded or oval, very irregularly, some
bipolar, most are very amoeboid, and they lie in strings and bunches and seem all
connected by fine processes. A. few examples are shown in PI. II, fig. 15 ; PI. Ill,
figs. 6, 7, 8, 9, 10. Often no nuclei are distinguishable : sometimes there are round
vacuoles. Some are almost granular. In the body- wall many are distinctly granular.

(iv) Ectoderm of body -wall. — High, well developed, with a smooth surface ; rather
higher than endoderm, where well developed, but neither as high as the mesogloea.
With a distinct and dense superficial narrow zone of more or less oval granular
gland-cells, whose “ ducts ” can often be traced inwards to the mesogloea. The inner
part is of a firm consistency.

(v) Actinopharynx. — With two large and well-marked siphonogiyphes in all cases.

(vi) Acontia. — The shape in transverse section of one of these is shown in PI. V,
fio-. 3. There is a three-lobed core of mesogloea ; the smaller of the two main
endoderm-divisions of the acontium is unspecialised ; but the large kidney-shaped
portion is crowded with gland-cells and cnidae. The gland-cells are granular and
elongate, and penetrate the whole thickness of the endoderm. The cnidae are
enormous. I have not observed such large ones in any part of any anemone except in
acontia and acrorhagi, to which they seem restricted. They are long and canoe-like,
blunt at both ends : there is a structure running along one side or down the middle
which looks like a keel, and which is probably the narrow spiral thread contained by
the cell, the individual coils not being visible : there is usually a twist, more marked
in some cases than in others, near one end. A rather similar but smaller and less
twisted cnida, from the acontium of another species, is figured in PI. Ill, fig. 18.

(vii) Tentacles. — Ectoderm very high, much thicker than mesogloea or endoderm.
Endoderm very narrow, folded ; endodermal circular muscle feeble ; endoderm dense
externally, almost hollow below, with a few large granular gland-cells superficially, and
here and there a keeled twisted cnida) similar in form to those of the acontia, but
much smaller, and about the size of an ectodermal spirocyst. The longitudinal muscle
is ectodermal, with a rather “ reversed ” appearance, in a dense layer about as thick as
the mesogloea. The processes are very high and much branched, and although there is
a fair amount of mesogloea in them, they are not nearly as stout as those of the
mesenteries and body-wall circular muscle (which has processes as stout as those of the
mesenteries). They are so numerous and closely packed that it is impossible to draw
individuals : I have given the general appearance under a low power in PI. Y, fig. 7.
Mesogloea a little thicker than endoderm. Ectoderm very regularly developed, deeply
folded : nerve-layer narrow, distinct. There are gland-cells and spirocysts throughout
the ectoderm more or less, but they are by far the most numerous superficially, where
they form a dense distinct battery : spirocysts long and broad, subcylindrical, tapering
a little more gradually internally than externally. The deeply embedded gland-cells are
mostly diffuse and poor : but among the spirocyst-battery they are large and densely

46

“TERRA NOVA” EXPEDITION.

granular, of various shapes, often with a duct inwards as well as outwards. Here and
there occurs a thick-walled cnida such as is mentioned above for the endoderm. The
oral disc has a similar structure to the tentacles.

(viii) Sphincter. — At the very top this almost fills the mesogloea, leaving only a
very narrow band between itself and the ectoderm on one side and the endoderm <»n
the other. These two bands gradually widen below, and the sphincter thus tapers off
and its narrow lower part lies in the middle of the mesogloea (PI. \ , fig. 5). A small
part of its exact structure much enlarged is shown in PI. V, fig. 6. It i- neither reticular
nor alveolar ; the processes are arranged in clumps of various sizes (shown as black
blotches), and these in turn in larger clumps. These larger clumps tend to be arranged
in transverse rows above, longitudinal ones below, but there is no distinct layering.
The clumps are very irregular for the most part.

This species seems to me extraordinarily beautiful and interesting, and at any rate
there should be no difficulty about its identification.

SAGARTIA, Gosse.

Sagartiinae wTitli the body not widely expanded above: surface smooth or with
small suckers: margin tentaeulate, not lobed: base not specialised for clasping
zoophytes : cinclides irregularly distributed.

11. Sagartia, sp. (PI. I, fig. 12 ; PI. II, fig. 19 ; PI. Ill, fig. 18 ; PI. V, figs. 12,
13, 15, 17.)

One specimen. Station 134, Spirits Bay, near North Cape, New Zealand, 11-20
fathoms, Aug. 31, 1911. Dredge. Bottom — shelly.

Measurements. — Diameter of oral disc and tentacles, l'l cm. ; of column, • 7 cm. ;
of pedal disc, '9 cm. Height of column, ’5 cm.

External characters.—- Colour wholly dull yellowish white in formalin. Pedal disc
exceeding the column, almost circular in outline, rather thin and flaccid, semi-
transparent, nearly smooth though a little irregularly radiately wrinkled and minutely
corrugated. No cuticle. Column a little wider than high, irregular in shape, broad
below, sloping inwards to a little below the margin, where there is a slight constriction,
and then becoming a little broader again so that the oral disc is really as wide as the
pedal. Wall thin but strong and tough, semi-transparent so that the insertions of the
mesenteries show through in places. Margin tentaeulate, with a very slight rim.
Surface smooth but for a few irregular longitudinal wrinkles and numerous fine
transverse striations. No verrucae, suckers, or visible cinclides. Oral disc small,
cup-like, smooth, thin, flaccid and semi-transparent. Mouth slightly open with very
distinctly tubercled lips. Tentacles rather short, conical, broad at the base, tapering to
a sharp point, stiff enough to support themselves out of spirit, finely and regularly
transversely striated. Inner much larger than outer on the whole, but irregularity
of size occurs on account of different stages of contraction. Their arrangement is rather

o O

ACTINIARIA— STEPHENSON.

47

irregular. The first three cycles are 6 + 6 + 12, and the total is about 80. The
arrangement is something like 6 + 6 + 12 + 24 + 40 = 88.

Structure, (i) Mesenteries. — These are very irregularly arranged. Two pairs are
directives. There are twenty-three perfect mesenteries arranged without any system
(the animal is not bilaterally symmetrical), there are five large imperfect mesenteries,
and the rest are very small, and occur singly or in pairs or triplets here and there.
Total number about fifty-five mesenteries, in three extremely uneven cycles. Those of
the first two bear pennons, which vary enormously in size and shape in individual cases.
The third cycle mesenteries have no pennons. The pennons occupy only about one-third
of the muscular surface of the mesentery, and are often very narrow. A large typical
one is figured on PI. V, fig. 17 : it has had to be partially restored, as my sections of this
species are poor and broken. As a rule the pennons are diffuse at the level of the
actinopharynx, and end rather abruptly on the side nearest the body-wall, tapering otf
gradually in the other direction. Below the actinopharynx level they are various, but
mostly broader and shorter, almost circumscribed. One pair of directives has better
pennons than any of the other mesenteries, ending abruptly at both extremities, but
the other pair has them reduced to practically nothing, so that they cannot really be
called pennons. The processes are numerous, rather coarse, crowded, a good deal
branched, the branches mainly short. The parieto-basilar muscle is very weak, forming
no fold as far as I can see. No gonads.

(ii) Sphincter. — This is well developed. It lies right against the endoderm, but is
throughout separated from the ectoderm by a narrow mesogloeal band, filling nearly
all the mesogloea above. It is long and rather narrow, broadest a little below the top,
tapering off above, but much more gradually below (PI. V, fig. 12). The body-wall is
wrinkled, and the sphincter, of course, follows the wrinkles. As it tapers, so does the
mesogloea. Its structure is coarsely reticular, not layered, and I have drawn part of it
as accurately as I can from poor sections in PL V, fig. 13.

(iii) Endoderm. — In simple filaments, the ciliated lobes are well developed and
bent back on themselves, and the glandular lobe has a fairly stout stem : in more
complex filaments the ciliated lobes are regularly folded. The glandular lobes contain
fairly elongate finely granular gland-cells and two sorts of thick walled cnidae, all
small ; some as in the body- wall, some much shorter and a little stouter so as to be
almost oval. Endoderm of mesenteries and body- wall ordinary, with here and there
a gland-cell, which is not visibly granular.

(iv) Acontia. — These have various shapes in transverse section. Most typically,
they are a rounded pear-shape, the cnidae crowded in the rounded broad end, the gland-
cells at the back on each side of the stem. The gland-cells are unusual : they are large
and of a fairly regular oval-oblong, not granular, and they stain a curious bright blue
with haematoxylin, with a black spot at the open end, which often has two horns.
Some even have a short neck, so that they look like flasks. One of these is shown in
PI. II, fig. 19. I take them to be mucus-cells : perhaps the black spot is escaping

48

“TERRA NOVA” EXPEDITION.

mucus. The cnidae are very large, long, broad and thick walled, keeled, with blunt
ends, straight or curved, or with a slight twist (PL III, fig. 18).

(v) Mesogloea. — Thin and rather fibrous, with numerous small cells, many of them
lying in clear spaces.

(vi) Actinopharynx. — I cannot give certain data about siphonoglyphes. If any, there
are two. The ectoderm has numerous medium-sized deeply embedded and superficial
granular gland-cells, and a few small thick walled cnidae like those of the body-wall.

(vii) Ectoderm of body-wall.- — Higher than endoderm. With numerous large,
elongate, irregularly shaped gland-cells, many of which run from one surface to the
other, and which, although fairly deeply stained with haematoxylin, are not visibly
granular. There are also a few small, short, blunt-ended, thick walled cnidae, narrow
but not needle-like.

(viii) Tentacles. — Ectoderm higher than mesogloea and endoderm together.
Longitudinal musculature ectodermal, on short simple or slightly branched processes
(PI. V, fig. 15). Nerve-layer well developed, fairly broad. The ectoderm contains a
few thick walled cnidae like those of the body-wall, but rather larger and longer, though
not nearly as large as those of the acontia. Here and there occur larger processes of
longitudinal musculature on hills of mesogloea, so that a tufted appearance is produced.
Spirocysts numerous in ectoderm, fairly broad, not large. There are scattered gland-
cells, mainly superficial, not visibly granular.

I have not attempted to identify this Sagartia as regards its species : it is practically
impossible to identify a small Sagartia on anatomical characters, from the very poor
descriptions at present existing. It is very rarely that a description is sufficient for
identification in the genus, as details and plates are necessary. McMurrich’s descrip-
tions are certainly an exception, especially in the case of 8. spongicola, for instance, as
of course are a few others.

Sub-Family AIPTASIINAE, Simon.

Sagartiidae in which the sphincter-muscle is either absent, or, if present, weak,
diffuse, endodermal. Not more than six pairs of mesenteries are perfect, and these are
often fertile. Cinclides present.

AIPTASIA, Gosse.

With the characters of the Sub-family. Tentacles usually long. Cinclides usually
in one or more horizontal rings round the middle of the body.

If Aiptasia couclii, Gosse, should prove to be provided with a mesogloeal
sphincter the name Aiptasioides* will probably lapse ; and that species, with
Aiptasioides prima and A. pallida , may be included in Aiptasia. Another generic name
may then have to be introduced for species (at present placed in Aiptasia ) with
the sphincter absent or endodermal.

* See below.

ACTINIAE I A— STEPHENSON.

49

12. Aiptcisia minima, n. sp. (PL I, figs. 2, 3 ; PI. VI, fig. 1.)

Fourteen small specimens, adherent to small pieces of stone. Bay of Islands, New
Zealand, Sept. 1912. Rock-pool.

Measurements, (i) Largest specimen. — Diameter of column, ■ 7 cm.; of pedal
disc, 1 • 1 cm. ; of oral disc and tentacles, • 7 cm. Length of a tentacle, * 4 cm.
Height of column, *4 cm.

(ii) Smallest specimen. — Diameter of pedal disc, about ‘15 cm.

External characters. — Two specimens, enlarged, are shown on PI. I, figs. 2, 3,
in different states of expansion. Pedal disc broad, much exceeding the column :
smooth, with traces of mucus but no cuticle upon it. Its outline is regular in some
specimens, irregular in others, in the largest case so much so that it is practically
lacerate. Column a low cone, broader than high, with coarse ring-like transverse
contraction-folds and fine longitudinal striations. Margin a distinct smooth parapet,
not notched, and with no acrorhagi, enclosing a deep fosse. Oral disc small, thin :
mouth with prominent ribbed lips, inner tentacles near mouth. Tentacles thin- walled :
in most specimens they are only partially protruded and not properly inflated, though
in none are they completely retracted. In one specimen, however, they are fully
expanded, and in this case they are very large and long for the animal’s size, and
completely hide the column, spreading in all directions. They are stout at the base,
tapering gradually to a point : they are flaccid out of spirit : the inner are much
larger than the outer. In this fully expanded specimen the column is rather less
conical than usual, naturally ; but even then the base well exceeds the oral disc. The
animal is so small that it is impossible to determine exactly the number and arrangement
of the tentacles : there are about fifty, in four or five cycles, as far as I can make out.

Structure, (i) Mesenteries. — I am sorry that I am unable to give more figures of
the anatomy of this species. It is a difficult one to sectionise, and the processes of the
muscle-pennon are not individually clear in any of my preparations, so I can only
indicate the shape of a mesentery generally (PI. VI, fig. 1). Moreover, there is no
sphincter, and the musculature of the tentacles is so excessively weak that it is hardly
visible. In the case examined in detail the mesenteries are arranged in three cycles,
not quite regularly. There are six large primary pairs, including two pairs of directives ;
these are all perfect, and all bear pennons, which are proportionately larger at the
level of the actinopharynx than below that level. Of the six primary exocoels, four
are normal, each containing a pair of imperfect secondary mesenteries, with a small pair
of the third cycle on each side of it : but each of the other two exocoels (and these two
are adjacent) contains two secondary imperfect pairs, and consequently three small
third-cycle pairs. So the arrangement comes to be 6 p. + 8 p. + 14 p. = 28 p., the
first cycle alone perfect, the first two cycles fertile. The mesenteries of the second cycle
bear pennons, but in all but two cases these are very feebly developed, and found on
the mesentery only at a certain level. The third-cycle mesenteries bear no pennons,
but some of them have simple filaments. The mesogloea of the mesenteries is very

VOL. V.

H

50

“TERRA NOYA” EXPEDITION.

thin, but the endoderm is unusually thick and regular, giving them a very characteristic
appearance. On the primary mesenteries the small pennon lies immediately behind
the gonad ; it only occupies a small part of the muscular surface of the mesenteries,
and is circumscribed and fan-shaped. The outline of it is shown in the muscular part
of a primary mesentery in PI. VI, fig. 1. The processes are few, high, a good deal
branched, and radiate rather as in a fan. The pennons of the directives are a little
larger than the others. The rest of the muscular surface of the mesentery has a feeble
fringe of low processes of the longitudinal muscle on one side, and of the very feeble
unspecialised parieto-basilar muscle on the other : this latter ends in no fold or
prominence of any sort. The specimen is a male.

(ii) Sphincter. — Quite absent. The general endodermal circular muscle of the
body-wall is fairly developed.

(iii) Endoderm. — That of the body-wall lower than that of the mesenteries, which
is a network in its inner part, rather denser in its outer, and contains a fair number of
medium and large, finely granular gland-cells which are variously shaped but mostly
deeply embedded and elongate. The body-wall endoderm is not interesting. The
glandular lobes of the filaments contain large elongate coarsely granular gland-cells,
and the nuclei of the supporting cells are mainly concentrated in two belts at the back,
one on each side of the stem. The glandular lobe is almost sessile ; the ciliated lobes
are well developed, with the nuclei thickest in the inner part, and they tend to become
folded in compound filaments.

(iv) Body-wall.— Mesogloea very thin, endoderm thicker, ectoderm thickest, very
well developed. Ectoderm of base lower than that of column, practically devoid of
gland-cells. Ectoderm thinnest in fosse. The ectoderm contains numerous gland-cells,
near the surface or fairly deeply embedded, with ducts to the exterior : they are hardly
granular. There are also, superficially, two kinds of thick walled cnidae, which are
shorter than those of the tentacles. One kind is short and very broad, blunt at the
ends — a long oval. The others are narrower and more cylindrical, similar to those of
the tentacles, but shorter, often with a twist. The inner part of the ectoderm contains
what may be small clear gland-cells.

(v) Actinopharynx. — Two siphonoglyphes. Ectoderm with many large and often
elongate granular gland-cells, and seemingly also crowded with small ones. It has also
thick walled cnidae, as large as the largest in the tentacles.

(vi) Acontia. — Oval in transverse section. They contain gland-cells and enormous
thick walled cnidae, far larger than any others in the body, which are boat-shaped,
keeled, with square ends and a hardly perceptible twist.

(vii) Mesogloea very thin, slightly fibrous, containing numerous small cells, often
in clear spaces. Too small to make out details.

(viii) Tentacles. — Mesogloea thin, endoderm thicker, ectoderm thickest, but not as
high as that of the body-wall. The endoderm contains zooxanthellae. The longitudinal
musculature is excessively feeble, the short processes appearing as mere dots. In the

ACTIXIARIA— STEPHENSON.

51

ectoderm the nuclei of the supporting cells are scattered : its inner part contains

zooxanthellae, its outer part spirocysts and cnidae. The spirocysts are fairly numerous,
of moderate length, stout, and, as far as I can see, rounded at the outer end and pointed
at the inner. The thick walled cnidae are of two kinds : some are rather short, very
narrow and needle-like : others are more various in length, the largest being as long as
the ectoderm is thick. These latter are thick and blunt at both ends, and many are
keeled and twisted.

Sub-family METRIDIINAE, Carlgren.

Sagartiidae with six (rarely a few more) pairs of perfect mesenteries, which are
sterile ; one or more siphonoglyphes ; body -wall relatively thin, without a cuticle ;
cinclides present. Sphincter mesogloeal.

AIPTASIOIDES, n. gen.

Metridiinae with thin body- wall and very feeble mesogloeal sphincter. Only six pairs
of mesenteries perfect. No collar. Disc not lobed. Base not specialised. Cinclides
in a narrow belt about half-way up the body ? Genotype, A. Prima, n. sp.

It seems necessary to establish a new genus for forms which resemble Aiptasia,
but possess a mesogloeal sphincter. Pax (49, p. 198) has limited the genus Aiptasia
to forms possessing no sphincter ; but of course A. mutabilis, if it does possess an
endodermal one, must be included. But when the sphincter is mesogloeal, that is
another matter, and forms possessing it can hardly be included in Aiptasia. An
endodermal sphincter is only the concentration of the ordinary circular muscle of the
body-wall, but a mesogloeal one is a more specialised structure. The new genus
Aiptasioides will include A. pallida and the species described below, whose cinclides I
cannot, unfortunately, exactly locate.

13. Aiptasioides prima , n. sp. (the Jirst species described under the name
Aiptasioides). (PL I, fig. 9 ; PI. V, figs. 11, 14, 16 ; PI. VI, fig. 11.)

Pour specimens. Station 36, shore collection, S. Trinidad Island, July 26-30,
1910.

Measurements, (i) Largest specimen. — Diameter of tentacular crown, • 9 cm. ; of
column, 1 • 2 cm. ; of pedal disc, 1 • 7 cm. Height of column, 1 ’ 4 cm.

(ii) Smallest specimen. — Height, 1 • 2 cm.

External characters. — This is a very flimsy and brittle species, and it is almost
impossible to obtain good preparations. I was unfortunately obliged to destroy two
and a half specimens in order to obtain any data at all.

Pedal disc exceeding column, very irregular and lacerate in outline, not very thin,
with bits of adherent brown cuticle. Column broader below than above, its wall
extremely thin and delicate, but even more so above than below. A circular
constriction (a temporary feature of the only sound individual remaining) divides the
column more or less into scapus and capitulum. There is no cuticle. The lower

H 2

52

“.TERRA NOVA” EXPEDITION.

part of the scapus is finely striated transversely. The wall of the capitulum is
excessively thin and flabby and semi-transparent, the mesenterial insertions showing
through clearly. There is no distinct margin. Colour absent in formalin. About the
top of the scapus are two irregular horizontal rows of small vesicles (clearly seen in PI. I,
fig. 9) : these look at first like verrucae or suckers, they are so small ; but they are merely
little thin-walled outpushings of the coelenteron, and in sections it is seen that they have
exactly the same thickness and structure as the rest of the wall. Some are perforated,
but the perforations seem to me to be due to damage. It is, of course, quite possible
that the cinclides, though they cannot be seen, are arranged on these little vesicles. I
think living specimens would be necessary to make certain. Disc small, thin, concave,
radially furrowed. Mouth with a smooth circular inconspicuous lip. Tentacles small,
short, slender, pointed, all about the same size. The formula is 6, 6, 12, 24, 48 = 96 :
in some specimens the last cycle incomplete, even in the largest some tentacles have
been removed from it by damage in some way.

Structure, (i) Mesenteries. — The coelenteron is a little decayed or diseased, and
here and there occur bundles of fibres slightly suggesting Sporozoan parasites, but this
is by no means certain. There are three cycles of mesenteries, the primaries alone (six
pairs) perfect. Two pairs of directives. The first two cycles bear pennons, the third
cycle mesenteries are very small, and do not. The third cycle is not complete. There
are no gonads developed. The pennons of the primary mesenteries arc broad and well
developed (PI. V, fig. 11); they fill more than half the mesentery at the level of
the actinopharynx, and have both ends rather abrupt. The processes arc fairly
numerous, much branched, and high and fine ; they are less crowded than usual. The
pennons vary in shape considerably in different mesenteries, often ending abruptly at
.both extremities, sometimes less so, sometimes even tapering. The parieto-basilar
muscle is feebly developed and forms no fold, and its processes (and those of the
opposite side of the mesentery just where it leaves the body-wall) are short and simple
or slightly branched.

(ii) Sphincter. — A very weak alveolar sphincter is present in the mesogloea. It
fills the whole of the excessively thin mesogloea, which is very slightly expanded at
the margin. It consists of a practically single row of large compressed alveoli fringed
more or less with fibres in transverse section, and extends a good way down the wall.
I have figured its upper part as accurately as possible in PI. V, fig. 14. The “ alveoli ”
in this case may conceivably be mere splits, in which case of course the sphincter could
not be described as “ alveolar,” but they look genuine.

(iii) Endoderm.- — That of the body-wall is fairly high, and is crowded with large
zooxanthellae. That of the mesenteries is also high and full of zooxanthellae, is loose
in texture, and here and there are also fairly large irregular granular gland-cells, which,
too, are present in the glandular lobes of the filaments, where the latter can be
identified. Circular muscle of the wall rather high and well developed, on branching
processes.

ACTINIARI A — STEPHEN SON.

53

(iv) Acontia. — Only degenerate remains of these and mesenterial filaments are
present. They contain large pen-like cnidae, thick walled, long and fairly wide,
straight or slightly twisted.

O O J

(v) Mesogloea. — So exceedingly thin in the body-wall that its structure cannot be
seen. It is faintly thicker at the origins of mesenteries and at the margin, and here it
is fibrous, and contains a fair number of cells, all of which are small.

(vi) Ectoderm of body -wall. — This is high in the lower part of the column,
but low in the upper part. It is loose in structure, and contains sparsely small,
short, blunt-ended, boat-shaped, thick-walled cnidae. It seems devoid of gland-cells.

(vii) Actinopharynx. — Siphonoglyphes ? Ectoderm firmer than that of the body-
wall, with numerous large coarsely granular gland-cells of many forms, but chiefly
club-like.

(viii) Tentacles. — Endoderm thicker than mesogloea : ectoderm thicker still.

Longitudinal musculature ectodermal, its processes very short and stumpy, simple
or a little divided (PI. V, fig. 16). Endoderm crowded with large zooxanthellae : the
structure of two of these is shown in PI. VI, fig. 11. Each contains what I take

to be a nucleus, a pyrenoid, and sometimes vacuoles. Ectoderm uot very well

preserved. Nuclei of supporting cells scattered. Rather short stout spirocysts of
moderate size are numerous : there are also a few thick- walled cnidae which are longer
and narrower (proportionately) than those of the body-wall, but far smaller than

those of the acontia. Small bits of scattered granular gland-cells are present. In

the oral disc the mesogloea is thicker than elsewhere ; the radial muscles are
ectodermal, and better developed than the longitudinal muscles of the tentacles.
The circular muscle is well developed, on short, stout, branched processes.

As far as I know, this species is not identical with any species previously
described under Aiptasia : of course it is distinct from all forms with the sphincter
endodermal or absent, but I have not been able to obtain notes of quite every species in
the genus. The same applies to the last described, Aiptasia minima, but that is
even less likely to be known.

CALLIACTIS, Verrill.

Metridiinae with smooth body-wall, except for the cinclidal tubercles in some
species. Cinclides in horizontal rows at the base of the column. Six pairs of perfect
mesenteries. No collar. Disc not lobed, though sometimes a little undulated in a
living animal. Usually commensal with hermit-crabs.

14. Calliactis reticulata, n. sp. (PL I, fig. 15 ; PI. V, figs. 9, 10 ; PI. VI, fig. 2.)

I have called this species reticulata from the appearance of the surface of the
body-wall.

Nine specimens. Station 42, 22° 56' S. ; 41° 34' W., 40 fathoms, May 2, 1913.
Agassiz trawl. Bottom — sandy.

54

“TERRA NOVA” EXPEDITION.

Measurements, (i) Largest specimen. — Diameter of tentacular crown, 1 1 4 cm.
Height of column, 2 cm. Diameter of pedal disc, 1 • 6 cm. ; of column, 1 * 7 cm.

(ii) Smallest specimen. — Height, 1*6 cm.

External characters. — With two exceptions the specimens were not attached to
anything : those two, one large and one small, were attached to a Gastropod shell
inhabited by a hermit-crab, and are showm, with the shell, in PI. I, fig. 15 : the crab is
practically hidden. In formalin, the colour is wholly dull greyish white, with a slight
blue tinge. Base well exceeding the column, with a thin and well marked edge, and
an entire outline, very little undulated. No trace of cuticle on the base, which is
smooth and thin, faintly wrinkled, especially radially. The column is not far from
cylindrical, being slightly barrel-shaped, considerably exceeded by the base, and slightly
by the oral disc. It is a very little narrower just below where the disc expands, and
becomes a little wider from here down, expanding more suddenly at the base. The wall
is thin, but tough and self-supporting. The surface has no verrucae, etc., but has a
very characteristic reticulate appearance. It is divided up into little flat, low,
rectangular and rounded or irregular papillae of various sizes, with slight furrows
between them. These are not mere areas isolated by longitudinal and transverse con-
traction-wrinkles, as in many species, but really papillae ; not verrucae or suckers or
tubercles, howrever. They are very fine near the limbus, most irregular at the margin,
which is tentaculate with a slight indistinct rim. The arrangement of cinclides is not
clear in many cases, though in all they are confined to a zone just above the limbus. In
the best specimen, they are in three circles : the upper circle contains six large cinclides,
the next four large cinclides, and the lowest one about seventeen smaller ones. The
three rows alternate irregularly with one another. In some specimens the two upper
circles appear more like one irregular circle. The number does not seem to be constant.
The cinclides are not mounted on tubercles or elevations of any sort, but appear merely
as small brown dots, flush with the surface of the column. The column-'4 papillae ” do
not show to any extent in the figure, as a glass is needed to see them clearly. The oral
disc slightly exceeds the column, but is almost entirely covered with tentacles, so that
the free area round the mouth is small. In all cases the inner tentacles are inturned
over the disc. Lips prominently tubercled by the apices of the oesophageal ridges.
Tentacles very small, short, conical, pointed, stiff, crowded, the inner a good deal larger
than the outer : they are rather like teeth. The tentacles are too small to be exactly
enumerated. But the smallest has about 100, the largest about 170, two others about
140-150. Arrangement probably 12, 12, 24, 48 + (48 to 96) as the animal grows,
reaching finally 192.

Structure, (i) Mesenteries. — The primary mesenteries have large internal stomata.
I can find no external ones. There are three cycles of mesenteries in small specimens,
four cycles in older ones. Only the six primary pairs are perfect. All bear filaments. In
one specimen the third and fourth cycles alone were fertile ; in another the second and
third cycles were fertile, the fourth cycle being too small to see. In all examined the

ACTIXIARIA— STEPHENSON.

55

primaries are sterile. The mesenteries are very thin, delicate, transparent membranes,
very hard to sectionise well. Those of the last cycle are very small, and bear no
pennons. They have a rather stout base, where they leave the body-wall, with a few
short stout processes, about the same on each side, and approaching almost, on a far
smaller scale, those of Leptoteichus : some are a little branched. The mesenteries of
the higher cycles all have a similar base, but they have pennons as well. The pennons
are diffuse, and taper off at both ends. They are low, and have numerous rather stout
processes with short rather stout branches. These cannot well be drawn individually,
so I have given the general appearance of one pair of directives in PI. VI, fig. 2. The
other large mesenteries are much larger, and so the pennon occupies less of the surface.
It lies about the middle of the muscular part in the first and second cycles, nearer the
body-wall in the third. The parieto-basilar muscle is feeble and forms no fold, its
processes nearly resembling those of the basal part of the longitudinal muscle.

(ii) Sphincter. — This is well developed. It is not very long, and though it nearly
fills the mesogloea above, it is separated by a narrow band from the endoderm, which
increases slightly and gradually in width lower down (PI. V, fig. 9). At the top it is
right up against the ectoderm, and gradually separates from it below, tailing off to a
fine point lying nearer endodern than ectoderm. Its shape is irregular, as it follows
more or less the ectodermal side of the mesogloea, which is wrinkled. In some sections
it is broadest at the top ; in some, about the middle. Its structure resembles that of
Artemidactis victrix in its detail (see PI. Y, fig. 6), the muscle-fibres being arranged in
groups, which are of various shapes and sizes, and larger above than below. The
bundles are arranged in more or less distinct transverse rows, and are densest in the
upper part.

(iii) Endoderm.— That of the body- wall has here and there a fair-sized granular
gland-cell. That of the mesenteries is thin, and has a few such gland-cells. The
filament-trefoils have a characteristic shape, being much broader than long. The
ciliated lobes are very large, and are not much exceeded by the glandular lobe, which
has a narrow stem and is fig-shaped. The ciliated lobes are sickle-shaped, their nuclei
chiefly concentrated in a dark zone in the inner part towards the tip of the sickle.
Gland-cells occur in them occasionally towards the base. In the compound filaments
they become beautifully folded, and the glandular lobes are better developed, rounded,
with many large coarsely granular gland-cells, and I think small, narrow, thick walled
cnidae. Endodermal circular muscle of wall well developed, on short, stout, slightly
branched processes.

(iv) Mesogloea. — Clear and homogeneous, with numerous cells of moderate size ;
these are very plentiful near the ectoderm at the margin, but in the disc they are
simply packed, so numerous are they. They are fairly uniform in appearance, having
large mealy nuclei surrounded by narrow clear protoplasm with fine processes. Shapes
very various : bipolar, tripolar, multipolar, or in chains ; some round, without
processes.

56

“TERRA NOVA” EXPEDITION.

(y) Ectoderm of body-wall. — Not nearly as thick as the mesogloea. It is firm,
and has numerous large, irregular, elongate, narrow, granular gland-cells, many of
which extend from base to surface. In the inner part there are also numerous little
cells which seem about to migrate into the mesogloea, and here and there one seems to
be in the very act of doing it.

(vi) Act inopl larynx. — Descending about half-way down the coelenteron, ribbed, with
two siplionoglyphes. Ectoderm high and well developed, with many gland-cells with
very large granules, of various shapes, many elongate, some deeply embedded, and some
superficial. There is a dense band of small bodies in the inner part which may be
nuclei of the supporting cells, but look more like small cells of some other kiud.

(vii) Acontia. — Kidney-shaped in transverse section. Front part with cnidae and
regularly scattered nuclei, back part with a concentration of granular gland-cells. I
cannot give details of the cnidae, except that they are much smaller than in the acontia
of any other species with which I have had to do.

(viii) Tentacles. — Mesogloea and endoderm about equally high : ectoderm a good
deal higher. Longitudinal musculature well developed, ectodermal, none of it
embedded in the mesogloea as far as I can tell, filling about one-third of the ectoderm
where best developed. Its general aspect is shown in PI. Y, fig. 10. It varies greatly in
height in different parts of the same tentacle : the processes fairly stout, branched.
Enclodermal circular muscle not conspicuous. Endoderm with zooxanthellae, and a fair
number of granular gland-cells. Ectoderm with broad well-developed nerve-layer : nuclei
of supporting cells scattered. In the outer part, a dense spirocyst-battery : spirocysts
of medium size, long and subcylindric. There are also a fair number of rather small
not usually elongate granular gland-cells, some slightly embedded, some superficial ( i.e .,
immediately below the ectodermal surface). Radial muscles of disc ectodermal.

A careful anatomical comparison (or, better still, a comparison of living examples !)
of the various species of the genus Calliactis is needed, as their structure is rather
similar. C. reticulata differs from C. tricolor ( = C. bicolor = C. eyletes) in the nature
of the sphincter, which in the latter is markedly reticulate, as well as in other details.
It is quite distinct from the British C. rondeletii and also from C. polypus. (For one
thing, in these last two, part of the longitudinal musculature of the tentacles is
embedded in the mesogloea.) The anatomy of C. miriam , with its extremely widely
expanded base and single circle of cinclides, is not known. Nor do we know the
structure of C. decorata and C. fusca, but both of them have prominent cinclidal
tubercles, which are totally absent in C. reticulata, which seems, further, characterised
by its small size and finely papillate skin. Cinclidal tubercles become more prominent
when a specimen is preserved— in fact, they are often quite absent in C. rondeletii when
it is alive, but appear after death : so that since they do not appear in C. reticulata
even when it is preserved, their absence may be taken as a specific character.
C. krtfyeri, although very large, has far fewer tentacles than C. reticulata, and from
the arrangement of cinclides may even belong to another genus.

ACTINIARIA— STEPHENSON.

57

Sub-family CHONDRACTINIINAE, Haddon.

“ Sagartiidae with only six pairs of perfect mesenteries, which alone of the well-
developed mesenteries are sterile ; two gonidial grooves and two pairs of directives ;
body-wall usually thick, with a cuticle, and often nodulated; cinclides absent (?) ;
acontia rarely emitted, and then by the mouth only ; strong mesogloeal sphincter-
muscle.” (Haddon, 1898, p. 458.)

LEPTOTEICHUS* n. gen.

Chondractiniinae with more than two cycles of tentacles, which are devoid of basal
swellings. Acontia rudimentary. Musculature of mesenteries tending to be peculiar.
Body-wall smooth, without tubercles or cuticle, not divided into scapus and capitulum.
Longitudinal muscles of the tentacles almost entirely ectodermal, radial muscles of the
disc meso-ectodermal. Genotype, L. insignis, n. sp.

This genus is easily distinguished from others in the same sub-family by its
rudimentary acontia, absence of cuticle, nodules, basal tentacular swellings, etc. To
describe the curious form of the muscle-processes of the mesenteries, I have borrowed
the word “ reversed ” from a letter from Prof. McMurrich, as a convenient one. See
below.

15. Leptoteichns insignis, n. sp. (PI. I, fig. 14 ; PL II, figs. 4, 5, 6 ; PI. Ill, fig. 20 ;
PI. IV, figs. 1, 2, 3, 5, 6.)

“ Insignis ” because it is a very handsome anemone, and if it could be seen alive
would no doubt be exceedingly so. There is very little resemblance between a live and
a dead Actinian : some of them when alive are among the most exquisite creatures in
existence.

One specimen. Station 338, 77° 13' S. ; 164° 18' E., 207 fathoms, entrance to

McMurdo Sound, Jan. 23, 1912. Agassiz trawl. Bottom — mud.

Measurements. — Diameter of tentacular crown, 10 cm. ; of oral disc, 5'2 cm. ;
of mouth, 4’1 cm. Length of a large tentacle, 2-7 cm. Height of column, 7 cm.
Diameter of column, 5*3 cm. ; of pedal disc, 4'5 cm.

External characters. — The single specimen is excellently preserved. Pedal disc
irregular in outline, smaller than the column, thinner than the body-wall, firmly
adherent to several small pieces of stone. There is a certain amount of dirt and loose
sloughed mucus about it. Column a little higher than wide, barrel-shaped, narrower
at base and just below margin than in centre. At the margin itself it expands again
very slightly, so that the disc exceeds the column a little. The margin is tentaculate,
no trace of parapet or fosse. The surface is smooth and firm, with no trace of verrucae,
etc., but it is deeply wrinkled transversely and less distinctly longitudinally, so that it
is broken up into numerous little roughly rectangular areas. Most of the furrows are

VOL. V.

*Ae7 tto5, and Ta^os, “ with a smooth wall.

I

58

“TERRA NOVA” EXPEDITION.

probably due merely to contraction. The longitudinal ones are more or less continuous
from top to bottom, but die out entirely in some places. Most of the ectoderm has
been removed, but it remains here and there in patches and in cracks, and is thick,
felty, much corrugated, and reddish. The mesogloea is pale, and a good deal of dirt is
in some of the cracks. The body -wall is stiff and firm, and very thick. The oral disc
is smooth, and thin in texture. From the base of each tentacle there runs a low
rounded ridge, towards the mouth, near which it dies out. These ridges are of course
bounded by furrows, those from the primary tentacles are very large, those from
the secondary ones medium, and from the tertiary narrow. So that large, small,
medium, etc., alternate in a regular manner. The mouth is nearly circular, and gaping,
so as to reveal the coelenteron, but the actinopharynx is not everted. The lip is
prominent, formed by the rather swollen apices of the ridges of the actinopharynx.
These are rather damaged, but there seem to be about nine, and each is divided into
two by a narrower ridge. There are exactly one hundred tentacles, in three cycles —
25, 25, 50 = 100. Those of the first cycle are mostly very large, as are some of the
second : those of the third are much smaller. They are rather irregular in size through
being irregularly expanded. They are very beautiful organs, very broad and thick at
the base, tapering gradually to an obtuse tip. They are thin walled, and flaccid when
out of water, but fairly strong, and stand up well in spirit. In colour they and the
disc are flesh-tint. They are finely transversely striated, but not fluted. The animal
is well expanded, and the manner in which it carries its tentacles may be seen from
PI. I, fig. 14. The tentacles should be 24, 24, 48 = 96 in arrangement, but a small
abnormality has occurred in one exocoel of the mesenteries, bringing with it four extra
tentacles.

Structure, (i) Mesenteries. — There are four cycles. 6 + 7 + 13 + 26 p. = 52
pairs. Only the six primary mesenteries are perfect, and they are much larger than
the others, so that the animal is very hollow. In all exocoels but one they are
normally arranged, each primary exocoel containing the secondary pair with a tertiary
pair on each side of it, and four quarternary pairs alternating with these. In one lateral
exocoel the arrangement is abnormal, causing the abnormality in the number of
tentacles. All mesenteries bear filaments. The primary mesenteries run all the way
down the actinopharynx. They bear immense knots of filament, and are sterile. The
secondary have also large masses of filament low down, but this is simpler higher up,
and here they bear gonads. The mesenteries of the other two cycles have simple
filaments and are narrow and almost entirely covered with even rows of gonads. A
shrimp was loose in the coelenteron. Internal stomata are present, but no external
ones. Etere and there a piece of filament has come loose from a mesentery, and
rather resembles an acontium, but it is always attached at one or both ends to the
edge of a piece of mesentery devoid of any other filament. These are not acontia, as I
have made certain by sectionising the loosest of them, but the animal does possess true
acontia, in a different position from these loose filaments. The mesenterial musculature

ACTINIARIA— STEPHENSON.

59

is extraordinary. I will first describe a typical mesentery of tlie fourth cycle, such as is
figured (the whole muscular portion, behind* the gonad, is shown) in PI. IV, fig. 5.
The mesogloea of the whole mesentery is very thick, but where the mesentery leaves
the body-wall it is extraordinarily so, and tapers off gradually. On each side it sends
out an equal development of enormously stout lobe-like processes, with short, thick,
rounded, lobe-like branches, which are fringed by the muscle-fibres. In the sections
there is so much mesogloea that it loohs as if the musculature were reversed, projecting
as narrow branched processes of endoderm into the mesogloea : this is of course only
apparently so, on account of the thickness of the true processes of mesogloea. On one
side this constitutes longitudinal, on the other parieto -basilar muscle, both being about
equally developed. The rest of the mesentery is short, and bears a low fringe of short
thick processes of the longitudinal muscle. There is no pennon. If we now take a
large mesentery of the second cycle a difference will be observed. The appearance of
the mesentery where it leaves the body-wall is very much as in the small one above
described ; the thick processes have the same character, but instead of being equally
developed on both faces, they are stronger and more numerous on the longitudinal than
on the parieto-basilar side. The rest of the mesentery is, of course, much larger than
before : its mesogloea is very thick, and one surface has a feeble fringe of parieto-basilar
muscle all along it. About one-third of the opposite surface, however, is occupied by a
weak diffuse pennon ; between the large processes of the mesentery just where it leaves
the body-wall, and the somewhat different processes of the pennon, is a weak low fringe.
The pennon has no regular and definite shape, and its processes are fairly high and
narrower than those near the body-wall, but they are closely packed against each other,
and it is here that the “ reversed ” appearance is most striking. A few processes are
exactly copied in PI. IV, fig. 6. It will be obvious from this figure that it looks as if
slender black processes were projecting in the ordinary way (from right to left) from
the layer which is marked en. But this is not the case : the letter m is placed in the
mesogloea, and the real processes are the great fingers of mesogloea projecting into the
endoderm (from left to right), with the muscle-fibres (shown as a continuous black line)
squeezed up between them and round their edges. The third-cycle mesenteries are
intermediate between the two types described.

(ii) Sphincter.-. — This has no very definite form. Its outline is shown in PI. IV,
fig. 1. At the top it nearly fills the mesogloea. It is separated from the endoderm
throughout by a band of mesogloea gradually widening downwards, and by a similar
band from the ectodermal side. It is not very long, and tapers to a point below, which
is about in the middle. The ectodermal side of the mesogloea is much folded, by
contraction-wrinkles of the body, and the sphincter follows the folds and so has a
serpentine outline. Spaces free from muscle-fibre, where the mesogloea is as usual,
occur here and there in the middle of the sphincter, and in the figure two are indicated.

I 2

* i.e., between the gonad and the body-wall.

60

“TERRA NOVA” EXPEDITION.

The muscle-fibres are small, and are arranged in little clumps of various sizes, endless
in number and various in shape, but all transversely elongate, and tending to be
arranged in transverse rows. A small portion of the sphincter is shown exactly in
PI. IV, fig. 2 : the portion shown covers nearly the whole width of a narrow section of
sphincter near the bottom. The upper part of the sphincter has a patchy appearance,
due to a very slight tendency to irregular vertical layering. The clumps are on the
whole rather larger and more numerous below than above, and are very feeble on the
ectodermal fringe of the muscle. It will be noted that here the structure of the
sphincter is neither reticular nor alveolar (see p. 6).

(iii) Endoderm. — That of the mesenteries contains many large gland-cells of various
shapes with large granules (PI. Ill, fig. 20). Mostly these stain with liaematoxylin as
usual, but some miss staining and stay yellow, or take only eosin. There are very few
of these cells in the body-wall endoderm. Here and there occurs a narrow needle-like
thick walled cnida. The mesenterial filaments possess well-developed ciliated lobes,
which in large filaments tend to be folded rather as in Artemidactis (see description of
that genus). The glandular lobes of the larger filaments contain very many coarsely
granular irregular gland-cells, mainly fringing their edges : in the smaller filaments
these are smaller and more regular, some rounded, some long and narrow. The simple
filaments of the smaller mesenteries have a trefoil form, the ciliated lobes being
rounded, the glandular lobe having a narrow stem. The glandular lobes contain long,
rather small, needle-like thick walled cnidae, and certain larger, broader, keeled,
unstained, square-ended bodies, which may be curious cnidae, but are not improbably
ingested diatoms : they are not very clearly visible.

(iv) Mesogloea. — Homogeneous. Contains many cells of various shapes and sizes.
Many are small and bipolar with a clear nucleus. Some are large and irregular, and
many are granular. They are not particularly interesting or much different from some
of those figured for other species.

(v) Ectoderm of body-wall. — This, where it remains, is poorly preserved and not
as high as that of the tentacles. Above its surface are bits of dirt in places and
adherent sloughed mucus, but no cuticle. It contains numerous gland-cells, some
superficial and some deeply embedded. In many cases the deep ones have a clearly
stained duct.

(vi) Actinopharynx. — This runs more than half-way down the coelenteron, and is
strong. It is grooved and wrinkled longitudinally and transversely, and has two well-
marked siphonoglyphes, descending well below the general level.

(vii) Tentacles. — Endoderm a good deal thinner than the other layers, forming a
very loose open network in its inner part, with a much denser crust containing here and
there a coarsely granular gland-cell. The ectoderm and mesogloea vary in thickness,
sometimes one and sometimes the other being the thicker. Endodermal circular muscle
feeble. Longitudinal muscles almost entirely ectodermal, with a fibre here and there
just embedded in the mesogloea (PI. IV, fig. 3), this being more noticeable in some

ACTINIARIA— STEPHENSON.

61

parts than in others. The processes are short and very stout, simple or a little branched.
The ectodermal nerve-layer forms a narrow distinct band just outside the musculature.
The nuclei of the supporting cells are scattered. The outer part of the ectoderm is
crowded with large subcylindrical spiroeysts, which are broadest in the middle.
Coarsely granular gland-cells like those of the mesenteries are present, but not in large
numbers, mainly just below the surface. For a figure of this ectoderm see PI. II, fig. 6.
In this figure it happens that at the particular point selected the nerve-layer is not
clear. The yellowish gland-cells which have missed staining may be seen about the
middle, and also the characteristic appearance of spiroeysts cut across.

(viii) Gonads. — The specimen is a male. The gonads are interesting in that each
testis is surrounded by a number of zooxanthellae. The forms of the testes are more
clearly visible than in many species. One is shown in PI. II, fig. 5. This shows the
structure of a typical Actinian testis. The spermatozoa are dark purple, and arranged
in a fan-like manner. The zooxanthellae round them are brownish, and the whole
testis is embedded in the mesogloea of the mesentery.

(ix) Acontia. — Attached to the edges of many of the large mesenteries, near their
bases, and just below their large knots of mesenterial filament, are small acontia. Each
consists of a tiny coil of several turns, attached to the mesentery at one end, free from
it at the other, and differing entirely in appearance from the mesenterial filament. The
largest coil is less than a quarter of an inch long, most are smaller. When sectionised,
the acontium proves to be completely different in its histological elements from the
mesenterial filament ; the two chief points being that it contains more and larger
granular gland-cells, and a dense battery of very large thick-walled nematocysts (rather
like the one shown in PI. Ill, fig. 18) which is absent from the filament. These
nematocysts are blunt at one end and more pointed at the other. This case of acontia
which are too small to be of any service to such a large animal, aDd must therefore be
regarded as rudimentary, is very interesting. Whether they are incipient or degenerate
I leave an open question, but I incline for various reasons to the idea that they are
incipient. I may say that Prof. G. C. Bourne agrees with me in believing that the
structures under discussion, though small, are truly acontia.

62

“TERRA NOVA’* EXPEDITION.

V— LITERATURE.

1. Alcock, A. — “ On some Actiniaria from the Indian Seas.” Journ. Asiat. Soc. Bengal, Vol. LXII,

Nat. Sci. 1893, p. 151.

2. Andres, A. — “ Le Attinie.” Mem. R. Accad. Lincei, Rome. Ser. 3. Vol. XIV, 1883, p. 211.
2a. Andres, A. — “ Le Attinie,” I. Fauna u. Flora G. von Neapel, ix. Monogr., 1884.

3. Annandale, N. — “ Coelenterates.” Fauna of the Chilka Lake, No. 1. Mem. Ind. Mus., Vol. V,

1915, p. 65.

4. Appellof, A. — “ Die Actiniengattungen Fenja, Aecjir, und Halcampoides , Dan.” Bergens Mus.

Aarbog, 1896, No. XI.

5. Appellof, A. — “ Ptychodactis patida, n. g. and sp.” Bergens Mus. Aarbog, 1893, No. IV.

6. Bourne, G. C. — “A Description of Five new species of Edwardsia, Quatr., from New Guinea, with

an Account of the Order of Succession of the Micro-mesenteries and Tentacles in the
Edwardsiae.” Journ. Linn. Soc. Vol. XXXII, Zool. 1916, p. 513.

7. Bourne, G. C. — “Anthozoa.” Lankester’s “Treatise on Zoology,” Part II, London, 1900.

8. Carlgren, O. — “ Studien liber Nordische Actinien.” K. Svenska Vet.-Akad. Handl. Bd. XXV,

No. 10, 1893.

9. Carlgren, O. — “ Beobachtungen liber die Mesenterienstellung der Zoantharien nebst Bemerkungen

liber die bilaterale Symmetrie der Anthozoen.” Zoolog. Studier, Festskrift Wilhelm Lillje-
borg, Upsala, 1896, p. 147.

10. Carlgren, O. — “Beitriige zur Kenntniss der Actinien-Gattung Bolocera, Gosse.” Ofv. K. Vet.-

Akad. Fork. No. 4, 1891.

11. Carlgren, O. — “ Rber Pentactinia californica, n. gen. n. sp.” Ofv. K. Vet.-Akad. Forh. No. 10,

1900.

12. Carlgren, O. — “ On the genus Porponia and related genera.” Scottish Nat. Antarct. Exped.

[“ Scotia”]. Trans. R. Soc. Edinb. Vol. L, Part I, 1913-14, p. 49.

13. Carlgren, O. — “ Ostafrikanische Actinien.” Mitth. Naturhist. Mus. Hamburg, XVII Jahrg.

1900, p. 21.

14. Carlgren, O. — “ Uber Dactylanthus ( Cystiactis ) antardicus (Clubb).” Wiss. Ergebn. Schwed. Siid-

polar-Exped. 1901-3. Bd. VI, Lief. V, 1911.

15. Carlgren, O. — “Zoantharien.” Ergebn. der Hamburger Magalhaensischen Sammelreise, Lief. IV,

No. 1, Hamburg, 1899.

16. Carlgren, O. — “ Actiniaria d. Olga-Expedition.” Wiss. Meersuntersuch. Komm. deutsch. Meere.

Bd. V, Abt. Helgoland, 1902, p. 31.

17. Carlgren, O. — “Die Brutpflege der Actiniarien.” Biol. Centralbl., Bd. XXI, 1901, p. 468.

17a. Carlgren, O. — “Anthozoa.” Bronn’s ‘ Klassen und Ordnungen d. Tier-Reichs.’ 1908.

18. Clubb, J. A. — “Actiniae.” Reports, Nat. Antarctic Expedition, 1901-4 [“Discovery”]: Nat.

Hist., Vol. IV, 1908.

19. Dana, J. D. — “Zoophytes.” U.S. Explor. Exped. Philadelphia, 1846.

20. Danielssen, D. C. — “Actinida of the Norske Nordhavs-Expedition, 1876-1878.” Bd. V,

Christiania, 1887.

21. Dixon, G. Y., and A. F. — “ Notes on Bimodes thallia, B. verrucosa, and Tealia crassicornis.”

Sci. Proc. R. Dublin Soc., Vol. VI, 1889, p. 310.

22. Dixon, G. Y. — “ Remarks on Sagartia venusta and S. nivea.” Sci. Proc. R. Dublin Soc.,

Vol. VI, 1888, p. 111.

23. Duerden, J. E. — “ Jamaican Actiniaria,” Part I. Sci. Trans. R. Dublin Soc., Ser. 2, Vol. VI,

1898, p. 329; Part II, op. cit. Vol. VII, 1900, p. 133.

24. Duerden, J. E. — “The Actiniaria around Jamaica.” Journ. Inst. Jamaica, Vol. II, 1897, p.449.

25. Farquhar, H. — “Preliminary Account of some New-Zealand Actiniaria.” Journ. Linn. Soc. Zool.

Vol. XXVI, 1898, p. 527.

26. Gosse, P. H. — “ A history of the British Sea Anemones and Corals.” London, 1860.

27. Gourret, P. — “ Topographie Zoologique des etangs de Caronte,” etc. Ann. Mus. Hist. Nat.

Marseille, Zool. Tome XI, 1907.

ACTINIARIA— STEPHENSON.

63

28. Haddon, A. C. — “The Actiniaria of Torres Straits.” Sci. Trans. R. Dublin Soc., Ser. 2, Yol. YI,

1898, p. 393.

29. Haddon, A. C. — “ A revision of the British Actiniae,” Part I. Sci. Trans. R. Dublin Soc. Ser. 2,

Yol. IY, 1889, p. 297. Part II (by A. C. Haddon and A. M. Shackleton), t. cit. 1891,
p. 609.

30. Haddon, A. C., and Duerden, J. E. — “ On some Actiniaria from Australia and other Districts.”

Sci. Trans. R. Dublin Soc., Ser. 2, Yol. VI, 1896, p. 139.

31. Haddon, A. C. — “ Report on the Actiniae dredged off the South-West Coast of Ireland.” Proc.

R. Irish Acad. Ser. 3, Yol. I, 1890, p. 370.

32. Hertwig, R. — “Report on the Actiniaria dredged by H.M.S. ‘Challenger.’” “Challenger”

Reports, Zool., Yol. YI, 1882 ; also Supplement, Yol. XXYI, 1888.

33. Kerb, H. “ Studien iiber die ungeschlechtliche Fortpflanzung der Gonactinia prolifer a, Sars.”

Bergens Mus. Aarbok, 1913, Nr. 3.

33a. Kirk, H. B., and Stuckey, F. G. A.— “ Two species of Actiniaria from Campbell Island.”
Article XVIII in C. Chilton’s “ Subantarctic Islands of New Zealand,” Vol. I, Wellington,
1909, p. 384.

34. Kwietniewski, C. R. — “Actiniaria von Ambon u. Thursday Island” (R. Semon’s “Zoolog.

Forschungsreise,” Bd. Y), Denkschr. Med.-Naturwiss. Ges. Jena, Band VIII, 1898, p. 385.

35. Kwietniewski, C. R. — “Actiniaria von Ost-Spitzbergen.” Zool. Jahrb. Syst., Band XI, 1898,

p. 121.

36. Kwietniewski, C. R. — “ Revision der Actinien, welche von Herrn Prof. Studer auf der Reise der

Korvette Gazelle um die Erde gesammelt wurden.” Jena. Zeitschr. Yol. XXX, 1896,

p. 583.

37. McMurricii, J. P. — “Notes on some Actinians from the Bahama Islands, collected by the late

Dr. J. I. Northrop.” Ann. New York Acad. Sci., Yol. IX, 1897, p. 181.

38. McMurricii, J. P. — “ Report on the Actiniae collected by the U.S. Fish Commission Steamer

Albatross during the winter of 1887-88.” Proc. U.S. Nat. Mus., Vol. XYI, 1893 [1894],
p. 119.

39. McMurrich, J. P. — “ Report on the Actiniaria Collected by the Bahama Exped. of the State

Univ. of Iowa, 1893.” Bull. Lab. Nat. Hist. State Univ. Iowa, Yol. IY, 1898, p. 225.

40. McMurricii, J. P. — “The Actiniae of the Plate Collection” (Fauna Chilensis, Vol. III). Zool.

Jahrb. Suppl.-Bd. VI, 1904, p. 215.

41. McMurrich, J. P. — “ Report on the Hexactiniae of the Columbia Univ. Exped. to Puget Sound

during the summer of 1896.” Ann. New York Acad. Sci., Yol. XIV, 1901, p. 1.

42. McMurrich, J. P. — “ Contributions on the Morphology of the Actinozoa. YI, Halcurias pilatus and

Endocoelactis.” Biol. Bull., Yol. II, 1901, p. 155.

43. McMurrich, J. P. — “ A Contribution to the Actinology of the Bermudas.” Proc. Acad. Nat. Sci.

Philadelphia, 1889, p. 102.

44. McMurrich, J. P. — “ The Actiniaria of the Bahama Islands.” Journ. Morph., Vol. Ill, 1889, p. 1.

45. McMurrich, J. P. — “On two New Actinians from the Coast of British Columbia.” Proc. Zool.

Soc., 1913, p. 963.

46. Maguire, K. — “Notes on certain Actiniaria.” Proc. R. Dublin Soc. (N.S.), Yol. VIII, 1898,

p. 717.

46a. Mangan, J. — “ The Entry of Zooxanthellae into the Ovum of Millepora, and some Particulars
concerning the Medusae.” Quart. J. Micr. Sci., Vol. LIII, 1909, p. 697.

47. Milne-Edwards, H. — “ Histoire Naturelle des Coralliaires ou polypes proprement dits.” Tome I,

Paris, 1857.

48. Pax, F. — “ Unterklasse : Hexacorallia.” Probe- Artikel zum Handbuch der Zoologie, Kiikenthal.

49. Pax, F. — “Studien an westindischen Actinien.” Zool. Jahrb. Suppl. Bd. XI, 1910, p. 157.

50. Pax, F. — “ Aktinien der Aru-Inseln.” (H. Merton, Ergebn. zool. Forschungsreise siidostl. Molukken.

Bd. 1.) Abhandl. Senckenb. Naturforsch. Gesellsch., Bd. XXXIII, Frankfurt, 1910,
p. 297.

51. Pax F. — “Vorarbeiten zu einer Revision der Familie Actiniidae,” Inaug.-dissertation, Breslau,

1907.

64

“TERRA NOVA” EXPEDITION.

52. Pax, F. — Les Actinies de la Cote du Perou.” Ministere de l’Instruction publique. Miss. . .pour

la mesure d’un Arc de meridien equatorial en Amerique du Sud, 1899-1906, Tome IX,
Fasc. 4, Paris, 1912, p. D. 1.

53. Pax, F. — “ Revision des types des Actinies decrites par Quoy et Gaimard.” Ann. Sci. Nat., Ser. 9,

Vol. XVI, 1912, p. 309.

54. Pax, F. — “ Die Aktinien der ostafrikanischen Inseln.” In A. Voeltzkow’s “ Reise in Ostafrika,

1903-1905.” Wiss. Ergebn. Band. II, Heft IV, Stuttgart, 1909, p. 399.

54a. Pratt, Edith M. — “ The digestive organs of the Alcyonaria, and their relation to the Mesogloeal
cell plexus.” Quart. J. Micr. Sci., Vol. XLIX, 1905-6, p. 327.

55. Rees, O. M. — “ On Eloactis mazeli.” Journ. Mar. Biol. Assoc., Vol. X, 1913, p. 70.

56. Rees, O. M. — “Notes on Actinostola callosa (Verrill) = Dysadis crassicornis (Hertwig).” Ann.

Mag. Nat. Hist., Ser. 8, Vol. XII, 1913, p. 382.

57. Rees, O. M. — “ Contributions to the Comparative Anatomy of some British Actiniae.” Journ.

Mar. Biol. Assoc., Vol. X, 1915, p. 521.

58. Stuckey, F. G. A. — “ On Two Anemones found in the neighbourhood of Wellington.” Trans. New

Zealand Inst., Vol. XLI, 1909, p. 370.

59. Stuckey, F .G. A.— “A review of the New Zealand Actiniaria.” Trans. New Zealand Inst.,

Vol. XLI, 1909, p. 374.

60. Torrey, H. B. — “ Papers from the Harriman Alaska Expedition,” XXX, “ Anemones, with discussion

of Variation in Metridium .” Proc. Wash. Acad. Sci., Vol. IV, 1902, p. 373.

61. Verrill, A. E. — “ Synopsis of the Polyps and Corals of the North Pacific Expl. Exped. 1853-56.”

Proc. Essex Inst., Vol. V, 1866-7, Communications, p. 17 [1868],

62. Verrill, A. E. — “Descriptions of imperfectly known and new Actinians, with critical notes on

other species.” Amer. Journ. Sci. and Arts, Vol. VII, 1899, pp. 41, 143, 205, 375.

63. Walton, C. L. — “ Kodioides borleyi, n. sp.” Journ. Mar. Biol. Assoc., Vol. IX, 1910, p. 85.

64. Walton, C. L., and Rees, O. M. — 1 u Some rare and interesting Sea Anemones from Plymouth.”

Journ. Mar. Biol. Assoc., Vol. X, 1913, p. 60.

65. Walton, C. L. — “Actiniae collected by the SS. ‘Huxley’ in the North Sea during the summer

of 1907.” Journ. Mar. Biol. Assoc., Vol. VIII, 1908, p. 215.

66. Wassilieff, A. — “Japanische Actinien” (in Dofiein, Beitr. Naturg. Ostasiens). Abhandl. K.

Bayer. Akad. Wiss. Suppl. Bd. I, 1908.

67. Wyragevitsch, T. A. — “ Sur une espece [ Halcampella oslrooumowi] du genre Halcamjjclla . . . trouvee

dans la Mer Noire.” Zool. Anz., Vol. XXIV, 1901, p. 246.

ACTINIARIA— STEPHENSON.

65

INDEX.

Figures iu thick type refer to the special accounts of the respective genera and species.

Acontia, definition, 6.

„ rudimentaiy, 57, 61.

Acrorhagi, definition, 6.

,, Pseudo, definition, 8.
Actiniaria, note on form of, 4.
Actiniina, 1, 8, 14, 19.

Actinopharynx, definition, 6.
Actinostola, 5, 31, 37.

„ callosa, 5.

,, intermedia, 13.

„ spetzbergensis, 31.

Actinostolinae, 27.

Actinostome, definition, 6.
africana, Bolocera, 22.

Aiptasia, 8, 48, 51, 53.

„ minima, 1, 49, 53.

„ mutabilis, 51.

., pallida, 48, 51.

Aiptasiinae, definition, 48.
Aiptasioides, 48, 51.

„ pallida, 48, 51.

„ prima, 2, 48, 51.

Alcyonium, 22, 44.

,, digitatum, 2.

alfordi, Anthopleura, 5.

Aliciidae, definition, 23.

Alloactis, 28.

Alveolar, definition, 6.

Ammophilactis, 28.

Amoeboid cells, 2, 3, 22.

Amphioxus, 3.
antarctica, Isotealia, 27.
antarcticus, Cystiactis, 19.

„ Dactylanthus, 1, 5, 19

Antheadae, 19.

Antholoba, 28.

Anthopleura alfordi, 5.

Archactis, 28.
arctica, Halcampa, 9.

Artemidactis, 4, 5, 40, 41, 60.

,, victrix, 1, 26, 40, 41, 55.
aspera, Halcampoides, 10.

Athenaria, 14.

badia, Epiactis, 25, 26, 27.

,, Leiotealia, 25, 26, 27.
Balanoglossus, 3.

Basilar muscles, note on, 3.

bicolor, Calliactis, 56.

Body-wall, musculature in ectoderm of, 6, 18.

,, spirocysts in ectoderm of, 6, 17.

Bolocera, 6, 18, 19, 20, 22, 25.

,, africana, 22.

,, brevicornis, 6, IS, 21, 22.

„ kerguelensis, 22, 25, 26.

„ longicornis, 2, 3, 17, 20, 22, 23, 26, 31.

„ multicornis, 22, 23.

,, multipora, 22.

„ norwegica, 23.

,, pannosa, 21, 22.

„ pollens, 8, 22.

„ tuediae, 23.

Boloceridae, 8, 19.

Boloceroides, 6, 20.

Boloceropsis, 20.

brevicornis, Bolocera, 6, 18, 21, 22.

Bunodidae, 20.

Cadosactis, 28.

Calliactis, 53, 56.

,, bicolor, 56.

,, decorata, 56.

,, egletes, 56.

„ fusca, 56.

„ kr0yeri, 56.

„ miriam, 56.

,, reticulata, 2, 53, 56.

„ rondeletii, 56.

,, tricolor, 56.

callosa, Actinostola, 5.

Capitulum, definition, 7.

carlgreni, Halcurias, 18.

Cells, amoeboid, 3, 22.

„ of mesogloea, note on, 2, 22.

Cereus, 4, 40.

„ pedunculatus, 40, 41.

Cerianthus, 8.

Ciliated lobes of mesenterial filament, definition, 7.

Cinclides, definition, 7.

Chiton, 3.

Cbondractiniinae, definition, 57.

chrysanthellum, Halcampa, mesogloeal sphincter
of, 9.

clavus, Halcampa, 9, 13.

„ Halcampoides, 9, 13.

,, Halcampomorphe, 9, 13.

VOL. v.

K

66

“TERRA NOVA’’ EXPEDITION.

clematis, Phymactis, 23.

Cnidae, definition, 7.

,, measurements of, -3.

Coelenteron, definition, 7.
crassicornis, Urticina, 5.

Cribrinidae, definition, 24.

Crustacea, Decapod, 3.

Crustacean parasite in mesentery, 13.
Cyathactis, 28.

Cymbactis, 28, 29, 35.

„ actinostoloides, 36.

,, faeculenta, 36.

,, maxima, 36.

„ selaginella, 1, 3, 5, 13, 31, 36.

Cystiactis, 19.

,, antarcticus, 19.

Dactylanthus, 19.

,, antarcticus, l, 5, 19.

Decapod Crustacea, 3.
decorata, Calliactis, 56.

Definitions of terms, 6.
digitatum, Alcyonium,. 2.

Distinguishing genera, note on, 4.
dubia, Epiactis, 25, 27.

,, Leiotealia, 25, 27.
duodecimcirrata, Halcampa, 9.

Ectoderm of body-wall, musculature in, 6, 18.

,, ,, spirocysts in, 6, 17.

Edwardsia, 8.

Edwardsiaria, 1.
egletes, Calliactis, 56.

Endocoel, definition, 7.

Endocoelactis, 14.

endocoelactis, Halcurias, 1, 6, 14, 16.

Endocoel actidae, definition, 13.

Endodermal sphincter, evolution of, 25.
endromitata, Halcampella, 9.

Enterostome, definition, 7.

Epiactis, 5, 24, 27.

„ badia, 25, 26, 27.

,, dubia, 25, 27.

,, fecunda, 27.

„ marsupialis, 25, 27.

,, novo-zealandica, 1, 24, 25, 26, 27.

,, nymphaea, 25, 27.

,, prolifera, 25, 27.

„ regularis, 27.

„ ritteri, 25, 27.

,, spetzbergensis, 25, 27.

„ thompsoni, 24, 25, 26, 27.

Exocoel, definition, 7.
faeculenta, Cymbactis, 36.
fecunda, Epiactis, 27.

Filament, definition of ciliated lobes of, 7.

Filament, definition of glandular lobe of, 7.

,, detailed description of a, 44.

„ restriction of, to part of a mesentery, 40,
41, 43.

Form of Actiniaria, note on, 4.

Fosse, definition, 7.
fusca, Calliactis, 56.

Genera, note on distinguishing, 4.

„ of Paractidae, key to, 27, 28.

General notes, 2.

Glandular lobe of mesenterial filament, definition,
7.

Gonads, note on, 5.

„ note on Zooxanthellae in, 3.

Halcampa, 8, 9, 10, 28.

,, arctica, 9.

„ chrysanthellum, 9.

,, clavus, 9, 13.

„ duodecimcirrata, 9.

„ kerguelensis, 1 3.

Halcampella, 9, 10.

,, endromitata, 9.

,, maxima, 9, 13.

,, minuta, 13.

„ ostrooumowi, 13.

| Halcampidae, 8.

Halcampinae, 8, 9, 10.

Halcampoides, 8, 9, 10.

„ aspera, 1, 5, 9, 10, 13.

„ clavus, 9, 13.

,, kerguelensis, 13.

,, maxima, 9, 13.

„ ostrooumowi, 13.

Halcampoidinae, 9, 10.

Halcampomorphe, 9, 10.

,, clavus, 9.

Halcampomorphidae, 8.

Halcampomorphinae, 8, 9, 10.

Halcurias, 14, 15, 17.

,, carlgreni, 18.

„ endocoelactis, 1, 6, 14, 16.

,, pilatus, 6, 18.

Halianthella, 9, 10.

Halianthinae, 9, 10.

Halianthus, 9, 10.

Haloclava, 13.

Hormosoma, 28, 29, 34.

„ scotti, 1, 29.

Ilyanthidae, 8, 9.

Ilyanthus, 8.

insignis, Leptoteichus, 57.
intermedia, Actinostola, 13.

Isotealia antarctica, 27.
kerguelensis, Bolocera, 22, 25, 26.

ACTINIARIA — STEPHENSON.

67

kerguelensis, Halcampa, 13.

,, Halcampoides, 13.

Kodioides, 4.
kr0yeri, Calliactis, 56.

Lacunae in mesogloea, 33, 35.
lacunifera, Lilliella, 1, 33.

Leiotealia, 24, 27.

., badia, 25, 26, 27.

,, dubia, 25, 27.

,, nymphaea, 25, 27.

,, spetzbergensis, 25, 27.

,, thompsoni, 24, 25, 26, 27.

Leptoteichus, 5, 31, 34, 43, 55, 57.

,, insignis, 1, 3, 57.

Lilliella, 28, 33.

„ lacunifera, 1, 33.

Limbus, definition, 7.

Lobes, glandular and ciliated, of mesenterial
filament, 7.

longicomis, Bolocera, 2, 3, 17, 20, 22, 23, 26, 31.
Marginal stomata, definition, 7.
marsupialis, Epiactis, 25, 27.

Marsupifer, 28.
maxima, Cymbactis, 36.

,, Halcampella, 9, 13.

,, Halcampoides, 9, 13.

Mesenterial filament, detailed description of, 44.
Mesenteries, Crustacean parasite in, 1 3.

„ note on muscles of, 5.

„ peculiarities of musculature, 29, 30,

31, 36, 38, 40, 43, 57, 59.

,, perfect, definition, 7.

„ perfect, note on, 5. '

Mesogloea, cells of, 2, 22.

,, lacunae in, 33, 35.

Methods of staining, 2.

Metridiinae, definition, 51.

Metridium, 30.

Millepora, 3.
miniata, Sagartia, 4.
minima, Aiptasia, 49.

,, Halcampella, 13.
miriam, Calliactis, 56.
multicornis, Bolocera, 22, 23.
multipora, Bolocera, 22.

Muscles, basilar, 3.

„ mesenterial, note on, 5.

,, parieto-basilar, 3.

„ sphincter, 5.

Musculature of ectoderm of body-wall, note on, 6,
„ 18.

,, of mesenteries, peculiarities of, 29, 30,
31, 36, 38, 40, 43, 57, 59.
mutabilis, Aiptasia, 51.

Xematocysts, definition, 7.

,, measurement of, 3.
norwegica, Bolocera, 23.
novo-zealandica, Epiactis, 1, 24, 25, 26, 27.
nymphaea, Epiactis, 25, 27.

„ Leiotealia, 25, 27.

Nynantheae, 14.

Oral stomata, definition, 7.
ostrooumowi, Halcampella, 13.
pallida, Aiptasia, 48, 51.

,, Aiptasioides, 48, 51.
pannosa, Bolocera, 21, 22.

Paractidae, 27.

„ key to genera of, 27, 28.

Paractinae, 27, 29.

Paractinia, 28.

Paractis, 5, 28, 29.

Parantheoides, 4, 5, 28.

Paranthus, 4, 5, 28, 29.

Parasite of mesentery, 13.

Parieto-basilar muscles, note on, 3.
pedunculatus, Cereus, 40.

Pennon, definition, 7.

Perfect mesenteries, note on, 5.

,, ,, definition, 7.

Phelliomorpha, 28, 33.

Phelliopsis, 28.

Phymactis, 23.

„ clematis, 2, 23.

Physa, definition, 7.
pilatus, Halcurias, 6, 18.

Plates, notes on, 2, 3.
pollens, Bolocera, 8, 22.

Polyopis, 20.
polypus, Calliactis, 56.

Polystomidium, 20.

Porponia, 18.

prima, Aiptasioides, 51.

prolifera, Epiactis, 25, 27.

Protactinina, 19.

Protanthea, 6.

Protan theae, 14.

Pseudo-acrorhagi, definition, 8.

Ptychodactidae, 18.

Pycnanthus, 28, 29.

Raphactis, 28.
regularis, Epiactis, 27.

Restriction of filaments to part of mesentery, 40,
41, 43.

Reticular, definition, 8.
reticulata, Calliactis, 53.
ritteri, Epiactis, 25, 27.
rondeletii, Calliactis, 56.

Rudimentary acontia, 57, 61.

68

“TERRA NOVA

EXPEDITION.

Sagartia, 5, 40, 46, 48.

„ miniata, 4.

„ spongicola, 48.

„ viduata, 5.

„ sp., 1, 46.

Sagartiidae, 6, 8, 27, 40.

Sagartiinae, 40.

Scapus, definition, 8.
scotti, Hormosoma, 1, 29.
selaginella, Cymbactis, 36.
spetzbergensis, Actinostola, 31.

,, Epiactis, 25, 27.

„ Leiotealia, 25, 27.

Sphincter, endodermal, note on evolution of, 25.

„ muscles, note on, 5.

Spirocysts, definition, 8.

,, in body- wall ectoderm, 6, 17.

,, measurement of, 3.

spongicola, Sargartia, 48.

Staining, methods of, 2.

Stichodactylina, 1, 14.

Stomata, marginal, definition. 7.

„ oral, definition, 7.

Sycyonis, 28.

Synanthus, 28.

Tealidium, 28.

Tentacles, note on, 5.

Terms, definitions, 6.
thompsoni, Epiactis, 24, 25, 26, 27.

„ Leiotealia, 24, 25, 26, 27.
tricolor, Calliactis, 56.
tuediae, Bolocera, 23.

Urticina crassicornis, 5.

Verrucae, definition, 8.

Vesicles, definition, 8.

victrix, Artemidactis, 1, 26, 40, 41, 55.

viduata, Sagartia, 5.

Zoanthinaria, 1.

Zooxanthellae in gonads, note on, 3.

LONDON : PRINTED BY WILLIAM CLOWES AND SONS, LIMITED, DUKE STREET, STAMFORD STREET, S.E. 1, AND GREAT WINDMILL STREET, W. 1.

PLATE I.

Fig.
Fig.
Fig.
Fig.
Fig.
Fig.
Fig.
Fig. 8.
Fig. 9.
Fig. 10.
Fig. 11.
Fig. 12.
Fig. 13.
Fig. 14.
Fig. 15.

-Halcampoides aspera, n. sp. Nat. size.

j Aiptasia minima, n. sp. A little enlarged.

-Halcurias endocoelaciis, n. sp. Nat. size.

- Bolocera longicornis, Carlgren. Small specimen. Nat. size.

-Phymactis clematis, Drayton. Nat. size.

-Cymbactis sclaginella, n. sp. Nat. size.

-Epiacfis novo-zealanclica, n. sp. Nat. size.

-Aiptasioides prima, n. gen., n. sp. Nat. size.

-Dactylanthus antarcticus, Carlgren. Small specimen. Nat. size.

- Cymbactis sclaginella, n. sp. Nat. size.

-Sagartia, sp. Nat. size or a little enlarged.

-Lilliella lacunifera, n. gen., n. sp. Nat. size.

-Lcptoteichus insignis, n. gen., n. sp. Nat. size or a little less.

-Calliactis reticulata, n. sp. Two specimens on gastropod shell. Nat. size.

1

Bnt.Mus. I Nat. Hist.)

Brit Antarctic (Terra Nova) Exped. 1910.
Zoology, Vol.V.

Coelenterata, Part 1, pj
Actiniaria.

••

r A Stephenson del.

HutK, Lithr Lon don .

A

PLATE IT.

Arternidactis victrix. Figs. 1, 3, 7, 15.
Halcurias endocoelactis. Figs. 8-11, 16.
Hormosoma scotti. Figs. 2, 17, 18.
Leptoteichus insignis. Figs. 4—6.
Lilliella lacuniferci. Figs. 12-14.

Sagartia, sp. Fig.

b. en., body-wall endoderm.

c. l., ciliated lobe of mesenterial filament.
c.m., circular muscle.

ec., ectoderm.

en., endoderm.

g.b., granular gland-cell.

g.c., granular gland-cell.

g. 1. , glandular lobe of mesenterial filament.

19.

l. m., longitudinal musculature.

m. , mesogloea.

m. f. , muscle-fibres.

n. , nucleus.
oo., ovum.
sp., spirocyst.

sp'., spirocyst in transverse section.
spz., spermatozoa.

Fig. 1.
Fig. 2.
Fig. 3.
Fig. 4.

-Arternidactis victrix, n. gen., n. sp. Small specimen. Nat. size.
- Hormosoma scotti, n. gen., n. sp. Nat. size.

-Arternidactis victrix.
-Lepiote i cli us i nsig n is.

Fig. 5.—

Fig. 6.—

Fig. 7. — Arternidactis victrix.

Fig. 8,
Fig. 9.
Fig. 10,
Fig. 11,
Fig. 12,
Fig. 13,
Fig. 14,
Fig. 15.
Fig. 16.
Fig. 17,
Fig. 18,
Fig. 19.

A whole third cycle mesentery. Oc. 3, obj. j.

Small portion of one side of the basal part of a fourth cycle mesentery,
showing the character of the musculature. Oc. 3, obj.

Single testis (in portion of mesentery). Oc. 3, obj.

A small portion of the ectoderm of a tentacle. Oc. 3, obj. TV,.

Small portion of the folded endoderm of the ciliated lobe of a large
mesenterial filament-trefoil. Oc. 3, obj. -J-.

Halcurias endocoelactis. Cells from mesogloea of body-wall. Oc. 3, obj. y1,,.

Lilliella lacunifera. Cells from mesogloea of body-wall. Oc. 3, obj. T\j.

-Arternidactis victrix. Cell from mesogloea of mesentery. Oc. 3, obj. TV
-Halcurias endocoelactis. Cell from mesogloea of body-wall. Oc. 3, obj. tl.

j Hormosoma scotti. Cells from mesogloea of body-wall. Oc. 3, obj. T’y.

-Sagartia, sp. Homogeneous gland-cell from acontium. Oc. 3, obj. Tx?.

Brit.Mus. (Nat. Hist)

Brit Antarctic (Terra Nova) Exped.1910.
Zoology, Vol Y.

Coelenterata, Part 1,
Actjniaria.

PI. II.

' 1 1

/ IN

j A . Ayl Y

V i "ss' J f ■ \

,4'

N -T

T A Steplienson del.

Huth Lith r L on don

PLATE III.

Arternidactis victrix. Figs. 6—10.

Bolocera longicornis. Figs. 11-16, 24, 27.
Epiactis novo-zealandica. Fig. 28.
Halcampoides aspera. Figs. 1-4.

Halcurias endocoelactis. Figs. 22, 23, 25, 26.
Hormosoma scotti. Figs. 17, 19, 21.
Leptoteiclius insignis. Fig. 20.

Lilliella lacunifera. Fig. 5.

Sagartia, sp. Fig. 18.

b.w., body-wall.
ec., ectoderm.
en., endoderm.

l.m., longitudinal musculature.
in., mesogloea.

n.l., nerve-layer.
pb.m., parieto-basilar muscle.
sph., sphincter.
t., tentacle.

1-4, cycles of mesenteries.

Fig. 1.
Fig. 2.
Fig. 3.
Fig. 4.

Fig. 5.
Fig. 6.
Fig. 7.
Fig. 8.
Fig. 9.
Fig. 10,
Fig. 11,
Fig. 12,
Fig. 13,
Fig. 14,
Fig. 15,
Fig. 16,
Fig. 17,
Fig. 18,
Fig. 19,
Fig. 20,
Fig. 21.
Fig. 22,

Fig. 23.
Fig. 24,
Fig. 25,
Fig. 26,
Fig. 27.
Fig. 28.

- Halcampoides aspera. A few processes of the sphincter muscle. Oc. 3, obj. g.

„ „ Parietal muscle of perfect mesentery. Oc. 3, obj. -§.

„ ,, Muscle-pennon of perfect mesentery. Oc. 3, obj. 1 Jj.

,, ,, Diagram of proportions of layers in part of transverse section of a

tentacle. Oc. 3, obj. |.

- Lilliella lacunifera. Network from mesogloea of body-wall. Oc. 3, obj.

Arternidactis victrix. Cell-types from mesogloea of mesentery. Oc. 3, ob. T*5.

i

T21

obj

Bolocera longicornis. Cells from mesogloea of tentacle. Oc. 3, obj. -jV

-Hormosoma scotti. Cell from mesogloea of body-wall. Oc. 3, obj. TL.

-Sagartia, sp. Keeled thick walled nematocyst from acontium. Oc. 3, obj.

- Hormosoma scotti. Cell from mesogloea of body- wall. Oc. 3, obj. T\j.

-Leptoteiclius insignis. Coarsely granular gland-cell from mesentery. Oc. 3,

-Hormosoma scotti. Cell from mesogloea of body-wall. Oc. 3, obj. T\>.

-Halcurias endocoelactis. Pennon of large mesentery, below level of actinopharynx. Oc. 3,

obj. f.

,, ,, Musculature of mesentery as it leaves body-wall. Oc. 3, obj. §.

- Bolocera longicornis. Muscular portion of a third-cycle mesentery. Oc. 3, obj. §.

-Halcurias endocoelactis. Diagram of part of transverse section of a tentacle. Oc. 3, obj. ’ .

„ ,, Diagram of arrangement of mesenteries in one-half of the specimen.

-Bolocera longicornis. Sphincter-muscle, and half tentacle-sphincter. Oc. 3, obj. j.

-Epiactis novo-zealandica. Diagram of part of transverse section of a tentacle. Oc. 3, obj.

Brit. Mus. (Nat. Hist.)

Brit. Antarctic (Terra Nova) Exped. 1910.
Zoology, Vol. V.

Coelenterata, Part I,
Actiniaria.

PI.

A. Stephenson del.

Adlard & Son & West Newman. Ltd.

Coelenterata, Part I. — Actiniaria, PI. IV.

PLATE IV.

Fig.

Fig.

Fig.

Fig.

Fig.

Fig.

Fig.

Fig.

Fig.

Fig.

Fig.

Fig.

%

Fig.

Fig.

Fig.

Cymbactis selaginella. Figs. 14, 15.
Hormosoma scotti. Figs. 4, 7-9.
Leptoteichus insignis. Figs. 1—3, 5, 6.
Lilliella lacunifera. Figs. 10-13.

b.en., body-wall endoderm.
ec., ectoderm.

e.c.m., endodermal circular muscle.
en., endoderm.

9; gonad.
lac., lacunae.

l.m., longitudinal musculature.

m., mesogloea.
m.p., muscle-processes,
j n.l., nerve-layer.

pb.m., parieto-basilar muscle.
sph., sphincter.
t., tentacle.

t.m., tentacle-mesogloea.

7.

8.

9.

10.

11.

12.

13,

14

-Leptoteichus insignis. Small-scale diagram to show shape of sphincter.

„ „ Exact detail of a small portion of sphincter. Oc. 3, obj. |.

,, ,, Part of a transverse section of a tentacle. Oc. 3, obj. }r.

-Hormosoma scotti. Exact detail of portion of sphincter. Oc. 3, obj. A.

-Leptoteichus insignis. Muscular portion of a last-cycle mesentery. Oc. 3, obj. 5.

,, , „ Some processes of the muscle-pennon of a large mesentery, to show how

they appear to project from the endoderm into the mesogloea. The
letter m is placed in the mesogloea. Oc. 3, obj. ^ .

-Hormosoma scotti. Diagram to show shape of sphincter. Oc. 3, obj. 11,-.

,, ,, Part of transverse section of tentacle. Oc. 3, obj. §.

,, ,, Muscular part of a medium-sized mesentery. Oc. 3, obj. 1|.

-Lilliella lacunifera.

Detail of small portion of middle part of sphincter (exact). Oc. 3, obj. |.
Small portion of the lining of a lacuna. Oc. 3, obj. yL.

Portion of body-wall mesogloea, showing lacunae : the dotted areas within

the lacunae are patches of the network-lining
. — ,, ,, Diagram to show form of sphincter.

, — Cymbactis selaginella. Part of transverse section of a tentacle. Oc

Oc. 3, obj. §.

3, obj. From a large

15.—

specimen.

Exact detail of small part of sphincter of small specimen. Oc. 3, obj.

PI. IV.

Brit. Mus. (Nat. Hist.)

Brit. Antarctic (Terra Nova) Exped. 1910-
Zoology, Vol. V.

Coelenterata, Part I,
Actiniaria.

T. A. Stephenson del.

Adlard & Son & West Newman, Ltd.

PLATE V.

Aiptasioides prima. Figs. 11, 14, 16.
Artemidactis victrix. Figs. 3, 5-8.
Calliactis reticulata. Figs. 9, 10.
Cymbactis selaginella. Figs. 1, 2, 4.
Sagartia, sp. Figs. 12, 13, 15, 17.

b. en.y body-wall endoderm.

c. e., firmer outer part of endoderm.
ec., ectoderm.

en., endoderm.

l. m., longitudinal musculature.

m. , mesogloea.

m.en., mesentery-endoderm.

n.l., nerve-layer.

oe.n.y actinopharynx-endoderm .

pb.rn., parieto-basilar muscle.

s. e., spongy part of endoderm.
sph., sphincter.

t. , tentacle.

z'., zone of spirocysts.

Fig. 1. — Cymbactis selaqinella.
Fig. 2.- „ _

Fig. 3. — Artemidactis victrix.

Fig. 4. — Cymbactis selaginella.

Fig. 5.-
Fig. 6.-
Fig. 7.-
Fig. 8.-

Fig. 9.
Fig. 10.
Fig. 11.
Fig. 12,
Fig. 13,
Fig. 14,
Fig. 15
Fig. 16
Fig. 17

-Artemidactis victrix.

Diagram showing shape of sphincter of small specimen.
Diagram showing form of sphincter of large specimen.
Diagram showing shape and layering of transverse section o
Oc. 3, obj.

Muscular part of a medium-sized mesentery. Oc. 3, obj. $.
specimen.

Diagram to show shape of sphincter.

Exact detail of a small part of sphincter. Oc. 3, obj.

Part of transverse section of a tentacle. Oc. 3, obj. §.
Processes from basal part of a directive mesentery. The letter
in the mesogloea. Oc. 3, obj. §.

Diagram showing form of sphincter.

Part of transverse section of a tentacle.

Muscle-pennon of a perfect mesentery. Oc. 3, obj. f.

Sagartia, sp. Diagram of shape of sphincter. Oc. 3, obj. §.

„ Detail of small portion of sphincter. Oc. 3, obj. T*7.

Aiptasioides prima. Upper part of sphincter. Oc. 3, obj.

Sagartia, sp. Part of transverse section of a tentacle. Oc. 3, obj.

Aiptasioides prima. Part of transverse section of a tentacle. Oc. 3, obj, j
Sagartia, sp. Perfect mesentery, partially restored. Oc. 3, obj. f.

— Calliadis reticulata.

— Aiptasioides prima.

Oc. 3, obj. §•.

f acontium.
From small

m is placed

3rit. Mus. (Nat. Hist.)

Brit. Antarctic (Terra Nova) Exped. 1910.

Zoology, Vol. V. Coelenterata, Part I,

Actiniaria.

PI. V.

1

. Stephenson del.

Adlard & Son & West Newman, Ltd

Coelenterata, Part I.— Actiniaria, PI. VI.

PLATE VI.

Aiptasia minima. Fig. 1.
Aiptasioid.es prima. Fig. 11.
Artemidactis victrix. Figs. 5-9, 12.
Calliactis reticulata. Fig. 2.
Cymbactis selaginella. Fig. 14.
Epiactis novo-zcalandica. Figs. 3, 4.
Halcampa chrysanthellum. Fig. 16.
Halcampoides aspera. Fig. 1 3.
Hormosoma scotti. Fig. 10.
Phymactis clematis. Fig. 15.

b. en., body-wall endoderm.

c. l., ciliated lobe of mesenterial filament.
c.w., column- wall.

ec., ectoderm.

e.c.m., endodermal circular muscle.
en., endoderm.

/., fosse.
g., gonad.

g.l., glandular lobe of mesenterial filament.
l.m., longitudinal musculature.

m., mesogloea.

m.en., mesentery-endoderm.

oe., actinopharynx.

pb.m., parieto-basilar muscle.

p.d., pedal disc.

s.ec., siphonoglyphe-ectoderm.

sph., sphincter.

z., Zooxanthellae.

1-6, cycles of mesenteries.

Fig. 1. — Aiptasi a minima. Diagram to show general structure of muscular part of a large mesentery.

Oc. 3, obj. -§.

-Calliactis reticulata. One pair of directive mesenteries. Oc. 3, obj. $.

-Epiactis novo-zealandica. Sphincter. Oc. 3, obj. 1 1 .

- ,, „ Muscular part of a large mesentery. Oc. 3, obj. i{.

\ Artemidactis victrix. Granular gland-cells from mesentery-endoderm. Oc. 3, obj. T'5.

Cells (amoeboid) from mesentery-endoderm. Oc. 3, obj. T'5.

Fig. 2.-
Fig. 3.-
Fig. 4.-
Fig. 5.-
Fig. 6.-
Fig. 7.-
Fig. 8.-
Fig. 9.-
Fig. 10 -
Fig. 11.-
Fig. 12.-

Fig. 13. — Halcampoides aspera.

i

T?'

-Hormosoma scotti. Cell (amoeboid) from tentacle-endoderm. Oc. 3, obj.

-Aiptasioides prima. Two Zooxanthellae from the tentacle-endoderm. Oc. 3, obj. T'2.
-Artemidactis victrix. Diagram to show the arrangement of the mesenterial filaments in distinct

zones on the mesenteries of the various cycles.

The tip of a muscle-process from the pennon, showing the fingers of
mesogloea, with the oblong muscle-fibres (in transverse section)
lying round them in the endoderm. Oc. 3, obj.

-Cymbactis selaginella. A small muscle-process from the pennon of a mesentery. Oc. 3, obj. T'5.
-Phymactis clematis. A small mesenterial filament to show the curious distribution of Zoo-
xanthellae in the mesogloea, and nuclei in the endoderm. Oc. 3,
obj. -}.

Longitudinal section of a tentacle, and the upper part of the
body-wall, to show the sphincter (from a Plymouth specimen).
Oc. 3, obj. §.

Fig. 14.
Fig. 15.

Fig. 16. — Halcampa chrysanthellum.

Brit. Mus. (Nat. Hist.'

Brit. Antarctic (Terra Nova) Exped. 1910.
Zoology, Vol. V.

Coelenterata, Part
Actiniaria.

A Stephenson del.

Adlard & Son & West Newman. Ltd,

.

.

(grtftef (Museum (QUfuraf gteforg).

This is No. V of 25 copies of

« Nova" Zoology , F?/. F, Coelcnterata ,

Part II., printed on Special paper.

69

COELENTERATA.

PART II— MADREPORARIA.

(a) On Favia conferta, Verrill, with Notes on other Atlantic

Species of Favia *

BY GEORGE MATTHAI, M.A.,

Professor of Zoology, Government College, Lahore, Punjab
( late Mackinnon Student of the Boyal Society and sometime Besearch Student of Emmanuel College, Cambridge).

WITH TWO FIGURES IN THE TEXT AND FOUR PLATES.

I. — Introduction .

II. — Explanation of Nomenclature

III. — Colony-formation .

IV. — Description of Favia conferta, Verrill

V. — Notes on other Atlantic Species of Favia

VI. — References ......

Index ......

FAGE

69

70
73
73

80

94

95

I.- INTRODUCTION.

The Astrseid corals collected by the British Antarctic (“ Terra Nova”) Expedition, which
Professor J. Stanley Gardiner, F.R.S., handed over to me for description, consisted of
six specimens of moderate size (none of these were entire colonies), and a few smaller
pieces in alcohol. These were all obtained at Station 37 (off South Trinidad, 20° 28' S.,
29° 25' W.), and, though at first I had regarded them as referable to Favia fraguin
(Esper), they are, on further study, found to belong to the related species Favia conferta,
Verrill. Three of the specimens were subsequently dried for their coralla (the largest
being No. B, in table on p. 90) ; one containing a few polyps on part of its calicinal
surface is still kept in alcohol, and the remainder have been decalcified for their polyps
in 2 per cent, solution of nitric acid in 75 per cent, alcohol. Since the pieces had been
plunged into alcohol without previous fixation, f the soft parts are not in as good

* This work was done by me as Mackinnon Student of the Royal Society, aided by grants from
Emmanuel College, Cambridge, and from the Francis Maitland Balfour Fund of the University.
[Manuscript received July 31, 1918. — S. F. Harmer.]

| According to information given by Mr. D. G. Lillie, member of the Expedition.

VOL. v.

L

70

“TERRA NOVA” EXPEDITION.

'condition for histological study as might be desired. Seven polyps were sectioned,
viz., five transversely (table on p. 78) and two longitudinally. The morphology of the
soft parts of this species is herein described for the first time.

Four additional species of Favia have been recorded from the Atlantic region,
viz., F. fragum (Esper), F. gravida, Verrill, F. leptophylla, Verrill, and F. ichitfiddi ,
Verrill, whilst a sixth species, which is an undoubted Favia, is Orbicella aperta, A errill.
In 1915, I was able to examine Verrill’s types of all the Atlantic species of Favia in
the Yale University Museum and in the American Museum of Natural History, New
York, as well as the large collections of F. fragum in the l nited States National
Museum, Washington. Among the exchange specimens for the University Museum,
Cambridge, entrusted to me by Dr. Wayland Vaughan, are a dried corallum of
F. conferta, and another of F. gravida from Brazil, described on p. 90.

In July, 1915, while at the Marine Biological Station, Tortugas, I had the
opportunity of collecting a large number of colonies of F. fragum at different stages of
growth, some thirty of which were fixed in 10 per cent, formaldehyde after narcotisation
in a solution of magnesium sulphate in sea-water in a partially extended condition, and
over a hundred were dried. Live colonies of this species were also kept under observa-
tion in aquaria of running sea-water, some of which extruded larvae, but not in any
abundance at that time of the year. In 1914, Professor Stanley Gardiner kindly placed
at my disposal a well-preserved colony of F. fragum, which he had received from
Dr. Wayland Vaughan, from South Bight, Andros Island, Bahamas. The whole of this
colony has been decalcified, and seven series of sections were made, six transverse
(table on p. 85) and one longitudinal. Thus, it has been possible to make a detailed
comparison of the morphology of the hard and soft parts of F. conferta with that of
F. fragum.

The probable relationships of the Atlantic species of Favia, based on data obtained
from the above sources, are also discussed in this Eeport.

II.— EXPLANATION OF NOMENCLATURE.

Bilateral symmetry. — The lateral compression of the stomodaeum of a polyp and
the presence of a directive couple of mesenteries at each of its ends.

Body-ioall. — The entire wall of a polyp surrounding the coelenteric cavity, viz.,
the oral disc with the tentacles above, the basal disc below, and between these laterally
the column-wall.

Bud. — A new polyp formed outside or inside the tentacular ring of an older polyp.

Calyx. — The cavity enclosed in a corallite.

Coenosarc. — The soft parts of a colony overlying the peri theca.

Colline. — The calcareous partition between two adjacent valleys.

M ADR EPOR ARIA — MATTHAI.

71

Colony. — The entire soft and hard parts of a Madreporarian coral containing more
than one stomodaeum.

Column-wall. — The vertical portion of the body-wall of a polyp, forming the
lateral boundary of its coelenteric cavity.

Coralhte. — The part of the corallum in which a monostomodaeal polyp lies.

Corail it e-icall. — The wall of a corallite inclusive of theca and any perithecal

deposit.

Corallum. — The hard parts of a Madreporarian colony.

Edge-zone. — The soft parts of a colony covering the outer free surface of the
wall of one of its corallites ; indistinguishable in some cases.

Enterostome. — The lower opening of the stomodaeum into the coelenteric
cavity.

Entoccele. — The part of the coelenteric cavity included within a couple of
mesenteries.

Exoccele. — The part of the coelenteric cavity included between two neighbouring
couples of mesenteries.

Hexameral symmetry. — The presence of six, or some multiple of six, mesenteries
around the stomadseum of a polyp.

Interstomodceal septum. — A septum under the enteroecele of an interstomodaeal
couple of mesenteries.

Mesenteries : —

Principal mesenteries meet the stomodaeum, subsidiary mesenteries do not.

Couple of mesenteries. — Two adjacent mesenteries of a polyp with the
pleats of their middle lamina vis-a-vis.

In a completely principal couple, both the mesenteries meet the
stomodaeum ; in an incompletely principal couple, one of the mesenteries meets
the stomodaeum whilst the other does not ; in a subsidiary couple , neither of
the mesenteries meets the stomodaeum.

An interstomodceal couple lies between two adjacent stomodaea in a
common tentacular ring, each mesentery being related to one of the
stomodaea. Such a couple is completely principal when the two mesenteries
are attached to their respective stomodaea, incompletely principal when only
one of the mesenteries is attached to a stomodaeum, or subsidiary when neither
of the mesenteries is attached, their inner margins being free.

Middle lamina. — A descriptive name for the supporting lamella present between
the ectoderm and endoderm of coral colonies, and forming the median core of
the mesenteries.

Nematocysts* — Minute structures occurring in the epithelia of Coelenterate
animals and said to be used for stinging.

* For a full account of nematocysts in Astrreid corals, vide 5, pp. 10-17.

L 2

72

“TERRA NOVA” EXPEDITION.

In type I , otherwise known as the “ tentacular type,” the sac of the nematocvst is
more or less trumpet-shaped and contains a filament spirally coiled against the inner
surface of its membranous w-all. The number of turns of the spiral varies in different
species. From the base of the sac a slender process can be traced to the middle lamina.
Nematocysts of this type attain their full size in the tentacular ectoderm, where they
are closely arranged to form batteries, but are found more or less scattered in the
edge-zone, oral disc, stomodseal ridges and the straight regions of the mesenterial
filaments.

In type II, the sac is larger, broad at one end and bluntly pointed at the other.
A narrow “ axis ” extends from the bluntly pointed end of the sac to about two-thirds
its length, consisting of an “axial core” enclosed in a thin membranous sheath, against
the inner service of which lies a close spiral. The axial core is usually straight,
sometimes slightly bent, and tapers towards its pointed outer extremity. Found
principally in the convoluted regions of the mesenterial filaments, where they may
form batteries.

In type Tib, the sac is much narrower and somewhat longer than in type II and
the axis does not extend beyond a third of the length of the sac. Found in the
terminal batteries of tentacles, interspersed among those of type I.

In type III, the sac is larger than in I or II, being broader in the middle than at
the ends, and it contains a protoplasmic core staining more deeply than the rest of the
contents. A long “ thread ” is spirally coiled along the inner surface of the membranous
wall, thicker but with fewer turns of the spiral than in I. The thread consists of an
“ axial strand ” enclosed in a thin membranous sheath, against the inner surface of
which lies a fine closely wound spiral. Occurring mainly in the coils of the mesenterial
filaments.

Oral disc. — The circumoral wrall of a polyp bounded by the outermost cycle of
tentacles.

Pali. — The upright calcareous rods arising from the basal plate of a corallite
external to the columella.

Peristome. — The circumoral space of a polyp enclosed by the oral disc and
tentacles.

Peritheca. — That part of the corallum of colonial Madreporaria which is deposited
outside and subsequently to the theca.

Polyp. — Any part of the soft parts of a colony which has a distinct circumoral
tentacular boundary.

Monostomodceal , distomodeeal, or tristomodeeal, when containing one, two
or three stomodsea respectively within a common tentacular boundary.

An intramural poly stomodeeal polyp contains more than three stomodsea,
arranged in a single row along its middle.

Mother-polyps are those polyps in which intratentacular buds are formed.

Valley.- — Part of the corallum in which a polystomodseal polyp lies.

MADREPORARIA — MATTHAT.

73

III.— COLONY-FORMATION.

It is now found that, in the Astrseidse, colony -formation takes place by extra-
tentacular and intratentacular budding, i.e., by the formation of new stomodaea, hence
of buds, outside or inside the tentacular rings of older polyps ; in both cases new
stomodaea arise afresh in diverticula (text-figs. 1 and 2) by invagination of the oral disc
or by union of the margins of the broader mesenteries, without involving the longitudinal
fission of existing stomodaea. In intratentacular budding one or more stomodaea arise
within the same tentacular ring, and one or two couples of mesenteries between every
two stomodaea (text-fig. 1, ax a2, b1 K, and cx c2). The bud is separated from the
mother polyp by the growth inwards and fusion of the parts of the column- wall over
the interstomodaeal septa (e.g., of those below the entocoeles of bx b2 and cx c2), and of
the latter to form a dividing wall between the bud and parent corallites (PI. I, figs. 18
and 23).

In the genus Favia iutratentacular budding results in distomodaeal and
tristomodaeal polyps, i.e., polyps containing two and three stomodaea respectively.
Tristomodaeal polyps are linear when the two buds are formed at opposite ends, usually
in the sagittal plane, of the mother-polyp (PI. I, figs. 20-22), triangular when the two
buds arise as a pair from neighbouring entocoeles of the mother-polyp (PI. I, figs. 10
and 17). Mono- and di-stomodaeal polyps are more numerous in colonies than tri-
stomodaeal ones. Intramural polystomodaeal polyps, i.e., polyps containing more than
three stomodaea in a median row, when present, are short and few, and finally break
up into mono-, di- or tri-stomodaeal polyps. In all these cases, the planes of intra-
tentacular buds not infrequently disagree with those of their mother-polyps (PI. HI,
fig. 6), although the planes of oral disc areas invariably correspond with those of their
respective stomodaea. While colony-formation in Favia is mainly by intratentacular
budding, extratentacular budding unaccompanied by bilateral and hexameral symmetries
occurs at the growing edges of colonies.*

o o o

IV.- DESCRIPTION OF FAVIA CONFERTA, Verrill.

Favia f conferta, Verrill. (Text-fig. 1’ ; PI. I, figs. 1-27 ; PI. II, figs. 1-3, 6, 10,
11 ; PI. Ill, figs. 2, 8 ; PI. IV, figs. 1, 3, 4, 7, 10.)

1868. — Favia conferta, Verrill, Trans. Conn. Acad., I, p. 355.

1902. — Mceandra conferta, Verrill, Trans. Conn. Acad., XI, pp. 84 and 188.

1903. — Mceandra conferta, Verrill, Zoology of Bermudas, I, pp. 84 and 188.

* A full account of Colony -formation in Astraeid Corals will be given in another paper which, it is
hoped, will be published in the near future.

f I have given a diagnosis of the Genus Favia on a previous occasion (5, pp. 77-79), to which I do
not propose to add anything further in the present state of our knowledge of the Genus.

74

“TERRA NOVA” EXPEDITION.

Localities. — “ Terra Nova," Stat. 37, off South Trinidad (six specimens). Other
sources, Abrolhos Reefs ; Bahia and Fernando Noronha ; Mar Grande, Bahia ;
Pernambuco. Cape Frio to Pernambuco, common in tide-pools (Yerrill).

CORALLUM.

Sub-hemispherical, small, compact. Corallites oval in outline, or laterally
compressed (when containing mono- and di-stomodaeal polyps), or triangular (when
containing tristomodseal polyps); width of their calices 2*5 mm. to 3 mm., depth
2*5 mm. In addition, a few short valleys which are irregularly linear or curved in
shape and up to 15 mm. or 20 mm. in length. Walls solid in transverse section, about
1 ’25 mm. in thickness, often with shallow median furrows.

Septa in two alternating series of broad and narrow ones, with vertical inner
margins ; the former about *75 mm. apart (about 20 such septa in corallites of mono-
stomodseal polyps), with spinulose sides and exsert to about • 5 mm. These broad septa
become perforated a short distance below corallite-rims, the perforations often breaking-
through inner margins, thus resulting in coarse irregular teeth which increase in length
towards upper level of columella ; the lower teeth tend to be obliquely raised, but
do not form a definite palicrown. Most of the broad septa meeting the columella,
the rest curve towards and fuse with the neighbouring septa meeting .columella ;
exsert ends arched, dentate, meeting in notches along middle of walls. Narrow
septa with irregularly toothed inner margins, slightly exsert, and often meeting-
over walls. Columella spongiform, composed of interlacing septal trabecula, and
occupying inner one-third of width of calices ; columellar centres often difficult
to distinguish, since, in valleys, columella extend as ridges of more or less even
thickness. Endothecal dissepiments concavo-convex (as seen in vertical section), ’5
to *75 mm. apart.

Yerrill at first described this species under Favia, subsequently reclassifying it
under Mceandra. But in all the specimens examined, corallites (of mono- and
di-stomodaal polyps) are far more numerous than valleys, the latter being invariably
short and tending to be divided into corallites by the formation of incipient transverse
partitions. In the Yale University Museum are five dried specimens assigned by Yerrill
to Favia confer ta, a list of which is given on p. 89. In addition to these, there are
two more specimens belonging to F. conferta ; one of them (without number) resembles
Verrill’s figured type but has thicker corallite-walls, in the other (Cat. No. 4519)
corallite-rims tend to be elevated, thus making the inter-corallite grooves conspicuous,
as in Favia gravida, Yerrill.

POLYPS.

Polyps are mostly mono-, di- and tri-stomodseal, polystomodseal polyps being-
few and containing four to six stomodsea (PI. I, figs. 1-25). Linear tristomodseal

MADREPORARIA — MATTHAI.

75

and polystomodseal polyps are not infrequently curved in shape (text-fig. 1,
PI. I, figs. 26, 27).

Body-wall. — The ectoderm of the oral disc is of more or less uniform thickness
(*025 mm. to *04 mm.)* over its entire extent, with a translucent free border, and it
contains numerous elongated nuclei (mostly granular) along its middle. The outer half
of the layer contains oval mucous vacuoles, whilst in the granular inner half a few round
and oval nuclei lie scattered. Short nematocysts I with about twenty-five turns of the
spiral are occasionally present in the ectoderm. The middle lamina is thin. The

Tkxt-fig. 1. — Fuvia conferia, Ver. Transverse section through the stomodreal region of a curved tristo-
modseal polyp. I, II, III, Stomodsea of polyp. «j a2, bl h2, cx c2, Interstomodaeal couples of mesenteries.
Opposite stomodseum I are two terminal intratentacular diverticula, containing seven and five
subsidiary couples of mesenteries separated by the mesenterial couple 1 , 1 2. Opposite stomodaeum III
is another terminal intratentacular diverticulum, containing four subsidiary couples of mesenteries.
The hard parts are represented by dots. (Constructed from Camera lucida tracings of serial
transverse sections of 400 VI, “Terra Nova.”)

endoderm is thinner ('01 5 mm. thick) than the overlying ectoderm, somewhat
lobulated but not vacuolated to the same extent as in F. fragurn. The nuclei are
round or oval, homogeneously stained dark in Heidenhain’s iron hsematoxylin, or with
a dark central spot. Algse are plentiful but are not massed.

* The histological measurements are all taken from sections.

76

TERRA NOVA” EXPEDITION.

Both entocoelic and exocoelic tentacles are uniformly short, being in sections about
•2 to '3 mm. long and ’2 mm. in diameter (PI. II, fig. 6). Their terminal batteries
arc nearly semicircular in transverse section and are about ’06 mm. in height. In the
outer half of the terminal battery nematoeysts I are massed. When fully formed these are
nearly ' 03 mm. long and having a somewhat larger number of turns of the spiral than in
F. fragum, viz., about 40. Among nematoeysts I are a few \\!> ones, ‘035 mm. to
*04 mm. in length and '0075 mm. in width, whose axis (stained dark in iron
haematoxylin) extends to about one-third the length of the sacs, in which a faint coiled
thread is visible. In the inner half of the battery lie numerous nuclei, the lower ones
being round or oval and the upper ones elongated. Fibrils are distinctly seen in the
lower half, becoming continuous with the underlying middle lamina. Sub-terminal
batteries are not visible in longitudinal sections through any of the tentacles. The
endoderm is swollen, tending to fill the lumina of the tentacles, and it contains
numerous algae.

The column-wall is quite thin in the upper half of the polyp. The calicoblastic
ectoderm continues to the base of the polyp ; in this region the endoderm is transparent
and reticular, being swollen to ‘07 or ‘08 mm. thick, and its nuclei are gathered
towards its periphery. Algae are scarce in the endoderm.

Stomoclamm (PI. II, fig. 10). — In a distomodseal polyp 3 mm. in height (as seen
in longitudinal section) the stomodaea are about ' 5 mm. long and are wider at the
enterostome than at the mouth-opening, whilst in a monostomodaeal polyp 2 * 5 mm.
high the stomodaeum is * 35 mm. long. The stomodaeal ectoderm is ciliated, as usual.
Although the polyps are small, the stomodaea have well-developed ridges, which are
about • 06 mm. broad and twice as thick, and whose sides arc sloping when wide apart or
vertical when the ridges are approximated. The intervals between the ridges vary in
the different stomodaea, doubtless due to the state of preservation and embedding. In
the lower half of the stomodaeum, shallow median grooves are present on the ridges.
The intermesenterial parts are usually folded outwards. The histology of the stomodaeal
ectoderm is typical. The middle lamina is not thickened in the ridges. The endoderm
is much thinner (about *01 mm.) than the ectoderm of the intermesenterial parts
(about ’025 mm.), and algae are rare.

Mesenteries (PI. II, fig. 10). — On an average nine to eleven mesenteries are present
around a stomodaeum. In the stomodaeal region, entocoelic pleats extend over the outer
two-thirds of width of principal mesenteries, not so close together as in F. fragum,
gradually increasing in width towards the inner one-third of pleatal region wdiere they are
broad, somewhat constricted at their bases and knobbed at the tips, often sub-divided
and slender in transverse section, then becoming narrower over the rest of the pleatal
region. The mesenteries decrease in width from above downwards, most of them
stopping short of the bases of polyps. The endoderm appears to increase in width
from the column-wall to stomodaeum. Below the stomodaeal region entocoelic
pleats extend over the greater part of the width of the mesenteries, as is usually

MADREPORARIA— MATTHAI.

77

the case in the Astraeidse. Numerous algae are present on the exoccelic side,
massed in places.

Mesenterial filaments (PL III, fig. 8) are more or less semicircular in transverse
section, in which granular vacuoles are seen. In their convolutions nematocysts III are
present, having long narrow sacs (‘06 mm. to '065 mm. long and ’01 mm. to
*0125 mm. broad), resembling those in Gonicistreci retiformis (5, PI. VI, fig. 61) with
15-20 turns of the thread which, in some nematocysts, is seen to be partially extruded.
In the present histological condition of the soft parts it is difficult to determine if these
nematocysts are arranged in specialised regions of the convolutions as in F. fragum,
although in sections they do not appear to be so numerous as in the latter species.
Xematocysts II occur in the convolutions as well as in the straight regions of the
filaments, but appear mostly broken in the sections. Some parts of the convolutions
are swollen, stained brown in Eosin, with nuclei lying somewhat wide apart along their
middle, hence are probably glandular in nature. In the principal and the broader
subsidiary mesenteries, the filament-epithelium usually extends down to the greater
part of the distance below the stomodseal region, after which it disappears. The
convolutions are not so scarce within the ccelenteric cavity as in F. fragum, perhaps
because they are not protruded into the peristome to the same extent.

Gonads (PL II, figs. 1-3). — In nearly every polyp sectioned, male, female, and
hermaphrodite mesenteries are present. In male mesenteries there are up to four or
five small spermaries within spaces in the middle lamina. The minute, somewhat
triangular heads of the spermatozoa are homogeneously stained dark in iron
hsematoxylin, but they are so massed in the spermaries that their tails cannot be
detected in the sections. One to three groups (oval or circular in transverse section)
of germ-cells are present in some of the mesenteries (about seven such mesenteries
in polyp series 400 VII). Each germ-cell has a comparatively large nucleus consisting
of a definite nuclear membrane enclosing somewhat granular contents, in which
lie one or two nuclei stained dark in iron hsematoxylin. When two nucleoli
are present they appear to be connected by a fine curved strand, also stained
dark. The cytoplasmic areas are not well defined in the existing state of fixation
of the tissues. The mesenterial endoderm surrounding the groups of germ-cells is
usually swollen and granular. Ova are small, about * 1 mm. in diameter, lying within
the middle lamina of mesenteries ; occasionally three ova are seen in a mesentery at
the same level pressed against one another. In mesenteries containing growing ova
the endoderm surrounding them is swollen and granular, whilst in those containing
ripe ova the endoderm returns to its original condition, being frequently thinned by
the distension caused by the ova, as is better seen in the polyps of F. fragum that
have been studied. Ova occur in some mesenteries containing spermatozoa, as well as
in some others enclosing groups of germ-cells. Sexual products do not, as in the
Astrseidse generally, appear in the mesenteries above the level of the enterostome,
except in the case of some subsidiary mesenteries.

VOL. V.

M

78

“TERRA NOVA” EXPEDITION.

Table of Mesenteries of Polyps Sectioned Transversely'.
nij = principal mesentery. ni2 — subsidiary mesentery. A, B, C, stomodaea of incipient buds.

Species.

Polyp series.

Locality.

Stomodoea.

No. of complete
principal couples.

No. of incomplete
principal couples.

Total No. of
principal couples.

No. of subsidiary
couples.

Total No. of couples.

Mesenteries of
inters tomodaeal
couples.

No. of principal
mesenteries.

No. of subsidiary
mesenteries.

Total No. of mesen-
teries.

G *
•a so
»- o

«*- o
°

6 ©
25 £

ajj

£ *
O

Intra tentacular
diverticula.

Favia

conferta,

Verrill

400

III

J)

JJ

>>

Off South
Trinidad,
“ Terra
Nova ”
Stat. 37
>>

J J
J )

I

II
A
B
C

3

4
3

4

0

1

7

4

4

3

3

1

10

7
5

8
7

lm, facing stom.
II, lm, it lm,
facing stom. A

lm, facing stom. I
2m, facing stom. I

12

9

9

11

6

3

23

15
12

16
14

\

>10

One terminal diver-
ticulum opposite
stom. I, and two
terminal diverti-
cula separated by
a principal couple
of stom. II have
just become buds.

Favia

400

Off South

I

5

3

8

8

16

13

19

32

13

At one end are

conferta,

IV

Trinidad,

five contiguous

V errill

“ Terra

subsidiary mesen-

Nova ”

teries, perhaps

Stat. 37

lying in an in-

cipient diverti-

culum.

Favia

400

Off South

I

4

3

7

5

12

lm, facing stom. II

11

14

25

\

Opposite stom. II

conferta,

V

Trinidad,

is the commence-

Verrill

“ Terra

)1 1

ment of a termi-

Nova ”

1 1

nal diverticulum

Stat. 37

containing three

J J

II

3

5

8

7

15

lm2 facing stom. I

11

20

31

contiguous sub-

sidiary couples of

mesenteries.

Favia

400

Off South

I

5

nil

5

13

18

lm, facing stom. II

10

27

37

\

Two terminal diver-

conferta,

VI

Trinidad,

ticula containing

V errill

“ Terra

seven and five

Nova ”

subsidiary couples

Stat. 37

9

separated by a

JJ

JJ

II

2

3

5

4

9

lm, facing stom. 1,

9

12

21

principal couple

2m, facing stom.

of stom. I, and a

Ill

terminal diverti-

>5

})

III

2

3

5

7

12

2m, facing stom. II

9

17

26

culum containing

four contiguous

subsidiary couples

opposite stom. Ill

(text-fig. 1).

Favia

400

Off South

I

3

3

6

4

10

lm, facing stom. II

9

12

21

Opposite stom. Ill

conferta,

VII

Trinidad,

are five contiguous

Verrill

“Terra

subsidiary mesen-

Nova ”

teries, perhaps

Stat. 37

I11

lying in an in-

>>

>5

II

3

2

5

5

10

lm, facing stom. I,

9

14

23

cipient terminal

lm, & lm, facing

diverticulum.

stom. Ill

J J

J )

III

4

5

9

8

17

lm, & lm2 facing

14

22

36

stom. II

MADREPORARIA— MATTHAI.

79

The differences of specific value between F. conferta and F. fragum may be
summarised as follows : —

Favia conferta. Favia fragum.

Polyps : —

1. A few intramural polystomodseal polyps
present in colonies, containing four to six stomodaea.

2. Linear tristomodseal and polystomodaeal polvps
not infrequently curved in shape.

3. Both entoccelic and exoccelic tentacles
uniformly short and without subterminal batteries.

4. Stomodaeal ridges without nematocysts ; middle
lamina not thickened in ridges.

5. In stomodaeal region, entoccelic pleats extend-
ing over outer two-thirds of width of principal
mesenteries, not so close together as in F. fragum,
gradually increasing in width towards inner one-
third of pleatal region.

6. Nematocysts III in convolutions of mesen-
terial filaments appearing to be less numerous
than in F. fragum.

7. In the endoderm generally, numerous algae
are present.

Corallum : —

1. A few short valleys present in Corallum.

2. A series of narrow septa regularly alternating
with broad ones, hence as many as the latter.

3. Broad septa exsert to • 5 mm.

4. Broad septa without definite paliform rods.

5. Broad septa not meeting columella, far fewer
than those that meet, the former curving towards
and fusing with the latter.

1. Polystomodaeal polyps almost absent.

2. Linear tristomodceal polvps, when present,
rarely curved in shape.

3. Entoccelic tentacles varying in length, the
larger ones with six or seven subterminal batteries
in a longitudinal section ; exoccelic tentacles
uniformly short with three or four subterminal
batteries.

4. Stomodaeal ridges containing nematocysts III,
up to ten, closely arranged in transverse section ;
middle lamina considerably thickened in ridges.

5. In stomodaeal region, entoccelic pleats not
usually extending beyond outer half of width of
principal mesenteries, arranged close together,
gradually increasing in width from column-wall to
middle of pleatal region.

6. Nematocysts III present in large numbers in
convolutions of mesenteries.

7. In the endoderm, algae are less numerous than
in F. conferta.

1. Valleys rare or absent.

2. Narrow septa not forming a regularly alter-
nating series, hence fewer than broad ones.

3. Broad septa exsert to less than • 5 mm.

4. Broad septa with rough upright paliform rods.

5. Broad septa not meeting columella, nearly as
many as those that meet ; the former curving
towards and fusing with the latter or ending off
with free margins.

NOTES ON POLYPS SECTIONED.

Favia conferta, Verrill.

Series of longitudinal sections : —

(1) 400 I, a monostomodaeal polyp.

(2) 400 II, a distomodseal polyp, oral disc region between the two mouth-openings
becoming constricted, being 4 * 5 mm. long and 2 mm. broad in alcohol.

Series of transverse sections

(3) 400 III, a discontinuous polystomodaeal polyp from edge of colony. Oral disc
region between mouth-openings I and II becoming constricted, being 8 mm. long and
2 • 5-3 mm. broad in alcohol. Only one interstornodaeal couple (completely principal)

m 2

80

“TERRA NOVA” EXPEDITION.

present between stomodsea 1 and II at this late stage, hence it is unlikely that a second
interstomodseal couple will be formed before division takes place wThere polyp is con-
stricted. At one end (opposite stomodseum I) an intratentacular bud has arisen by
the formation, of a new stomodseum A to wrhich nine mesenteries appear to have been
joined. Stomodsea I and A appear to be connected by two interstomodseal couples
(one completely principal and the other incompletely principal). At the other end
(opposite stomodseum II, towards the edge of colony) a pair of intratentacular buds
have arisen by the formation of two new stomodsea B and C. It is difficult to
determine the number of mesenteries meeting stomodsea B and C, but obviously they
wTill be related to stomodseum II on the triangular tristomodseal mode.

(4) 400 IV, a monostomodseal polyp.

(5) 400 V, a distomodseal polyp. Measurements in alcohol are : total length of oral
disc, 6 mm. ; greatest width, 2 mm. ; oral disc region between the two mouth-openings,
2 -25 mm. long and constricted to a width of 1 mm. Only one interstomodseal couple
between stomodsea I and II (as between stomodsea I and II of 400 III), now sub-
sidiary, which may or may not become principal before the polyp divides.

(6) 400 VI, a linear tristomodseal polyp having a curved shape. Distance between
adjacent mouth-openings, 2 ’25 mm. Two interstomodseal couples (both completely
principal) between stomodsea II and III, but only one such couple (subsidiary) between
stomodsea II and I. As yet no indications of stomodseal formation in any of the three
intratentacular diverticula (text-fig. l).

(7) 400 VII, a linear tristomodseal polyp. A separate tentacular ring has beeu
formed around stomodseum I, although the column- wall has not as yet begun to be
divided. Only one interstomodseal couple (subsidiary) between stomodsea I and II, as
between stomodsea I and II of 400 VI. The sagittal plane of stomodseum I (hence
also of its oral disc region) lies more or less at right-angles to that of stomodsea II
and III. — (Best specimen for histology.)

V.-NOTES ON OTHER ATLANTIC SPECIES OF FA VIA.

1. FA VIA FRAGUM (Esper).

(Text-fig. 2; PI. I, figs. 28, 29; PI. II, figs. 4, 5, 7, 8, 9; PI. Ill, figs. 1, 3-7, 9;

PI. IV, figs. 5, 6, 9, 12.)

(For a discussion of the synonymy of this species, vide 7, pp. 34-40.)

The skeletal characters subject to variation in F. fragum are the growth-form of
coralla, size of calices, thickness of corallite-walls, elevation of corallite-rims and
consequent presence of intercorallite furrows, thickness of septa, dentition of inner
margins of septa, shape of exsert ends of septa, and the degree of development of
columellse. These points are illustrated by three representative colonies E, F and G
in table on p. 90, which are connected by numerous gradational forms in my collection

MADREPORARIA— MATTHAI.

81

from Tortugas. Two to four colonies are sometimes united by their contiguous sides,
thin epithecal rims being then visible over the line of union.

The largest example of F. fragum in the Cambridge University Museum is an
entire more or less hemispherical colony from St. Vincent, 4‘ 5 cm. high, 7 cm. long, and
5 • 5 cm. broad ; corallites being present on its entire exposed surface. On one side of the
specimen, corallite -walls are thin (1 mm. to V25 mm.) and ridged above, with septa
meeting over the walls in gable-form (as in specimen E on p. 90). On the opposite side,
corallite-walls are much thickened (up to 3 mm.), and
more or less flat (as in specimen F on p. 90). Calices
are shallower and paliform rods are well formed.

There is a regularly alternating series of narrow septa
in the corallites of this specimen, as in F. conferta.

Examples of F. fragum with thin ridged walls,
pentagonal corallites and conspicuous paliform rods
(as in No. E on p. 90) bear some resemblance to
certain colonies of the Indo-Pacific F. pentagona
(Esp.) (5, PI. XXIV, fig. 3) ; but in the latter
species, septa are farther apart, hence are fewer in
corallites, calices deeper, septal sides much less rough,
and the corallum appears to attain to a larger size.

Body-wall (PI. II, fig. 9 ; PI. Ill, figs. 4 and 6).

Incompletely formed nematocysts II and short
nematocysts I with about twenty turns of the spiral
are occasionally present in the ectoderm. The
nuclei of the endoderm, though generally scattered

be

gathered towards its free border. Algae are neither
abundant nor massed in the endoderm.

Tentacles* (PI. II, fig. 7) are usually found to
decrease in length towards the ends of the longer
diameter of the oral disc, being as short at the
poles as the exoccelic tentacles. Nematocysts I of
terminal batteries have about thirty-five turns of the
spiral, among which are many nematocysts II b. Sub-terminal batteries are arranged
more or less in rings around the tentacles, and are ‘03 mm. to *05 mm. in height;
short nematocysts I, having about twenty-five turns of the spiral, are gathered in the
middle of the outer half of each of these batteries, forming an outwardly diverging
group, below which the ectoderm is granular and translucent, containing numerous
elongated nuclei ; on each side of the group of nematocysts the ectoderm is vacuolated,

* In live colonies the tentacles appear to be transparent except the batteries which are white and
opaque.

in its highly vacuolated protoplasm, tend to

1mm

Text-fig. 2. — Favia fragum (Esp.).
Transverse section through the
stomod;eal region of a monostomo-
daeal polyp with a terminal intra-
tentacular diverticulum, containing
five subsidiary couples of mesen-
teries. The hard parts are repre-
sented by dots. (Constructed from
Camera lucida tracings of trans-
verse serial sections of 500,- VI.,
South Bight, Andros Island,
Bahamas.)

82

“TERRA NOVA” EXPEDITION.

clear and with few nuclei. The endoderm is somewhat lobulated and thinner than the
overlying ectoderm ( • 02 mm. to • 03 mm. thick) in the sub-terminal region, in which
algse are not abundant. Both ectoderm and endoderm decrease in thickness towards
the base of tentacles.

Owing to the better fixation of the polyps of this species, it was possible to make
a more detailed study of the histology of the column-wall than could be done with
F. conferta, but doubtless in this respect the latter species will be found to correspond
with F. fragum. The column-wall is ’003 mm. to ’005 mm. thick in the upper half
of polyp. The middle lamina is considerably attenuated, being absent in places where
the calicoblastic ectoderm merges into the endoderm. The calicoblastic nuclei are wide
apart and arranged tangentially in a row. Where the mesenteries join the column-
wall, the calicoblastic layer frequently becomes vacuolated and swollen, and stages in
the formation of the processes of attachment are commonly seen ; here the calicoblast,
in which nuclei lie scattered, is up to ‘03 mm. in thickness and the endoderm is about
*01 mm. thick. The nuclei of the endoderm are round or oval and arranged
horizontally. Algae are almost absent everywhere from the column-wall endoderm.

Stomodceum (PI. II, figs. 4 and 8 ; PL III, figs. 4 and 6). In some distomodaeal
polyps of colonies from Tortugas, new stomodaea ate recognisable on a surface-view
below the oral discs without corresponding mouth-openings. In such cases the new
stomodaea are probably formed by the union of the margins of the broader mesenteries
of diverticula and secondarily acquire external openings. In a distomodaeal polyp,
3 ’ 25 mm. in height (as seen in longitudinal section), the stomodaea which are 2 mm.
apart are ‘3 mm. and ‘25 mm. long, becoming wider from mouth-opening to
enterostome ; in alcoholic specimens mouth-openings of distomodaeal polyps are
2‘5 mm. to 3‘5 mm. apart. Nematocysts III in ridges are about ‘033 mm. long and
•01 mm. broad, extending from below the striated border of the ridge to the greater
part of its width. Nematocysts II are occasionally present in the ridges. Nuclei of
the endoderm are gathered to form a peripheral layer.

Mesenteries (PI. II, figs. 5, 8 and 9 ; PI. Ill, fig. 5). The endoderm is of more or
less even thickness over the greater part of the width of mesenteries, getting much
attenuated for a short distance from column-wall, where the middle lamina is
quite thin.

The endodermal pad behind the filament is well developed, being as large as, in
transverse section, or larger than the filament itself. In the straight region of
filaments nematocysts II ( ‘ 04 mm. long) are present, whose axes extend beyond the
middle of sacs, but not to two-thirds their length. Nematocysts III also occur, similar
to those in the stomodseal ridges, but are fewer in number. In certain regions of the
convolutions the filament-epithelium is vacuolated, swollen, somewhat transparent, and
with comparatively few nuclei — resembling, in histological condition, the endoderm
behind. In such parts of the filament-epithelium long narrow nematocysts III (similar
to those described in F. conferta ) are closely arranged, their threads having remained

M ADR EPOR ARIA— MATTHAI.

83

coiled within the sacs ; among these, nematocysts II are sometimes included. In some
parts of the convolutions the swollen filament appears to consist of a row of narrow
columnar bodies, radiating from its base to periphery, stained brown in Eosin and
containing small globules. A somewhat similar condition has been noticed in F.
conferta, being probably glandular in both cases. Sometimes dumb-bell shaped
transverse sections are seen with this glandular epithelium at one end, and the swollen
filament containing closely arranged nematocysts III at the other end (PI. II, fig. 5).
In mesenterial coils with normal filament (as in the straight region) nematocysts II
are seen, nematocysts III, when present, resembling those in stomodseal ridges. In
most of the polyps examined, convolutions of mesenteries are protruded through
edge-zone, oral-disc, and mouth-opening, and lie in knots on the surface. Perhaps as
a result of this, convolutions are scarce within the polyps, and are almost absent towards
their base (PI. Ill, figs. 4 and 6), the intermesenteric spaces appearing more or less
empty in transverse section. In longitudinal sections the three layers are seen to end
off at the openings through which convolutions are protruded, the ectoderm and
endoderm narrowing towards them ; if they were permanent pores (and not due to
rupture), the ectoderm would be continuous through the openings with the endoderm
(PI. II, fig. 9).

Ova are present below the stomodseal region in up to ten mesenteries in a polyp,
a single ovum occupying nearly the entire width of a mesentery, whose endoderm is
then quite thin ; rarely two ova occur in a mesentery at the same level. Spermaries
were not found in any of the mesenteries, although Duerden records the presence of
hermaphrodite mesenteries in this species (I, p. 572).

The histology of the middle lamina of Astrseid Corals is clearly exemplified in
the polyps of F. fragum that have been sectioned. The wavy unbranched fibres
forming the mass of the middle lamina are evident in tangential sections (also in
F. conferta). The plexus of branching fibres is well seen in the thickened regions at
the stomodseal ridges, in which numerous nuclei lie scattered ; each of these nuclei
usually has a thin investment of granular cytoplasm, which is drawn out into
filamentous processes, finally merging in the substance of the middle lamina (PI. II,
fig. 4). The plexus and nuclei are also apparent in sections through tentacles.

Solid embryos or planulse have been found to lie free in the coelenteric cavities of
the sectioned polyps of F. fragum from Bahamas (PI. Ill, figs. 4, 7 and 9). About
eight or ten such embryos have been counted in polyp series 500/ III as well as in 500/
IV. In longitudinal section each of them is dome-shaped, perhaps owing to contraction
(cf. 2, PI. I, fig- 6). The middle lamina has appeared as a thin continuous fibrous
membrane. An outer ectoderm is thus marked out which is of even thickness along the
flat or somewhat concave aboral surface, but increases in thickness towards the oral end,
and is provided with long cilia over its entire surface. The inner mass lies subjacent
to the middle lamina, and consists of an endodermal stratum (thinner than the
ectoderm) surrounding a central reticulum. Both the ectoderm and the endoderm are

84

TERRA NOVA” EXPEDITION.

stained pink in Eosin, nucleated and containing numerous particles of yolk obscuring
the cytoplasm and nuclei. In contrast to these, the central reticulum appears more or
less clear and sparsely nucleated, and is evidently breaking down to form a coelenteric
cavity. Algse occur everywhere in the embryo, mainly in the endodermal stratum, whilst
those in the ectoderm will presumably either disintegrate or migrate into the endoderm,
since both in the free-swimming larva and in the polyp, algae are entirely restricted to
the endoderm lining the coelenteric spaces. At the narrow end of the embryo is a blind
invagination into the central reticulum, being the commencement of the stomodseum.
Nowhere in the protoplasm of the embryo is there the slightest indication of cell-
structure. Nematocysts have not arisen at this stage.

Some of the embryos in the polyps mark a stage before the commencement of
invagination, but in every embryo examined the middle lamina has appeared as
described above, and seems to increase in thickness at the expense of the adjacent
protoplasm. In Manicina areolata, according to H. V. Wilson, the endodermal mass is
formed by delamination from a hollow blastula (10, pp. 195-199) — a stage that has
not been found in any of the sectioned polyps of F. fragum — whilst the correspondence
between the central reticulum of the embryo and the vacuolated condition of the ripe
ovum suggests that the former might be the result of the withdrawal of the egg-
protoplasm towards the periphery, in which also come to lie the products of the repeated
division of the egg-nucleus without corresponding division of the cytoplasm. By the
complete vacuolisation of the central reticulum a coelenteric cavity would result (PI. Ill,
figs. 1 and 3).

NOTES ON POLYPS SECTIONED.

Series of longitudinal sections : —

(1) 500/ V, a distomodseal polyp.

Series of transverse sections : —

(2) 500/ I, a piece from the edge of colony, containing a distomodseal polyp and
two extratentacular buds, the latter being situated towards edge of colony. In
the distomodseal polyp a thin ridge has appeared across the oral disc, between the two
mouth-openings ; the tentacles on each side of this ridge are getting directed towards
the adjacent mouth-opening preparatory to the formation of two distinct tentacular
rings. As many principal couples of mesenteries are present around stomodseum A as
around I. Column-walls of extratentacular buds separate from each other and from
that of distomodseal polyp. In each extratentacular bud, nearly as many principal
mesenteries have appeared as around each stomodseum of distomodseal polyp.

(3) 500/ II, a triangular tristomodseal polyp from the edge of colony, the pair of
buds being at the edge. Buds subequal in size ; each of them though smaller than the
mother-polyp has as many principal couples as the latter. Entocoelic tentacles appear
(in alcohol) to decrease in length towards the two arms of the oral disc, being as short
at the ends of the arms as exoccelic tentacles.

MADREPORARIA — MATTHAI.

85

Table of Mesenteries Sectioned Transversely.

mj = principal mesentery. m2 = subsidiary mesentery. A, B, stomodjea of incipient buds.

! © CQ

® m

TO

5 o5

Species.

02

<D

©

1 ®

1 o

; p-*

>5

'c3

o

o

*

c3*

8

o

a

o

•+3

m

-2 rE,

a §

o °

1 £ 'cis

O 3
• o

ft

&

S o

go

3

.&
o '5

• 3

o -a
'A »

O 'ft
. 3
O O
£ *

J .S4

o ’©
EH .2

ft

2 «
C/2 ©

-S ft
« O

o °

6

Hir

CH

O

o’ CO-
_ ft

Jg

o o
EH

Mesenteries of
interstomodaeal
eouples.

ft •

.-V in
o .©

a

■p ©

<4-1 ©

O g

6 g
A

c3 .
m

"in

©

5 "2
©

^r* W

O ©

d s
A

<4-1 B

O co

. ©

o-rj

A S

"c3 §

o o

* 3

ft’C

^ ©
O -+3

. a

o ©
h m
©

© 9
to H
eg ’-p
© ft
> ’©
<1

Intratentaeular

diverticula.

Favia

500,.

Andros

i

3

3

6

9

15

1 m, & 1 m2 facing

10

22

32

On one side of stom.

fragum,

I

Island,

stom. A

I are three ad-

Esper

Bahamas,

» 1 0

jacent subsidiary

Brit. Mus.

couples, and on an-

(.Nat. Hist.)

other side five

5?

yy

A

3

3

6

5

11

1 m, & 1 m., facing

10

14

1 24

contiguous subsi-

stom. I.

diary mesenteries

yy

III

4

2

6

4

10

10

10

20

which may mark
the beginnings of

yy

yy

IV

3

3

6

4

10

9

11

20

two diverticula.

Favia

500,

Andros

I

2

3

5

5

10

2 m, facing stom.

10

14

■24

Nil.

fraqum,

II

Island,

A ; 1 mj & 1 m.

Esper

Bahamas,
Brit. Mus.

facing stom. B

'

10

(Nat. Hist.)

yy

A

3

3

6

2

8

lm, 4 1 m„ facing

10

8

18

stom. I

•

y y

yy

B

3

2

5

3

8

1 m, & 1 m.2 facing

9

9

18

stom. I

Favia

500,

Andros

I

4

3

7

10

17

1 m, & 1 m. facing

12

24

36

Three contiguous

fragum,

III

Island,

stom. II

subsidiary couples

Esper

Bahamas,

■ 12

of mesenteries on

Brit. Mus.

one side of stom.

(Nat. Hist.)

I, perhaps form-

yy

yy

II

5

2

7

6

13

1 mt & 1 m„ facing

13

15

28

ing the commence-

stom. I

ment of a terminal
diverticulum.

Favia

500,

Andros

I

3

1

4

6

10

2 mj facing stom. II

9

13

22

Opposite stom. I is

fragum,

IV

Island,

an incipient ter-

Esper

Bahamas,

' 8

minal diverticu-

Brit. Mus.

lum containing

(.Nat. Hist.)

three contiguous

yy

yy

II

2

1

3

5

8

2 mj facing stom. I.

7

11

18

subsidiary Couples
of mesenteries.

Favia

500,

VI

Andros

I

4

2

6

9

15

1 m2 & 1 m2 facing

11

21

32

At one end of stom.

fraqum ,

Island,

stom. II

Ill is a terminal

Esper

Bahamas,

> 9

diverticulum con-

Brit. Mus.

taining five con-

(Nat. Hist.)

tiguous subsidiary

yy

II

3

nil

3

6

9

1 m, 4 1 m2 facing

7

13

20

couples of mesen-

stom. I

teries (text-fig. 2).

III

5

2

7

12

19

12

26

38

On one side of stom.

II are three con-

tiguous subsidiary
couples, perhaps
forming the com-
mencement of a
terminal diverti-
culum.

Favia

500,

Andros

I

3

4

7

8

15

10

20

30

10

At one end of polyp

fragum,

VII

Island,

is an incipient di-

Esper

Bahamas,

verticulum con-

Brit. Mus.

taining three con-

(Rat. Hist.)

tiguous subsidiary
couples of mesen-

1

teries.

VOL. v. N

86

“ TERRA NOVA” EXPEDITION.

(4) 500/ III, a distomodaeal polyp from the edge of colony. The sagittal plane of
stomodaeum II (also of its oral disc region) at right angles to that of stomodaeum I as
noted in 400 VII. Two interstomodaeal couples (one completely principal and the
other subsidiary) between the two stomodaea.

(5) 500^ IV, a distomodaeal polyp. The incipient intra tentacular diverticulum
with an oblique direction of growth, w'hich is indicated on surface by a similarly
directed oral disc region. The latter with a bluntly pointed extremity towards which
entocoelic tentacles decrease in length. Each of the tentacles over the interstomodgeal
couples being directed towards one of the mouth-openings. Two interstomodaeal couples
(both completely principal) connecting the two stomodaea. (Best specimen for
histology.)

(6) 500/ VI, a piece containing a distomodaeal and a monostomodaeal polyp from
the edge of colony. In the distomodaeal polyp the sagittal plane of stomodaeum II is
nearly parallel with the transverse plane of stomodaeum I ; oral disc narrowing to a bluntly
pointed end beyond mouth-opening II. Two interstomodaeal couples (one completely
principal and the other subsidiary) between stomodaea I and II. There is no sign as
yet of the formation of a new stomodaeum in the terminal intratentacular diverticulum
next to stomodaeum III.

(7) 500/ VII, a monostomodaeal polyp.

Localities. — Azores (Quelch) ; Bermudas (Verrill) ; West Indian and Caribbean
region, abundant on the Florida Reefs (Vaughan). Fossil in Curagao ; Arube ; also in
the elevated reefs of other West Indian Islands (Vaughan).

2. FA VIA GRA VIDA, Verrill.

Verrill (8, p. 354, 9, pp. 91, 188, PI. XIII, fig. 3), differs from F. conferta in
having fewer valleys, somewhat wider (3 to 3‘5 mm.) and deeper (3 mm.) calices.
The corallite-rims are elevated up to 3 mm. and the vesicular perithecal matter is
increased up to a thickness of 3 or 4 mm., hence the intercorallite grooves are more
conspicuous and neighbouring thecae are not fused. The septa are somewrhat thicker,
with a greater number of teeth on their inner margins, rougher sides and are exsert to
•75 or 1 mm. The costae are alternately broad and narrow (corresponding to the two
series of septa within) with transversely extending granulations and meeting in
intercorallite -furrows where the latter are not too wide. The columellae are formed of
closer texture. But these differences appear to be varietal and if, as is probable,
future examination of the soft parts of F. gravida should reveal no greater differences,
the two species would have to be combined under the earlier name F. gravida, as
Vaughan (7, p. 40) suggested from a study of the hard parts of these species.

When corallite-rims are conspicuously elevated and are wide apart, the corallum
of F. gravida appears to assume the facies of Dichocoenia. The elevated corallite-rims
bounding deep intercorallite-grooves, the two alternating series of broad and narrow

MADREPORARIA — MATTHAI.

87

septa, and the two such series of costse (with transversely extending granulations) meet-
ing in the grooves, recall the Indo-Pacific species F. wakayana (Gard.) and F. solid tor
(Ed. and H.) (5, pp. 105 and 106).

Localities. — Abrolhos Reefs ; Pernambuco and Bahia ; Cape Frio to Pernambuco,
common in tide-pools (Yerrill).

It would appear from the above list of localities that F. conferta and F. gravida,
and F. fragum, have a mutually exclusive distribution, i.e., the first two have been
recorded only from the east coast of Brazil between Pernambuco and Cape Frio, whilst
F. fragum has not been known south of the Caribbean Sea ; St. Vincent being, up
till the present, the southernmost limit of its recent members.

It is necessary to add that there is considerable difficulty in determining whether
the Favia fragum of Vaughan and Verrill or the Favia gravida of Verrill corresponds
to Esper’s Madrepora fragum. For example, the statements in Esper’s diagnosis that
“ die Sterne sincl sehr erhoht ” and “ die Lamellen wecliseln mit grossern und kleinern
ab” (3, p. 79) would seem to be more applicable to F. gravida * as also the appearance
in his figures (PI. LXIV, figs. 1 and 2) of conspicuous furrows between the raised
corallite-rims. Since Esper remarks, in regard to his specimens, that “ Ihr Aufenthalt
ist in den siidlischen Americanischen Meeren ” (p. 80), whilst F. fragum, as defined by
Vaughan and Verrill, has not yet been recorded south of the Caribbean Sea, in contrast
to F. conferta and F. gravida, neither of which has been found north of Pernambuco on
the Brazilian Coast, it still remains an open question whether Esper was not dealing
with one of the twro species subsequently constituted by Verrill. Till this point has
been settled by actual reference to Esper’s types, any discussion of the synonymy of
these species cannot have more than provisional value.

The three Atlantic species, F. conferta, F. gravida and F. fragum, form a
homogeneous sub-section within the genus Favia. The characters common to them
may be stated as follows : —

Corallum.

1. Maximum growth-size comparatively small. f

2. Corallites never far apart, hence perithecal matter never abundant.

3. Calices of mono- and di-stomodseal polyps not more than 4 • 5 mm. wide and
3 mm. deep.

4. Both broad and narrow septa present.

5. Broad septa "7 5 to 1 mm. apart, many of them meeting columella.

6. Columella spongiform, not more than one-third the width of calices.

* In one half of a small colony of F. fragum 2 ’5 cm. high, 3 '5 cm. long, and 2- 5 cm. broad, from
Tortugas, the corallite-rims are raised to 1'5 or 2 mm., and cost* are present as in F. gravida, but the
specimen is apparently an instance of abnormal skeletal variation.

f Colonies presumably die when the maximum growth-size has been reached, which in F. fragum
appears to be somewhat less than in F. conferta. According to Verrill, colonies of F. conferta are two or
four inches in diameter (8, p. 355). In all three species a strongly wrinkled epitheca is present on the free
part of the under surface of colonies.

N 2

88

“TERRA NOVA” EXPEDITION.

Polyps (as seen in F. conferta and F. fragum).

1. Polyps small, mostly mono-, di- and tri-stomodseal ; polystomodaeal polyps
when present, few, short and temporary.

2. Although polyps are small, stomodsea have well-developed ridges which, in
transverse section, are twice as thick as broad.

3. Average number of principal mesenteries around a stomodaeum, nine to eleven.

4. Entocoelic pleats of principal mesenteries as seen in transverse sections
extending to the middle or outer two-thirds of their width ; gradually increasing in
breadth towards the middle or inner one-third of the pleatal region where they are
broad, somewhat constricted at their bases and knobbed at the tips, often subdivided and
slender in transverse section, then becoming narrower over the rest of the pleatal region.

5. Convolutions of mesenteries never abundant. '

6. Nematocysts III in coiled regions of mesenterial filaments long and narrow,
each with 15-20 turns of the thread.

The soft parts of F. fragum , and to a less degree of F. conferta, bear some
resemblance, in the following respects, to the Indo-Pacific species F. pentagona (Esper)
(5, pp. 96 and 97). The polyps, though small, possess well-developed stomodaeal ridges.
The appearance of principal mesenteries in transverse section is somewhat similar —
entocoelic pleats not extending beyond the outer two-thirds of the width of mesenteries,
being broad, thin and sometimes subdivided in the middle of the pleatal region,
elsewhere narrow. The comparative thinness of the endoderm and the fewness of
algae. Convolutions of mesenteries never abundant in the coelenteric spaces, being
almost absent towards the base of polyps. F. fragum agrees with the Indo-Pacific
Goniastrea retiformis (Id., p. 119) in the thinness of the endoderm, scarcity of
mesenterial convolutions, number of principal mesenteries around a stomodaeum, and
in the presence of long narrow nematocysts III occurring in the coiled regions of
mesenterial filaments.

In my study of Astraeid corals different sexual conditions have been found, viz.,
(l) the presence of either ovaries or spermaries in any polyp of the same colony, as in
Cyphastrea serailia (5, pp. 31 and 40); (2) the presence of both ova and spermatozoa in
the same polyp within separate male and female mesenteries or within hermaphrodite
mesenteries, all three kinds of mesenteries occurring in the same polyp, as in Facia
conferta; (3) the presence of ova only in all polyps of the same colony,* as in Facia
fames, F. hululensis (5, pp. 82, 88), etc.; (4) the absence of reproductive elements from
all polyps of the same colony as noticed in many species of Facia, Echinopora, etc.
From these records it would appear that in an Astraeid colony there is no differentiation
into sexual and asexual polyps, every polyp being capable of performing both the nutritive
and generative functions. Nor is it unusual to find, in the same species, colonies of
different sizes containing gonads, which may mean that in the same colony the sexual
and nonsexual phases alternate more than once, or as Gardiner suggested that “ in an

* I have not up till now found a colony in which only male elements are present in its polyps, although
such a condition is not improbable.

MADREPORARIA — MATTHAI.

89

area — owing probably to some change in the physical conditions of their environment —
all the corals of some single species have been stimulated to ripen and dehisce their
generative products at the same time, the act leading ultimately to their weakening
and death ’ (4, p. 471). All available evidence seems further to indicate that in any
one Astrseid species there is a maximum size beyond w'hich colonies do not grow, as is
exemplified in F. conferta, and F. fragum, and as observed by Gardiner in Flabellum
and other Madreporarian corals, but we have no direct evidence of the factors which
determine the limit of growth of colonies.

List of Verriel’s Specimens of F. coxferta, F. gravida and F. fragtdi in Yale University Museum.

Species.

Catalogue

Number.

Locality.

j Height
in cm.

Length
in cm.

Breadth
in cm.

Remarks.

Favia conferta,
Veri-ill

1466

Abrolhos Reefs,
Brazil

2-5

6-5

4-5

Typical. Longest valley, 17 mm.
in length. Edge - region of
this specimen figured by Verrill
(9, PL XIII, fig. 6).

Favia conferta,
Verrill

1466

Abrolhos Reefs,
Brazil

4-5

6-0

4-0

•

Corallum entire. Corallite- walls
somewhat thicker than in
the above specimen.

Favia conferta,
Verrill

1466

Abrolhos Reefs,
Brazil

—

6-0

3-5

Corallum incrusting.

Favia conferta,
Verrill

4520

Mar Grande,
Bahia, Brazil

3-0

»o

o

r-H

7-0

Favia conferta,
V errill

4546

Pernambuco,

Brazil

3-5

5-0

4-0

Favia (jravida,
Verrill

1465

Abrolhos Reefs,
Brazil

5-0

4-75

Corallum entire, incrusting ; one,
sigmoid valley, in which be-
ginnings of two transverse
partitions are present. Edge-
region of this specimen figured
by Verrill (9, PL XIII, fig. 3).

Favia (jravida,

V errill

4549

Pernambuco,

Brazil

3-0

4-0

3-5

No valleys, but three or four
corallites which appear to have
contained tristomodseal polyps.

Favia f ragum,
(Esper)

6747

Bermudas

2-0

4-5

4-0

Corallum entire. Corallite-walls
thick, with shallow median
furrows (9, PL XIII, fig. 2).

Favia f ragum,
(Esper)

6748

Bermudas

—

4-75

4-5

Corallum entire and incrusting.
Corallite-walls thin (9, PI.
XIII, fig. 1).

90

“TERRA NOVA” EXPEDITION.

* Table showing Skeletal Variation in F. conferta, F. gravida, and F. fragum.

Specimens. Valleys. Corallites of mono-

AND DI-STOMOD.EAL POLYPS.

Species.

Number.

Locality.

Shape.

Measure-
ments in
cm.

Shape.

Measure-
ments in
mm.

Shape.

Average
measure-
ments in
mm. of
calices.

Favia conferta,
Verrill.

(PI. IV, figs. 1,
7 and 10.)

A

Off South
Trinidad.
“Terra
Nova ”
Stat. 37.

Corallum broken on one
side along entire height,
narrowing from top to
bottom. Calicinal surface
convex, restricted to head
of specimen, with a narrow
shallow depression along
its middle.

h, 7'5
1, 5

greatest
b, 4

Irregularly
linear or
curved.

Greatest
1, 15
av. w.,
2-5 to 3
av. d, 2 5

Oval or
laterally
compressed.

av. 1, 4 to 6
av. w, 2-5 to 3
av. d, 2 5

Favia conferta,
Verrill.

(PI. IV, figs. 3
and 4.)

B

Ofi South
Trinidad.
“Terra
Nova ”
Stat. 37.

Corallum broken on every
side, broader at bottom
than at top, to which cali-
cinal surface is restricted.

h, 7
1, 7

greatest
b, 7

Two more
or less
linear
valleys, the
shorter
bifurcated
at one end.

1, 15 and
13-5

av. w, 2-5
av. d, 2-5

As in A.

As in A.

Favia conferta,
Verrill.

C

Brazil.
Exchange
specimen
from U.S.
Nat.

Museum,

Washing-

ton.

Corallum entire, but
greater part dead.

h, 4-5
1, 7-3
b, 3-7

Linear.

Greatest
1, 20
av. w, 2
to 2-5
av. d, 2-5

As in A.

As in A.

Favia gravida,
Verrill.

(PI. IV, figs. 2,
8 and 11.)

D

As in C.

Corallum entire, more or
less hemispherical, calicinal
surface partly destroyed in
patches and with four or
five bore - holes. Some-
what concave attached sur-
face.

h, 4-5
1,7
b, 6

As in A.

Greatest
1, 14
av. w, 3
to 3-5
av. d, 3

As in A.

av. 1, 4 to 6
av. w, 3 to 3-5
av. d, 3

Favia fragum
(Esper).

(PI. IV, fig. 5.)

E

Tortugas.

Corallum entire, narrow-
ing towards one end, slight
depression across middle of
calicinal surface, flat

attached surface.

h, 3
1,7

greatest
b, 6

No val

leys.

Often penta-
gonal.

av. 1, 4 to 6
av. w, 3'5
to 4-5
av. d, P5

Favia fragum
(Esper).

(PI. IV, fig. 6.)

F

As in E.

Corallum entire, incrust-
ing, convex calicinal sur-
face with a bore-hole.
United with one side is a
much smaller corallum,
with thin epitheoal rim
visible over line of union.

1,5
b, 3

One linear
valley.

1, 12
w, 2-5
d, 2-5

Oval or
laterally
compressed.

av. 1, 4 to 6
av. w, 2'5
av. d, 2-5

Favia fragum
(Esper).

G

As in E.

Corallum entire, exposed
surface evenly convex and
covered with corallites.

h, 2-5
1, 5-5
b, 4-5

No val

leys.

Often penta-
gonal.

av. 1, 4 to 6
av. w, 3 to3'5
av. d, 2'5 to 3

I

h = height ; 1 — length ; b = breadth ; w = width ; d = depth ; av. 1 = average length ;
av. w = average width ; av. d = average depth.

* The specimens referred to in this table (pp. 90, 91) are in the collection of the British Museum (Natural History).

MADREPORARIA — MATTHAI.

91

Table showing Skeletal Variation in F. coypERTA, F. gravida, and F. fragum.

COBALLITES OF
MONO- AND DI-
STOMOD.EAL POLYPS

Septa.

Walls.

Shape and Size.

Inner margins.

Exsert

Ends.

Columella.

Kemarks.

Average thick-
ness, 1-25 mm.

Greatest thick-
ness, 2 mm.

Solid, owing to
fusion of neigh-
bouring thecse.

Median grooves
when present
shallow.

Two alternating
series of broad and
narrow septa ; most
of the former meet-
ing columella, the
rest curving towards
and fusing with
principal septa.

Inner margins of
both series vertical.

With coarse ir-
regular teeth in-
creasing in length
towards level of
columella, lower

teeth obliquely
raised.

Narrow septa also
with irregular mar-
ginal teeth.

•5 mm.
Arched, den-
tate, meeting
in notches
over middle
of collines.

Of spongy
texture, about
a third of
width of
corallites.
Columellar
centres not
often distin-
guishable
in valleys.

A species of Colonial Ascidian
incrusting on parts of the dead
surface of corallum.

In one half of
specimen as in A ;
the other half
having gravida

facies — i.e., coral-
lite - rims raised
up to 1 mm., and
up to 2'5 mm.
apart, with con-
spicuous grooves
between them.

As in A.

As in A.

As in A.

Feeble,
composed of
a few loosely
interlocking
trabeculae,
almost absent
in a few
corallites.

Distinct corallite-rims with
conspicuous grooves between
them are characteristic of this
specimen.

Incrusting on one side is a thin
piece of a Porites, also a species of
Colonial Ascidian and a few
small Actinians.

Largest specimens of F. con-
ferta in “ Terra Nova ” collection.

Thickness as in A.

Corallite - rims
more distinct than
in A, hence grooves
better marked, but
not so deep as in
F. gravida.

Somewhat thicker
than in A.

Larger number of
teeth than in A.

•75 mm.
Arched,
dentate.

Well formed,
of closely
twisted
trabeculae.

Three valleys stretch across
existing calicinal surface, but
transverse partitions being formed
in them.

This specimen approaches D in
the presence of corallite-rims and
grooves between them and in
septal characters.

Owing to in-
crease of perithecal
matter, av. thick-
ness 1'5 mm. to
2 mm. Greatest
thickness 4 mm.

Corallite - rims
raised to 3 mm.,
with correspond-
ingly deep grooves
between them.

Somewhat thicker
than in A.

Larger number of
teeth than in A,
owing to greater
depth of calices ;
lower ones tending
to be raised oblique-
ly upwards, hut not
forming a palicrown.

•75 mm. to
1 mm.
Arched,
dentate.

Of closer
texture than
in A, but not
to the same
extent as in C.

Corallum lighter than in A
owing to increase of vesicular peri-
thecal matter. Corallites coarser
owing to septa being somewhat
thicker, with rougher sides, and
to more marginal teeth.

One region having conferta
facies, since corallite-rims are
slightly raised, walls 1 mm. thick,
inter-corallite grooves narrow and
shallow.

Av. thickness,
1 mm.

Mostly thin and
ridged above.

Broad septa slop-
ing from walls for
some distance into
calices, then falling
vertically to colu-
mellee.

Somewhat more
irregular than in A.

Bough upright
paliform rods on
broad septa.

Not quite
so exsert
as in A.

Meeting over
walls in
gable-form.

Spongiform,
one third
width of
calices, or
somewhat
less.

Walls thin and ridged, septa
slightly exsert and meeting over
walls in gable-form, and prominent
paliform rods are characteristic of
this specimen.

Thin variety.

Av. thickness,
T5 to 2 mm.

Greatest thick-
ness, 3 mm.

Upper surface
more or less flat,
with a median row
of low tubercles.

Somewhat thicker
than in E.

Broad septa fall-
ing vertically to
columellee.

Less irregular

than in E, with
2-4 short spiniform
teeth.

Paliform rods less
conspicuous than in
E.

More or less
flat, with
transversely
extending
granulations.

Less developed
than in E,
loosely spongy,
but composed
of thicker
trabeculae.

Walls considerably thickened
and flat above, exsert ends of
septa more or less flat, calices
deep but narrow, columellse of
open texture, septal margins more
or less definitely dentate are
characteristic of this specimen.

Thick variety.

Av. thickness,
1‘25 or 1-5 mm.

Shallow median
grooves, except in
certain parts.

Broad septa fall-
ing vertically to
columellas.

In some corallites
short spiniform
teeth as in P, in
others longer and
irregular.

Paliform rods

better formed than
in F, but less than
in E.

Arched.
Some of the
broad septa
more exsert
than others,
up to '5 mm.

Better
developed
than in F,
but somewhat
less than
one-third
width of
calices.

Intermediate between E and F
in most characters.

92

“TERRA NOVA” EXPEDITION.

3. FA VIA LEPTOPIIYLLA, VerrilL

Verrill (8, p. 353, 9, p. 91, PL XIII, figs. 4 and 5). The type from Ahrolhos
Reefs, Brazil (Cat. No. 1517), in Yale University Museum, measures G'5 cm. iu
height and 11 cm. iu diameter. A piece (3 '5 cm. broad) has been cut off from
this specimen aud mounted ou a separate stand for the vertical view. In general
appearance Ver rill’s type somewhat resembles dried colonies of the Indo-Pacific Favia
hululensis, Gard., although it forms an entirely separate species (5, p. 87, and cf. PI. XXII,
fig. 6 and PI. XXXV, fig. 1, with VerrilTs figure). The principal characters of this specimen
are as follows : calicinal surface convex, peritheca vesicular, hence corallum light,
corallites narrower than in F. hululensis, corallite-rims distinct, the thin thecae being
elevated to a short height, septa thin, few and with highly exsert arched <>r truncated
ends, costae often meeting over the perithecal regions except when the latter are wide ;
an alternating series of very narrow septa present which have no costae or only feebly
marked costae, columella loosely spongy. The vertical section of this specimen resembles
that of the Indo-Pacific F. clouei, Val. (5, PI. XXXI V, fig. l), since the peritheca is
highly vesicular in both species, and consequently the corallum is light.

4. OB BICE LI A APERT A, Verrill.

Verrill (8, p. 356, 9, p. 103, PI. XXXIII, figs. 1 and 1„). The specimen in
Yale University Museum on which this species is based is a large corallum from
Abrollios Reefs, Brazil, 15 ’5 cm. high, 20 ‘5 cm. long, and 10 ‘5 cm. broad (Cat.
No. 1518), which appears to be only a skeletal variety of F. leptophylla , the main
difference being that in 0. aperta there is much more perithecal matter, hence the
calices are wider apart. Such variation is found to occur in many species of Favia,
e.g., F. doreyensis, Ed. and II. (9, pp. 84-87). On a humpy region at one end of the
specimen the condition, owing to the narrower perithecal regions, Is similar to that of
Verrill’ s type of F. leptophylla. Towards the opposite end the perithecal regions are
much wider, hence calices are far apart, and the cost® of neighbouring corallites become
continuous over the perithecal areas as in some examples of F. doreyensis. The vertical
section of 0. aperta is identical with that of F. leptophylla.

5. FA VIA WHI TFIELDI, Verrill.

Verrill (9, p. 132, PI. XXV, fig. 5). The two specimens from Nassau, New
Providence, now in the American Museum of Natural History (Cat. No. 543),
which have been referred to this new species, have convex calicinal surfaces, the
smaller one being Verrill’s figured type. These coralla, while quite different from
those of any other Atlantic species of Favia, so greatly resemble some examples of the

M A DREPOR ARIA — MATTHAI.

93

Indo-Pacific Faria fams, Forsk. (5. PI. XXI. tig. l). that they would be brought under
the latter species if the Atlantic corals were not. as a rule, specifically different from
those of the Indo-Pacific region. In Verniks types of F. whitfield i the eorallites are
circular in outline when small, but become laterally compressed or somewhat triangular
with growth ; there is no definite pali-crown, the lower broader one-third of principal
septa simulating such an appearance as in F. far us, but unlike the latter species some
of the principal septa are thicker and more exsert than others. In the figured specimen
the corallite-rims are more sharply marked and elevated than in the other. In the
same Museum there is a third specimen (from "West Indies or Central America”)
identical with the larger of Yen-ill's types of F. whitfield i, and bearing the same number
and specific name.

94

TERRA NOVA

EXPEDITION.

VI —REFERENCES.

1. Duerden, J. E. — 1902. “West Indian Madreporarian Polyps.” Nat. Acad. Sci., U.S.A., VIII.,

p. 403.

2. Duerden, J. E. — 1904. “ The Coral Siderastrea radians and its Postlarval Development.” Carnegie

Institution, No. 20, Washington, U.S.A.

3. Esper, E. J. C. — 1797. “ Fortsetzungen der Pflanzenthiere, etc.” Niimberg.

4. Gardiner, J. Stanley — 1902. “Some Notes on Variation and Protandry in Flahellum rubrum, and

Senescence in the same and other Corals.” Proc. Camb. Phil. Soc., XI, Pt. IV, p. 463.

5. Matthai, G. — 1914. “A Revision of the Recent Colonial Astraddle possessing distinct Corallites.”

Trans. Linn. Soc., XVII, p. 1.

6. Quelch, J. J. — 1886. “Report on the Reef-Corals.” Challenger Reports, Zool., Vol. XVI, Part

XLVI.

7. Vaughan, Wayland T. — 1901. “ Some Fossil Corals from the Elevated Reefs of Curasao, Arube,

and Bonaire.” Sammlungen des Geologischen Reic-hs-Museums in Leiden. Ser. II, Bd. II,
Heft 1.

8. Verrill, A. E. — 1868. “Notice of the Corals and Echinoderms collected by Prof. C. E. Hartt,

at the Abrolhos Reefs, Province of Bahia, Brazil, 1867.” Trans. Conn. Acad., I, p. 351.

9. Verrill, A. E. — 1902. “Variations and Nomenclature of Bermudian, West Indian and Brazilian

Reef Corals.” Trans. Conn. Acad., XI, p. 63, and “Comparisons of the Bermudian, West
Indian, and Brazilian Coral Faunae,” i. cit., p. 169.

10. Wilson, H. V. — 1889. “On the Development of Manicina arcolata.’’ Jour. Morph., II, p. 191.

MADREPORARIA — MATTHAI.

95

INDEX.

aperta, Orbicella, 70, 92.
areolata, Manicina, 84.

Astrseidae, 73.

Astrseid corals, 69 ; sexual conditions in, 88.
Bilateral symmetry, 70.

Body-wall, 70.

Bud, 70.

Budding, extratentacular, 73; intratentacular, 73.
Calyx, 70.
clouei, Fa via, 92.

Ccenosarc, 70.

Colline, 70.

Colony, 71.

Colony -formation, 71.

Column-wall, 71.

conferta, Favia, 69, 70, 73-SO, 87-90.

„ Mfeandra, 73.

Corallite, 71.

Corallite-wall, 71.

Corallum, 71.

Cyphastrea serailia, 88.

Dichoccenia, 86.
doreyensis, Favia, 92.

Echinopora, 88.

Edge-zone, 71.

Enterostome, 71.

Entoccele, 71.

Exoccele, 71.

Favia, 69.

„ clouei, 92.

„ conferta, 69, 70, 73-80, 87-90; corallum,
74; polyps, 74, 79; localities, 74; skeletal
variation, table of, 90-91 ; mesenteries, table
of, 78.

,, conferta and F. fragum, table of specific
differences, 79.

,, conferta, F. gravida, and F. fragum,
characters common to, 87 ; list of Yerrill’s
specimens, 89.

,, doreyensis, 92.

,, favus, 88, 93.

„ fragum, 69, 70, 80-90 ; morphology, 80-86 ;
localities, 86; polyps, 81 ; mesenteries, table
of, 85 ; skeletal variation, table of, 90-91.

„ gravida, 70, 74, 86, 87; corallum, 86;
skeletal variation, table of, 90-91.

Favia, hululensis, 88, 92.

,, leptophylla, 70, 92; corallum, 92.

,, pentagona, 81.

,, solidior, 87.

,, wakayana, 87.

,, whitfieldi, 70 ; corallum, 92.

favus, Favia, 88, 93.

Flabellum, 89.

fragum, Favia, 69, 70, 80-90.

,, Madrepora, 87.

Goniastrea retiformis, 77, 88.
gravida, Favia, 70, 74, 86, 87.

Hexameral symmetry, 71.
hululensis, Favia, 88, 92.

Interstomodieal septum, 71.
leptophylla, Favia, 70, 92.

Madrepora fragum, 87.

Mseandra, 74.

,, conferta, 73.

Manicina areolata, 84.

Mesenteries, couple, 71 ; completely principal
couple, 71 ; interstomoda?al couple, 71 ; in-
completely principal couple, 7 1 ; principal, 7 1 ;
subsidiary couple, 71.

Middle lamina, 71.

Mother-jwlyps, 72.

Nematocysts, 71.

Nomenclature, 70.

Oral disc, 72.

Orbicella aperta, 70; corallum, 92.

Pali, 72.

pentagona, Favia, 81.

Peristome, 7 2.

Peritheca, 72.

Polyp, 72 ; distomodmal, 72; intramural polysto-
modieal, 72, 73; monostomodaeal, 72; tristo-
modaeal, 72; linear, 73; triangular, 73.
retiformis, Goniastrea, 77, 88.

Septum, insterstomodaeal, 71.
serailia, Cyphastrea, 88.
solidior, Favia, 87.

Symmetry , bilateral, 7 0 ; hexameral, 7 1 .

Valley, 72.
wakayana, Favia, 87.
whitfieldi, Favia, 70.

LONDON : PRINTED BY WILLIAM CLOWES AND SONS, LIMITED, DUKE STREET, STAMFORD STREET, S.E. 1, AND GREAT WINDMILL STREET, W. 1.

’

PLATE I.

Views of polyps from decalcified colonies.

a. Oral view of polyp ; b. basal view of same. X 3.

Figs. 1-27. — Favia conferta, Ver. Off South Trinidad. (“ Terra Nova,” Stat. 37.)

Fig. 1. — Monostomodseal polyp.

Figs. 2-7. — Monostomodseal polyps showing stages of development of intratentacular diverticula. Fig. 7
showing curved oral disc.

Figs. 8-13.— Distomodseal polyps, each with a single drawn out tentacular ring. In each polyp the
mouth-openings lie towards the ends of the sagittal plane of the common oral disc.

Fig. 14. — Distomodseal polyp in which a distinct tentacular ring has been formed around each mouth-
opening, but column-wall still undivided.

Fig. 15. — Distomodseal polyp with an intratentacular diverticulum at each end.

Figs. 16 and 17. — Triangular tristomodseal polyps.

Fig. 18. — Distomodseal polyp in which the common tentacular ring is becoming constricted in the middle.

Fig. 19. — A short discontinuous polystomodseal polyp showing dichotomous branching at one end.

Figs. 20-22.- — Short discontinuous linear polystomodseal polyps.

Fig. 23. — Three monostomodseal polyps formed by the separation of mother-polvp and intratentacular
buds of a linear tristomodseal polyp.

Figs. 24 and 25. — Short discontinuous polystomodseal polyps showing lateral branching.

Figs. 26 and 27. — Curved polystomodseal polyps.

Figs. 28 and 29. — Favia fragum (Esp.). Tortugas. Early stages of colony, tig. 28 being younger than
fig. 29.

Brit. Mus. (Nat. Hist.)

Brit. Antarctic ( Terra Nova) Exped. 1910.
Zoology, Vol. V.

Coelenterata. Part II, PI. I.
Madreporaria .

W. Tams, photogr.

London Stereoscopic Co. imp.

►S

Coelenterata, Part II.— Madreporaria, PI. II.

PLATE II.

Lettering employed : — alg- Algal bodies, cil. Cilia. edg.z. Edge-zone. end. Endoderm. <j.e.

Germ-cells. Ig.f. Longitudinal fibres (unbranched) of the middle lamina, m. Mesentery, m.f. Mesenterial

filament, m.l. Middle lamina, m.l.nu. Nuclei in middle lamina. n3. Type III nematocyst. or.d. Oral

disc. ov. Ovum, ov.nu. Nucleus of ovum. «p. Spermary. st.r. Stomodaeal ridge, sub.b. Subterminal

battery, ter.b. Terminal battery.

Fig. 1. — Favia conferta, Yer. Transverse section through a $ mesentery bearing four spermaries.
Off South Trinidad (“ Terra Nova ”). (400 VII, Sn 1,.) * x *300.

Fig. 2. — Favia conferta, Yer. Transverse section through a 9 mesentery bearing an ovum. The
nucleus of the ovum is large and granular, containing a nucleolus stained uniformly dark
in Heidenhain’s iron luematoxylin. Off South Trinidad (“Terra Nova”). (400 VII, S9 2„.)

X 300.

Fig. 3. — Favia conferta, Ver. Transverse section through a mesentery containing three groups of
germ-cells. Off South Trinidad (“Terra Nova”). (400 VII, S9 17.) X 300.

Fig. 4 .—Favia fragum (Esp.). Transverse section through a stomodaeal ridge and part of the mesentery
meeting it. Part of an adjacent ridge is also figured to show the deep groove between the
two ridges. Note the thickened middle lamina containing nuclei at the attachment of the
mesentery to the stomodaeum. Andros Island, Bahamas. (500, TV, S10 2,.) X 300.

Fig. 5. — Favia fragum (Esp.). T ransverse section through the convoluted region of a mesentery. At one
end of the dumb-bell shaped structure the mesenterial filament (m.f., resembling mesenterial
endodei’m) contains a battery of nematocysts III (ws), while at the opposite end it consists
of granular columns. (500/ V, Sn 24.) Andros Island, Bahamas. X 300.

Fig. 6. — Favia conferta, Ver. Longitudinal section through an exoccelic tentacle. Off South Trinidad

(“ Terra Nova ”). (400 VII, S4 12.) x 160.

Fig. 7. — Favia fragum (Esp.). Longitudinal section through an entoccelic tentacle. Andros Island,
Bahamas. (500y IV, Ss 24.) x 100.

Fig. 8. — Favia fragum (Esp.). Transverse section through a principal mesentery and part of the

stomodaeal wall. Andros Island, Bahamas. (500rIV, S10 12.) X 90.

Fig. 9. — Favia'fragum (Esp.). Longitudinal section through the edge-zone ( edg.z .) and part of the oral
disc (or.d.) of a polyp to show convolutions of mesenteries protruded through the edge-
zone. Andros Island, Bahamas. (500 f V, S,3 23.) x 50.

Fig. 10. — Favia conferta, Ver. Transverse section through a principal mesentery and part of the

stomodaeal wall. Off South Trinidad (“ Terra Nova ”). (400 VII, S6 14.) X 65.

Fig. 11. — Favia conferta, Ver. — Tangential section through a mesentery to show the wavy longitudinal
(unbranched) fibres of the middle lamina. Off South Trinidad (“Terra Nova”). (400
I, Su 2U.) x 120.

*

400 VII, No. of series. Su llf eleventh slide and first section on first row.

Brit. Mus. (Nat. Hist;

Brit. Antarctic ( Terra Nova') Exped. 1910.
Zoology, Vol. V.

Coelenterata, Part II, PI. II.
Madreporaria..

London Stereoscopic Co. imp.

PLATE III.

Lettering employed : — alg. Algal bodies, c.o. More deeply staining core of nematocyst III. col.iv.

Column-wall of polyp, ect. Ectoderm, edg.z. Edge zone. evib. Embryo, end. Endoderm. gr.v. Granular

vacuole, m. Mesentery, m.f. Mesenterial filament, m.l. Middle lamina, muc.v. Mucous vacuole. n.rm.

Membranous wall of nematocyst III. or.d. Oral disc. ov. Ovum. ret. Central reticulum of embryo.

st. Stomodseum. ten. Tentacle. th. Thread of nematocyst III. th.sh. Sheath of thread of nema-
tocyst III.

Fig. 1. — Favia fragurn (Esp.). Longitudinal section through a larva thirty-six hours after extrusion,
showing stomodaeal invagination (st.). The section has not passed through the thickened
aboral region of the ectoderm. Tortugas. (A4 3, Sj 3S.) x 185.

Fig. 2. — Favia conferta, Ver. A nematocyst, type III, from the convoluted region of a mesenterial
filament. In some of the turns of the thread the thin membranous sheath (th.sh.), with its
closely wound inner spiral and the axial strand (stained dark in Iron Haematoxylin), are
visible. (Cf. 5, P1- MI, fig. 61.) Off South Trinidad (“ Terra Nova”). (400 VII, S9 1,.)
X 1250.

Fig. 3 .—Favia fragurn (Esp.). Transverse section through a larva thirty-six hours after extrusion, in
which six mesenteries (m.) have joined the stomodaeum (st.). Tortugas. (A4 2, S, 2n.) x 185.

Fig. 4. — Favia fragurn ( Esp.). Longitudinal section through a distomodaeal polyp. Note convolutions of
mesenterial filaments extruded through one of the stomodaea (st.), three embryos (emh.) in
the ccelenteric cavity, and a mesentery bearing an ovum (ov.). Andros Island, Bahamas.
(500, V, S17 13.) x 13.

Fig. 5. — Favia fragurn (Esp.). Transverse section through part of a mesentery below the stomodaeal
region, showing mesenterial filament (m.f), and the swollen endodermal pad behind.
Andros Island, Bahamas. (500, IV, S15 15.) x 250.

Fig. 6. — Favia fragurn (Esp.). Transverse section through a distomodaeal polyp. Note that the sagittal
planes of the two stomodsea do not agree. Andros Island, Bahamas. (500, VI, Ss 14, vide
table on p. 85.) X 13.

Fig. 7. — Favia fragurn (Esp.). Longitudinal section through an embryo in which a stomodaeal invagination
is taking place at the upper end. Andros Island, Bahamas. (500, V, SJ6 la.) x 225.

Fig. 8. — Favia conferta, Ver. Transverse section through part of a mesentery below the stomodasal
region, showing the mesenterial filament. Off South Trinidad (“ Terra Nova ”). (400 VII,

S7 14.) X 250.

Fig. 9. — Favia fragurn (Esp.). Transverse section through another embryo at the same stage of
development as the one represented in fig. 7. Andros Island, Bahamas. (500, VI, SM 3S.)
X 265.

mnc.v.

mnc.v.

W, Tarr»3 & J Charles, photogr.

London Stereoscopic Co. imp.

Brit. Mus. (Nat. Hist)

Brit. Antarctic (Terra Nova] Exped. 1910.
Zoology, Vol V.

Coelenterata, Part II, PI. III.
Madreporaria.

PLATE IV.

Views of Coralla.

Figs. 1, 3, 4, 7 and 10 are taken from “ Terra Nova” specimens of Favia conferta.

Off South Trinidad, Stat. 37.

Fig. 1. — Favia conferta , Ver. Type, specimen A on p. 90. Off South Trinidad (“Terra Nova”). X If.
Fig. 2.— Favia gravida, Ver. Specimen D on p. 90. Brazil. X If.

Fig. 3. — Favia conferta, Ver. Specimen B on p. 90. Photographed in alcohol with the soft parts on
its calicinal surface to show incrusting Tunicates, a Poritid coral, and an Actinian on
the corallum. Off South Trinidad (“ T erra Nova ”). X .

Fig. 4. — Favia conferta, Ver. Calicinal view of fig. 3. x If.

Fig. 5. — Favia fragum (Esp.). Thin variety. Specimen E on p. 90. Tortugas. X If.

Fig. 6 .—Favia fragum (Esp.). Thick variety. Showing also fusion of two coralla. Specimen F on p. 90,
Tortugas. X If.

Fig. 7.— Favia conferta, Ver. Transverse section through fig. 1. X 2f.

Fig. 8. — Favia gravida, Ver. Transverse section through fig. 2. x 2f.

Fig. 9 .—Favia fragum (Esp.). Transverse section through a corallum intermediate between fig. 5 and

fig. 6. x 2f.

Fig. 10. — Favia conferta, Ver. Vertical section through fig. 1. X 2f.

Fig. 11. — Favia gravida, Ver. Vertical section through fig. 2. X 2f.

Fig. 12. — Favia fragum (Esp.). Vertical section through another corallum of thick variety. X 2|.

Brit. Mus. (Nat. Hist)

Brit. Antarctic (Terra Nova) Exped. 1910.

Zoology, Vol. V.

Coel enter ata, Part 11, PI. IV.
Madreporaria.

V. Tams, photogr.

London Stereoscopic Co. imp.

BRITISH MUSEUM (NATURAL HISTORY).

mf

iH ANTARCTIC (“TERRA NOVA”) EXPEDITION, 1910

NATURAL HISTORY REPORT.

ZOOLOGY. VOL. V, No. 3. Pp. 97-120.

Si

COELENTERAT,

PART III. — ANTIPATHARIA (and their Cirripede Commensals).

BY

Kfc , A. KNYVETT TOTTON,

Assistant in the Zoological Department , British Museum (Natural Histoiy).

WITH 18 FIGURES IN THE TEXT AND TWO PLATES.

B. M.

(N.H.) .

LONDON

PRINTED BY ORDER OF THE TRUSTEES OF THE BRITISH MUSEUM.

W "7 Loweirijfs, Gmss A Co., 89 Paternoster Row, E.C. 4; E. Quaritch, Ltd., 11 Grafton Street, Now Bond
feet, W. 1 ; Dut*,u A Co., Ltd., 34- 3G Margaret Street, Cavendish Square, W. 1 ; Oxford UNXVERKirr Press, Amen
Corner, E.C. 4 j Wheldor & W esley, Ltd., 2, 3, & 1 Arthur Street, Now Oxford Street, W.G. 2 ;

! Outer & Bovd, Tweeddolo Court, Edinburgh ;

AND AT TOE

British Museum (Natural History), Cromwell Road, Loud on, S.W. 7.

1923.

[AH rights reserved. j

Price Five Shillings.

v^dSr

Ptt I

,/ . J;. 4

QjSriftefl Qfftueeum ((Uafuraf ^XGtov^).

This is No. <? of 25 copies of
“ Terra Nova" Zoology , Vol. V., Coelenterata,
Part III. , pritited on Special paper.

COELENTERATA.

PART III.— ANTIPATHARIA (AND THEIR CIRRIPEDE

COMMENSALS).

BY A. KNYVETT TOTTON,

Assistant in the Zoological Department, British Museum (Natural History).

WITH EIGHTEEN FIGURES IN THE TEXT, AND TWO PLATES.

PAGE

I. — Introduction 97

II. — Descriptions of Species 98

III. — Structure and Growth of Axis, Branches, and Spines • . 110

IV. — Polychaete Commensals of Antipatharia ; with Observations on

Coelenterate GalLstructures 113

V. — Cirripede Commensals of Antipatharia; with Description of a new

species of Pedunculate Cirripede 115

VI. — List of Publications referred to 118

I.— INTRODUCTION.

The Antipatharia collected during the voyage of the “ Terra Nova” were all obtained
at Station 96, seven miles east of North Cape, New Zealand, at a depth of 70
fathoms. They belong to seven species, five of which are described below as new to
science.

In the present state of our knowledge it is impossible to say what are the natural
affinities of some of these species ; in assigning them to genera I have followed the
methods of recent workers, particularly Pesch.

Some new observations on the growth of the axis, branches, and spines, and on
the origin of the axial canals, are recorded below.

The material includes examples of a new species of commensal Cirripede, which
is described ; some general notes on the commensals of Antipatharia are given.

When studying Antipatharia it has to be borne in mind that : —

1. The spines undergo considerable change in form with the addition of successive
laminae of chitin. For this reason those at the tip of a branch generally differ in
form from those nearer the base, or on the older parts of the colony.

2. When the polyps are restricted to one side of the axis, it is generally found

VOL. v.

p

98

“TERRA NOVA” EXPEDITION.

that the spines on the polyp side are longer and are arranged in fewer longitudinal
rows than those on the other side.

3. The number of longitudinal rows of spines that can be seen from one aspect of
an axis may be quite different from that seen from the opposite side. In this report,
when the number of longitudinal rows of spines is mentioned, it is to be understood
that every row, counting right round the circumference, has been taken into account.
This is the only satisfactory method.

II.— DESCRIPTIONS OF SPECIES.

1. Bathypathes scopctria, sp. n. (PI. II, figs. 3, 4.)

Material. — Four pieces of the axis and branches of a large specimen, representing
a length of over 42 cm. At the basal fracture the stem measures 3 mm. by 3-5 mm. ;
at the apical fracture it has a diameter of 0-5 mm.

Specific description. — Colony with main stem bearing four series of incurved *
branches, an outer or more lateral pair set in planes which form an angle of about
100°, and an inner pair which meet at an angle of 35° (fig. 2, B, p. 100). The outer
branches are set about 6-8 mm. apart in each series, and alternate with those of the
other side ; the inner branches also alternate, each being inserted 2 mm. below the
outer branch of the same side. The outer branches are from 110 mm. to 180 mm.
long, and the inner from 60 mm. to 155 mm. Each outer branch bears a unilateral
series of about six smaller branches, from 110 mm. to 170 mm. in length, directed
forwards, inwards, and upwards ; at angles of from 15° to 20° from the branch that
bears them, these branches of the second order may bear others, of third and even fourth
orders, of a maximum length of 70 mm., arranged in a similar way, but in number
only two or three. The inner series of paired branches are also compound, generally
bearing two or three smaller branches from 50 mm. to 140 mm. long ; these are mostly
simple. Axial chambers f from 0-50 mm to 2-1 mm. (generally about 115 mm.)
long. Towards the pointed tips of these chambers there appear, on the inner wall,
certain circular ridges or incomplete septa. The septa dividing the chambers persist
for a long time, but eventually break down and give rise to axial canals, which
increase in diameter from 0T75 mm., in a branch 0-515 mm. in diameter, to 0-24 mm.
in a branch 0-865 mm. in diameter.

Spines on stem and branches in ten longitudinal rows (text-fig. 1), short, simple,
triangular, from 0-36 mm. to 0-075 mm. in length, and 0-450 mm. apart in the same
row. Polyps borne on that side of the stem which bears branches, and on the inner-
upper sides of the branches, elongated in the direction of the axes which bear them,

* The curvature of the branches may be in part due to the small size of the jar in which the
specimen was packed. Pesch (1914) is inclined to think that the rolled-up condition of the type
specimen of A'phanipathes reticulata may be due to a similar cause,
t See p. 110.

ANTIPATHARIA— TOTTON.

99

2-8 mm. to 3-4 mm. long, 0-3 mm. to 0-6 mm. apart, with tentacles in three pairs
about 2-2 mm. long (exclusive of basal bulbous parts).

Remarks. — This species appears to have certain affinities with Schizopathes
conferta, Brook, which bear numerous, simple, long, somewhat irregularly arranged
antero-lateral branches. It is by no means certain that these branches all belong to
one series. Indeed, on the unique specimen of S. conferta there are indications that
an original series of alternate outer branches may have become masked, as the growth
of the colony proceeded, by the addition of a more irregularly disposed secondary
inner series. Externally the polyps of the two forms are similar, and are intermediate
between the extreme “ Schizopathes-type ” found in S. crassa, Brook, where they are
very much crowded, and the extreme “ Eathybathes-type ” found in B. patula, Brook,
where the pairs of tentacles are widely separated. The most marked difference
between the two forms is the compound nature of the branches found in B. scoparia ;
but this may not be a very profound one, and in my opinion is much less important
than the difference between the branching-
systems of B. scoparia and of a complexly-
branched form such as Taxipctthes recta , that
has a complicated system of small off-shoots,
which are developed into big branches that
repeat the structure and arrangement of the
main stem.

Three other forms are known to produce

two paired series of antero-lateral branches, ... , ,

i 7-11^7 ,7 7 -7 7-. of spines in longitudinal rows on branch

namely, Eubathypathes quadribrachiata, ^ ^ ^

Pesch, Aphanipathes alata, Brook, and

Antipathes lilliei, sp. n. The only specimen of E. quadribrachiata at present
known is one, complete with basal plate, only 2-8 cm. long and, therefore, most
probably a young specimen. It would be inadvisable to base any conclusion on
a comparison between a mature colony and such a young specimen. The obvious
differences are that in E. quadribrachiata the branches are set very much more
closely together, the inner branches are comparatively small, and all the branches are
simple (text-fig. 2). Further growth might alter the characters of the specimen
considerably. In A. lilliei, sp. n., and Aphanipathes alata, Brook, the branches remain
simple throughout life, except the few that are developed into big branches.

One other form, Bathypathes lyra, Brook, must be mentioned because it develops
on the anterior side of the stem a series of relatively short (3 mm. to 5 mm. long)
branches, which are hollow and bear spines. Pesch has stated that they are not
more than a large sort of spine, but no spines bear other spines, nor are any hollow,
except for a very short distance at their bases. That they are developed much later
than the long paired branches can be assumed from the fact that the thickness of
chitin which separates their bases from the axial canal of the main stem is much

s\ /\ - ^ txi.

„ - v v v- ^ - v V V ^

^ w s? C? V? ^ V v—

Fig. 1.— Bathypathes scoparia. Arrangement

100

“TERRA NOVA” EXPEDITION.

greater than that which separates the bases of the long paired branches from the
axial canal of the stem. These short branches on B. lyra appear to be arranged in
one median row, but only an examination of a long series of specimens will show
whether they are really arranged in two rows separated by a small angle. In
B. lyra all the branches remain simple.

B. scoparia differs from all other species of the genus, as emended by Pesch,
in having compound branches, which are arranged in two paired series. It
agrees with them in that none of the branchlets are ever hypertrophied into large
branches which, in the arrangement of the branchlets borne on them, repeat the
character of the main stem, and in the shape and distribution of the polyps.

Fig. 2. — A. Eubathypathes quadribracliiata, Pesch (nat. size). Diagrammatic reconstructions from
original description : Projections of branches on (a) a vertical, (6) a horizontal plane.

B. Bathypathes scoparia : Projection of branches on a horizontal plane.

As Gravier remarks (1921), the genus Bathypathes seems to be confined to the
great depths (1,055 fathoms to 2,975 fathoms), except for B. patula and B. alternata,
which have been recorded in 55 fathoms on the Burdwood Bank, south of the
Falklands, where temperature conditions approximate to those in the depths of the
Atlantic. Bathypathes scoparia and the following species B. platycaulus, coming
from a depth of only 70 fathoms on the fringe of the tropical zone, are as well
distinguished by their habitat as by their structure from the other species of this
genus.

2. Bathypathes platycaulus, sp. n. (PL I, fig. 6.)

Material. — 139 mm. of the apical part of a specimen, and 105 mm. of a more basal
part of what was probably the same specimen, preserved in spirit. The specimens
appear to have been partially dried at some time.

ANTIPATHARIA — TOTTON.

101

Description. — At the basal fracture the axis of the main stem is nearly
cylindrical, with a diameter of 1-5 mm. It increases in width to 1-75 mm. in the next
7 cm., after which it tapers away gradually, the diameter at the distal fracture of the
apical part being 0-28 mm. In the middle part the stem is flattened on one side,
bearing on either edge of this flattened region a row of simple branches, the distance
apart of which averages 3 mm. Branches alternating, ascending at an angle of 30°,
in planes which intersect at an angle of 120° ; distal portions curved downwards
towards the base. The longest branch measures 8-8 cm. Signs of regeneration
appear at irregular intervals. The central canal of the branches measures about
0T5 mm. in diameter. There are no septa except at points where regeneration has
taken place. Within the canal are foreign particles.* Spines in 10 or 11
longitudinal rows, 6 or 7 on one side and 4 on the other, 0T3 mm. apart in the same
row, 0-04 mm. long. The number of rows at the apex of a branch is small. New rows
are intercalated at a distance of 1 mm. or 2 mm. from the tip. Polyps about 2 mm.
long in the direction of their supporting axes, 1 mm. apart, sharply divided into three
regions, with pairs of tentacles 0-85 mm. apart and 0-5 mm. long, and possessing
prominent oral cones.

Remarks. — This species differs from those previously described in the relative
position of the branches, and in the angle between them and the stem. The
broadening of the main axis in the middle region is peculiar.

3. Antipathes lilliei, sp. n. (PI. I, figs. 1-3.)

Material.- — A magnificent dried colony, 90 cm. by 70 cm. by 50 cm.

Description. — Colony bilaterally symmetrical, multipinnate, with main branches
directed to one side of the stem. Axis with four series of branches (PI. I, fig. 3),
mostly simple, slender, recurved, from 9 cm. to 1 1 cm. long. Outer series opposite,
set in planes which intersect at an angle of nearly 180°, alternating, + 3-2 mm. apart
on each side, forming an angle of + 60° with the axis. Inner (anterior) series
(undeveloped in youngest branches), in planes which form nearly a right angle,
inserted 1-5 mm. below the branch of the outer series on the same side. Branches of
either series may be hypertrophied, and may even be as stout and as long as the stem
beyond their point of origin. These hypertrophied branches bear series of branchlets
arranged like the branches on the stem ; some of these again may be enlarged and
bear branchlets. The diameter of the simple branches increases from 0-077 mm. at
the tip (exclusive of spines) to 0-585 mm. proximally. The diameter of the axial canal
increases from 0-066 mm. at the tip to 0-128 mm. in a branch with an external diameter
(exclusive of spines) of 0-571 mm. Spines smooth, triangular (text-fig. 3), in from 7
(on younger parts of branchlets) to 10 longitudinal rows, 0-15 mm. to 0-26 mm. (usually
0-195 mm.) apart, connected in some degree by ridges (text-fig. 16, p. 112) ; 0-065 mm.
long on the polyp side and 0-05 mm. long on the other. Polyps slightly elongated in

* See footnote, p. 105.

102

“TERRA NOVA” EXPEDITION.

the direction of the axis which bears them, from 1-86 mm. to 2-33 mm. long, and about
3-21 mm. apart. The length of the lateral tentacles (macerated in water) is 0-6 mm.

Remarks. — The curved axis, which is 1-5 metres long and 2 cm. in diameter at
its thickest, springs from a bell-shaped base (PI. I, fig. 2). The hypertrophied branches,
which may reach a length of 1-29 metres, occur at irregular intervals. Two pairs of
hypertrophied inner branches and a pair and a half of outer ones arise close together
at a point half a metre from the base. Lower down two more inner branches have
been enlarged, one (PI. I, fig. 2) bearing branchlets of the second, third, and fourth
orders. Most of the other branches are simple or bear only small branches. Seventeen
polyps were counted on a length of 3-5 cm. of a branchlet. A metre’s length of an
enlarged branch carries about 1,200 branchlets in all, and, in the case of a well-grown
one, would carry as many as sixty thousand polyps. It has been calculated that the
present colony, when alive, consisted of over 400,000 polyps.

On the distal ends of some branchlets is well illustrated the homology of spines

Fig. 3. — Antipathes lilliei. Young branchlet, showing hollow, chambered spines ( x 219).

and branches (text-fig. 16). In their early stages the spines (text-fig. 3) are hollow,
just like the branches. As usual in dried Antipatharia, during the process of drying
many of the branchlets have become stuck together at the tips (PI. I, fig. 1). The
condition described by F. Cooper (1909), and considered to be concrescence, may be
nothing more than an adhesion brought about by temporary drying.*

At rare intervals there are present on the branchlets curved, horn-shaped,
chitinous projections. They prove to be the chitinous sheaths laid down by the
tissues of the antipatharian round the peduncles of commensal cirripedes of the genus
Oxynaspis. Only two of the capitula of these cirripedes were preserved, and they
form the subject of a special section of this report. The lower part of the stem and
some of the branches bear numerous specimens of other cirripedes, but they are not
dealt with because they appear to fall into the category of chance associations, and
not of true commensals or definite associations.

Affinities. — The only mature specimen of an antipatharian known to the writer
that has a habit at all similar is Aphanipathes alata, Brook, from Mauritius ; but the
present species is a giant in comparison, and the spination is quite unlike that of A.

* F. Cooper says that “ concrescence ” is accompanied by a good deal of breaking down of tissue
elements, but gives no histological figures.

ANTIPATHARI A -TOTTON.

103

alata. There are certain resemblances between A. lilliei and a young specimen,
2-8 cm. long, described by Pesch (1914) as Eubathypathes quadribrachiata. This
young colony differs from the present one chiefly in the relative shortness of the inner
series of branches, and in that they form an acute angle (about 25°) with one another.
The branchlets are set more closely in E. quadribrachiata (2-9 mm. apart) than in
A. lilliei (3-2 mm. apart). The spines in E. quadribrachiata are 0-015 mm. and
0-03 mm. long and are not connected, whereas in A. lilliei they are twice the length
and are connected by longitudinal ridges. It is possible that a mature specimen of
E. quadribrachiata might show all the characters of A. lilliei, but on the other hand
a mature specimen might be found retaining all the characters of the young E. quadri-
brachiata, and it has seemed best to form a new species A. lilliei, about the sum of
the characters of which there is no doubt.

It is convenient for the present to assign this species to the genus Antipathes,
because that one seems to be the most heterogeneous, and the species does not appear
to belong to any of the others recognized by Pesch. In particular it is excluded
from Bathypathes, Brook, because the species described by him are all relatively small
forms with simple branches, and never have larger branches that repeat the characters
of the main stem.

4. Antipathes aperta, sp. n. (PI. II, figs. 1, 2.)

Material. — Five smaller fragments, and one larger, 13 cm. over all, broken from
the basal part of a colony, and without the base of attachment.

Description. — Main Stem irregularly branched in many planes, multipinnate.
Ultimate branches 1 cm. or 2 cm. long, alternate, occasionally opposite, 2-5 mm.
to 3 mm. apart on same side ; diameter of axial canal from 0-014 mm. (at tip of
branch) to 0-05 mm. (in branchlet 0-57 mm. diam.). Stretching across the axial canal
in younger parts of the colony are frequent domed septa, indicating discontinuous
growth. Spines in from 7 longitudinal rows (at tip of branch, text-fig. 4, b) to 10
(on branchlet 0-25 mm. in diameter, text-fig. 4, A) ; on older branches very numerous
and irregularly arranged, dichotomously branched, antler-like, and 1-1 mm. in length
(text-fig. 5) ; at tips of ultimate branches bracket-shaped and 0-017 mm. in length
(text-fig. 4, b) ; intermediately (branch 0-61 mm. in diameter) needle-shaped and
0-15 mm. long, interspersed with secondary, shorter spines of very irregular form
(text-fig. 6). These secondary spines are irregularly distributed, and sometimes
appear at an early growth-stage of a branch. Polyps slightly elongated, 0-80 mm.
long, 0-50 mm. apart, and with tentacles (contracted) 0-13 mm. long.

Remarks. — Presumably the colony at first was simply pinnate. By the sub-
sequent hypertrophy of certain branches on which developed yet others again, came
about the irregular branching system of the old colony. Two specimens of an
Oxynaspid Cirripede, one large and the other smaller, are attached in the region of
the lower branches (PI. II, fig. 2). On many parts of the colony there is an irregular

104

“TERRA NOVA” EXPEDITION.

/\ j\/i y\A A A SLA /va AL/L^Sk-Str/iAn

A

B

Fig. 4. — Antipathes aperta. A. Branchlet, showing regular arrangement of spines in longitudinal
rows with slight dextrorse spiral twist ( x 30). The spines on the other side are arranged irregularly
in three main rows. A few scattered secondary spines may be seen. B. Tip of an ultimate branchlet,
with an alternating series of spines ( x 230). The chambered structure of the axial canal is shown.
Proximally two regenerations have taken place.

Fig. 5. — Antipathes aperta. Antler-like spines from a branch 2-4 mm. in diameter (exclusive

of spines) ( x 50).

Fig. 6. — Antipathes aperta. Spines from a branch 0-61 mm. in diameter ( x 151). Scattered
amongst the needle-shaped primary spines are numerous irregular secondary ones.

ANTIPATHAKIA — TOTTON.

105

investing sheath of chitin laid down over some organism, whose nature has not been
determined.

In shape the specimen recalls Aphanipathes somervillei, Forster Cooper (1909),
which also has spines 1 mm. long and closely set together on the stem. The
description does not give the shape of any of the spines ; but a figure (1909, 7, a)
appears to represent them as needle-shaped and not antler-shaped.

5. Antipathes glutinata, sp. n. (PI. I, fig. 5.)

Material. — A loose plexus of the distal pieces of coalescing branches and
branchlets, measuring 6 cm. by 2 cm. ; evidently separated from a larger colony,
which presumably was flattened in one plane. Coenosarc macerated.

Fig. 7. — Antipathes glutinata. A. The tip of a branchlet ( x 138). B. A branchlet, showing
arrangement of spines in longitudinal rows ( x 19-5). C. A coalescence between two branchlets ( x 48).
D. A branchlet, showing the incomplete septa of the axial canal ( x 251).

Description. — Multipinnate (mode of branching obscured by coalescence), flat-
tened in one plane ; branchlets usually coalescing wherever contact between them
takes place (text-fig. 7, c). Regeneration frequent. Ultimate branchlets directed to
one face of the colony, 1 cm. to 2-5 cm. long, 0 125 mm. in diameter. Axial canal
0-0113 mm. in diameter (at tips of branchlets) to 0-018 mm. (in branchlets 0-203 mm.
in diameter) ; very irregular in outline (text-fig. 7, d) and traversed by complete and
incomplete septa at regular intervals. The lumen is partly or wholly occluded at
frequent intervals by crystalline bodies.*^. Spines in 6 or 7 longitudinal rows (text-

* These bodies appear to be very line-grained aggregates. Under crossed Nicols the distribution
of interference colours indicates a radial structure. It is possible that they are sea-salts.

VOL. v. Q

106

“TERRA NOVA” EXPEDITION.

fig. 7, b), 0-23 mm. to 0-32 mm. apart in same row, maximum length 0-10 mm. ; at tips

of branchlets inclined distad at an angle of 35° to
40° from the axis (text-fig. 7, a), proximally at a
more obtuse angle (60° on branchlets 0-18 mm.
in diameter), on older branches (0-45 mm. in
diameter) projecting for 0 08 mm. at right angles
to the axis (text-fig. 8). Diameter of axial canal
at tip of branchlet about one-fifth of that of
branchlet.

Remarks. — In general form and in some details

of spination this species resembles the form described

by Pesch (1914) as Euantipathes plana (Forster

Cooper) ; in that form, however, the ultimate branch-

lets, which also are all directed more or less towards

Fig. 8.—Antipathes glutmata. A anterior side of the colony, are quite short (a
branch with spines that project at .

right angles ( x 48). ^ew mm* in length).

6. Parantipathes tenuispina, Silberfeld. (PI. II, fig. 5.)

Material. — Two specimens without polyps ; a larger imperfect one 18 cm. in
length, with remains of a basal plate of attachment, which is about 2 mm. thick ; and
a smaller, more perfect one, 12 cm. in length, with stem 1-14 mm. in diameter, complete
except for basal plate. Both specimens contain worm-runs. The following descrip-
tion is based on the smaller, more perfect specimen.

Description. — Colony bottle-brush shaped, oval in section, in the middle region
measuring 16 mm. by 25 mm. In the apical region pinnate ; in the middle and basal
region with a second series of paired branches. Branches of the first series alter-
nate, from 1-7 mm. to 2-8 mm. (average 2-18 mm.) apart, set on to the stem at angles
of from 70° to 85°. These branches themselves are alternately pinnate in the middle
regions of the colony. Those of opposite sides set on to the stem in planes which
intersect at angles of from 120° to 150°. The inner proximal branchlets of the main
series of branches unite, either by themselves or by their smaller branchlets (text-
fig. 10) with those of the other side to form arches — the framework of a worm-run.*
Small irregularly disposed branchlets, running roughly parallel with the main axis,
link up these arches to form a complicated lattice-work tube (text-fig. 9), 3 mm. in
diameter at the base, increasing to 4 mm. at the tip. On the opposite side of the
stem is the second series of paired branches, which appears to develop a little later.
The planes in which the rows of these branches of the second series lie intersect at
angles of from 60° to 90°. These branches may be simple at the base of the colony,
but in the middle region may develop a maximum of six alternate branchlets, which

* The arrangement in Aphanipathes barbadensis, Brook, is similar (PI. I, fig. 4), but there is little
coalescence, and the individual arches are not linked up by branchlets running parallel to the main axis.

ANTIPATHARIA — TOTTON.

107

may again bear two or three smaller branchlets. These small branchlets of the
second and third order do not always appear to be alternate, because supernumerary
irregular branchlets are developed. In both series of larger branches — the main one,

Fig. 9. — Parantipatlies tenuispina. Posterolateral view of part of colony, showing the interlacing

of branchlets to form the “ worm-run ” ( x 8).

Fig. 10. — Parantipathes tenuispina. A. A diagrammatic projection of the branches on a horizontal
plane. B. A diagrammatic elevation of part of the stem and one branch of the second series, to
illustrate the tilting of the branching planes.

which forms the worm-run, and the series on the opposite side of the stem — the
planes in which the alternate or sub-opposite branchlets of the second order lie are
not quite horizontal. The branches themselves are inclined upwards (distad), and

108

“TERRA NOVA” EXPEDITION.

the upper surfaces of the branching-planes of those of opposite sides are tilted towards
each other, each plane generally intersecting the main axis of the stem at an angle
of about 80° in the main series, and 50° in the series on the opposite side of the stem
(text-fig. 10, b). The length of the branches of the main series is from 9 mm. to
12 mm. The branchlets borne by them are, proximally 7 mm., distally 1-75 mm., to
3-5 mm. long. The branches of the second series on the side of the stem opposite
to the worm-run have a maximum length of about 9 mm. In both series branchlets
of second or third orders, after growing out at an obtuse angle, curve upwards (distad)
and run more or less parallel with the branches (text-fig. 10, a). This arrangement
has been noted by Pesch in Euantipathes plana, F. Cooper.

f\ — a A — J\ 'I— -—rr — ^ cr- — ^ — w— -N — ^ ^

0 l/~* l/-**-

Fig. 11. — Parantipathes tenuispina. A. Tip of a branchlet ( x 180). B. The spines on an older
branchlet ( x 38). C, D. Two views of part of branchlet with spines in six longitudinal rows.

The axial canal increases in diameter from 0 01 mm. at the tips of branchlets
(0-05 mm. in diameter) to 0-0225 mm. in branchlets measuring 0-3 mm. in diameter.
Septa and incomplete septa traverse the lumen of the canal at irregular intervals.
Spines. — At the extreme tips of branchlets the spines are inclined distad at an angle
of from 25° to 30° from the axis. They are longest at a little distance (1 mm. to
2 mm.) from the tips, being 0*12 mm. long and inclined distad at an angle of 45°
(text-fig. 11, a). At the bases of the branches only 0-075 mm. of the length of the
spines projects, and that almost at right angles (text-fig. 11, b). At the base of the
main axis this length is reduced to 0-04 mm. This is about the length of the largest
spines on the small supernumerary branchlets of 0-075 mm. diameter. On the
branchlets the spines are arranged in six equidistant longitudinal rows (text-fig. 11,
c-d) and are from 0-18 mm. to 0-28 mm. (average 0-23 mm.) apart in the same row.

No details of the macerated ccenosarc can be seen.

ANTIPATH ARIA — TOTTON.

109

7. Stichopathes variabilis, Pesch, var. lissispina, Pesch.

Material. — Five fragments of a simple, spirally coiled stem.* What appears to
be the more basal fragment, though without the plate of attachment, is 17 cm. long
and forms a segment of a circle with a radius of 8 cm. On to this join two separate
coils 11 cm. in diameter, a smaller irregular coil 6-5 cm. in diameter with a peripheral
length of 34 cm., and the most distal part, which is 53 cm. long when straightened
out and forms three irregular coils from 4 cm. to 6 cm. in diameter. The apical
portion of the colony is missing.

Description. — The specimen must have had a total length of over 150 centi-
metres— this being the combined length of the fragments, which increase in diameter
(exclusive of spines) from 0-32 mm. to 1 mm. There is a corresponding increase in
the diameter of the axial canal, from 0-26 mm. distally to 0-59 mm. proximally. The
spines (text-fig. 12) are arranged in from 12 to 16 longitudinal rows, which are

— — — ^ —

•> ..
.

- ^

'C- — —

y —

"\T ^ ^3 ^ -xt

Fig. 12. — Stichopathes variabilis, var. lissispina, showing twelve longitudinal rows of spines ( x 48).

wound in a very gradual and hardly perceptible dextrorse spiral, those on the same
side as the polyps being more distant from one another and consisting of shorter
spines than those of the other side. Their distance apart in one and the same row
is from 0-38 mm. to 0-46 mm. ; and their length 0-14 mm. on the polyp side, but only
0-07 mm. on the other. The smaller spines especially are often bifid. The polyps
are on one side only, crowded, occasionally as much as 0-75 mm. apart, of various
sizes — largest 1-65 mm. long ; mouth elongated in a sagittal direction ; sagittal
tentacles at a lower level than the laterals. In the larger polyps the sagittal
tentacles are about T5 mm. long and the laterals 0-71 mm. The state of preservation
is bad, the specimens having been allowed to dry at some time.

Nomenclature. — It has been a matter of great difficulty to find a name that
could satisfactorily be applied to this specimen. Old specific descriptions do not
mention the total number of longitudinal rows of spines, nor do they give adequate
measurements of spines and axial canals. The types of S. gracilis, Gray, S. echinu-
lata, Brook, and S. occidentals, Brook, have been closely examined afresh and have
yielded the following data, which diverge considerably from those given by Pesch
(1914).

* The spiral is sinistrorse, seen from the centre.

110

“TERRA NOVA” EXPEDITION.

Stichopathes.

gracilis, Gray.
Type.

occidentalis, Brook.
Type.

echinulata, Brook.
Type.

variabilis, Peach.
“ Terra Nova ”
Specimen.

Diameter of stem (ex-
cluding spines)

1*71 mm.

0-488 mm.
0-625 mm.

0-776 mm.

0- 6 mm.

1- 0 mm.

Diameter of axial canal *

Q-43 mm.

0-375 mm.
0-425 mm.

0-475 mm.

0-35 mm.
0-59 mm.

No. of rows of spines
(longitudinal)

16

15

18

16

Distance between spines,
same row

0-38-0-81 mm.
0-8 mm.

0-375 mm.

0-45-0-675 mm.
0-575 mm.

0-38-0-46 mm.

Length of spines —

Polyp side

0-18 mm.

0-15 mm.

0-15 mm.

0-14 mm.

Other side

0-13 mm.

0-075 mm.

0-075 mm.

0-07 mm.

* Where two values are given for diameter of stem and of axial canal, the first of one corresponds
with the first of the other.

This table shows that great caution must be observed when using numerical
values as a criterion for specific distinction. This is borne out by Pesch’s diagnosis
of his very variable species variabilis. As yet we know very little about the
value to be attached to the shape of the colony, whether it be straight, sinuous, a
dextrorse or a sinistrorse spiral, and whether the spiral coils have a large or small
diameter. Much may be learned in tbe future from detailed examination of groups
of specimens attached to one small area such as the group of 120 specimens attached
to a stony mass measuring 10 in. by 5 in. mentioned by Johnson (1899).

III.— STRUCTURE AND GROWTH OF AXIS, BRANCHES,

AND SPINES.

It is probable that spines and branches are homologous. They are both produced
by series of hollow, superimposed chambers ; but at a very early stage the spines
become solid owing to the fact that there ceases to be any chamber between the
successive laminse of chitin.

The mode of growth appears to be very similar to that observed in Gorgonia
cavolinii by von Koch (1887). After settling down and secreting a basal chitinous
plate (text-fig. 13), the base of an antipatharian planula probably becomes invaginated,
thereafter secreting a fresh chitinous dome-shaped plate (text-fig. 13, a). The ectoderm
lining the invagination withdraws by successive stages from the last chitinous deposit,
whilst the upper end of the animal continues to grow. By a continuation of this type
of growth (text-fig. 14) a cylindrical axis, formed of domed chambers, is built up.

ANTIPATHARIA— TOTTON.

Ill

n

When, at certain points on the wall of this axis, the ectoderm withdraws, either
spines or branches are developed by a similar
process.*

Although very early growth-stages of Anti-
patharia do not appear to have been observed,
some of the specimens contained in this collec-
tion, particularly those of B. scoparici, sp. n., very
clearly demonstrate the method of growth, espe-
cially the origin of the axial canals. In this species
forty-one chambers were counted in the distal
51 mm. of a branch (text -fig. 15), so that each
chamber is relatively very long. As a rule the
transverse walls appear to break down eventually,
leaving a continuous axial canal, which increases
slightly in diameter with age.* The canals in the bases of spines and in branches
(text-fig. 16) never communicate with one another in any specimens that have been

i

Fig. 13. — Hypothetical stages in the
early development of Antipatliaria. A.
The planula has secreted a basal plate of
chitin. B. The base of the animal has
become invaginated, and has secreted a
dome-shaped septum. C. A second sep-
tum and a further part of the cylindrical
wall of the axis has been laid down.

E

Fig. 14. — Diagrams to illustrate hypothetical growth of forms like B. scoparia. A. The coenosarc
is in contact with the tip of the axis. B. Incipient withdrawal of coenosarc from the tip of the
axis. C, D. Formation of the next axial chamber. E. A pause in the growth of the coenosarc,
which is now forming the apical part of the chitinous axial chamber.

Dantan (1920) states that the axis is secreted by mesenchym, and is originally solid.

112

“TERRA NOVA” EXPEDITION.

examined by the author ; but Forster Cooper (1909) makes a contrary statement in
his description of A. heterorhodzos. The hollowness of spines is well shown in A. lilliei,
sp. n. (text-fig. 3). When examining a thick antipatharian pinnule mounted as a
transparency, it may often be noticed that there are a number of dark spots on the
walls of the central canal, corresponding to the spines. These are the hollow spaces
in the bases of the spines.

Fig. 15. — Bathypathes scoparia. A. Young branchlet ( x 40), showing spindle-shaped swellings.
B. Branchlet, showing cylindrical outline, after further deposition of chitin ( x 16).

Fig. 16. — A. lilliei. Branchlet ( x 135), showing discontinuity between cavities in the hollow spines
and the axial canal. The spines are connected by longitudinal ridges.

Fig. 17. — Diagrammatic section of the chitinous axial wall of an antipatharian to illustrate the
cause of the varying degrees of diminution of the diameter of the axis following regeneration. The
black sections show the thickness of the axial wall at the times of two different fractures.

Antipatharia, whose young branches show no traces of axial chambers, must be
supposed to grow continuously, laying down chitin internally from the sides of the
growing cylinder of coenosarc, without a check in growth and consequent formation of
end-walls or septa.

Roule (1905) observed and figured at the extreme tip of the conical end of

ANTIPATH ARIA— TOTTON.

113

a branch of Stichopathes, sp., an open thin-walled cylindrical expansion of the axis.
It is probable that the ccenosarcal cap had been worn away from the tip of this axis.

Fracture of off-shoots takes place frequently, and is followed by regeneration.
The truncated tube of ccenosarc covering the truncated axis probably forms an un-
differentiated end-cap, which then grows as before and adds a new series of axial
chambers to the truncated axis. Depending on the thickness of the axial wall at the
time of fracture the decrease in the diameter of the regenerated off-shoot is more or
less marked (text-fig. 17).

IV.— POLYCHyETE COMMENSALS OF ANTI PATH ARIA ;

WITH OBSERVATIONS ON COELENTERATE GALL-
STRUCTURES.

An examination of numbers of Antipatharici from all seas shows that there are
two classes of animals which commonly live domiciled on them, namely, Cirripedia
and Polychceta : Steenstrup (1851) has described a mollusc, Rhizochilus antipathum,
which attaches itself to an antipatharian, and thereafter becomes much modified ; but
this mollusc does not appear to be so common as the other two classes of commensal.
Forming a secondary association, and living in the worm-” run” of certain forms, may
be found small porcellanid crabs.* The type specimen of Aphanipathes barbadensis,
Brook, bears, beside two specimens of these crabs, both cirripede and polychmte
commensals.

Pourtalks (1867) described an association between a polychmte, Marphysa anti-
pathum, and a pinnate antipatharian, Antipathes Jilix, along whose stem the worm
builds a parchment-like tube armed with sponge spicules. In 1878 he described
another antipatharian, A. abietina, which was constantly associated with a worm-
tube. Brook (1889) described a further form, A. cylindrica, resembling in growth
A. abies (Linn.), with horny worm-tubes somewhat like those of Eunice tibiana
incorporated in the stem. In none of these does the antipatharian form reticulated
worm-” runs”. In 1870 Duchassaing described a species, Arachnopathes (. Pciranti -
pathes) columnaris, from Guadeloupe, which possessed a cylindrical reticulum, the
nature of which he appears not to have understood. Pourtalbs (1874) gave a further
description and figure of a specimen from Barbados, pointing out that the reticulum
was the habitation of an annelid and noting similar occurrences of “ parasitic ” annelids
in Lophohelia, Stylaster, and Allopora. The collections of the U.S. S. “ Blake ” from
N.E. of Cuba included more specimens of A. columnaris, which were described by
Pourtales (1878). He says, “Two specimens in the number are destitute of the
parasitic worm and of the tube produced by it ; their branchlets are more spiny, but
the general shape is the same.” Brook (1889) described and figured Antipathella

* Hornell (1922) says that Porcellanids occur also as commensals on the Alcyonarian Spongodes in
Indian seas, along with Galatheids, Synalpheids, and a Dromiid crab,
von. v.

K

114

“TERRA NOVA” EXPEDITION.

contorta and Tylopathes cnspa, along parts of the stems of which are reticular
“runs” inhabited bypolychaetes. Silberfeld (1909) gave descriptions and photographs of
worm-“ runs ” in Parantipathes tenuispina from Sagami Bay, and Pesch (1914) gave
further details of the “ runs ” in P. columnciris (Duchass.), Brook, mentioning that
he observed several annelids in the lower half of one tube in this species and also in
a specimen of P. tenuispina which he examined.

There is no evidence of parasitism ; and it is difficult to imagine that the Anti-
patharians profit by the presence of the worms, which they are powerless to repel.

The question of interest in connexion with these reticulated worm-“ runs ” is
whether or not they are merely abnormal growths solely due to the presence of com-
mensal epizoites, and, if so, what the normal growth forms of the Antipatharians
concerned is like. Pourtal&s (1878) described two specimens of A. (. Parantipathes )
columnaris which lacked both commensal and recticulum, the general shape of the
colony being otherwise the same as in those with reticula. In Tylopathes crispa,
Brook (1889), the reticula on a specimen are localized, and we can see at once that
they are solely due to the commensal worms : where the reticulations are absent
there is merely a clear passage-way along the branches. In A. myriophylla, and
other species which do not appear to have commensal worms, this passage-way is
converted into an open channel, walled in at the sides by short processes standing out
on each side from the lateral branchlets : in A. spinescens , Gray, these outstanding
processes give rise to others, some of which arch over the channels, turning them into
covered-in tunnels ; as there is no sign of commensal worms, we suppose that this is
the natural growth form of the Antipatharian.

Probably, in nature, active wandering polychaetes roam about on the sea-bottom
and explore the stems of organisms growing on it. In creeping about the branches
of Antipatharia and Alcyonaria they find a very favourable habitat. In many Anti-
patharia there is a clear passage-way along the branches, which in some species is
converted into an open channel or even a covered-in tunnel. On such forms the
worms have made permanent homes * and have produced the reticulated runs, which
may be regarded as abnormal growths or galls.

Similar associations between polychaete worms and Primnoid Alcyonaria have been
noted from time to time, and are more common than is generally known. Analogous,
too, are the tubular structures in Solenocaulon inhabited by Alpheid Crustacea,
structures which Hickson (1903) suggests are of the nature of galls ; the similar
tubular structures on the stems of Pennatula jimhriata, Herklots, occupied by
polychaete worms and described by Hickson (1916) ; and also the coral galls formed
by females of the small crab Hapalocarcinus marsupialis on branches of Seriatopora
and Pocillopora, described by Semper (1883).

In all these instances there seems to be an abnormal growth of the coelenterate

* Compare the occurrence of polychsetes in the gastral cavities of Euretid and other sponges, where
the association has led to the remarkable modification of the worm seen in Syllis ramosa, McIntosh (1879).

ANTIPATH ARIA -TOTTON.

115

associated with the presence of a commensal, which is analogous to the gall- formation
in plants associated with insects, mites, and fungi. The difference between plant
galls and the analogous structures in Ccelenterates is that in most of the latter the
domiciled organism is constantly extending the field of stimulation ; for example,
there is in the collections of the British Museum a specimen of a Parantipathid with
a reticulum no less than 35 centimetres long. In the case of coelenterate gall-
structures we do not know by what means the commensal influences the growth of
its host ; but in view of the fact that both are bathed in sea-water the influence is
more likely to be mechanical than chemical, unless, indeed, mucus may prevent toxic
substances from being washed away.

V.-CIRRIPEDE COMMENSALS OF ANTI PATH ARIA ; WITH
DESCRIPTION OF A NEW SPECIES OF PEDUNCULATE
CIRRIPEDE.

Darwin (1851) described the first cirripede commensal with an Antipatharian,
naming it Oxynaspis celata. Four other species of this genus have been described,
namely, 0. patens, Aurivillius (1894), 0. aurivillii, Stebbing (1900), 0. inclica, Annan-
dale (1914), and 0. bocki, Nilsson-Cantell (1921). There are further species in the
collections of the British Museum, in addition to the new species 0. terrce-novce
described below. Species of Oxynaspis occur at Madeira, West Indies, Red Sea,
Mauritius, Ceylon (Thomson and Simpson, 1905), Bay of Bengal (Annandale, 1909),
Penang, Kei Islands, Tizard Bank (China Sea), Japan (Nilsson-Cantell, 1921), New
Britain (Stebbing, 1900), Torres Straits, and off North Cape, New Zealand. It is
noteworthy that, with the exception of Dichelaspis tydemani, Hoek * (1907), all the
Cirripedia Pedunculata so far found upon Antipatharia belong to the genus
Oxynaspis, species of which have not been found attached to anything but an
antipatharian. Species of Oxynaspis are not confined to one genus of Antipatharia,
but there are certain genera, such as Schizopathes, on which they have not yet been
found.

Darwin’s statement that the spiny membrane which covers the external surface
of specimens of Oxynaspis is a continuation of the “ horny muricated bark ” of the
Antipatharian has been doubted by several authors (Aurivillius 1894, Annandale
1909, Nilsson-Cantell 1921), but I can confirm in Aphanipathes alata, Brook, the
statement by Annandale (1914) that the membrane may give off small, but normal
branches of the Antipatharian, so that its nature seems to be beyond doubt.

* It is almost certain that this species was found upon an Antipatharian. The description reads,
“ The specimens are attached to long, yellowish-brown hairs or spines, which are hollow and bear
triangular pointed teeth placed in irregular rows. They were taken, October 1899, near Saleyer
anchorage, at a depth of 10-25 m.”

116

“TERRA NOVA” EXPEDITION.

Many of the branchlets of Antipathes lilliei at their proximal ends bear speci-
mens of an Oxynaspid, which seems to be distinct from the five species hitherto
known.

Oxynaspis terrce-novce, sp. n.

Description. — Valves partly separated (text-fig. 18, a), a distinct interspace
between middle lobe of scutum and basal part of tergum, and between scutum and
upper (tergal) limb of carina. Tergum with basal (scuto-carinal) angle rounded.
Scutum trilobed ; rostral angle produced into a short spur, which forms the basal
lobe ; middle lobe largest, rounded, undivided, reaching lower (scutal) limb of carina ;
umbo slightly above middle of occludent margin ; part above the umbo narrowed to
form upper (tergal) lobe, which meets the tergum for a quarter of the length of its
scutal margin ; a furrow extending from the umbo to a notch which separates the
rostral angle (basal lobe) from the middle lobe. Carina with umbo well below
middle, distal (tergal) limb being 2f or more times the length of basal limb.
Peduncle about the same length as capitulum, finely corrugated, springing from
flared opening of tubular sheath, formed on chitinous axis of a branchlet of Antipathes
lilliei. Capitulum and peduncle overgrown by antipatharian tissues, which cover the
cirripede with a thin chitinous membrane bearing spines. Caudal appendages
vestigial (text-fig. 18, c), each with two setse. Filamentary appendages : dorsal
absent, one pair at base of first pair of cirri (text-fig. 18, b). Mandible (text-fig. 18, e)
with four teeth and an acute, serrated proximal angle. Maxillula (text-fig. 18, f)
with inner edge hardly emarginate, with group of strong spines distally, and with very
prominent spinose proximal angle. Cirri with long setae on protuberances on basal
segments except on basal parts of those of first pair, where they are reduced to short
spinules (text-fig. 18, d). Two small conical bosses covered with short spinules
present between bases of first cirri. Ovigerous frena finger-shaped (text-fig. 18, b)
provided with a row of processes.

Remarks. — Only two complete specimens were found, of which one, the holo-
type, has been dissected, the valves and body being preserved in spirit, and
certain appendages being mounted as microscopical slides. The paratype remains
dry, and attached to a separated branch of the antipatharian. The size of some
of the sheaths, from which the cirripedes have become detached, indicates that the
holotype may be only half grown. The gross measurements of the holotype are as
follows : —

Length over all (exclusive of sheath), 15-6 mm.

Length of peduncle, from sheath to middle lobe of scutum, 6-8 mm.

Length of capitulum, from apex of tergum to base of carina, 10-6 mm.

Width of capitulum (maximum), 6 mm.

Externally this species most nearly resembles 0. hocki, Nilsson-Cantell, from
which it differs in having an undivided and relatively larger central lobe to the
scutum. The presence of filamentary appendages is noteworthy. Darwin (1851) said

Tig. 18. — Oxynaspis terrcc-novcc. A. Holotype ( x 5f). The spiny chitinous antipatharian sheath investing
t • peduncle and capitulum is only partly indicated. B. The body removed from the valves ( x 9), to show the
r< itive proportions of the cirri, filamentary appendage, and ovigerous frenum of the right side. C. Hind view of
tl body of the holotype ( x 85). Each caudal appendage bears two setae. D. The region of the bases of the first
ci i ( x 32), showing the general arrangement of the mouth-parts, the short spinules on the basal part of the pro-
ti erance (a) on the proximal segment of the first cirrus, the bosses (&) between the bases of the first cirri, and the
fi.tnentary appendages (c). E. The right mandible ( x 125), inner surface, showing groups of setae set in depressions
oi he surface. F. Right maxillula ( x 192), inner face, showing double row of enlarged setae on the edge.

118

“TERRA NOVA” EXPEDITION.

that in 0. celata there appeared to be no filamentary appendages ; but examples of
this species from an antipatharian in the British Museum, which in all probability is
the specimen from which Darwin removed his types, possess each a pair at the bases
of the first cirri. On specimens of an unnamed species from the holotype of Aphani-
pathes alata, Brook, two dorsal appendages are present in addition. Accordingly
the diagnosis of the family Oxynaspidce given by Nilsson-Cantell (1921) needs
revision.

VI— LIST OF PUBLICATIONS REFERRED TO.

Allen, E. J. — 1921. “ Regeneration and reproduction of the Syllid Procerastca." Phil. Trans. R. Soc.,

London, B. 211, p. 231.

Annandale, N. — 1909. “An account of the Indian Cirripedia Pedunculata . . . Lepadidae.” Mem.
Indian Mus., Calcutta, II, p. 61.

„ — 1914. “New and interesting pedunculate Cirripedes from Indian Seas.” Rec. Ind.

Mus., Calcutta, X, p. 278.

Aurivillius, C. W. S. — 1892. “ Neue Cirripeden aus dem Atlantischen,Indischen undStillen Ocean.”

Ofversigt af K.-Vet.-Ak. Forhandlingar, Stockholm, XLIX, No. 3.

,, — 1894. “ Studien iiber Cirripeden. I. Morphologie und Systematik neuer oder wenig

bekannter Cirripeden.” K. Svenska Vet.-Ak. Handl., Stockholm, XXVI, No. 7.

Broch, H. — 1922. “ Studies on Pacific Cirripedes.” Vidensk. Meddelelser Kebenhavn, LXXIII,
p. 215.

Brook, G. — 1889. Aniipatharia. Report Sci. Res. H.M.S. Challenger . . . Zoology XXXII.

Cooper, C. F. — 1909. Rep. Percy Sladen Trust Exp. Indian Ocean, 1905. Antipatluxria. Trans.

Linn. Soc., London, XII, p. 301.

Dantan, J. L. — 1920. “Recherches sur les Antipathaires.” Arch. anat. microsc., Paris, XVII,

p. 137.

Darwin, C. — 1851. “A Monograph on the Sub-Class Cirripedia . . . the Lepadidae. . .” London,

The Ray Society.

Duchassaing de Fonbressin, P. — 1870. Revue des Zoophytes et des Spongiaires des Antilles. Paris.
Gravier, C. — 1921. “Antipathaires provenant des campagnes des yachts Princesse- Alice et Hiron-
delle II ” (1903-13). Res. camp. sci. Monaco, Fasc. LIX.

Hickson, S. J. — 1903. The Alcyonaria of the Maldives, Part I. Gardiner, J. S., “ The fauna and
geography of the Maidive and Laccadive Archip.,” II (1), p. 473.

,, — 1916. “ The Pennatulacea . . Siboga-Expeditie, Monogr. 14.

Hoek, P. P. C. — 1907. “ Cirripedia Pedunculata''' Siboga-Expeditie, Monogr. 31 a.

Hornell, J. — 1922. “Some commensals of Indian Alcyonarians and crabs.” J. Bombay Nat. Hist.
Soc., XXVIII, p. 926.

Johnson, J. Y. — 1899. “ Notes on the Antipatharian corals of Madeira.” Proc. Zool. Soc., London,

p. 813.

Koch, G. von. — 1887. Die Gorgoniden. Fauna und Flora des Golfes von Neapel . . . , Monogr. XV.
McIntosh, W. C. — 1879. “On a remarkably branched Syllis, dredged by H.M.S. Challenger.

J. Linn. Soc. London, Zool., XIV, p. 720.

Nilsson-Cantell, C. A. — 1921. “ Cirripeden-Studien . . .” Zoolog. Bidr., Uppsala, VII, p. 75.

Pesch, A. J. van. — 1914. “ The Antipatliaria of the Siboga Expedition.” Siboga-Expeditie, Monogr. 17.
PouRTALks, L. F. de. — 1867. “ Contributions to the fauna of the Gulf Stream at great depths.” Bull.

Mus. Comp. Zool., Cambridge, Mass., I, No. 6, p. 103.

ANTIPATHAEIA— TOTTON.

119

Pourtales, L. F. de. — 1874. “Zool. Ees. . . . Hassler Exp., 1 . . . Deep-sea Corals.” Illustrated

Cat. Mus. Comp. Zool., Cambridge, Mass., No. 8.

„ — 1878. “Dredging . . . U.S.C. S.S. ’Blake.’ 2, . . . Corals . . .” Bull. Mus. Comp.

Zool., Cambridge, Mass., V, No. 9, p. 197.

Eoexe, L. — 1905. “ Description des Antipathaires . . . Atlantique nord.” Ees. camp. sci. Monaco,

XXX.

Semper, K. — 1888. “ The natural conditions of existence as they affect animal life.” 3rd edition,

London.

Selberfeld, E. — 1909. “ Japanische Antipatharien.” Abh. bayer. Ak. Wiss., Miinchen, 1. Suppl.-

Bd., 7. Abh.

Stebbing, T. E. E. — 1900. “On Crustacea brought by Mr. Willey from the South Seas.” Willey, A.,
“ Zoological Eesults . . . Part V, p. 605.

Steenstrttp, J. J. S. — 1851. “ Beskrivelse over nogle maerkelige, tildeels nye Dyrarter.” Oversigt

over det Kgl. danske Videnskabernes Selskabs Forhandlinger, Kjobenhavn, 1850.

Thomson, J. A. and Simpson, J. J. — 1905. “ Eeport on the Antipatharia . . . Ceylon.” Herdman, W. A.,

“ Eeport . . . Pearl Oyster Fisheries . . .,” Part IV, Suppl. Eept. 25.

Printed in England at the Oxford University Press by Frederick Hall

PLATE I.

Fig. 1. — Antipathes lilliei, sp. n. (nat. size) ; young branch, showing only one series of branchlets.

Fig. 2. — A. lilliei ; ( x ^). Lateral view of holotype.

Fig. 3. — A. lilliei ; end view of branch ( x §) to show the four series of branchlets.

Fig. 4. — Aplianipatlies larbadensis, Brook ; end view of the stem of the holotype ( x 2^), to show the
formation of a ‘ worm-run ’ by overarching of the branchlets. (Cf. text-fig. 10, a).

Fig. 5 .—Antipathes glutinata , sp. n. ; holotype (nat. size).

Fig. 6. — Bathypatlies platycaulus, sp. n. ; the two holotype fragments ( x §). The tip of the apical
portion on the right is recurved.

Brit.Mus. (Nat. Hist.)

Brit, Antarctic ( Terr a Nova) Exp e d.1910 .

Z oology ,Voi.V.

Coelenter ata,P art III, PI. I.

Ccelenterata, Part III.— Antipatharia, PI. II.

PLATE II.

Fig. 1. — Antipathes aperta, sp. n. ; fragments of the holotype (nat. size).

Fig. 2. — A. aperta ; reconstructed holotype ( x f). Two commensal cirripedes may be seen at
bases of the lower branches.

Fig. 3. — Bathypathes scoparia, sp. n. ; fragment of the stem with four series of branches ( x |).

Fig. 4. — B. scoparia ; fragment of stem with branches ( x |) ; lateral view.

Fig. 5. — Parantipathes tenuispina, Silberfeld ; ( x £§).

Brit. Antarctic ( Terr a N ova) Exp e d 1910 .

Huth coll.

BRITISH MUSEUM (NATURAL HISTORY).

BRITISH ANTARCTIC (“TERRA NOVA”) EXPEDITION, 1910

NATURAL HISTORY REPORT.

ZOOLOGY. VOL. V. No. 4. Pp. 121-130.

COELENTERATA.

PART IV.— MADREPORARIA.

(b) Turbinolidae and Eupsammidae.

BY

J. STANLEY GARDINER, M.A., F.R.S., F.L.S.

WITH ONE PLATE.

LONDON i

PRINTED BY ORDER OF THE TRUSTEES OF THE BRITISH MUSEUM

SOLD BY

B. Quaritch, Ltd., 11 Grafton Street, New Bond Street, W.l ; Dulau & Co., Ltd., 32 Odd Bond Street,
London, W.l; Oxford University Press, Warwick Square, London, E.C.4
Oliver & Boyd, Tweeddale Court, Edinburgh

and at

The British Museum (Natural History), Cromwell Road, S.W.7

1929

[All rights reserved.]

Price Two Shillings and Sixpence.

[Issued 17 th August, 1929J

British fIDnseum (IRatural Ibistors).

This is No. ^ 0/2 5 copies of

“ Terra Nova ” Zoology , Vol. V No. 4,
Coelenterata , printed on Special paper.

COELENTERATA.

PART IV— MADREPORARIA.

(b) TURBINOLIDAE AND EUPSAMMIDAE.

BY J. STANLEY GARDINER, M.A., F.R.S., F.L.S.

WITH ONE PLATE.

PAGE

I. Introduction ........... 121

II. Notes on collecting and preservation . . . . . . 122

III. Systematic notes and description of new species . . . . 122

IV. Note on the anatomy of Gardincria antardica . . . . . 128

I. INTRODUCTION.

Of the Madreporarian corals obtained by the “ Terra Nova ” Expedition, those from
the Atlantic Ocean and belonging to the genus Favia have been reported on by Prof.
George Matthai (Zoology, Yol. V, No. 2, 1919). The remainder are discussed in this
report.

The specimens were obtained in two widely distant areas, the Antarctic (Glacial)
area of the Ross sea, and the temperate area north of New Zealand.

The following is a list of the species : —

Antarctic

Gardineria antardica , sp. n.

,, lilliei, sp. n.

Balanophjllia, sp. ?

Theco'psammia, sp. ?

v. 4.

121

1

122

“TERRA NOVA ” EXPEDITION.

New Zealand

Flabellum harmeri, sp. n.

Gardineria, sp. ?

Desmophyllum crista-galli E. & H.

Caryophyllia profunda Moseley.

Trochocyathus [ Thecocyathus ], sp.

Dendrophyllia japonica Rehb.

NOTES ON COLLECTING AND PRESERVATION.

The examination of large numbers of specimens of these deep sea and mostly
solitary corals is exceedingly desirable, so that the criteria for species determination
can be ascertained in each genus. Hence collectors should obtain as many specimens
as possible. Indeed, in little-known seas I would submit that it would frequently
be of more scientific value to put down the nets a second time on an area proved to be
rich in organisms than to seek to explore a new line. Isolated hauls of dredge or trawl
allow no certain deductions as to the nature of the bottom.

Solitary corals may all be preserved in the flesh, while for colonial corals I recom-
mend drying, with a single piece of each mass, and with duplicate label, preserved.
In the field, killing and preserving is usually one process. Clean and dirty alcohol
and, to a lesser degree, different percentages of alcohols show different histological
results, while iron salts produce abnormal appearances in the fibrillation, perhaps
different coagulation appearances. Formalin tends to get acid and damages the
corallum. On the whole I prefer spirit, and I think it as well to drop all corals into
pure spirit 75-80 per cent., the bulk of spirit at least three times that of the corals,
changed once before being finally stored. The flesh alters little or not at all in 30 years,
provided that the corals are kept in the dark.

The research on corals of the last 30 years, particularly the lately issued catalogue
of the British Museum, emphasizes the importance of the preservation and examination
of the coral polyps. As there is a danger of going too far — and hence becoming
unpractical — I record my opinion that, if the limits of one species of a genus are worked
out on both soft and hard parts, the limits of all other species in the same genus will
be found deducible with considerable certainty from hard parts alone. Pure histological
studies had better be made in equipped laboratories rather than on the collections of
expeditions not preserved for such purposes.

SYSTEMATIC NOTES AND DESCRIPTION OF NEW SPECIES.

Flabellum, harmeri, n. sp. (PI. I, figs. 19, 20.)

There are three specimens of which only one is adult. It has the same number
of septa as F. curmtum Moseley but a very different shape. In this respect it is much

COELEXTERATA.

123

closer to F. transuersale Moseley, blit the latter is dense and heavy. The young forms
have points of resemblance to F. jpatagonichum Moseley, but the adult is more com-
pressed and there are 5 cycles of septa. These with F. multifore Gardiner and certain
other species, the types of which I have not seen, seem to represent a form of growth,
other forms being represented by F. pavoninum, F. rubrum and F. japonicum. Of
these we have some knowledge of the rubrum and of the pavoninum forms,* but all
others require to be correlated by a re-examination of the types and all other available
material. A tentative description of the present species follows : —

The corallum is elongated, compressed, conical with rounded ends, the angle of
the ends to the base of attachment being about 45°. Epitheca thin, transparent, with
transverse growth rings outside and with little emphasis of septal ridges.

The mouth of the calicle is oval, the longer axis about double the shorter and the
summit of the side wall about 1 cm. higher than at the ends. The septa are not exsert
and slope inwards at an angle of less than 45° before they drop perpendicularly to the
columellar junctions of the septa, which are situated about half-way down in the calicle.
There are 5 complete cycles of septa with some septa of cycle VI. Three cycles fuse
to form the trabecular columella, together with such septa of cycle IV as have septa
of cycle VI in their part of the corallum. The septa are thin and delicate and tend to
chip, giving an appearance of irregular teeth. Growth lines are visible on the septa,
parallel to their edges and each covered with a line of spines.

Largest specimen 50 mm. high, with calicle 42 x 21 mm. A second dead specimen
was clearly a little larger. Two young specimens 16 and 12 mm. high, with calicles
about 13 x 9 nun. and 3x6 mm., have 12 septa fusing to form the columella, and
most of the septa of cycles 3 and 4 are developed.

Station 91. Near Great King Island, N.Z., 300 fms.

Genus GARDINERIA Vaughan.

Duncania, Pourtales, Illustr. Cat. Mus. Comp. Zool., Harvard, VIII, 1874, p. 44 ; Duncan,
Journ. Linn. Soc. Zool., XVIII, p. 32, 1885 ; Gardiner, Marine Investigations in S. Africa,
III, p. 120, 1904. Gardineria, Vaughan, U.S. Nat. Mus., Bull. LIX, p. 65, 1907.

I accept Vaughan’s statement that de Koninck in 1872 proposed the name Duncania
for a carboniferous coral and only regret that Vaughan did not suggest a new generic
name for Pourtales’ species. He unfortunately described a new genus, Gardineria,
on a single specimen, giving as a character that “ new coralla arise from the old by
internal gemmation,” a phenomenon which may be found in any of these solitary corals
but presents rather a different appearance in the epithecate-walled coralla of
Rhizotrochus and Flabellum to the coralla of Desmophyllum, etc., which are covered
over by living tissues and are therefore thecate. I have already expressed my views
on “ stereoplasma.” Any Turbinolid coral if growing under unfavourable conditions

* See especially Gardiner, Marine Investigations in S. Africa, II, pp. 114-154, 1902, and Vaughan,
U.S. Nat. Mus., Bull, 59, pp. 48-64, 1907.

124

“TERRA NOVA” EXPEDITION.

on the calicular surface tends to show such deposition of corallum filling in the inter-
septal chambers. The contrast of the “ multiple pillared columella " of Pourtales
with the “ weak false columella ” of Vaughan breaks down when the present specimens
are examined. Clearly Vaughan’s name must be adopted for the genus, with which
Rhizotrochus, Flabellum and Haplophyllia are related, so far as the coralla are concerned.

Gardineria antarctica, n. sp. (PI. I, figs. 11-18.)

The corallum is cylindrical with a widely open rounded calicle. It is sometimes
obliquely attached and sometimes freed by the wearing away of its wall and base of
attachment. The wall is a relatively thick epitheca, sometimes with circular growth
markings (often overgrown by Polyzoa, worms, etc.), but where bare commonly showing
low ridges over the attached edges of the septa. By secondary deposition of corallum
within the cup, the wall and all calcareous structures are thickened as growth
proceeds.

Twelve septa meet in the centre, each having 2 or 3 blunted teeth, otherwise their
edges being smooth and straight. These are alternated by other 12 septa, which reach
about half-way to the centre and then slope perpendicularly to the bottom of the cup.
Between these there are a further 24 smaller septa, and there are some others of cycle V
(20 in a corallum of 19 mm. diameter, while this cycle is nearly complete in a specimen
of 26 mm.). The septa do not project above the edge of the epitheca, so that the saucer
shape of the surface defined by their oral edges is conspicuous.

In the younger stages the first 6 septa meet in the centre of the corallum and may
be joined by their twisted and folded edges. In the older stages there are 12 septa so
joined and on this trabecular columella arise about 6 rather irregular pillars.

Station 314. Near Inaccessible Island, McMurdo Sound, 222-241 fms. 4 speci-
mens measuring 26, 20, 19 and 17 mm. in diameter, by 28, 26, 22 and 21 mm. high, the
first and last dead when obtained, the last an unattached cone.

Station 349, McMurdo Sound, 80 fms. 3 specimens 19, 17 and 11 mm.
diameter by 20, 15 and 12 mm. high ; the third dead, the second obviously immature
in spite of its size with as yet no septa of cycle V and the columella rods not properly
formed. I have also examined a broken-up specimen from the upthrust sea-muds
S.E. of Mt. Larsen, collected in 1907 (?) by Prof. David (G 12103) and 2 fine specimens
collected by Mr. B. Priestley at Evans Cove, Victoria Land, in a glacier 30 feet above
the sea.

The coralla show considerable variation correlated with shape of growth, and this
may affect the internal septa causing irregularities in their edges, particularly, as in a
cornute form, where peculiarly rapid growth is required on one side, so that the polyp
mouth may remain horizontal and rise above the bottom of the sea and its deposits,
etc. The first teeth of the larger septa will probably be found at times to simulate
pali, but the specimens before me are all clearly distinct from D. capensis, the type of

COELEXTERATA.

125

which is placed with them in the British Museum. The rings on the outside seem to
be largely due to spurts in growth. An account of the polyp's anatomy is given below.

Gardineria lilliei, n. sp. (PI. I, figs. 3-10.)

The corallum is similar to the last but the epitheca is an exceedingly thin wall,
without much internal thickening by deposition of corallum, and the circular growth
markings are more conspicuous and numerous.

The maximum number of septa meeting in the centre is 10 and the maximal total
number is 34. They slope at an acute angle into the cup, the columella in which lies
at a depth of the half at least of its diameter. The edges of the septa are generally
smooth with few or no teeth, but they may be wrinkled. The sides of the septa are
as in the last species smooth.

The fusion of the larger septa in the centre of the cup, forming a twisted mass of
plates, may have upstanding plates or pillars ; the whole forms a trabecular columella.

Station 194. Off Oates Land, 180-200 fms. The specimens are as follows :
13 mm. diameter x 17 mm. high, 34 septa ; 11 x 17, 28 ; 10 x 11, 32 ; fix 22, 16 ;
12 x 17, 31 (top partially broken) ; 10 x 13, 32.

Gardineria, sp.

There are 3 further specimens from Station 96, New Zealand, 70 fms., of similar
sizes and of similar delicacy. They differ in having the 4th cycle of septa more com-
plete, the septal sides studded with low blunt spines and the septal edges at the level
of the upstanding trabecular columella much twisted, the whole thus presenting a
peculiarly rough appearance. There is also a specimen from Station 90 off Three Kings
Islands, 100 fms., a bent horn 32 mm. maximal height with diameter of calicle about
11 mm. A further specimen from the same station, now much broken, was approxi-
mately 23 mm. in diameter by the same in height. It has similarly rough, twisted and
spiny septa, and of these there are four full cycles (48) and a few septa of cycle V. 1
regard these 5 specimens as likely to prove to belong to the same species, and that may
be an undescribed one for there are no clear intermediates to D. lilliei. I am not,
however, going to give them a name for they may be 2 species — or even, a most unlikely
supposition so far as the large form is concerned, growths under unfavourable conditions
of D. lilliei.

Desmophyllum crista-galli Ed. & H.

Milne Edwards and Haime, Ann. Sci. Nat., Paris, IX, p. 253, pi. 7, fig. 10, 1848. See also
for figures : Duncan, Trans. Zool. Soc., London, VIII, p. 321, pi. XLI, figs. 10-15, 1873 ;
and Marenzeller, “ Steinkorallen,” Wiss. Ergebn. “ Valdivia,” VII, p. 267, pi. XV, fig. 2,
1904.

Two dried specimens, the second dead when dredged and a block growing with a

species of Stylaster, retained in spirit. The largest calicle is 43 mm. long by 32 mm.

1 a

126

“TERRA NOVA” EXPEDITION.

broad and about 65 mm. high. Both specimens show a number of small attached
coralla. Some of these are Desmophyllum, collectively showing stages from 6 to 48
septa, 1 to 4 cycles ; I think they are all D. crista-gall i, believing the great exsertness
of its septa to be a gradual and later development. Other attached forms are young
CaryopJiyllia. There is no columella of any sort, this with size being the chief difference
from D. ingens Moseley.

Station 91. Near Great King Island, N.Z., 300 fms.

Caryophyllia profunda Moseley.

Report Sci. Res. H.M.S. “ Challenger,” Zoology II. Part III. 1881, pi. I. fig. 6.

There is a single corallum 24 x 16 mm. by 32 mm. high, which clearly belongs to
the clavus-growp. Perhaps it is that fossil species, but here it is assigned to C. profunda.

There is a number of young coralla from this station, some of which are young
Caryophyllia, which at times is said to be exceedingly abundant, particularly between
100 and 200 fms. A large collection of any species would be of immense value so as
to obtain the species-criteria for the genus.

Station 91. Off Great King Island, N.Z., 300 fms.

Trochocyathus [Thecocyathus], sp. ?

A single little cone, 10 mm. high, with calicle 7*5 mm. in diameter must be referred
to this genus. It has theca and epitheca with growths of other organisms on the latter.
1 cannot judge whether this epitheca is a normal structure or not. Any polyp may
form corallum for protection over any part of its ectoderm outside the tentacular ring.
If it lias a theca, a foraminiferan, a worm or a polyzoan may settle and kill the polyp
tissues outside the same, the power of resistance to such varying very greatly in different
polyps. Where the costae are conspicuously high, there are endoderm-lined tubes
extending down between them, and the ectodermal tissues, for protection from such
sedentary organisms, may form an extra epithecal wall, as perhaps may be the case here.

Station 91, off Three Kings Islands, N.Z., 300 fms. 1 specimen.

Thecopsammia, sp. ? Balanophyllia, sp. ’(

There are six specimens from Station 91, three of which are covered by epitheca
extending close up to the edge of the calicle ( Thecopsammia ), while the other three are
smaller, more turbinate in shape, with only basal epitheca ( Balanophyllia ). A true
epitheca, if theca be present, shows intercostal spaces, and this is true of some
Thecopsammia. In Balanophyllia it is not unusual to find the base of the polyp killed
by mud, etc., which sweeps over its tissues. The tissues join up and retract inwards
depositing corallum between them and the mud, etc., usually retiring within the theca,
the holes in which are completely blocked by the new corallum thus formed ; the

COELENTERATA.

127

resulting surface appearance of the corallum is such as to simulate true epifkecal
structures. The basal part of the corallum is, by a similar process, blocked up with
solid corallum as the polyp grows and raises itself on its cup.

Unfortunately, neither species is identifiable with certainty, the species of both
genera requiring thorough revision.

Station 191. Bay of Whales, 194-250 fms.

Dcndrophyllia japonica Behberg. (PI. I, figs. 1, 2.)

Rohberg. Xcue unci wenig bekannte Korallen," Abh. Naturw. Hamburg, XII. 1892, p. 28,
pi. IV, fig. 1 ; van cler Horst, Trans. Linn. Soc. Zool., XIX, p. 44, pi. 3, figs. 4-5, 1926.

The chief specimen is a colony 20 x 16 cm. diameter x 13 cm. high and consists
of about forty corallites, which vary in size from 5 mm. diameter to 36 mm. diameter.
It is formed entirely by low branchings on the outer surface of a giant corallite, and
subsequently additional branchings from the corallites first formed. It was apparently
lying freely on the bottom, its base being dead and largely overgrown by worm tubes.
The photograph here published rejdaces further description.

There were also 3 single corallites obtained in the same dredging, one dead and two
either dead or moribund ; these were sent to Dr. van cler Horst, who identified them
as above, having compared them with Rehberg's type. He remarked, “ In type speci-
mens the calices are more oval and ribs better than in these dead pieces " ; and these
are characters which apply to the large colony. The septa number 6, 6, 12, 24, 48, by
cycles, and seldom exceed these numbers. In the smallest calicles there are 6 large
septa and in the largest 24 of almost equal size, ordinary calicles showing about 12.
The smaller calicles, even up to 12 mm. diameter, would be described as having no
columella, whereas in the largest oval-shaped calicles it extends to about 40 per cent,
of the length of the calicle, being 2-4 mm. broad and of a spongy texture.

These and other points in the skeletal structure lead me to doubt whether this
species is not a synonym of a previously described species, but such points can only
be determined on a large number of specimens. The polyp characters as seen in one
scries of sections show that it belongs to the same genus as Dendrophyllia aurea (Q. &
G.) = Coenopsammici willeyi Gardiner, but the specimen was so hardened and stained
by iron from the can that I am uncertain whether certain histological differences are
not due to peculiar coagulation products resulting from the iron salts present in the
spirit.

Station 91. Off Great King Island, N.Z., 300 fms.

128

“TERRA NOVA” EXPEDITION.

NOTE ON THE ANATOMY OF GARDINERIA

ANTARCTICA.

I have had cut serial transverse sections of a polyp of the above from station 314.
222-241 fms. ; these are deposited in the British Museum for subsequent reference.
The coral was 12 mm. across the calicle and the cut polyp measures 10 mm. in diameter.
The general description of the polyp in relation to the corallum given by me for Duncania
{now Gardineria) capensis * applies to this specimen. The differences lie in that the
present species has 24 instead of 12 tentacles, and there are rather diffusely scattered
circular muscular fibres, not a massed sphincter muscle. The common characters are
no edge zone, mesenteries twice the number of septa, which hence are all entocoelic,
12 mesenteries reaching the stomodoeum, similar nematocysts to those found in the
external ectoderm and in the ectoderm of the mesenterial filaments, no “ nematocysts "
in the endoderm, acontia present and generative organs in mesenteries of cycles I, II
and III ; most of these should prove to be generic characters.

The general structure of the polyps as well as its relationship to the corallum is so
close to that of Flabellum that detailed description and figures are unnecessary. The
12 septa of cycles I and II are of equal size and are only separable in the fact that the
mesenteries bounding half of them (cycle II) do not reach the stomodoeum ; the
directive mesenteries are clear but with fewer muscular fibres and less pleating than
the others. There are 12 septa and 12 couples of mesenteries of cycle III and 5 septa
and 5 couples of cycle IV. The tentacles lie over the septa and would seem to form
2 rings of 12 each, those over septa I and II slightly external in the expanded polyp.
The couples of mesenteries concerned with each tentacle clearly extend for some
distance, which I should estimate at something under 25 per cent, of its length, into
each tentacle when fully expanded. Thickenings with nematocysts suggest the com-
mencing formation of tentacles over the 5 septa of cycle IV. The polyp is female, but
there are few eggs, usually 1 or 2 in a mesentery, most mesenteries without eggs ; how-
ever, they occur in cycles I, II and III.

The polyp was not sufficiently well preserved to justify histological description,
but it may be mentioned that the calicoblastic layer is everywhere well defined and
similar to that described in Flabellum ,j* as are also the processes of attachment of the
mesenteries to the corallum. The ridges in the stomodoeum over the mesenterial
attachments of cycle I and the mesenterial filaments have the same histological elements
and hence presumably the same functions, though in different degrees ; their nemato-
cysts are very sparingly present. The muscular fibres are well defined and the pleated
mass of muscle in the middle of each mesentery is large ; there are more muscular fibres
on the opposite sides of the mesenteries than is usual in corals. The histological

* Gardiner, Marine Invest. S. Africa, III, pp. 106-127, 1904.
f Gardiner, Marine Invest. S. Africa, II, pp. 137-143, 1902.

COELEXTERATA.

129

appearance of the structureless lamella or mesogloea is such that 1 was impelled to
re-examine this “ layer ” in a series of genera of solitary corals. I find, even in the
same species, wide differences in the appearances after different methods of preser-
vation. Cells that have taken up isolated positions in it are found in all the Zoantharia,
but I can find no connectives between them. Some of the “ fibres " seem to me to be
tendinous in function, but others to be only explicable as coagulation products and
hence not existent in the living polyps ; I await further histological examination based
on living corals before accepting Matthai’s views.* I think the technique elaborated
in the Madrid school of Ramon v Cajal is likely to yield valuable results.

There are throughout the endoderm of this polyp (not mesenterial filaments) round
bodies, which are of the shape and size of the commensal algae found in similar positions
in the reef corals, giving to these their characteristic green coloration and assisting in
their respiration and feeding. They are generally round, and they possess a stained
nucleus. The one or two vacuoles, characteristic of the reef commensals, are not
usually present, but there is generally indication of one such space in a lighter area
as the organism is focussed. For the rest their contents are little granular. While
most are round, odd ones look as if they were disintegrating, being irregular, often
rather oval. They are found everywhere in the above layer, and they are literally
scattered not massed, their numbers in any part being proportional to the bulk of the
endoderm that is present. I have found none in any part of the ectoderm, with which
I include stomodoeum and mesenterial filaments, though a very few odd round similar-
sized cells were found as inclusions in the latter.

There is no possibility of confusing the above bodies with nuclei or other tissue
elements. Accordingly, we must regard them either as parasites, or as food cells or
organisms carried into the coelenteron of the animal and absorbed into its endoderm.
Their whole appearance is of plant cells, and they do not look like plant organisms or
cells, which were in suspension in the water and have sunk about 200 fms. from the
light zone so as to be available as food for our polyp. Such ingested food in reef and
solitary corals is usually, if in the solid state, massed in the thick endodermal ridges
behind and supporting the filaments f ; and this is not the case here. While a very
few of these cells may possibly be being digested, this is not a general but a very
rare phenomenon. \

Under these circumstances I am impelled to suggest that these small bodies are
algae related to or the same as the commensal forms found in so many corals. If so,

* Quart. J. Micr. Sci. London, 67, pp. 103-122, 1923.

t Large masses of tissue are commonly held in the stomodoeum or between the mesenterial filaments,
where they gradually break up into fine particles which are ingested by the endoderm of any part but
mainly near the attachments of the filaments.

X In reef corals, if zooxanthellae are to be digested, this may be effected in situ in the endoderm, but
in two cases I have seen they were cast out into the coelenteron as a preliminary to this process ; the first
change was the loss of their chlorophyll ; the second a loss of their vacuole ; and the third a crinkling and
irregularity of their wall.

130

“TERRA NOVA” EXPEDITION.

they have lost their chlorophyll and have hence become parasites upon our polyp.
There is no difficulty in respect to the infection of new generations, since commensal
algae in corals usually penetrate the young polyp either as an egg or in the early stages
of its development, these being passed in the coelenteron of the mother. They are
clearly not “ the yellowish- green corpuscles of irregular shape ” mentioned by Boschma
as occurring in the endoderm of Dendrophyllia micranthus (Ehrb.) and D. coccinea
(Ehrb.) * ; I have failed to find these corpuscles or bodies similar to those now men-
tioned in D. aurea (Q. & G.) = Coenopsammia willeyi Gard., which is widely distributed
over coral reefs, living generally in dark spaces under stones and in crevices of surface
reefs, often between tide marks. As against this suggestion that these are algae, I
have not found them in any of the Flabellum and other Turbinolidae from S. Africa
( loc . tit.), amongst which Heterocyathus aequicostatus (Ed. and H.), 36-90 fms., is
included. However, in one set of sections of the latter from the Ceylon Pearl Banks,
about 5 fms., I have found similar bodies ; they are nowhere numerous or massed, but
they are mainly in the endoderm which immediately underlies the external ectoderm,
the proper position for commensal algae.

* Boschma, Proc. Koninklijke Akad. Amsterdam, XXVII, p. 21, 1924 ; Gardiner, Willey’s Zool.
Results, pp. 357-80, 1900 ; van der Horst, Trans. Linn. Soc. Zool., XIX (I), p. 47, 1926.

PRINTED IN GREAT BRITAIN BY WILLIAM CLOWES AND SONS, LIMITED, LONDON AND BECCLES.

Coelenterata, Part IV, PI. I.

PLATE I.

Dendrophyllia japonica Rehberg.

Pig. 1. — View of whole corallum. Xy7o

„ 2. — Oral view of the uppermost calicle in Fig. 1. Below it is a small calicle in which the columella
is not yet developed. Nat. size.

Gardineria lilliei, sp. nov.

Figs. 3-6. — Lateral view of four specimens. Nat. size.

„ 7-10. — Oral view of same specimens in same order, x 2.

Gardineria antarctica, sp. nov.

Figs. 11-12. — Oral view of specimens from Station 349. x 2.

„ 13-14. — Oral view of specimens from Station 314. X 2.

„ 15-16. — Lateral view of specimens shown in Figs. 11 and 12 respectively. Nat. size.

„ 17-18. — Lateral view of specimens shown in Figs. 13 and 14 respectively. Nat. size.

Fig. 19. — Oral view. Nat. size.

,, 20. — Lateral view. Nat. size.

Flabellum harmeri, sp. nov.

Brit. Mus. (Nat. Hist.).

Brit. Antarctic (“ Terra Nova”) Exped. 1910.
Coelenterata, Vol. V. No. 4.

PI. I.

*>vv

BRITISH MUSEUM (NATURAL HISTORY).

BRITISH ANTARCTIC (“TERRA NOVA”) EXPEDITION, 1910

NATURAL HISTORY REPORT.

ZOOLOGY. VOL. V, No. 5. Pp. 131-252.

COELENTERATA.

PART V.— HYDROIDA.

BY

A. KNYVETT TOTTON, M.C.

ASSISTANT KEEPER IN THE ZOOLOGICAL DEPARTMENT, BRITISH MUSEUM (NATURAL HISTORY).

WITH SEVENTY TEXT-FIGURES AND THREE PLATES.

7

v «• /

LONDON i

PRINTED BY ORDER OF THE TRUSTEES OF THE BRITISH MUSEUM

SOLD BY

V.Qiariich, Ltd., 11 Grafton Street, New Bond Street, W.l ; Dulao & Go., Ltd., 32 Old Bond Street,
London, W.l ; Oxford University Press, Warwick Square, London, E.C.l

AND AT

The British Museum (Natural History), Cromwell Road, S.W.7

1930

[All rights reserved.}

Price Fifteen Shillings.

d 2 '2nd February , 1930J

British flDuseum (IRatural Ihistors)

This is No. 0' of 25 copies Oj
“ Terra Nova ” Zoology , Vol. V., No. 5
Coelenterata , printed, on Special paper.

HYDROIDA.

BY A. K. TOTTON.

Assistant Keefer in the Zoological Department, British Museum ( Natural History).
WITH SEVENTY FIGURES IN THE TEXT, AND THREE PLATES.

PAGE

I. Introduction ............ 131

II. List of Species ........... 133

III. Station List . . . . . . . . . . . .136

IV. Description of Genera and Species. ....... 138

V. Reference List of Works ......... 243

VI. Index 249

I. INTRODUCTION.

The Hydroida collected during the voyage of the “ Terra Nova ” in the years 1910-13
were obtained in three widely distant areas : the Temperate area of Northern New
Zealand; the Subantarctic (Magellan) area west of the Falkland Islands; and the
Antarctic (Glacial) area of the Ross Sea and off the coast of Oates Land.

The dredge, trawl or tangle were used at 25 stations in all : Hy droids were taken
at 16 of them. At these the bottom was either shingle, rock, rock and sand, shell,
“ undecomposed animal remains ” or, as in the five McMurdo Sound stations, mud.
At four of the remaining eleven stations where hvdroids were not taken the bottom
was mud, and at two others, sand.

Descriptions of the hyclroids of the Ross Sea area have been given by Hickson and
Gravely (1907), Ritchie (1913), and Jaderholm (1926). Although the hydroid fauna of
the Antarctic is characteristic of, and to a great extent restricted to that area, we have
not found as evidence for the long duration of this isolation that the hydroice have
split up into a large number of greatly different types.

In general the Antarctic hydroid species are known with certainty only from the
Antarctic or Subantarctic areas. Three of the Lafooid species are alleged to be more
V. 5. 1

]32

“TERRA NOVA” EXPEDITION.

nearly cosmopolitan, but tlie accurate delimitation of many Lafooid species is not at
the moment possible, and the statement may not be correct.

The Expedition brought back only one new Antarctic hydroid, though it has been
found necessary to give new names and specific descriptions for three forms already
known from that area, but assigned to species to which they do not belong. Much
additional information is given about most of the known species.

The cosmopolitan genus Symplectoscyphus is well represented in the Antarctic and
Subantarctic areas, but other Sertulariid genera are absent.* There are no Antarctic
Statoplean Plumulariidae, and the Eleutheroplea are represented only by the primitive
genus Oswaldella, which is endemic and has two well separated species, and the cosmo-
politan genus Schizotricha.

References to all the hitherto known kydroids of New Zealand, about 120 in number,
can be found in four papers by Busk (1857), Jaderholm (1917), Bale (1924), and Trebil-
cock (1928). To this list have to be added twenty-seven new species and three old
species. The Museum is fortunately rich in types of New Zealand species of Busk
and Allman, several of them having been found and recognised for the first time during
the preparation of this report.

As a result of the study of the collection it appears that sexual dimorphism is
common amongst hydroids. In the genus Symplectoscyj)hus, for instance, it is important
from a systematic point of view to note the shapes of the gonothecae of the two sexes.
It is possible that in the past the two sexes of a single species have been described as
distinct species.

It is desirable in making systematic studies of Hydroida to take measurements of
the chitinous perisarc, and the points of reference used in making measurements should
conform to a standard for each genus. The fact that all the repeated parts that go
to form a colony may not be of one size must be allowed for. Presumably this variation
in size is correlated with change of rate of growth from region to region. Again, in some
species polyps periodically drop from their pedicels to be replaced by successively larger
ones ; whilst in racemose species the terminal polyps, which increase in size as growth
proceeds, are much larger than those budded laterally from them. In pinnate species
there is often a graded change in form and size from the proximal to the distal parts
of any particular pinna. The basal theca of the hydroclade of a much ornamented
species is often simpler and of ancestral type.

Attention may be directed here to the structure and homologies of corbulae.
Billard (1913) rightly expressed the opinion that they were secondarily branched
hydrocladia, but confirmed Allman’s statement that the distal part of the costa was
a replacement of the median nematotheca found below the theca (when present) on the
basal part of the costa. Further study indicates that the primary hydroclade of the
gonosome, which may be termed the gonocladium, may bear several orders of branches.

* Selaginopsis pachyclada, Jaderholm, appears to be a species of the Antarctic genus Staurotheca.

HYDROIDA— TOTTON.

133

First of all there are the gonohydrocladia, known also as the cretes basales, which often
bear hydrothecae, and whose distal parts may be prolonged into small leaflets bearing
nematothecae. The branches of the next order may be termed the costae. They arise
from the gonohydrocladia at points between the intrathecal nematothecae and the
hydrothecae. They form the main part of the corbnlae and are provided with numerous
nematothecae as a rule. The gaps between the bases of these costae are often filled
up by a final order of branch, the costal apophyses, at the bases of which may be seen
traces of the usual three nematothecae which surround a hydrotheca. The significance
of these is that branches as a rule arise from the bases of hydrothecae.

Observations on the situation of nodes in the stems of Sertularia marginata have
suggested an hypothesis which makes it possible to link np primitive species of Sertularia
with those which show more advanced colonial characters.

In the course of a discussion of many questions of nomenclature in Part IV of this
report, a selection of the genotypes of the following nine genera has been made : — -
Atractylis, Wright, 1858 (p. 139) ; Stegopoma, Levinsen, 1893 (p. 154) ; Grammaria,
Stimpson, 1854 (p. 161) ; Lictorella, Allman, 1888 (p. 166) ; Abietinella, Levinsen, 1913
(p. 166) ; Perisiphonia, Allman, 1888 (p. 166) ; Hincksella, Billard, 1925 (p. 151) ;
Pycnotheca, Stechow, 1919 (p. 214) ; and Thecocaulus, Bale, 1915 (p. 217).

II. LIST OF SPECIES.

In the following list the areas Antarctic (Glacial), Antarctic (Magellan), and Northern
New Zealand are indicated by the initial letters A, M, and N.Z. respectively.

ATHECATA.

BO UGAINVILLIIDAE .

A. Hydronema anguslum (Hartlaub, 1904).

A. Gravelya antarctica (Hickson and Gravely, 1907).

Eudendriidae.

A. Eudendrium antarcticum, nom. nov. (For E. ramosum, Hartlaub,

1904, not E. ramosum (L.) 1758).

N.Z. Eudendrium, sp.

THECATA.

Haleciidae.

A. Ophiodissa arbor ea (Allman, 1888).

N.Z. Ophiodissa armata, sp. nov.

A. Halecium ovatum, sp. nov.

A. Halecium antarcticum, Vanhofien, 1910.

134

“TERRA NOVA” EXPEDITION

Campanulariidae.

N.Z. Tulpa diverticulata, sp. nov.

N.Z. Clytia compressa, sp. nov.

A. Campanularia hicksoni, nom. nov. (For C. laevis, Hickson and G.,

1907. not C. laevis, Hartlaub, 1905.)

N.Z. Billardia novae-zealandiae, sp. nov.

M. ; A. Billardia subrufa (.Taderholm, 1904).

C AMPANULINID AE .

A. Campanulina belgicae, Hartlaub, 1904.

A. Stegella grandis (Hickson and Gravely, 1907).

N.Z. Stegolaria irregularis, sp. nov.

N.Z. Stegopoma fastigiatum (Alder, 1860).

Lafoeidae.

M. Hebella striata, Allman, 1888.

A. Hebella plana (Ritchie). (For H. striata, Alim. var. plana, Ritchie,

1907.)

M. Lafoea fruticosa, Sars, 1862.

A. Lafoea dumosa (Fleming, 1828).

A. Lafoea gracillima (Alder, 1856).

A. Lafoea gaussica, VanhofEen, 1910.

A. B.eticularia antarctica (Hartlaub, 1904).

N. Z. Acryptolaria minima, sp. nov.

N.Z. Acryptolaria conferta (Allman, 1877), var. australis (Ritchie, 1911).
N.Z. Zygophylax unilateralis, sp. nov.

N.Z. Zygophylax sibogae, Billard, 1918.

Syntheciidae.

N.Z. Synthecium ramosum, Allman, 1885.

N.Z. Synthecium carinatum, sp. nov.

N.Z. Synthecium robustum, sp. nov.

N.Z. Synthecium longithecum, sp. nov.

A. Staurotheca dichotoma , Allman, 1888.

A. Staurotheca antarctica, Hartlaub, 1904.

SeRTULARIID AE .

N.Z. Parascyphus simplex (Lamouroux, 1816).

N.Z. Symplectoscyphus columnarius (Briggs, 1914).

N.Z. Symplectoscyphus constrictus, sp. nov.

N.Z. Symplectoscyphus johnstoni (Gray, 1843).

N.Z. Symplectoscyphus delicatulus (Hutton, 1872).

N.Z. Symplectoscyphus spiritualis, sp. nov.

N.Z. Symplectoscyphus confusus, sp. nov.

HY DROID A— TOTTOX.

13r>

X.Z. Symplectoscyphus epizooticus, sp. nov.

X.Z. Symplectoscyphus tuba, sp. nov.

A. Symplectoscyphus vanhoejfeni, nom. nov. (For Sertularella subclicho-

toma, Vanh., 1910. not S. subdichotoma, Kirchenpaner, 1884.)

A. Symplectoscyphus glacialis (Jaderholm).

A. Symplectoscyphus curvatus (Jaderholm. 1917).

A. Symplectoscyphus plectilis (Hickson and Gravely, 1907).

M. Symplectoscyphus filiformis (Allman, 1888).

X.Z. Sertularella robust a, Coughtrev, 1876.

N. Z. Sertularella geodiae, sp. nov.

X.Z. Sertularella spiralis , Hickson and Gravely, 1907.

A. Sertularella elongate, Jaderholm, 1904.

A. Sertularella biformis, Jaderholm, 1905.

X.Z. Sertularella edentula , Bale, 1924.

X.Z. Dictyocladium moniliferum (Hutton, 1873).

X.Z. Sertularia unguiculata, Busk, 1852.

X.Z. Sertularia marginata (Ivirchenpauer, 1864).

N.Z. C rater itheca zelandica (Gray, 1843).

Plumulaeiidae .

A. Oswaldella bifurca (Hartlaub, 1904).

A. Oswaldella antarctica (Jaderholm, 1904).

N.Z. Antennella ritchiei, sp. nov.

N.Z. Antennella serrate, sp. nov.

N.Z. Pycnotheca mirabilis (Allman, 1883).

N.Z. Halopteris constricta, sp. nov.

N.Z. Halopteris heterogona (Bale, 1924).

N.Z. Halopteris campanula (Busk), var. zelanica, var. nov.

N.Z. Plumularia pulchella, Bale, 1862.

N.Z. Plumularia diploptera, sp. nov.

N.Z. Plumularia spirocladia, sp. nov.

N.Z. Plumularia triangulate, sp. nov.

N.Z. Plumularia brachiata, sp. nov.

N.Z. Plumularia tenuissima, sp. nov.

N.Z. Nemertesia elongate, sp. nov.

A. Schizotricha unifurcata, Allman, 1883.

N.Z. Aglaophenia laxa, Allman, 1876.

N.Z. Thecocarpus ctenatus, sp. nov.

N.Z. Thecocarpus rostratus (Bale, 1924).

N.Z. Thecocarpus spiralis, sp. nov.

N.Z. Thecocarpus chiltoni, Bale, 1924.

N.Z. Halicornaria regalis, sp. nov.

136

“TERRA NOVA” EXPEDITION.

III. LIST OF STATIONS.

The following is a list of the localities at which Hydroida were collected by the Expedi-
tion, with the nature of the bottom and the gear used at each.

Station 38. West of Falkland Islands, April 13, 1913 ; lat. 52° 23' S., long. 63' 50' W. ; depth 125
fathoms (229 m.). Agassiz trawl.

Billardia subrufa.
Hebetta striata.
Lafoea fruticosa.

Stegella grandis.

Staurotheca antarctica.
Symplectoscyph us fl iform is.

Station 90. Off Northern New Zealand, July 25, 1911 ; from Summit, Great King, Three Kings
Islands, S. 14° W., 8 miles ; depth 100 fathoms (183 m.) ; bottom, rock. Dredge.

Billardia novae-zealandiae .
Symplectoscyphus columnari us.
Symplectoscyphus tuba.
Sertularella geodiae.

Sertularella edentula.
Dictyocladium moniliferum .
Crateritheca zelandica.
Antennella serrata.

Pycnotheca mirabilis.
Halopteris heterogona.
N enter tesia ekmgata.
Thecocarpus rostratus.
Thecocarpus spiralis.
Thecocarpus chiltoni.
Halicornaria regalis.

Station 91. Off Northern New Zealand, July 26, 1911 ; from Summit, Great King, Three Kings
Islands, S. 10° W., 25 miles ; depth 300 fathoms (549 in.) ; bottom, rock. Dredge.

Ophiodissa armata.

Tulpa diverticulata.
Stegopoma fastigiatum .
Reticular ia, sp.

Acryptolaria conferta.
Zygophylax unilateralis.
Zygophylax sibogae.
Synthecium longithecum.
Symplectoscyphus constrictus .

Symplectoscyphus confusus.
Syniplectoscyphus epizooticus.
Antennella ritchiei.

Halopteris campanula, var. zelanica.
Plumularia triangulate.

Plumularia tenuissima.

Thecocarpus ctenatus.

Thecocarpus chiltoni.

Station 96. Off Northern New Zealand, August 3, 1911 ; 7 miles East of North Cape ; depth 70
fathoms (128 m.) ; bottom, sand and rock. Agassiz trawl.

Stegolaria irregularis. Sertularella geodiae.

Acryptolaria conferta var. australis. Plumularia brachiata.

Station 134. Off Northern New Zealand, August 31, 1911 ; Spirits Bay, near North Cape ; depth
11-20 fathoms (20-37 m.) ; bottom, shelly. Dredge.

Eudendrium, sp.

Clytia compressa.
Acryptolaria minimi .
Synthecium carinatum.
Synthecium robustum.
Symplectoscyphus sp iritualis .
Sertularella robusta.
Dictyocladium moniliferum .
Parascyphus simplex.

Sertularia unguiculata.
Sertularia marginata.
Pycnotheca mirabilis.
Halopteris campanula , var.
Plumularia pulchella.
Plumularia diploptera.
Plumularia spiroclad ia.
Aglaophenia laxa.

zelanica.

H YDROIDA — TOTTON.

137

Station 144. OH Northern New Zealand, September 13, 1911 ; from Cape Maria van Diemen, West
by South, 7 miles (true bearing! ; depth 35-40 fathoms (64-73 m.l ; bottom, rock. Dredge.

Billardia novae-zealandiae.
Acryptolaria minima.
Synthedum carinatum.
Synthecium ramosum.
Symplectoscyph us jolmston i.
Sy mplectoscyph us del icatulus.
Symplectoscyphus spirit ualis.
Sertularia unguiculata .
Crateritheca zelandica.

Antennella serrata.
Pycnotheca mirabilis.
Halopteris constricta.
Halopteris heterogona.
Plumularia diploptera.
Plumularia spirocladia.
Aglaophenia laxa.
Thecocarpus rostratus.
Thecocarpus chiltoni.

Station 194. Off Oates Land, February 22, 1911 ; lat. 69° 43' S., long. 163° 24' E. ; depth 180-200
fathoms (329-366 m.) ; bottom, undecomposed animal debris. Agassiz trawl.

E udendrium antarcticum .
Hebella striata.

Lafoea dumosa.

Lafoea gaussica.

Staurotheca dichotoma.
Staurotheca antarctica.
Symplectoscyphus (?) glaeialis.
Sertularella spiralis.

Station 220. Off Cape Adare, January 3, 1912 ; mouth of Robertson’s Bay ; depth 45-50 fathoms
(82-92 m.) ; bottom, shingle. Agassiz trawl.

Gravelya antarctica.

Halecium ovatum.
Campanularia hicksoni.
Lafoea gradllima.

Reticularia antarctica.
Staurotheca dichotoma.
Symplectoscyphus vanhoeffeni.

Symplectoscyphus glaeialis.
Symplectoscyphus curvatus.
Symplectoscyphus biformis.
Sertularella spiralis.
Oswaldetta antarctica.
Schizotricha unifurcata.

Station 294. Ross Sea, January 15, 1913 ; lat. 74° 25' S., long. 179° 3' E. ; depth 158 fathoms (289 m.) ;
bottom (?). Agassiz trawl.

Halecium antarcticum.

Station 316. McMurdo Sound, February 9, 1911 ; off Glacier Tongue, about 8 miles North of Hut
Point ; depth 190-250 fathoms (348-457 m.) ; bottom, mud and undecomposed animal remains.
Agassiz trawl.

Gravelya antarctica. Symplectoscyphus glaeialis.

Reticularia antarctica. Symplectoscyphus plectilis.

Staurotheca dichotoma. Schizotricha unifurcata.

Staurotheca antarctica.

Station 331. McMurdo Sound, January 14, 1912; off Cape Bird Peninsula; depth 250 fathoms
(457 m.) ; bottom, mud. Dredge.

Ophiodissa arborea.

Symplectoscyphus plectilis.

Station 338. McMurdo Sound, January 23, 1912 ; lat. 77° 13' S., long. 164° 18' E. ; depth 207 fathoms
(379 m.) ; bottom, mud. Agassiz trawl.

Billardia subrufa. Staurotheca antarctica.

Campanulina belgicae. Sertularella spiralis.

Hebella plana.

138

“TERRA NOVA” EXPEDITION.

Station 339. McMurclo Sound, January 24, 1912 ; lat. 77° 5' S., long. 164° 17' E. ; depth 140 fathoms
(256 m.) ; bottom, mud. Agassiz trawl.

E udendrium antarcticum .
Halecium antarcticum.
Billardia subrufa.

Lafoea gaussica.
Rcticularia antarctica.

Slaurotheca antarctica.
Symplectoscyphus vanhoeffen i.
Sympledoscyph us plecl il is.
Sertularella elongata.

0 steal della bifurca.

Station 340. McMurdo Sound, January 25, 1912 ; lat. 76° 56' S., long. 164" 12' E. ; depth 160 fathoms
(293 m.) ; bottom, mud. Agassiz trawl.

Stegella grandis. Symplecloscyphus vanhoejfeni.

Staurotheea antarctica. Osiealdella bifurca.

Station 355. McMurdo Sound, January 20, 1913 ; lat. 77° 46' S., long. 166° 8' E. ; depth 300 fathoms
(549 m.) ; bottom (?). Agassiz trawl.

Ophiodissa arborea. Hebella plana.

Station 356. Entrance to McMurdo Sound, January 22, 1913 ; off Granite Harbour ; depth 50 fathoms
(92 m.) ; bottom, mud. Agassiz trawl.

Hydronema angustum.
Gravelya antarctica.
Eudendrium antarcticum.
Ophiodissa. arborea.
Campanularia hicksoni.
Campanulina belgicae.

Lafoea gracillima.

Staurotheea dichotoma.
Symplectoscyphus glacial is.
Symplectoscyphus vanhoejfem.
Symplectoscyphus pledilis.

IV. DESCRIPTION OF GENERA AND SPECIES.

BOUGAINVILLIIDAE.

Hydronema angustum (Hartlaub).

Hydractinia angusta, Hartlaub, 1904, p. 7, pi. IV, figs. 1-7 ; H. dendritica, Hickson and Gravely,
1907, p. 9, pi. II, figs. 7-10; Hydronema dendriticum, Stechow, 1921, p. 252; 1923 c,
p. 67, fig. F. ; Hydronema angustum, Stechow, 1923 c, p. 67.

Material. — A small quantity without gonophores, encrusting Symplectoscyphus
vanhoejfeni from Station 356, McMurdo Sound, 50 fathoms, Jan. 22, 1913.

Remarks. — The material is badly preserved, and most of the polyps are retracted.
Some small blastostyles with two or three tentacles are present, but there is no trace
of gonophores or dactylozooids. Since Hickson and Gravely ’s “ specimen A " and
“ specimen B ’’ of H. dendritica are so different from one another, though indubitably
belonging to the same species, it seems reasonable to suppose that Hartlaub's H. angusta
falls within the range of variation and is conspecific. I have re-examined the specimens
of H. dendritica. They are very well preserved, though Hickson and Gravely’s
interesting figured specimen which links the two facies together, and which I hereby

HYDROIDA — TOTTON.

139

select as the holotype of H. dendritica, H. and Cf., has become badly discoloured by cork.
The basal encrusting piece (" facies A ”) and free branching piece (:c facies B ’’) have
become separated, though there are still portions of the two facies united.

In several polyps there has been a partial division of the oral region, some having
two mouths, and one at least having three heads. The blastostyles may have as few
as two tentacles, and there does not appear to be a complete discontinuity between small
blastostyles and large gastrozooids.

Parts of " specimen A ” are encrusting a slender branching undescribed Axinellid
sponge. I regard it as an encrusting facies. Stechow (1923) is of the same opinion.

Genus GRAVELYA, gen. nov.

Genotype. — Perigonimus antarcticus, Hickson and Gravely, 1907, p. 4, pi. I,
figs. 1-3, pi. IV, fig. 32.

Description. — Monoecious Bougainvillidae with numerous dimorphic cryptome-
dusoid gonophores scattered over the hydrorhiza.

Remarks. — The genotype was referred by Vanhoffen (1910) to the genus Atractylis,
Wright, 1858, genosyntypes : Eudendrium ramosum, van Bened., 1844 (non E. ramosum,
L.) = Bougainvillia ramosum (v. Bened.) ; Eudendrium repens, Wright, 1858 = Peri-
gonimus repens (Wright) ; and Eudendrium sessile, Wright, 1857 = Perigonimus sessilis
(Wright). I select Eudendrium ramosum, van Beneden, 1844, as the genotype of Atractylis,
Wright, 1858. Atractylis is now sunk in the synonymy of Bougainvillia, Lesson, 1836.
Atractylis arenosa, Alder, 1862, needs a new generic name. Allman (1872) had restricted
Atractylis to this same species which is not a genosyntype. At the same time he need-
lessly rejected Atractylis as preoccupied for a plant by Linnaeus, and replaced it by
Wriglntia. But this name also was preoccupied by Agassiz, 1862. It can be used only
for one of its three genosyntypes, Sertularia syringa, L., Laornedea acuminata, Alder,
and L. lacerta, Wright.

Perigonimus antarcticus, H. and G., 1907, does not appear to be congeneric with
either P. muscoides, Sars, the genotype of Perigonimus, or Atractylis arenosa . Alder. 1
take it as the genotype of a new genus Gravelya, and as the holotype I select a
specimen marked Brit. Mus. Keg. No. 07.8.20.1.

Gravelya antarctica (Hickson and Gravely). (Text-fig. 1, a-b.)

Perigonimus, sp., Hartlaub, 1904, p. 8, pi. I, fig. 2 ; P. antarcticus, Hickson and Gravely, 1907,
p. 4, pi. I, figs. 1-3, pi. IV, fig. 32 ; Atractylis antarctica, Vanhoffen, 1910, p. 283, fig. 8 ;
Atractylis antarcticus, Ritchie, 1913, p. 11.

Material. — Numerous hydranths and male gonophores arising from stocks that
creep over Ophiodissa, Symplectoscyphus and other hydroids from Station 356, entrance
to McMurdo Sound, 50 fathoms ; and from Station 220, off Cape Adare, 45-50
fathoms. A few hydranths and gonophores on S. glacialis from Station 316, McMurdo
Sound, 190-250 fathoms,
v. 5.

o

140

“TERRA NOVA” EXPEDITION.

Description. — Mature male gonophores have a length of 0-61 mm., half of which
represents the pedicel, and a greatest diameter of 0-24 mm. In comparison with that

figured by Hickson (1907) the pedicel
is longer and the passage of pedicel
into gonophore is more abrupt, often
being marked by a distinct shoulder.
The ectoderm bears numerous nema-
tothecae and is thickened to form
an apical pad. The pedicel is sur-
rounded by a loose folded perisarc,
which thins out and becomes in-
visible distally.

In young stages what appears
to be a “ glockenkern 5’ is visible.
In maturer forms a thimble- shaped
spermary can be seen seated on the
endodermal spadix, which is short
and broad with a distinct cavity.
In optical section there appears to
be a continuous endodermal lamella surrounding the spermary and there are indica-
tions of a slit-like subumbrella cavity.

The male gonosome, according to this interpretation of its structure, would there-
fore be a transition between a lowly cryptomedusoid and a styloid form ; but the
material is hardly well enough preserved for the study of its minute structure.

EUDENDRIIDAE.

Eudendrium antarcticum, nom. n.

E. ramosum, Hartlaub, 1904, pi. I, fig. 3 ; E. capillar e, Billard, 1906, p. 4 ; E. insigne, Hickson
and Gravely, 1907, p. 7, pi. I, fig. 4 ; E. ramosum, VanhofEen, 1910, p. 288, fig. 13 ;
E. ramosum, Ritchie, 1913, p. 12 ; E. ramosum, Billard, 1914, p. 2 ; E. capillarc, Billard,
1914, p. 2 ; ? E. rarneum, Jaderholm, 1905, p. 10.

Material. — Three branched fragments up to 1*5 cm. in length, with contracted
hydranths but no gonophores, from Station 356, entrance to McMurdo Sound,
50 fathoms ; a few specimens from Station 339, McMurdo Sound, 140 fathoms ; and a
small branched specimen 3 cm. in height from Station 194, off Oates Land, 180-200
fathoms.

Description. — Stem fascicled by addition of stolons, except in young specimens
not exceeding 2-5 cm. in height. Branching irregular and lax. Perisarc regularly
ringed at bases of branches, and irregularly for short distances in other parts of stem
and branches. An annular constriction in the base of the hydranth, with a narrowed
pedicel joining hydranth to ramulus. Ramuli of hydranths that bear female gono-

Text-fig. 1. — Gravely a antarctica (Hickson and Gravely).
Male gonopliores showing thimble-shaped spermary s,
and endoderm lamella l. X 32.

HYDROIDA— TOTTON.

141

phores short and irregularly ringed. Female gonophores irregularly distributed on
hydranth, which retains its tentacles. Hydranths with from 15 to 30 contractile filiform
tentacles.

Measurements :

Hydranth, without tentacles — - mm.

length . . . . . . . . . . . . . . . . . . 037-056

diameter . . . . . . . . . . • • . . . . 0-17-0-22

Hydrocaulus, diameter . . . . . . . . . . . . . . 0-10-0-12

Remarks. — There appears to be only one delicate little species of Eudendrium so
far recorded from the glacial Antarctic waters. Although it has been assigned to
several species it is very doubtful whether it is conspecific with any of them. From
Vanhoffen’s summary of our knowledge of it the above description has been drawn.
He assigns it to E. ramosum (L.) because of its fascicled creeping stem, the restriction
of perisarcal rings to short stretches, especially at the bases of branches, the loose
branching system and the presence of tentacles on the hydranth that bears the female
gonophores. But E. ramosum (L.) seems to be a much more robust and more regularly
branched species.

It will be convenient therefore to give to the present form a new name and descrip-
tion. I select as holotype of E. antarcticum a specimen bearing the British Museum
register number 07.8.20.3. It was taken by the “ Discovery ” expedition of 1901-04
and described by Hickson and Gravely as E. insigne, Hincks.

An annular groove near the base of the hydranth, first noticed in E. insigne by
Hincks and later referred to by Hickson and Gravely in this Antarctic species is present
also in other species. Further investigation of well preserved specimens will probably
show that this is a generic character.

In giving measurements of the hydranths of Eudendrium it must be remembered
that recently budded hydranths are very much smaller than the terminal hydranths
of older branches.

Eudendrium, sp.

? E. novae-zelancliae, Markt. T. 1890, p. 201, pi. Ill, fig. 21.

Material.- — A young sterile specimen 5-5 cm. in height from St. 134, near North
Cape, N.Z., 11-20 fathoms.

Remarks. — It is very difficult to select characters for the critical diagnosis of species
of this easily recognisable genus.

The form under report has a polysiphonic stem, which is not annulated throughout.

Only one record of a specimen of Eudendrium appears to have been made from
N. Zealand, and that in 1890 by Marktanner-Turneretscher for a new species E. novae-
zelandiae. It was of the same size as the “ Terra Nova ” specimen, which may perhaps
belong to the same species. It is difficult, however, to extract from the original descrip-
tion any diagnostic specific characters.

142

“TERRA NOVA” EXPEDITION.

HALECIIDAE.

Ophiodissa arborea (Allman, 1888). (Text-fig. 2, a.)

Halecium robustum, Allman, 1888, p. 10, pi. IV, figs. 1-3 (not 11. robustum, Verrill, 1873, p. 9 ;
Pieper 1884, p. 166 ; Ritcliie, 1907, p. 524 ; Vanhoffen, 1910, p. 319, fig. 35) ; H. arboreum,
Allman, 1888, p. 98 ; H. arboreum, Hickson and Gravely, 1907, p. 27, pi. IV, figs. 27-29,
not H. arboreum, Jaderholm, 1905, p. 11, pi. V, fig. 4 ; Ophiodes arboreus, Billard, 1910,
p. 4 ; Ritchie, 1913, p. 15, figs. 2-3 ; Billard, 1914, p. 8 ; Ophiodissa arborea, Stechow,
1919 a, p. 42.

Material. — A quantity of broken stems and branches up to 25 cm. in length and
5 mm. in diameter, with several ovoid “ scapus ” masses of gonothecae measuring

4 cm. x 1*5 cm., from Station
331, at the entrance to McMurdo
\ Sound, 250 fathoms ; a quantity

of broken branches up to 9 cm.

'] / / in length from Station 355,

/ / McMurdo Sound, 300 fathoms ;

.o \ f/ 7 / and fragments from Station 356,

\ \ \ / entrance to McMurdo Sound, 50

v \ o h fathoms.

fl\ s>\ \

// \ \ | — -V Remarks. — The species has

\ V\ \ / /\\ been recorded four times by

\ y / / Allman (material redescribed

\ \ / / by Billard), Hickson and

Gravely, Ritchie, and Billard ;
but no record has been published
of a supplementary “ Chal-
lenger ” specimen from Prince
Edward Island, which bears a
characteristic “ scapus ” mass.
The known distribution of this
species is as follows : ofi Ker-
guelen Island, 105 fathoms; off
Prince Edward Island, 310
fathoms; McMurdo Sound, 10-

Text-fig. 2. — a, Ophiodissa arborea (Allman), X 39 ;

b, 0. armata, sp. n„ x 39 . 250 fathoms ; Cape Royds, 20-80

fathoms ; Palmer Archipelago,

off Graham Land, 50 fathoms. Gonangia were found in McMurdo Sound on July 6

and Jan. 14, off Cape Royds on Aug. 20 and off Prince Edward Island on Dec. 27.

Ophiodissa armata, sp. n. (Text-fig. 2, b.)

Material. — Five specimens up to 5 cm. in length, without gonothecae ; from
Station 91, off Three Kings Islands, N.Z., 300 fathoms.

HYDROIDA — TOTTON.

143

Description. — Stems fascicled, branched mostly in one plane, with nematothecae
on accessory tubes. Hydrothecae shallow expanding, margins not everted but
markedly oblique to axis of internodes, primary ones sessile. A nematotheca on
proximal part of each internode on side opposite hydrotheca, and others scattered
irregularly.

Remarks. — Resembles 0. arboreum, Alim, of Kerguelen and Antarctic area in having
nematothecae ; but they are much smaller, and less restricted. Differs in having
relatively shorter adnate hyclrophore wall, with scarcely any free part below the level
of diaphragm, shallower thecae with more oblique margins, and generally more slender
proportions.

Measurements :

Internode,

width at base
length

Hydrotheca,

depth (margin to line of puncta)
mean

mm.

0-18-0-21

0-60-0-76

0-02-0-05

0-03

The dried holotype, Brit. Mus. Reg. No. 29.10.10.15, consists of two stems 5 cm.
long growing from a common root stock.

Halecium ovatum, sp. n. (Text-fig. 3.)

Material. — Numerous specimens up to 2-5 cm. in length,
with both spent and young gonothecae, growing in dense tufts
on the stems of other hydroids and worm tubes ; from
Station 220, off Cape Adare, 45-50 fathoms.

Description. — Stems scorpioid cymes, a pair of branches
similar to stem arising below each hydrotheca, themselves
rebranching in a similar way. Each successive series of
branches and thecae of smaller proportions than the preceding.

Hydrothecae deep, with flared opening, with a circle of
puncta in the middle region. Polyps with about 26 tentacles
in two rows alternately raised and depressed. Gonothecae
replacing all polyps of triple branched hydrophores, mature
ones (? female) ovate, their terminal circular apertures with thickened collars, immature
ones (? male) flattened, kidney shaped.

Measurements :

Text-fig. 3. — Halecium
ovatum, sp. n., X 35.

Hydrocaulus, diameter,
Primary series,

proximal internodes . .
mid-stem ,,
secondary series
tertiary series

mm.

0-150

0-195-0-225

0-175-0-200

0-150

144

“ TERRA NOVA EXPEDITION.

Hydrotkeca,

limbus diameter,
primary series,

terminal theca
other than terminal theca
secondary series
tertiary series
margin to line of puncta

Gonotheca (mature)
length

diameter (maximum)

mm.

0-325

0-300

0-275

0-250

0-100

1-09-1-22

0-78-0-83

Remarks. — The species somewhat resembles H. cymiforme, Allman, in habitus, but
is more robust and of larger proportions. The mature gonothecae are not compressed.
The type specimen of H. cymiforme is imperfect, and the branching system was not
well figured. Its gonothecae are of two types, the smaller male ones measuring
0-75 X 0-5 mm. with aperture 0-05 mm. in diameter, and the females R25 X 0*65 mm.
with aperture 0-1 mm. in diameter.

I select as the holotype of H. ovatum a tuft of stems, Brit. Mus. Reg. No. 29.10.10.1,
bearing many spent gonothecae and a few immature ones, growing on a piece of a stem
of Oswaldella antarctica 4*5 cm. in length.

Halecium antarcticum, Vanhoffen. (Text-fig. 4.)

Halecium antarcticum, Vanhoffen, 1910, p. 317, fig. 34 ; ? II. flexile, Allman, 1888, p. 11, pi. V,
figs 2, 2a.

Material.— Many straggling anastomosed delicate specimens up to 3 cm. in height,
with immature gonothecae, on a worm tube from Station 339, McMurdo Sound, 140
fathoms ; a few small specimens with immature gonothecae on a polyzoan, from Station
294, Ross Sea, 158 fathoms.

Description. — Stem fascicled below, a helicoid cyme with long irregular unpaired
anastomosing branches. Hydrocaulus smooth, a few constrictions at bases of inter-
nodes. Hydrophores well separated from hydrocaulus, with flared and recurved
margin. Gonothecae smooth, compressed, on short stalks from below hydrophores.

Measurements :

Hydrocaulus, diameter,
secondary series . .
Internode, length
Limbus to internode . .
Limbus, diameter,

secondary series . .
Margin to line of puncta

mm.

0- 17-0-20

1- 10-1-34
0-37-0-41

0-22-0-25

0-05-0-07

Remarks. — The habitus of this species with its long anastomosing unpaired branches
distinguishes it from H. ovatum , which has regularly paired branches. The proportions
of the hydrophore are different too, being distinctly less in the present form. None
of the gonothecae are mature, but they are of the same type as those of H. ovatum,

HYDROIDA — TOTTON.

145

H. eymiforme and H. flexile. The measurements correspond so closely with those of
Yanhoffen’s species that I assign the specmiens to it. An important measurement in
any species of Haledum, the distance from the place of attach-
ment (line of puncta) of the polyp to the margin of the
hydrophore, was not given by Vanhoffen, but judging from
his figure the “ Terra Nova ” and “ Gauss ” specimens are of
similar proportions. The gonothecae of the Gauss ” speci-
mens were likewise immature.

I have made a careful study and measurements of
H. flexile, Alim., 1888, which is a closely allied, if not identical,
species. I select as the holotype of H. flexile , Allman, the
specimen, Brit. Mus. Reg. No. 88.11.13.6, from Marion Island.

The syntype from Patagonia is similar to it but the hydranths
are larger and the distal part of the hydrophore is deeper,
measuring 0-075 mm. compared with 0-050 mm. ; and it lacks
gonothecae. We know the mature male gonothecae only of
H. flexile, and neither mature male nor female gonothecae of
H. antarcticum. It seems premature therefore to sink
H. antarcticum in the synonymy of H. flexile.

Hartlaub (1905) considers H. gracile, Bale, 1888, and
H. flexile synonymous ; but by deduction from Bale’s figures
H. gracile is of smaller proportions, and the male gonothecae
are considerably smaller and of different shape. I therefore
reject Hartlaub’s suggestion, basing my opinion on the type

figures of H.

gracile and leaving out of account specimens

assigned to this species by others, which do not agree amongst
themselves in measurement. There is need of more critical
and statistical work on these and similar species of Haledum.

Text-fig. 4. — Haledum
aniarclicum, Vanhoffen,
with immature gono-
thecae, X 24.

CAMPANULARIIDAE.

Tulpa diverticulata, sp. n. (Text-fig. 5, a-c.)

Material. — Two dozen specimens arising from a stock that creeps over a fragmentary
hydroid stem. Gonothecae absent. Station 91, off Three Kings Islands, N.Z., 300
fathoms.

Description. — Pedicels straight, becoming irregular just below thecae. Thecae
very large, slightly curved, narrowing below aperture, walls having from 9-11 longi-
tudinal ridges with concave facets between. Distally the ends of these fluted facets
appear as slight bosses that rise above the level of the entire, everted, slightly oblique
margin. A thickened ring below forms a diaphragm.

146

“TERRA NOVA” EXPEDITION.

Measurements :

Pedicel

„ length

Hydrotheca „

diameter,
below
above
at margin

mm.

0- 2-0-23

1 - 5-9-0
3-5-4-0

1-2

1-0

1-2— 1-4

Text-fig. 5. — Tulpa diverticulata, sp. n. a, X 39 ; b, c, X 24. To the left of a one of the external

diverticula is shown in optical section.

Clytia compressa, sp. n. (Text-fig. 6, a-c.)

? Clytia johnstoni, Hartlaub, 1901, p. 350, not Clytia johnstoni (Alder, 1856), p. 359, pi. XIII,
fig. 8.

Material. — A dozen pedicels and a single gonotheca, on a stock that creeps over an
alga in association with Sertularella robusta and Sertularia marginata ; and numerous
other pedicels and gonothecae growing on a small branched alga attached to a Gasteropod
shell, from Station 134, near North Cape, New Zealand, 11-20 fathoms.

Description. — Hydrorhiza flattened on attached side. Pedicels very variable in
length, with a dozen or more fine annulations at base and usually three or four coarser
ones forming “ spherules ” below thecae : the distal “ spherules ” better marked off
and more flattened. The middle part of the pedicels normally smooth, but often

HYDROIDA — TOTTON.

147

bearing irregular annulations. Hydrothecae campanulate or cylindrical, not compressed,

wall smooth, margin slightly folded and
very slightly everted, with from seven to
thirteen pointed teeth, the ends of teeth
being noticeably less rounded than the
emarginations. Polyps with trumpet-shaped
proboscis.

Ctonothecae not regularly amralated,
generally smooth, often irregularly folded,
compressed ; in broad aspect barrel-shaped,
slightly constricted just below the trun-
cated oval or circular distal end, tapering
below to a short conical pedicel.

Measurements :

Pedicel length
Hvdrotheca,
length

diam. at margin
Gonotheca,
length
breadth
diam. at end

Text-fig. 6. — Clytia compressa, sp. n. a, hydro-
theca, X 39 ; b, c, gonothecae ; X 39.

mm.

1-6

0-46-0-81

0-18-0-56

0-76-1-0

0-38

0-33-0-36

Mean length and diameter in millimetres of hydrothecae with several numbers of
marginal teeth :

Xo. of Teeth.

Length.

Diameter.

7

0-46

0-18

8

0-49

0-25

9

0-61

0-31

10

0-56

0-33

11

0-62

0-36

12

0-68

0-40

13

0-75

0-46

One theca only of the first four categories was available for measurement, and
seven, nine and eight respectively of the others.

Two abnormal eight-toothed thecae of large size 0-66 mm. X 0-49 mm. with much
broader teeth were observed.

Remarks. — The species is clearly very closely related to C. johnstoni (Alder), from
which it is to be distinguished readily by the shape of the marginal hydrothecal teeth
and of the gonothecae. These points may appear to be insignificant but they are of
the kind of specific differences to be expected,
v. 5.

3

148

“TERRA NOVA” EXPEDITION.

Marktanner’s ? Clytia elongata 1890 seems to be distinct. I suspect that Hartlaub
(1901) bad this new species before bim from French Pass, New Zealand.

In living specimens of C. johnstoni I bave counted two wborls of from nine to
thirteen tentacles alternately raised and depressed. It seems likely that the number
of marginal teeth is correlated with the number of tentacles in the lower whorl.

It is probable that the differences between the numbers of teeth of various hydro-
thecae, and the discrepancy in size are to be explained as due to enlargement of
successively produced polyps and not to specific variation.

Campanularia hicksoni, nom. n. (Text-fig. 7, a-e.)

Campanularia laevis; Hickson and Gravely, 1907, p. 25, pi. IV, fig. 26 ; Campanularia laevis,
Ritchie, 1913, p. 19, fig. 5 ; Campanularia laevis, Vanhoffen, 1910, p. 298, fig. 18 (not
C. laevis, Hartlaub, 1905, p. 565, fig. P. 1) ; ? C. volubilis var. antarctica, Ritchie, 1913,
p. 22, fig. 6 ; ? C. antarctica, Stechow, 1922a, p. 96.

Material— Hundreds of stalked thecae up to 35 mm. in height, with a few gono-
thecae, growing on Schizotricha , and a young specimen on the back of a Caprellicl, from

Station 220, off Cape Adare,
45-50 fathoms ; numerous speci-
mens of a smaller race, with
gonothecae, from Station 356, off
entrance to McMurdo Sound,
50 fathoms.

Description. — Normal unre-
generated pedicels of very variable
length, smooth or slightly sinuous,
generally with from one to six
broad annulations at origin from
hydrorhiza, and a single well
marked spherule below the theca .
Hydrothecae of very variable size,
deep, cylindrical ; larger ones
with longitudinal folds passing
down from the emarginations of
the folded and sinuous margin,
which has up to 20 rounded teeth
of the same shape as that of the
depressions between. Smaller
specimens with margin less folded ; teeth much larger than intervening spaces with
flatter tops and from 9 upwards in number. Gonothecae on short twisted pedicels,
very large, elongated, cylindrical, smooth walled, a little narrowed and truncated at
distal end tapering off to pedicel below. There is sometimes a slight annular dilatation
of the wall just below the margin.

Text- fig. 7. — Campanularia hicksoni, sp. n. a, young hydro-
theca, with truncated teeth, X 24 ; b, margin of old hydro-
theca, X 24 ; c, d, gonothecae, X 10 ; e, hydrotheca and
pedicel, X 10.

HYDROIDA— 1 TOTTON.

149

Measurements in millimetres :

Pedicel,

Station 220.

Station 356.

length

0-68-35-0

0-65-5-0

diameter

0-16-0-36

0-06-0-10

Hydrotheca,

length

1-07-3-53

0-68-1-22

diameter (over all at margin)

0-48-1-4

0-30-0-61

Gonotheca,

length

-6-6

2-68-3-30

breadth

-1-2

diameter of aperture

0-65-0-85

0-29-0-31

Number of Hydrothecal teeth . .

9-20

11-18

Mean lengths and diameters in millimetres of hydrothecae, with several numbers of
marginal teeth.

Specimens from Station 220.

Pedicel.

Theca.

Length.

Diameter.

No. of Teeth.

Length.

Diameter.

0-85

0-17

9

1-07

0-48

1-41

0-17

10

1-51

0-61

1-22

0-17

11

1-7

0-66

1-75

0-18

12

1-7

0-63

7-0

0-22

15

3-0

1-12

S-81

—

17

3-53

1-36-1-4

—

0-24

18

2-8

1-22

Specimens from Station 356.

1-08

0-07

11

0-73

0-33

5-0

0-10

16

1-10

0-55

—

18

1-22

0-61

Remarks. — Hartlaub’s species C. laevis with which Hickson identified the present,
appears to be a smaller one with differently shaped teeth and gonothecae. It appears,
too, to have no annulations at the bases of the pedicels. Hartlaub did not give the
minute detail of the marginal teeth, but he figures adult hydrothecae with truncated
teeth ; and no folding of the wall is shown. It seems best therefore to give a new name
to this species Avhich was first dealt with by Hickson. In dealing with Campanularians
of this type we have to remember that the pedicels and thecae are of various sizes and
that the number of teeth increases generally with size. The smallest forms with nine
square teeth and short pedicels are solitary, have large discs of attachment and have
evidently developed from planules. The largest forms arise from previously developed
hydrorhizal tubes. Some pedicels show signs of regeneration. It seems probable
that the thecae fall from time to time with the polyp, and that new and larger ones
are developed after increase of length of pedicels.

150

TERRA NOVA” EXPEDITION.

The gonothecae were first figured at all well by Ritchie.

The species is known only from McMurdo Sound and doubtfully from off Kaiser
Wilhelm II Land.

I have included under this name two races, a large one from Station 220 and a smaller
one from Station 356. They both bear female gonothecae, and were taken in January
of 1912 and 1913 respectively. Apart from differences in size, and judged stage for stage
according to the number of teeth, there is little reason for regarding them as distinct
species. The forms Ritchie (1913) described as C. volubilis var. antarctica I am inclined
to regard as the male colonies of this small race. Referring to his measurements it
will be seen that the trophosome in the 10-13 tooth stage corresponds in size as well as
shape with those of a similar stage from Station 356. The gonangia are a little smaller
and have narrower openings, features that might be expected of males. It remains for
a future expedition to say whether or not the two forms can be linked up, and if they
can whether they are to be regarded as a separate species.

Genus BILLARDIA,* gen. nov.

Genotype. — Billardia novae-zealandiae, sp. nov.

Simple or branched Campanulariidae. Older stems fascicled. Hydrothecae from
one to three on an internode, alternating, rising from short apophyses on sympodial
stem and branches. Branches alternate, one from below proximal theca of each stem
internode. No sarcothecae. Hydrothecae long, abruptly curved at base, inoperculate,
with basal septum, adnate for a very short distance or entirely free. Margins entire
and circular.

Gonothecae replacing hydrothecae smooth and compressed, or annulated.

Species included : Billardia novae-zealandiae , sp. n. ; Campanularia subrufa,
Jaderholm, 1905.

Billardia novae-zealandiae, sp. n. (Text-fig. 8, a.)

Material. — Two branched stems 2-5 cm. and 3 cm. in length rising from a few
rooting tubes without gonothecae, from Station 90, Three Kings Islands, New Zealand,
100 fathoms ; and two dozen simple shoots up to 1-7 cm. in length with gonothecae, on
polyzoa from Station 144, off Cape Maria van Diemen, New Zealand, 35-40 fathoms.

Description. — Hydrothecae tubular with abrupt bend near base, without perisarcal
thickening at this point, slightly curved up at distal end ; aperture at slight angle to
thecal axis.

Gonothecae f in place of hydrothecae, compressed, twice as long as hydrothecae,

* Dedicated to Professor Armand Billard of Poitiers.

f It is possible that these are only parasitised and hypertrophied hydrothecac.

HYDROIDA — TOTTON.

151

appressed to hydrocaulus, with slit-like irregular distal aperture and a basal flexure
similar to that of hydrothecae.

Measurements :

Stem,

diameter
internode length
Branch,

diameter
internode length
Hydrotheca,
length
diameter

mm.

0- 39

1- 83

0-22

1*51

0-98

024:

Two branched stems, Brit. Mus. Reg. No. 29.10.10.12, form the holotype of the
genotype.

Remarks. — The resemblances between the trophosomes of B. novae-zealandiae and
Campanularia subrufa, Jader-
holm, are so striking that in
spite of the differences between
gonothecae it seems certain that
they are closely related. The
abrupt, basal bend of the
hydrothecae is characteristic.

In the second species this angle
is filled in ’with a perisarcal
thickening and disguised.

Specimens from shallow
water are unbranched. The
affinities of the genus are not
at present clear. Perhaps it is
related to Hincksella, Billard,

1925, as the genotype of which genus I select the species H. sibogae, Billard.

I have named this genus after Professor Armand Billard of Poitiers, whose critical
work on the Hydroida during the past quarter of a century has greatly extended our
knowledge of this group.

Text-fig. 8. — a, Billardia novae-zealandia, sp. n. Hydrotheca,
X 34 b, B. subrufa ( Jaderholin), X 34.

Billardia subrufa (Jaderholin). (Text-fig. 8, b.)

Campanularia subrufa, Jaderholm, 1904, p. v ; 1905, p. 15, pi. VI, figs. 4-6.

Material. — A few branching stems up to 8 cm. in length with gonothecae, on a
worm tube from Station 339, McMurdo Sound, 140 fathoms. A few small non-fascicled
shoots without gonothecae from Station 338, McMurdo Sound, 207 fathoms ; and a
number of fine branched specimens growing on other organisms, with both polyps and
gonothecae, from Station 38, S.W. of the Falkland Islands, 175 fathoms.

152

“TERRA NOVA” EXPEDITION.

Description. — The abcauline wall of the hydrotheca shorter than the adcauline
with a short internal transverse chitinous ridge of varying thickness at its base. Theca
and apophysis joined by thin- walled narrow neck. In contraction hyclrotheca generally
telescoped onto apophysis, which has a subterminal thickened ring. Polyp has thirty
solid tentacles with thickened bases and a rounded dome shaped proboscis. There do
not appear to be any processes of attachment to the theca. Both branches and gono-
thecae replace hydrothecae.

Gonothecae spindle shaped, with from 12 to 20 circular thickened rings without
free frills. There is much variation in size and shape. The majority have a smooth
conical apex without visible opening. The largest gonothecae have 20 annulations and
are without the conical apex, having instead a dished apical segment 0*4 mm. in diameter.
This is sometimes 3 or 4 lobed, instead of annular. There is not sufficient evidence to
show whether the gonothecae are sexually dimorphic.

Measurements :

Hyclrotheca, mm.

length . . . . . . . . . . . . . . . . . . 086-093

diameter at margin . . . . . . . . . . . . . . 039-046

distance apart . . . . . . . . . . . . . . . . 1-59

Internode,

diameter . . . . . . . . . . . . . . . . . . 0-17-0-30

Gonotheca,

length . . . . . . . . . . . . . . . . . . 1-15-3-70

diameter . . . . . . . . . . . . . . . . . . 0-80-1-30

This species was taken once before, though not recorded, from McMurdo Sound at
the east end of the Great Ice Barrier by “ Discovery ” in 100 fathoms. The only
other records are off Joinville, Seymour and Snow Hill Islands in the Graham Land
Region ; and from Shag Rocks, W. of South Georgia (Swedish Southpolar Expedition).

CAMPANULINIDAE .

Campanulina belgicae, Hartlaub. (Text-fig. 9.)

Camjpanulina belgicae, Hartlaub, 1904, p. 10, pi. I, figs. 8-9.

Material. — Numerous specimens without gonothecae growing on Halecium, sp.
from Station 338, McMurdo Sound, 207 fathoms ; a few dozen specimens, also without
gonothecae from Station 356, entrance to McMurdo Sound, 50 fathoms.

Remarks. — Although in lateral view the operculum sometimes appears to resemble
that of Stegopoma spp. as illustrated by Vanhoffen (1910) and remarked on by Ritchie
(1913), this is an illusion, for the segments are large and about eight in number. The
pedicel is generally ringed, and often shows signs of regeneration.

HYDROIDA — TOTTON.

153

Measurements :

Pedicel,

length.

diameter

Hydrotheca,

length over all . .
greatest diameter

mm.

0-15-0-54

0-04-0-05

0-32

0-12

This record shows that the species is not confined to
deep water as suggested by Vanhoffen.

Text-fig. 9. — Campanulina

Stegella grandis (Hickson and Gravely). (Text-fig. 10, a-c .) belgicae, Hartlaub, x 39.

Campanularia verticillata, var. grandis, Hickson and Gravely, 19G7, p. 23, pi. IV, fig. 25 ; C. lobctta,
Vanhoften, 1910, p. 294, fig. 15 ; C. lobatci, Ritchie, 1913, p. 21 ; Stegella grandis, Stechow,
1919, p. 852 ; 1920, p. 17 ; 1923, p. 119, fig. P ; Stegella grandis, Jaderholm, 1926, p. 3.

Material. — A single badly preserved fragment 3-4 cm. in length without gonothecae,
from Station 340, Ross Sea, 160 fathoms.

Remarks. — This interesting form has been taken by four other Expeditions, viz.,
by “ Discovery ” in McMurdo Sound in 20 fathoms, Feb. 20, 1902, and Jan., 1903
(with gonothecae) * ; by “ Gauss ”
off Kaiser Wilhelm II Land in
385 metres, June 14, 1902,

Dec. 17, 1902 (with gonothecae),
and Feb. 8, 1903; by “Nim-
rod ” in McMurdo Sound in 7-8
fathoms, June, 1908, and 20-30
fathoms, July, 1908 ; and by
C. A. Larsen’s Ross Sea Exped.

1923-4 in Discovery Inlet, Ross
Sea, 550 metres, March 5, 1924.

On most occasions the specimens
have been fragmentary.

A re-examination of the
types shows that the gonothecae are of the form illustrated by Vanhoffen (1910). They
have a terminal slit-like opening between the two parallel flattened and rounded
terminal plates ; the hydrothecae have four-lobed margins beyond which the polyps
can be greatly extended. The rather small mouth is on a dome-shaped prominence.
Measurements have been given by Ritchie (1913). Stechow has pointed out that this
species is a Campanulinid and not a Campanulariid.

Text-fig. 10. — Stegella grandis (Hickson & Gravely), a, hydro-
theca ; b, aperture of gonotheca in lateral view ; c, the
same in frontal view. All figures magnified, X 39.

* The specimens were not kept separate.

154

“TERRA NOVA” EXPEDITION.

Genus STEGOLARIA, Stechow.

Stegolaria, Stechow, 1913, p. 137 ; ? Stegopoma, Levinsen, 1893, p. 35.

Genotype. — Cryptolaria geniculata, Allman, 1888, p. 41, pi. XX, fig. 1.

Undoubted species of Stegolaria have been taken from Fiji (H.M.S. “ Challenger ”
and Nutting) ;• from eight localities in the Malay area between 7° 53' N., 92° 59' E.
(S.W. of Nicobars) and 11° O' S., 122° E. (Timor Sea) ; and from off Dar-es-Salaam
(Eastern and Associated Telegraph Companies’ Cable Ships) in deep water from 100 to
1500 fathoms. Material is scarce and the species cannot be fully worked out yet.

The New Zealand forms harbour many polychaetes in large, partly discontinuous
runs in the stems and branches. This should be borne in mind when examining the
so-called gonothecae which are alleged to be large and sack-like.

No undoubted gonothecae have been seen in the New Zealand forms.

There seems little reason for refraining from sinking Stegolaria, Stechow, 1913, in
the synonymy of Stegopoma, Levinsen, 1893. The distinction that Stechow drew was
that in Stegopoma the gonotheca was free and pedicellate. In order to define our
subject I select Stegopoma plicatile (M. Sars) as the genotype of Stegopoma. In this
genus the gonotheca may sometimes be adnate, as figured by Levinsen (1893) and
described by Kramp (1913), so that the two genera cannot be distinguished by this
character. But although in Stegopoma the orifice of the gonotheca has been described
by Bonnevie as circular, and by Kramp and also Broch (1918) as covered by the usual
operculum, the only gonothecae that I have observed coming from the Indo-Pacifie
area have had circular orifices, and the gonothecae have been incorporated into the
stem and branches by overgrowth of accessory tubules. This statement applies only
to branched forms with fascicled stems. A simple New Zealand species of Stegopoma
described on p. 155 has a free pedicellate gonotheca with pent-roof operculum.

Stegolaria irregularis, sp. n.

Material. — Two stems irregularly branched in one plane, 9 cm. and 4-5 cm. long,
and fragments; from Station 96, 7 miles East of North Cape, New Zealand, in 70 fathoms.

Description. — Branching mostly in one plane, irregular. Stem and branches
fascicled, enclosing polychaete worm “ runs.” Hydrocaulus fine, ascending last
formed hydrotheca ; distal hydrothecae alternate, proximal ones on all sides of stem
and branches, borne by subsidiary tubules. Hydrotliecae cylindrical at base, flattened
“ dorso-ventrally ” at mouth to form two lateral teeth with folded opercular flap in em-
bayment both above and below, slightly curved at first, becoming longer and more curved
with reduplications, about half the size of those of S. geniculata, Allman. Hydranth
with from ten to twelve simple tentacles, with annular attachment to theca near base.

Measurements : mm.

Hydrothecae, distance apart . . . . . . . . . . . . 1*13

Length, (distal end to ring of attachment) . . . . . . . . . . 0-75

Diameter . . . . . . . . . . . . . . . . . . 0*19

Width at mouth . . . . . . . . . . . . . . . . 0-22

Stem diameter . . . . . . . . . . . . . . . . . . 3-0

HYDROIDA — TOTTON.

155

I select the larger specimen, Brit. Mus. Reg. No. 29.10.10.16, as holotype of the
species.

Stegopoma fastigiaium (Alder, 1860). (Text-fig. 11.)

Campanularia fastigiata, Alder, 1860, p. 73, pi. Y, fig. 1 ; Stego-
poma fastigiaium , Stechow, 1911, p. 135, fig. 9.

Material. — Two dozen specimens growing on Plumularia
tenuissima and Halecium armatum from Station 91, off Three
Kings Islands, N.Z., 300 fathoms.

Remarks. — There is much variation in the size of the thecae
and length of pedicels, as was pointed out by Stechow (1914)
in material from Trondhjem Fjord. This common North
Atlantic species has been reported also in Hawaii and Japan,
but is new to New Zealand. The free pedicellate gonotheca has
an operculum of the pent-roof type.

LAFOEIDAE.

Genus HEBELLA, Allman.

Hebella, Allman, 1888, p. 29 ; Syn. Hebellopsis, Hadzi, 1913.
Genotype. — H. striata, Allman, 1888, p. 30, pi. XV, figs. 3, 3 a.

Text- fig. 1 1 . — Stegopoma
fastigiaium (Alder).
Hydrotheca, x 39.

Hadzi (1913) described in Kroatian a new genus Hebellopsis which differs from
his conception of Hebella in having short-pedicelled hydrothecae, with no annular
thecal thickening but a stout diaphragm. He says that in Hebella there is a ring and
a thin diaphragm. I re-examined whole mounts of the holotypes of the genotype
H. striata, Alim., but saw no trace of either ring or diaphragm. The thecal walls are
of almost uniform thickness in the basal region. It seems, then, permissible to retain
Hebellopsis in the genus Hebella, Alim.

The gonosomes of five species of Hebella have been partially described. In
H. calcarata both A. Agassiz (1865) and Nutting (1901) have described the medusa, but
neither of them described the margin of the gonotheca. The medusa appears to have two
long tentacles and six tentacular buds, all with basal pigment spots, but no cirrhi. Bale
(1888) described the gonotheca of H. scandens as having 3-4 emarginations of the rim,
but did not mention the operculum. There were two medusae. Pictet (1893) figured
three medusa-buds in the gonotheca of II. cylindrica, v. Lend, but did not describe
emarginations or operculum. Ritchie (1909) described the immature medusae of
H crater oides as having 4 stout tentacles, but there is no mention of emarginations of
rim of gonotheca or operculum.

Levinsen (1913) described the 4-6 opercular valves of the gonotheca of H. contorta,
Markt., but did not mention the medusa-buds ; and finally Stechow and Muller (1923)
figured the 4-valved operculum and large medusa of the gonothecae of ? Hebellopsis
v. 5. 4

15G

“TERRA NOVA” EXPEDITION.

contorta (Markt.). The medusa appears to have four small tentacles and four rows of
reproductive organs on the radial canals.

Hebelia striata, Allman.

Hebella striata, Allman, 1888, p. 30, pi. XV, figs. 3-3a.

Material. — Many specimens without gonothecae, adhering to specimens of Lafoea
fruticosa and Symplectoscyphus filiformis , Alim., from Station 38, SAV. of Falkland Is.,
125 fathoms. A few specimens on fragments of Symplectoscyphus (?) glacialis, from
Station 194, off Oates Land, 180-200 fathoms.

Description. — The polyps have a conical or rounded proboscis and about 16
tentacles. No chitinous ring has been observed at the base of the hydrotheca. Occa-
sionally the striation of the thecal wall is very indistinctly marked.

Measurements :

(10 specimens)

Pedicel, mm,

length 0-23 (0-07-042)

Hydrotheca,

length 0-79 (0-61-0-97)

diameter at margin . . . . . . . . . . . . 0-23 (0-17-0-25)

Remarks. — Allman’s figures give a reliable idea of the form. The gonothecae of
this species have not yet been found. The specimens have been compared and agree
with the type material.

Hebella plana, Ritchie. (Text-fig. 12, a-b .)

Hebella striata (Alim.) var. plana, Ritchie, 1907, p. 530, pi. I, fig. 8.

I raise Ritchie’s variety to specific rank. The hydrothecae are not striated ; and

are longer and have longer pedicels than those of
H. striata.

Material. — Innumerable hydrothecae and several
gonothecae growing on Halecium arboreum from
Station 355, McMurdo Sound, 300 fathoms; some
specimens on Staurotheca antarctica, from Station
338, McMurdo Sound, 207 fathoms.

Description. — Gonothecae borne singly on hydro -
rhiza, much larger than hydrothecae, cylindrical,
tapering gradually below, with a square, slightly
embayed margin provided with a three- to five-
flapped operculum. The blastostyle bears two,
possibly three, medusae which escape from the gono-
theca. While still attached the medusa appears to
have 4 larger inrolled tentacles and ? 4-12 tentacular
buds and short manubrium. No pigment spots observed.

Text-pig. 12. — Hebella plana, Ritchie,
sp. n. a, hydrotheca, X 40 ; b,
gonotheca containing a single me-
dusa bud, X 24.

HYDROIDA — TOTTON.

157

Measurements :

Pedicel,

length

Hydrotheca,

length

diameter at margin
Gonotheca,
length

diameter at margin

mta.

0-20-0-74

1-26-1-36

0-27-0-33

1-98-2-08

0-66-0-74

Remarks. — This is the first Antarctic hydroid in which production of medusae has
been observed. It has been recorded before by Jaderholm (1926) from Ross Sea,
550 metres, without gonothecae on March 5, 1924 ; by Billard (1914) from off Graham
Land, 176 metres, without gonothecae on Jan. 20, 1909 ; by Ritchie (1907) from the
Weddell Sea, 161 fathoms, without gonothecae on March 12, 1904 ; and by Vanhoffen
(1910) from off Kaiser Wilhelm II Land, 350-400 metres. The present specimens
with gonothecae were taken on Jan. 20, 1913. The number of gonothecae is small,
but the medusa buds will be available for examination by workers on Antarctic medusae.

Lafoea f ruticosa, Sars. (Text-fig. 13.)

Lafoea fruticosa, Sars, 1831, p. 138 ; L. fruticosa, Allman, Report Sci. Res. H.M.S.“ Challenger ”
. . . Zoology XXIII, 1888, p. 34, pi. XVI, fig. 2.

Material— A. few small specimens, branched and unbranched, without gonosome,
from Station 38, W. of Falkland Is., 125 fathoms.

Description. — Hydrothecae slender, tubular, doubly curved on both sides, with
aperture often slightly oblique. Pedicel with two twists.

Measurements :

Hydrotheca,

length

diameter at margin
Pedicel,
length

mm.

0-50 (0-44-0-52)
0-12

0-22 (0-20-0-25)

Text-fig. 13. — Lafoea

fruticosa, Sars. Hydro-
theca, x 39.

Remarks. — The form described and figured by Allman (1888)
from Magellan Strait, 53° 37' S., 70° 56' W., 9-15 fathoms, is
identical in every way with this one. Allman’s figure 2 a, plate
XVI, is magnified about 35 times and not 15 as stated. The
hydrotheca generally has a distal upward bend with an oblique
aperture, as illustrated here for the “ Terra Nova ” specimens ; and the pedicel twists
are not so open as Allman’s figures suggest.

Large coppinia masses, unnoticed by Allman, are present on the “ Challenger
specimen. Many of the closely packed pentagonal or hexagonal gonothecae bear
acrocysts.

158

TERRA NOVA” EXPEDITION.

Measurements :

Gonothecae, mm.

length . . . . . . . . . . . . . . . . . . 0-49-0-61

diameter . . . . . . . . . . . . . . . . . . OlO-OTl

mouth tube,

length . . . . . . . . . . . . . . . . 0-05-0-08

diameter . . . . . . . . . . . . . . . . 0-08

Accessory tubes,

length (projecting) . . . . . . . . . . . . . . 1-23-1-35

diameter . . . . . . . . . . . . . . . . . . 0-08-0-10

Lafoea dumosa (Fleming). (Text-fig. 14.)

Sertularia dumosa, Fleming, 1820, p. 83.

Material. — A fine irregularly branched specimen measuring 7 cm. x 6 cm., with
coppinia masses on some of the older branches, from Station 194, off Oates Land,
180-200 fathoms.

Text-fig. 14. — Lafoea du-
mosa (Fleming) . Hydro-
thecae, X 39.

Measurements :

Hydrotheca,

length

diameter at margin
Pedicel,

length . .

mm.

0-45-0-47

0-16-0-17

0-19-0-22

Remarks. — The shortened, little twisted pedicels lead
me to assign the specimens to L. dumosa. The classification
and nomenclature of species of Lafoea, which appear to be
very variable, is in an unsatisfactory condition.

The specimen may possibly be conspecific with those
called L. gracillima, Alder, by Vanhoffen (1910) and Billard
(1914).

Lafoea gracillima (Alder). (Text-fig. 15, a-b.)

Campanularia gracillima, Alder, 1856, p. 361, pi. XIV, figs. 5-6.

Material. — A branched fragment 15 mm. in length over all without gonosome,
from Station 220, off Cape Adare, Ross Sea, 45-50 fathoms ; and a small specimen,
creeping over a Staurotheca, with two short simple erect stems, but no gonosome, from
Station 356, Granite Harbour, McMurdo Sound, 50 fathoms.

HYDROIDA — TOTTON.

159

Measurements in millimetres :

Hydrotheca,

length

diameter at margin
Pedicel,
length

Station 220. Station 356.

0-59 0-61-0-72

0-17-0-19 0-15-0-20

0-29-0-39 0-15-0-22

Remarks. — I have examined a specimen of L. gracillima that Alder sent to the
Museum in 1856, just after establishing the species. In ignorance of the designation
of any other, I select this specimen, Brit. Mus. Reg. No. 57.8.3.51, as the liolotype of the
species. This holotype is indistinguishable from British specimens that have beeu
assigned, perhaps erroneously, to L. fruticosa, M. Sars. Broch (1918) states that
G. 0. Sars’s (1873) figs. 17-18, Tab. IV, are undoubtedly based on M. Sars’stype speci-

Text-fig. 15 .—Lafoea gracillima, Alder, a, Part of Holotype (Brit. Mus. Reg. No. 57.8.3.51) ; b, Part of
specimen from Station 220. Both figures magnified, X 40.

mens of L. fruticosa, Sars. If this is so they differ a good deal from the holotype of
L. gracillima, Alder, which closely resembles G. 0. Sars’s fig. 20. For this reason I
hesitate to sink L. gracillima, Alder 1856, with established holotype, in the synonymy
of L. fruticosa, M. Sars, of which no holotype appears to have been selected, and of
which our ideas are not very clear.

The hydrothecae of the holotype are a little smaller than those of the “ Terra
Nova ” specimens.

Lafoea gaussica, Vanhoffen. (Text-fig. 16, a-b.)

Lafoea gaussica, Vanhoffen, 1910, p. 312, fig. 32.

Material. — An irregularly branched specimen measuring 3 cm. x3 cm., without
gonosome, from Station 339, McMurdo Sound, 140 fathoms ; and fragments from
Station 194, off Oates Land, 180-200 fathoms.

160

“TERRA NOVA” EXPEDITION.

Description. — The pedicel may be adnate to the hydrocaulus for more than half
its length. The margin of the hydrotheca is slightly everted.

Measurements :

Hydrotheca,

length

diameter at margin
Pedicel,
length

mm.

0-83-0-90

0-26-0-31

0-38-0-93

Remarks. — This species is readily distinguishable from other Antarctic species of
Lafoea by its large, almost symmetrical hydrothecae, and adnate pedicels.

Text-fig. 16. — Lafoea gaussica, Vanhoffcn, x 24,
to show adnate pedicels.

Reticularia antarctica (Hartlaub). (Text-fig. 17.)

Text-fig. 17. — Reticularia antarctica,
(Hartlaub). X 41.

Lafoea antarctica, Hartlaub, 1904, p. 11, pi. II, fig. 2 ; L. antarctica, Vanhoffen, 1910, p. 311,
fig. 31 ; ? F Helium antarcticum, Stechow, 1913, p. Ill ; 1925, p. 214.

Material. — Innumerable hydrothecae with some coppinia masses attached to other
hydroids, from Station 220, off Cape Adare, Koss Sea, 45-50 fathoms ; a few hydro-
thecae from Station 339, McMurdo Sound, 140 fathoms; from Station 316, McMurdo
Sound, 190-250 fathoms ; and from Station 38, West of Falkland Islands, 125 fathoms.

Description. — The hydrothecae, wdiose margins are 0-15 mm. in diameter, project,
excluding reduplications, 0-30-0-40 mm. from the basis to which they are attached,
the length of the adnate part being about the same. Many hydrothecae enclose pooriy
preserved hydroid polyps. The gonosome consists of small coppinia aggregates, in
which the fused accessory tubes surround and arch over the gonothecae and their
acrocysts, so as to form a marsupium or corbula.

Remarks. — There is little, if anything, to distinguish the hydrosome from that of
Campanularia serpens, Hassal, 1848. The systematic position of these species has not

HYDROIDA — TOTTON.

161

been stabilised. Wyville Thomson (1853) established the genus Reticularia for Hassal’s
species which is the genotype. Hincks (1868) needlessly substituted a new name
F Helium because Reticularia had been used for a fungus. More recently authors have
used for HassaTs species the generic name Grarnmaria, Stimpson, 1854, as genotype of
which I select G. robusta, Stimpson. There are a few hydrothecae of a species of
Reticularia on the base of specimens of Halecium armatum from Station 91, off Three
Kings Islands, N. Zealand, 300 fathoms.

Genus ACRYPTOLARIA, Norman.

Acryptolaria, Norman, 1875, p. 172 ; Scapus, Norman, 1875, p. 173 ; Cryptolaria, auct., nec Busk,
1857, p. 173 ; Oswaldaria , Stechow, 1923 (c),p. 147 ; not Perisiphonia, Allman, 1888, p. 43.

Genotype. — Acryptolaria andersoni nom. nov. (=A. exserta, Norman, 1875, p. 172,
pi. XII, figs. 1-2, not Cryptolaria exserta, Busk, 1858, p. 130, pi. XIX, fig. 3).

Nomenclature. — Norman (1875), when he instituted Acryptolaria, showed that
Cryptolaria exserta, Busk, 1858, was not congeneric with C. prima, Busk, 1857, the
genotype of Cryptolaria. At the same time he was 'wrongly assigning specimens to
C. exserta, Busk. Stechow (1923 c.) pointed out that C. prima was not congeneric with
any of the later authors’ Cryptolaria spp., three of which he then took as the genotypes
of new genera Stegolaria, Cryptolarella and Paratliecium. For the remainder he instituted
Osiedldaria, on the assumption that Norman’s Acryptolaria was congeneric and
synonymous with Perisiphonia, Allman, 1888, an assumption which proves to be
unfounded.

The holotype of C. prima is in the British Museum. The species has been re-
described recently by Trebilcock (1928) under a new name Perisiphonia quadriseriata.
It is certainly not congeneric with the specimens wrongly assigned by Norman to
C. exserta, Busk, the type of which species unfortunately is unknown. I take as the
holotype of the species which Norman described, but wrongly named C. exserta, Busk,
a specimen, Brit. Mus. Reg. No. 99.5.1.218. To this species, the genotype of the mono-
typic genus Acryptolaria, Norman, I give the new name andersoni in place of the one
wrongly applied to it by Norman in 1875. It is possible that Acryptolaria andersoni
(nom. nov. =M. exserta, Allman, nec Busk) is conspecific with A. humilis, Allman,
1888, but since the short description of A. humilis can hardly be regarded as a specific
diagnosis and the figured holotype specimen of that species is such a poor mutilated
one, and finally since many of Allman’s figures of hydroida have proved to be in-
accurate, this question of conspecificit}' cannot be settled with any degree of satis-
faction.

The specimen of Acryptolaria andersoni came to the Museum with the Hincks
collection, and is labelled as having been attached to one of the Atlantic cables 150
[miles] off Lands End, in 200 fathoms. It bears Sir James Anderson’s name and a
note initialled by Hincks “ a form allied to Salacia, Cryptolaria exserta. ’ ’ There is
little doubt that it is one of the specimens described by Norman as coming from the

162

“TERRA NOVA” EXPEDITION.

Falmouth and Lisbon Cable between N. Lat. 47° 58' and 47° 35', and in W. Long. 7° G'
at a depth between 89 and 205 fathoms. Norman pointed out that the specimens
were much less regularly branched than in Perisiphonia exserta (Busk), a characteristic
difference between the genera Acryptolaria and Perisiphonia. Norman’s figures on
plate 12 of his publication leave no doubt at all as to the generic association of this
species. He made a mistake in identifying the specimen as C. exserta, Busk. As in
other cases of this sort, we recognise as the genotype of the newly instituted genus the
renamed species to which the described but misidentified specimen belonged rather
than the species to which it was first of all wrongly assigned.

Stechow’s name Osivaldaria, 1923, must be sunk in synonymy, since Acryptolaria
andersoni, nom. nov. and Oswaldaria crassicaulis, Allman, the two genotypes, appear to
be congeneric, Acryptolaria having priority.

Amended Generic diagnosis. — Lafoeidae without a thecal diaphragm (Sub-fam.
Lafoeinae).

Multipinnate. Branches of hydrocaulus alternate from every fourth theca on
each side, in one plane, with irregular branches in other planes from accessory tubules.
Hydrocaulus of stem and branches with a pair of main accessory tubes, one dorsal and
the other ventral. On branches accessory tubes arise from level of proximal theca,
and later give off from point of origin descending tubes, ventral ones descending stem,
dorsal turning up stem. Branches of hydrocauli originate at level of mouths of thecae ;
adnate to level of mouth of next distal theca they then diverge from stem. A foramen
and coenenchymal connection formed at point of divergence. Mature specimens have
stems composed of a great number of accessory tubes. These give rise to gonosome
(Scapus, Norman, 1875) in form of closely packed flask-shaped gonothecae with short
tubular mouths, surrounding stem for distances of two or three centimetres. No
nematothecae, but smaller simple hydrothecae produced by accessory tubes.

Remarks. — We may perhaps trace the origin of the small nematothecae borne by
the accessory tubes on genera of the Sub-family Zygophylacinae, Stechow, 1921, in the
smaller simple hydrothecae found on the accessory tubes of the more primitive genus
Acryptolaria.

One New Zealand species of the genus A. gracilis (Allman, 1888) has been described
from 700 fathoms off East Cape. The “ Terra Nova ” species is of much smaller
proportions and is regarded as new.

Acryptolaria minima, sp. n. (Text-fig. 18, a-b.)

Material. — Two larger fertile specimens 7 cm. and 9 cm. in height, and two smaller
specimens, from St. 134, near North Cape, N.Z., 11-20 fathoms ; one fertile specimen
8*5 cm. in height from St. 144, off Cape Maria van Diemen, New Zealand, 35-40
fathoms.

Description. — Hydrothecae overlapping, short, little curved ; margin not everted
the plane in which it lies turned out at 45° to the horizontal. There is a distinct con-

HYDROIDA — TOTTON.

163

vexitv in the abaxial wall of the pedicel opposite the
“ arris,” * and a row of bright puncta marks the
division between hydrotheca proper and the pedicel,
about 0-05 mm. above the lateral part of the
“ arris.”

Measurements :

Hydrothecae,

distance apart
length,
total

free part . .
diameter at mouth

Gonothecae,

length

diameter

neck, length . .
mouth, diameter

mm.

. . 0-68 (061-0-73)

.. 0 52
. . 0-07-0-11
. . 0-13-0-15

. . 0-61-0-73

. . 0-12
. . 0-12
. . 0-05

Type. — The holotype is the largest fertile specimen,
Brit. Mus. Reg. No. 29.10.10.3, from Station 134.

Text-fig. 18. — Acryptolaria mini-
ma, sp. n. Part of Holotype.
a, X 41 ; b, X 25.

Acryptolaria conferta, Allman, var. australis, Ritchie. (Text-fig. 19, c-e .)

1 Cryptolaria conferta, Allman, 1877, p. 17, pi. XII, figs. 6-10 ; ? Cryptolaria conferta, Clarke,
1879 (in part), p. 244, pi. Ill, fig. 20; Cryptolaria conferta var. australis, Ritchie, 1911,
p. 826, pi. LXXXIV, fig. 2, pi. LXXXVII, fig. 1.

Material. — A small specimen without gonothecae from Station 91, off Three Kings
Islands, N. Zealand, 300 fathoms.

Description. — The hydrothecae are less curved than in Ritchie’s schizotype, Brit.
Mus. Reg. No. 10.10.4.19, and the plane of the hydrothecal margin less nearly vertical.

Measurements :

Hydrothecae, mm.

distance apart . . . . . . . . . . . . . . 0-72-0-90

length . . . . . . . . . . . . . . . . 0-72

diameter at margin . . . . . . . . . . . . . . 0-15

Remarks. — The specimen resembles Ritchie’s in having an evenly curved concave
abaxial wall from the row of puncta to the margin, and in that the lateral parts of the
“ arris ” extend down below the level at which the preceding hydrotheca becomes
free from the hydrocaulus, and in having only a very slightly emarginated rim. Allman
(1877) does not figure the adaxial wall fully as it stops at the line of puncta, but I
estimate that the termination would have been just below the point of emergence of
the hydro theca. In Clarke’s figure of the type the wall comes down just about as far

* The term “ arris ” is introduced for the internal perisarcal ridge formed at tfie intersection of the
distal wall of the thecal pedicel and the hydrocaulus. It is close to this spot that the thecal floor is laid
down in those forms that have one.

V. 5.

5

164 “TERRA NOVA” EXPEDITION.

as this level or a little short of it. The abaxial hydrothecal wall in Clarke’s figure of
Allman’s type appears to be slightly convex above the line of puncta, and to make an
angle with the upper part. There is an Atlantic species (Madeira, Brit. Mus. Reg.
No. 19.8.15.4) with hydrothecae not overlapping, and lower abaxial hydrothecal wall
convex, making an angle with the upper part, growing alongside a species of much the
same dimensions (Brit. Mus. Reg. No. 19.8.15.4a) which has overlapping hydrothecae
and evenly curved concave abaxial hydrothecal wall.

Text-fig. 19. — Acryptolaria conferta (Allman), a, specimen (Brit. Mus. Reg. No. 19.8.15.4), from Porto
Santo Island, 60 faths. ; b, variety from same locality ; c, var. australis, from “ Terra Nova ” Station
96 ; d, var. australis (Brit. Mus. Reg. No. 10.10.4.19), from “ Thetis ” Station 42 ; e, the same. All
figures magnified, X 29.

It is therefore desirable to have a more critical specific description of Allman’s type
material before we can with certainty cite the Australian and New Zealand specimens
as varieties of this species. Two dozen small branched specimens up to 1*5 cm. in
height on a fistule of a sponge Phloeodictyon, sp. from Station 96, off North Cape, New
Zealand, 70 fathoms, possibly belong to this species. Their measurements are :

Hydrothecae, mm.

distance apart . . . . . . . . . . . . . . . . 0-68

length . . . . . . . • • . . . . . . . . . 0-55-0-60

diameter at mouth . . . . . . . . . . . . . . 0-10

If these two groups of specimens belong to the same species the variation in size
may possibly be correlated with depth. The specimens with the largest hydrothecae
come from the greater depth.

Genus ZYGOPHYLAX, Quelch.

Genotype. — Zygophylax profunda, Quelch, 1885, p. 4, pi. I, fig. 4.

Specimens of two species of this rare genus were taken in New Zealand waters,
Z. sibogae, Billard, and a new one which is obviously related to “ Lafoea ” convallaria,
Allman, “ Lictorella ” cervicornis, Nutting, and “ Lictorella ” concinna, Ritchie. Speci-
mens of the gonosome of Z. sibogae, briefly described by Billard and lent to the Museum

HYDROIDA— TOTTON.

165

by him, furnish, the clue for connecting the majority of Zygophylax species with the
genotype Z. profunda, Quelch — a form that has a peculiar gonosome, which at first
sight is very unlike that of other species.

A fragment without gonosome mounted as a microscopical slide is all that appears
to remain of the type specimen of the genotype Z. profunda, Quelch. However, other
Atlantic specimens have been taken which seem to be conspecific. Of these one, Brit.
Mus. Reg. No. 19.8.15.2, has a gonosome consisting of conjoined gonothecae of irregular
shape and with a varying number of tubular openings provided with flared mouths.
It is assumed, because of resemblances between the respective trophosomes, that this
is the type of gonosome borne by the genotype. Z. sibogae, Billard, has a similar gono-
some in which, however, some of the gonothecae are quite separate ; and these show
obvious resemblances to the normally separated gonothecae of such species as “ Lictorella”
cervicornis, Nutting, “ Lictorella ” concinna, Ritchie, “ Lcifoea ” pinnata, Sars, and
Lcifoea ” convallaria, Allman. Eighteen species of this genus have been described as
follows : pinnata (Sars), 1874 ; convallaria (Allman), 1877 ; profunda, Quelch, 1885 ;
tizardensis, Kirkpatrick, 1890 ; cervicornis (Nutting), 1905 ; biarmata, Billard, 1906 ;
armata (Ritchie), 1907 ; levinseni (Saemundsson), 1911 ; Carolina (Fraser), 1911 ; concinna
(Ritchie), 1911 ; curvitheca, Stechow, 1913; sibogae, Billard, 1918 ; brevitheca, Jaderholm,
1919 ; pacifica, Stechow, 1920 ; recta, Jarvis, 1922 ; africana, Stechow, 1923 ; valdiviae,
Stechow, 19236 ; brownei, Billard, 1924.

The first, Lcifoea pinnata, Sars, has never been included in the genus because the
criterion has always been held to be the presence of nematothecae. Apart from this
difference from the genotype and most other species of the genus, L. pinnata cannot be
excluded ; and since both a more southerly species Z. brownei , Billard, and a more
northerly one Z. levinseni (Saemundsson), which are very closely related to Tj. pinnata,
both have nematothecae, the absence of them in R. pinnata and in L. concinna, Ritchie,
cannot be regarded as a reason for exclusion from the genus.

Since the generic name Lictorella has been used for species of Zygoplnjlax it will
be as well to look into the history of this and other closely related genera which are
somewhat ill-defined. About Lictorella there is great confusion. No genotype appears
to have been fixed. When instituting the genus in 1888 Allman had before him
specimens of two species which appear to be congeneric. He did not notice the gono-
some of either, nor did he observe the nematothecae of the hydrothecal peduncles of
one of them, L. cyathifera. Ever since then the absence of nematothecae from forms of
this group in which there is a floor to the hydrotheca has been held to be the criterion
for inclusion in the genus Lictorella ; but I have seen nematothecae in L. antipathes (Lk)
and L. cyathifera, Alim. ; and Bale has described them in what is probably a third
species, L. rufa (Bale), 1884. To the “ Challenger ” Torres Straits specimens Allman
wrongly applied the name halecioides which he had given in 1874 to an Atlantic
species, probably the Lafoea pinnata of Sars. These Torres Straits specimens,
however, appear to agree in all respects with the Indo-Pacific Sertularia antipathes

166

“TERRA NOVA” EXPEDITION.

of Lamarck and I select this species, L. halecioides, Alim. 1888=L. antipathes
(Lk) (non L. halecioides, Alim. 1874 =L. pinnata, Sars), as the genotype of Lictorclla.
Allman. The gonosome, first noticed by Trebilcock, is a coppinia mass of closely
packed and conjoined gonothecae having hooded apertures much as in Perisiphonia,
but without nematotheca bearing ramules. Broch (1909) says that the genus Lielorella
was set up for Lafoea pinnata, Sars.:;:

Abietinella, Levinsen, 1913, is the only one of the related genera in question in which
an operculum has been observed. I select Zygophylax operculata, Jaderholm, 1903, as
its genotype. Whether, as suggested by Vanhoffen (1910), any species of Zygophylax
besides Z. operculata, Jaderh., and Z. grandis, Vanhoffen, have opercula is uncertain,
though no such structures appear to have been seen in the many specimens examined.
The gonosome of Z. operculata is unknown.

Another closely related genus is Perisiphonia, Allman, 1888. As its genotype I
select P. pectinata, Alim. 1888, with liolotype, Brit. Mus. Reg. No. 88.11.13.34, from New
Zealand. The species seems to agree very closely with one that occurs at Madeira, and
has been described and figured by Stechow (1925). The gonosome of the genotype
was first noticed by Trebilcock whilst examining “ Challenger ” types during the War.
Stechow’s figures are characteristic. The gonosome figured by Stechow is sexually
dimorphic, but in each sex is a coppinia composed of adherent gonothecae, with hooded
apertures.

The genotype of Brucella has a compact gonosome of the coppinia type with short
tubular apertures. We do not know whether the gonosome is sexually dimorphic.

Still another genus that appears to be related must be considered here, namely
Cryptolaria, Busk (nec auctt.). The holotype of the genotype C. prima, Busk, is Brit.
Mus. Reg. No. 57.1.2.86. It appears never to have received a critical examination. f
I find that the axial tube bears four rows of hydrothecae in decussating pairs, each
hydrotheca separated off proximally by a floor. The four rows are so arranged that
two rows of apertures are more or less on the upper, and the other two on the lower
side of the hydrocladia. Lying next the axial tube, one on each of its two opposite
sides, are a pair of accessory tubes furnished with nematothecae which possibly remain
at the surface : and the whole is overlain on all sides by a number of other accessory
tubes, some of which at any rate originate as branches of the axial tube. In the unique
dried type specimen from New Zealand the hydrothecal orifices are flush with the
general surface of the stem and hydrocladia. The branches are sub-opposite. I know
of no other species obviously congeneric with C. prwm, Bk ; and consider it would be
premature to say that the species of Perisiphonia are congeneric.

* We take the view that Allman was establishing a new genus for the species to which belonged
specimens lying before him. To them he wrongly applied an existing specific name. In consequence,
instead of taking for the genotype the species rightly bearing that misapplied name, we must take the
species of which specimens were in front of Allman ; at the same time giving it its first available name
antipathes, Lamarck.

f The species has been described by Trebilcock (1928) as a new species Perisiphonia quadriseriata.

HYDE OIDA— TOTTON.

167

Zygophylax unilateralism sp. nov. (PI. I, figs. 1,2 : text-fig. 20.)

Material. — A dozen stems, tlie largest 6 cm. in lieiglit and 1 mm. in diameter, on
two small masses of calcareous sand, associated with, a creeping alcyoiiarian ; from
Station 91, off Three Kings Islands, New Zealand, 300 fathoms.

Description. — The species generally resembles Lictorella concinna, described and
figured by Ritchie (1911), but differs from it in two respects. Nematophores are
present, one on the outer side of each stem or branch
apophysis at base of hydrothecal pedicel ; and the
hydrothecae are more or less unilateral, larger than
in concinna, relatively longer than pedicels, and open-
ings are turned up more, facing both upwards and
outwards, so that those of one side are at right angles
to those of the other, as if hydrothecae had been
bilateral and straight instead of unilateral and curved.

Polyp with an abcauline blind sac and about ten
tentacles. There is usually a node between every three
hydrothecae.

mm.

.. 60-0
.. 1-0
. . 0-12
. . 0-10 '

. . 0-49
. . 0-20
.. 0-17

Zygophylax .sibogae, Billard. (Text-fig. 21.)

Zygophylax sibogae, Billard, 1918, p. 21, fig. 1.

Material. — A single slender specimen with three branches, 15 mm. in height, with-
out gonosome, from Station 91, near Three Kings Islands, New Zealand, 300 fathoms.

Description. — Colony pinnately branched, stem fascicled, accessory tubes bearing-
numerous narrow tubular nematothecae. Hydrothecae sharply recurved and flattened
like a cobra’s hood, shorter than the long pedicels which carry them, constricted at
bend, wall of re-entrant angle thickened ; mouth circular without operculum, facing
upwards and inwards towards stem. Generally there is one short nematotheca at
base of each pedicel on the inner side, pedicels not being strictly lateral, but having a
tendency to be inclined to one, the inner side.

Measurements :

mm.

Height of colony . . . . . . . . . . . . . . . . 15-0

Stem, diameter . . . . . . . . . . . . . . . . 0-2

Axial tube, diameter . . . . . . . . . . . . . . . . 0-07

Branch, diameter . . . . . . . . . . . . . . . . 006

Measurements :

Height of colony
Stem, diameter
Axial tube, diameter
Branch, diameter
Hydrotheca,
length

diameter at mouth
peduncle, length

168

“TERRA NOVA” EXPEDITION.

mm.

Hydrotheca, length . .
diameter at mouth
peduncle, length. .

Nematotheca, length . .

diameter

Remarks. — The species appears to be related to Zygo-
phylax tizardensis, Kirkpatrick, to Z. curvitheea, Stechow,
and to a fourth undescribed species from the Timor Sea
that has short pedicels and less curved hydrothecae, more
than to other species that has been included in this
genus. Z. sibogae differs from Z. tizardensis in having
no second re-entrant angle and thickening on the convex
side of the hydrotheca ; and in having the mouth turned
upwards and towards the stem. It differs from the other
species in having very long hydrothecal pedicels.

Two unfascicled stems 5 mm. in length growing on
Acryptolaria conferta var. australis from Station 91 have less
curved hydrothecae, but probably belong to the same species.

SYNTHECIIDAE.

Genus SYNTHECTUM, Allman.

Genotype. — Synthecium elegans, Allman, 1872, p. 229, fig. 1.

There are two facts which at present make the identification of Australian and
N. Zealand species of Synthecium a very difficult matter. In the first place there is
reason to believe that the dimensions of male and female specimens of any species are
different, the female being sometimes the larger and sometimes the smaller ; and,
secondly, the types of the five first known species whose names are so often mentioned
in synonymy lists : S. elegans, Alim., S. patulum (Busk), S. orthogonium (Busk),
S. ramosum, Alim., and S. campylocarpum, Alim., have been lost or are in a poor state
of preservation.

No type specimen of S. patulum (Busk) is known to exist. We must therefore
base our ideas of the species on Busk’s sketch which is reproduced here. It resembles
Bale’s fig. 10, pi. V (1884), but unfortunately is not accompanied by any indication of
the scale to which it was drawn.

The specimen of S. elegans figured by Allman in 1876 is not known to exist.
Therefore I select as lectotype of this genotype another dried New Zealand specimen
from the Busk Collection, Brit. Mus. Keg. No. 99.7.1.6746, which is almost certainly
part of the original material. The gonangia figured by Allman (1876) were artificially
flattened from side to side.

Text-fig. 21. — Zygopliylax
sibogae, Billard. X 33.

0-34

0-12

0-34-0-44

0-05-0-1

0-02

HYDROIDA— TOTTON.

169

The only original material of S. orthogonium known to exist is a slide in the Busk
Collection, Brit. Mus. Beg. No. 99.7.1.6367. This specimen is the holotype of the
species.

The syntypes of S. campyJocarpum, Allman, consist of a figured group of six or
seven male stems arising from a hydrorhiza, a group of four female stems attached to
barnacles, and five detached specimens. One of these, 3-5 cm. in length, consists of
two secondarily connected female stems arising from a hydrorhizal tube overgrown by
a polyzoan. I select it as the holotype of the species. The female specimens are of
smaller dimensions.

Text-fig. 22. — a, Synthecium elegans, Allman, lectotype (Brit. Mus. Reg. No. 99.7.1.6746), (?) X 33 ;
b, enlarged reproduction of Busk’s sketch of same species with scale, X 32 (?) ; c, S. ■patulum
(Busk), enlarged reproduction of Busk’s sketch (scale unknown, sketch enlarged half as much as
original of b).

Specimens of Synthecium species were taken at Stations 91, 134 and 144 off Northern
New Zealand, and fall into six categories according to dimensions.

Two groups of Terra Nova ” material from Stations 134 and 144 form a category
with the smallest dimensions. The size of the specimens is about the same as that of
S. orthogonium. They are regarded as types of a new species S. carinatum. Two other
specimens from Station 144 fall into a category of specimens about the size of S. ramosum,
one probably being female and the other male. Four others from Station 134 form a
group resembling none of the types mentioned very closely, three specimens of one

170

TERRA “NOVA” EXPEDITION.

size being of one sex probably and the remaining one representing the other. A single
fragment of larger dimensions from Station 91 agrees with an unnamed specimen taken
by Haddon in Torres Straits and is regarded as a new species. The species of Synthe-
cium require working out in detail. Meanwhile the “ Terra Nova ” material is dealt
with as tentatively suggested above.

It is usual to find in specimens of Synthecium spp. that the twro hydrothecae
forming the basal pair of any hydroclade are not alike. The base of the upper one is
nearer the axis of the stem, and its free adcauline wall is noticeably shorter than that
of others. Also the length of the free adcauline wall of the proximal pair of each
successive series of stem hydrothecae is less than that of the following pair or pairs.

Synthecium ramosum, Allman. (Text-fig. 23, a-c.)

Synthecium ramosum, Allman, 1885, p. 137, pi. XII, figs. 3-4.

Material. — Three branched stems, without gonothecae, one 1*5 cm. in length with
larger unreduplicated hydrothecae, and two 2T cm. and 2-5 cm. in length wdth smaller
reduplicated thecae, from Station 144, off Cape Maria van Diemen, N.Z., 35-40 fathoms.

Text-fig. 23. — Synthecium ramosum, Allman, a, Scliizotype (Brit. Mus. Reg. No. 86.2.19.12-13), $ ;
b, (?) $ from Station 144 ; c, (?) from Station 144. All figures magnified X 33.

Description. — The specimens of smaller dimensions correspond closely with the
dried and flattened holotype of S. ramosum, Brit. Mus. Reg. No. 86.2.19.12-13,
both in shape and size. Judging by S. campylocarpum these smaller specimens
are males.

HYDROIDA — TOTTON.

171

Measurements in millimetres :
Hydro theca,

length,

6

$

fused part

0-74

0-62

free part

0-32

0-29-0-32

over all . .

0-93

0-80-0-85

diameter

at margin

0-27

0-21

Remarks. — The hydrothecae of the holotype of S. ramosum differ from those of
the type of S. elegans in having a larger free adcauline wall, a wider orifice and the plane
of opening less nearly vertical. Billard (1910), who did not state what particular
specimens he was examining, found himself unable to keep the two species apart. In
view of the unsatisfactory nature of the dried and distorted types I prefer to leave the
question open.

Synthecium carinatum, sp. n. (Text-fig. 24, a-d.)

Material. — A flattened tuft of 4 cm. long stems, about 5 cm. in width, growing on
a sponge from Station 134 in 11-20 fathoms; another irregular tuft of 2 cm. long

Text-fig. 24. — Synthecium carinatum, sp. n. a, nematocysts, X 255 ; b, lateral view of gonotheca in
optical section, X 33 ; c, gonotheca, (?) <J, x 55 ; d, a pair of hydrothecae from holotype (Brit. Mus.
Reg. No. 29.10.10.10), from Station 134, showing internal teeth, X 158.

stems growing on a sponge and a polyzoan from Station 144, off Cape Maria van
Diemen, N. Zealand, 35-40 fathoms,
v. 5.

G

172

“TERRA NOVA” EXPEDITION.

Description. — The specimens are slender compared with those of S. elegans or
S. campylocarpum. The distal third part of the hydrotheca curves gradually away
from the axis, so that the plane of the orifice makes an angle of 20° with the axis, and
its axial wall is generally flattened. The gonothecae which appear to be male have
folded walls and are very much flattened in a plane at right angles to the plane of
branching. With the exception of the proximal two or three circular ones, the twelve
folds coalesce at the edges, where they are either opposite, sub-opposite, or alternate.
In some specimens there is on the adcauline side of the distal part of the hydrothecal
wall a well-marked keel-like internal tooth. Occasionally there are traces of two
rounded lateral internal teeth on the abcauline side. In other specimens there are no
traces of these teeth, and in others again rudimentary teeth are formed in some hydro-
thecae whilst the majority of the hydrothecae show no trace of teeth at all. A number
of banana-shaped nematocysts are scattered through the ectoderm of the hydrocaulus
and of the distal part of the gonophores t and ten or twelve are present in the bases of
some of the polyps.

Measurements ( holotype ) :

Hydrotheca,

length,

fused part
free part
chord (over all)
diameter at orifice
Gonotheca,
length
breadth
thickness

number of folds . .
Nematocysts

mm.

0-29

0-12

0-37-0-39

0-11

0-83-1-02. Av. (19) 0-9
0-64
0-27
12

0-05X0-01

Remarks. — Two species of Synthecium having internal teeth have been described
before : S. dentigerum, Jarvis (1922), and S. singulare, Billard (1925). From the former
S. carinatum is distinguished by being only two-thirds of its size. Of S. singulare it
appears to be about half the size ; and in comparison the free part of the hydrotheca
is relatively shorter. The dimensions of the group of specimens from Station 144 are
a little greater than those of the holotype which is the group of stems, Brit. Mus. Reg.
No. 29.10.10.10, from Station 134.

Some comparable data for a fragment of the holotype of S. dentigerum , Jarvis, are
given below : —

Measurements, of S. dentigerum, Jarvis (schizo-holotype) : —

Hydrotheca,
length,

fused part
free part
chord (over all)
diameter at orifice

mm.

0-49

0-25

0-70-0-73

0-22

HYDROIDA — TOTTON.

173

Reduplications, which considerably lengthen the free part of the hydrotheca, have
not been taken into account.

The possibility must not be overlooked that, in view of the disparity in size
between the sexes in S. samauense, Billard, S. carinatum and S. dentigerum or S. singular e
might prove to be the two sexes of one and the same species.

Synthecium robustum, sp. n. (Text-fig. 25, a-d .)

Material. — A specimen complete with gonothecae, 5-5 cm. in height, having larger
hydrothecae ; and two specimens 2 cm. and 2-5 cm. in height without gonothecae and

Text pig. 25. — Synthecium robustum , sp. n. a, b, d, Schizoholotypes, (?) $, from Station 134 ; c, Paratype.
(?) cj, from Station 134. All figures magnified X 33.

with smaller hydrothecae, the larger of these two consisting of four stems arising from
a common root stock ; from Station 134, off North Cape, N. Zealand, 11-20 fathoms.

174

“TERRA NOVA” EXPEDITION.

Description. — One-fourth part of adcauline wall of hydrotheca free, distal end of
theca bending away at angle of 40° from axis. No internal hydrothecal teeth. Gono-
thecae ovoid, compressed at right angles to plane of branching, with from eleven to
twelve transverse thickened folds, upturned at the free edge, and apical aperture.

Measurements :

nyarotneca,

Holotype.

Paratype.

length,

mm.

mm.

fused part

0-82-0-92

0-57

free part

0-25-0-30

0-17-0-20

diameter at orifice

0-25-0-26

0-20-0-21

Gonotheca,

length

1-45-2-0

breadth

1-15-1-25

thickness

0-70-0-80

Remarks. — The large complete specimen, Brit. Mus. Beg. No. 29.10.10.5, is the
holotype. If the paratype belongs to the same species it is probably of the other sex.

Synthecium longithecum, sp. n. (Text-fig. 26.)

Material. — A branched fragment 12 mm. in length, without gonothecae, from

Station 91, off Three Kings Islands, N. Zealand,
300 fathoms.

Description.- — Hydrothecae strongly curved out
so that the orifice is nearly vertical, two-thirds
of length being free.

Measurements :

Hydro theca,

length, mm.

fused part . . . . 0-84-0-90

free part . . . . 0-47-0-57

diameter at orifice . . 0-25-0-26

The measurements of three stem hydrothecae
with their positions on the internodes are : —

. , Proximal. Intermediate. Distal,

lengtn, mm. mm. mm.

fused part . . 0-90 0-84 0-76

free part . . 0-25 0-45 0-79

The diameter of the orifice of the stem hydro-
thecae may be as great as 0-31 mm.

Remarks. — There is a very similar species at
Murray Island, Torres Straits. A specimen, Brit.
Mus. Reg. No. 20.2.27.2, bears a single smooth
cylindrical gonotheca 1-25 mm. in length and
0-6 mm. in diameter.

Text-pig. 26. — Synthecium, longithecum,
sp. n. Holotype, X 30.

HYDROIDA— TOTTOK

175

Genus STAUROTHECA, Allman.

Genotype. — S. dichotoma, Allman, 1888, p. 76, pi. XXXVI. figs. 1-la.

Billard (1914) has given as a definition of the genus, a slight modification of
Ritchie’s (1907). StaurotJieca is dioecious ; and, as Billard long ago suspected, sexual
dimorphism is a generic character : the gonothecae are occasionally ornamented

instead of simple.

It is interesting to recall that the first specimen of any StaurotJieca species known
to have been taken was a female S. dichotoma, Allman, taken in Antarctic waters by
H.M. Ships “ Erebus ” and “ Terror ” in the years 1839-43.

Allman (1888) did not record all the specimens taken by H.M.S. “ Challenger.”
From off Marion Island came both male and female (unrecorded) specimens of the
genotype. I select the male specimen figured and described by Allman as the holotype
of the genotype. The female specimens were first noticed by Trebilcock about ten
years ago. Billard (1914) assigned to S. affinis, Jaderholm, the female specimens taken
by H.M.S. “ Challenger ” off Prince Edward Island. The same vessel took a male
specimen of Stauroiheca, sp. from Station 232 off Japan and a batch of male specimens
hitherto unrecorded from off Heard Island.

The following are the characters upon which classification has been based : —

1 . Nature of stem, whether fasciculated or not.

2. Grouping of hydrothecae, whether in decussating pairs or in whorls of

three or more.

3. Shape of gonangia.

4. Size.

5. Nature of coenosarc, whether formed of one or many tubes.

6. Relative length of free portion of hydrotheca.

Taking these seriatim : — 1. Fasciculation appears to depend on whether the
colony has reached a sufficient size to need secondary support in the stem.

2. On one branch of a specimen of the type of Dictyocladium fuscum the hydro-
thecae are arranged in decussating pairs : this branch grows out of another on which
the thecae are whorled ; and vice versa. This variation probably depends on a change
of rate of metabolism : the whorled arrangement is often correlated with luxuriant
growth.

3. Taking all species, two main types of gonangia have been described. Specimens
hitherto described have either been sterile or else borne gonothecae of one or the other
type. The only histological evidence of the sex of any of these specimens is that given
by Billard (1914), who has observed ova in the coenosarc of the “ Challenger ” material
coming from Prince Edward Island. The gonothecae on these specimens are of the
urn-shaped type and have relatively large openings. No further histological evidence
will be brought forward in this report, but the circumstantial evidence is almost con-

176

“TERRA NOVA” EXPEDITION.

elusive for saying that the known species of Staurotheca are sexually dimorphic. The
fact that the gonothecae of any specimen are urn-shaped or spindle-shaped is not a
specific criterion.

4. With regard to the size of specimens there is a marked division between those
of larger and those of smaller dimensions. In the first category come those attributed
to S. dichotoma, S. affinis, Jaderholm, and Dictyocladium fuscum, Hickson and Gravely.
In the second comes S. antarctica, Hartlaub. This latter species can be distinguished
from other species by its relatively small dimensions, as well as by its peculiar female
gonothecae. A study of the comparative table of dimensions which follows will show
that the specimens from the Graham Land region and Cumberland Bay, S. Georgia,
named by Jaderholm “ dichotoma ,” are of smaller size than those of that species ; and
that they should probably be assigned to S. antarctica, Hartlaub.

Selaginopsis pachyclada, Jaderholm, 1904, may prove to be a Staurotheca species.
It has neither operculum nor marginal teeth.

The “ Terra Nova ” Staurotheca material comes from three localities : one off
Oates Land to the North of Lillie Glacier Tongue and Mount Bruce, another some
seven or eight degrees further South, between Granite Harbour and Glacier Tongue,
McMurdo Sound, and the third West of the Falkland Islands.

The following is a list of the specimens : —

Station. No.

Species.

Sex.

Remarks.

38

antarctica

(sterile)

fragments

194

dichotoma

antarctica

male and female

99 99 99

dioecious

99

220

dichotoma

(sterile)

316

dichotoma

antarctica

(sterile)
( „ )

1 fragment only
fragments only

338

antarctica

male

339

antarctica

female

340

antarctica

female

356

dichotoma

sterile

overgrown fragments

SUMMARY OF MEASUREMENTS OF RECORDED SPECIMENS OF STAUROTHECA.

Staurotheca dichotoma, Allman.

HYDROIDA— TOTTON.

177

t -J

0

0

-*-> ©

05

00

'©©J

r-

O

CO

O

1

1

0 O

CD

O

3 '3

AO

O

£S

O

0

O

00

©1

IO

©

,0

0

CM

O

AO

a

—

|

0

O

1

©

£

CO

00

O

>-0

£

0

CM

ft

0

Ol

C-l

O

!>•

0

is

O

00

0

CO

©

AO

AO

0

c

O ©

O

O

_

O

■4J ©

0

w

AO

c? •—*

5 3

O

0

*+-

0

CO

O

“-1

0

O

00

£—

00

O

O

0

t—

0

1

1

O

O

1

s

£

0

O

£

fc

r-

tS

AO

O

0

AO

0

--

0

AO

Ol

si

y

0

0

CM

—

-f

O

(-H

©

0

10

AO

CO

C5

<m

<M

0

CM

©

—

CO

CO

CO

CO

CO

g

O

0

0

0

01

CO

0

0

t—

r-

H O

CO

<M

CO

Cl

01

0

0

O

AO

0

-f1

AO

C5

CO

01

AO

ft

CO

CO

f-4

©

0

O

0

0

0

AO

£"•

r*>

w

CD

a

©

ft!

M

O

0

AO

C5

ft

10

r-

-t

0

AO

AO

O

t"

1

OD

1

1

U-

1

0

co

O

O

O

<M

a

O

co

OO

■0

I"

AO

lO

-3<

©

5 „ ■ .2

O

1

- S3 0 s

CO

v'vjz a

0

O

CO

y >>0 0 - ^
ci—

1

0

AO

lO

O

AO

0

H

0

T*

n

a?

u

0

0

>>

a

bfi

0

bu

G

bfi

0

17

Cj

0

0

3

"0

^0

d

§S

0

0

0

O

O

Q

H

r-J *

03 '.

q

0

0

0

co

HH

CO

h- 1

O

CO

0

0 72

03 co

ft,

0

q -e

u

*rt

©

2

s §

S a

u

> 0
73 2

W 2

g

O

S3

5z5

. 0

0

co

O-1

q co

02

O

r=3

GO

? g

§ g

Oh

cj g

.a a

1

ft

O

S-3

CO

CO

0 -Tj

"S

0

0

0

iri

0

a

IS

1 3

>>

0

a

0

w

PH

p

£

£

ft

ft

o

ft

ft

ft

o

ft

pQ

c3

I

O

o

c3

w

I

o

ao

I

o

»o

o

fc.

s

e

S3

0

Ph

ft

q

o

0

0

o

Ph

H

GQ

fcq

o 0
£ <??

e

GQ

2

bo

eft

e q
73 ©

§ a

°3

,=3 e

03

=3

*fc!

:0

-a

ca

<3

t>

53

P=<

Figures placed within brackets relate to reduplicated thecae. The unit of measurement is one micron.

178

“TERRA NOVA” EXPEDITION.

Staurotheca dichotoma, Allman. (PI. II, fig. 9 ; text-fig. 27, a-c.)

Staurotheca dicliotoma , Allman, 1888, p. 76, pi. XXXVI, figs. 1-la ; Selaginopsis affinis, Jader-
holm, 1904, p. XI ; Dictyocladium fuscum, Hickson and Gravely, 1907, p. 21, pi. Ill, fig. 22.

Material. — A number of branched fragments up to 4 cm. in length, with male and
female gonothecae, from Station 194, off Oates Land, 180-200 fathoms ; part of a sterile
colony 9 cm. in length from Station 220, off Cape Adare, 45-50 fathoms ; a minute
sterile fragment from Station 316, off Glacier Tongue, McMurdo Sound, 190-250 fathoms ;
and some overgrown sterile fragments from Station 356, entrance to McMurdo Sound,
50 fathoms.

Description. — Dioecious ; female gonothecae vase-shaped, tapering below, with
aperture at end of short wide cylindrical collar that arises from broad shoulders.

Text-pig. 27. — Staurotheca dichotoma, Allman, a, hydrotheca from Station 316 ; b, specimen from
Station 220 to show three rows of hydrothecae ; c, male and female gonothecae from Station 194. All
figures magnified X 24.

Remarks. — This appears to be a very variable species, but is easily distinguished
from the only other well defined one by its greater size and plain vase-shaped female
gonothecae.

The specimen from Station 220 closely resembles Hickson and Gravely’s types of
Dictyocladium fuscum. Since the gonothecae of this form were not found it may best
be included in S. dichotoma.

Staurotheca antarctica, Hartlaub. (PI. II, fig. 6 ; text-fig. 28.)

Staurotheca antarctica, Hartlaub, 1904, p. 16, pis. I, fig. 4, II, fig. 4 ; S. dichotoma, Jaderholm,
1905, p. 33, pi. XIV, figs. 1-2 (not S. dichotoma, Allman, 1888, p. 76, pi. XXXVI, figs. 1-la) ;
? S. reticulata, Ritchie, 1907, p. 538, pi. I, figs. 1-1 b.

Material. — Fragments up to 2 cm. in length with male and female gonothecae
from Station 194, off Oates Land, 180-200 fathoms ; a few fragments 1-5 cm. in length

HYDE OIDA— TOTTON.

179

without gonothecae from Station 316, McMurdo Sound, 190-250 fathoms ; a tangle
of anastomosing specimens 11 cm. in height bearing male gonothecae, from Station 338,
McMurdo Soimd, 207 fathoms ; branched fragments 3-4 cm. in length bearing female
gonothecae from Station 339, McMurdo Soimd, 140 fathoms ; branched fragments up
to 5-5 cm. in length with female gonothecae, from Station 340, McMurdo Sound,
160 fathoms ; and a fragment 3 cm. in length from Station 38, West of Falkland Islands,
125 fathoms.

Description. — Dioecious ; male gonothecae like those of S. dichotoma but smaller.
Female gono theca covered in its distal half with longitudinal rows of irregular or finger-
like processes except on the outer side, where there is a round aperture, which is guarded
by a curved and often bifid finger-shaped process standing out horizontally from the
hydrocaulus, from the upper side of which process the gonotheca itself arises. One
large rounded ovum or embryo lies against the operculum.

Remarks. — Measurements for this species have been given on p. 177. It is a
much smaller species than the only other well defined one, S. dichotoma, Allman, and
is distinguished readily by its peculiar female gonothecae.

Parascyphus simplex (Lamouroux). (Text-fig. 29, a-b.)

Laomeclea simplex, Lamouroux, 1816, p. 207. For synonymy see : Stechow (1925), p. 224 ;
and Splettstosser (1929), p. 126.

Material. — Five or six dozen stems from a stolon creeping over an old Gasteropocl
shell, from Station 134, off North Cape, New Zealand, 11-20 fathoms.

Description. — Nine or ten of the specimens bear a single short branch in the
middle region. Branches have not been seen before. There are no gonothecae.
v. 5. 7

(

Text-fig. 28. — Slaurot/ieca anlarclica,
Hartlaub. Female gonotheca, X 34.

Text-fig. 29. — Parascyphus simplex (Lx.), a, from
“ Terra Nova ” Station 134 ; h (Brit. Mus. Reg.
No. 10.10.17.2-7), from Victoria-Tasman cable. Both
sketches magnified X 53.

SERTULARIIDAE.

180

“TERRA NOVA” EXPEDITION.

* Measurements :

Colony, height
Stolon, diameter
. Hydrocaulus, diameter
Internode, length
Hydrothecae,
distance apart
length (to apophysis)
maximum breadth . .

mm.

8-12

0-18-0-25

0-12-0-22

0- 42-0-55

1- 0

0-55

0-20

Remarks. — This little-known hy droid has been recorded before from the West
Coast of Scotland, Gough Island, South Coast of Australia, Tasmania, and New Zealand ;
but never in any abundance. Neither when taken in December/ January (New Zealand
and Tasmania), in April (Gough Island), nor at the end of August (New Zealand), were
any gonothecae found. The gonothecae have been described once by Bale (1915) ;
but the season when they were found was not indicated. This genus represents the
survival of a stage in the evolution of Symplectoscyphus. The diaphragm or floor of
the hydrotheca, which appears late in the ontogeny of species that possess it, is small
and incomplete.

Symplectoscyphus columnarius (Briggs). (PI. I, fig. 10 ; text-fig. 30, a-c.)

Sertularella columnaria, Briggs, 1914, p. 293, fig. 1 ; Sertularella columnaria. Bale, 1924, p. 239.
Material. — A branched polysiphonic fragment four centimetres long bearing
gonothecae, from Station 90, off Three Kings Islands, N.Z., 100 fathoms.

Text-pig. 30. — Symplectoscyphus columnarius. Briggs, a, b, gonothecae, X 11 ; c, hydrotheca, X 33.

Description. — The dimensions of branch internode diameter and of hydrothecae
correspond with those given for the species by Briggs. The gonothecae, which have
not been described before, are much like those of S. meridionalis, Nutting, except that
the terminal portion flares open gradually, there being no distinct intermediate tubular
portion. They are somewhat flattened, and have three distal corrugations.

* See Ritchie (1909«, 1911), Billard (1909), Briggs (1914), Bale (1915).

HYDR 01DA— TOTTON.

181

Measurements :

Gonotheca,

length

greatest diameters
opening, diameter
terminal tube, length . .

mm.

3-19

1-30 X 1-68

0-45

0-55

Remarks. — The fragment looks like a detached branch, in which case the original
description of the stem applies also to the branches, which may bear secondary branches
15 mm. long. The species has already been recorded from an undetermined locality
in New Zealand by Bale (1924).

Symplectoscyphus constrictus, sp. n. (PI. I, fig. 3 ; text-fig. 31.)

Material. — Parts of four specimens, the largest 7-5 cm. long, from Station 91, off
Three Kings Islands, New Zealand, 300 fathoms.

Description— A branched species, with a
trophosome resembling that of Sertularella
macrocarpa, Billard, and S.pedunculata, Billard,
only larger. Stem with a few accessory tubes
in the basal part and with regularly alternate
pinnae. Separating each alternating pair of
pinnae is a pair of alternating hydrothecae,
and in the axil of each pinna a hydrotheca.

Hydrotheca tubular, free part about
twice the length of adnate, with three small
teeth. On the adcauline side of the free
portion there is a characteristic constriction
close to the cauline wall. Internodes of
pinnae fine, equal in length to hydrothecae or
up to a third longer, forming a zig-zag.

Gonothecae absent. Polyps with prominent
hypostome and about sixteen tentacles.

Measurements :

Hydrothecae,

Text-fig. 31.

—Symplectoscyphus constrictus.
sp. n. X 35.

length

orifice diameter

mm.

0-76

0-24

Symplectoscyphus johnstoni (Gray). (Text-fig. 32, a-c.)

Sertularia johnstoni, Gray, 1843, p. 294 ; Sertularella johnstoni, Allman, 1876, p. 261, pi. XIII,
figs. 1-2 ; Sertularella johnstoni , Bale, 1924, p. 239 ; ? Sertularella australis, Marktamier-
Turneretscher, 1890, p. 226 ; not Sertularia delicatula, Hutton, 1872, p. 256.

Material. — A small tuft of stems one and a quarter inches in height attached to
each other by tendrils. The specimens are covered with Polyzoa and a Campanularian

182

“TERRA NOVA” EXPEDITION.

hydroid. Very few gono thecae are present ; from Station 144, off Cape Maria van
Diemen, N.Z., 35-40 fathoms.

Description. — The nodes are scarcely visible. The liydrothecae have three internal
teeth — a very rare phenomenon for species of Symplectoscyphus. These internal teeth
are small and papilliform, one abcauline and two lateral.

Measurements :

Hydrotkeca,

length

diameter

Gonotheca,

length

width

aperture, diameter
annulations

mm.

0-3

045

1-15

0-34

0-07

14

Text-fig. 32. — Symplectoscyphus johnstoni (Gray), a, hydrotheca with internal teeth, X 163 ;

b, gonotheca, X 56 ; c, hydrotheca, X 55.

Remarks. — Two of Gray’s sixteen dried syntypes, collected by Dr. A. Sinclair,
have been prepared as microscopic transparencies. Internal hydrothecal teeth are
present, though not so well seen as in fresh material. The measurements of the hydro-
thecae agree with those given above. Some of the gonothecae are rather larger than
those whose measurements are given above, and the aperture is usually 0-1 mm. in
diameter. The annulations number twelve or thirteen. The compressed gonothecae
with corrugations and stout, tubular, wide-mouthed apertures are quite distinct from
those of allied species. The best figures have been given by Allman (1876) and Bale
(1914, pi. II, fig. 10).

The species is quite distinct from Sertularella divaricata, Busk, the holotype of
which I found in 1922. Symplectoscyphus divaricatus (Busk) is a species allied to
Symplectoscyphus subarticulatus (Coughtreyj, and has a fascicled stem.

HYDROIDA — TOTTON.

183

Symplectoscyphus delicatulus (Hutton). (Text-fig. 33.)

Sertularella delicatula, Hutton, 1872, p. 256 ; Sertularella capillar is , Allman, 1885, p. 133,
pi. VIII, figs. 1-3 ; not Sertularella johnstoni (Gray), 1843, p. 294.

Material. — A tuft of intertwining stems one and a half inches high, with gono-
thecae, from Station 144, off Cape Maria van Diemen, N.Z., 35-40 fathoms.

Description. — Stems monosiphonic, pinnate or doubly pinnate, interconnected by
tendrils. Nodes sometimes slightly or not at all marked. Hydrothecae inclined to
one side. Three-toothed orifices inclined slightly upwards, free part of adcauline side
equal to fused part, no internal teeth. Gonothecae adnate except at distal end which
is inflated, with from ten to twelve annulations not produced into free frilly margins,
and not always marked on the proximal part ; apertures on short truncated cones,
eccentric.

Measurements :

Hydrotheca,

maximum length
maximum width
Gonotheca,
length
width

diameter of external aperture
number of annulations . .

mm.

. . 0-27-0-29

. . 0-15

1-1 -1-22
0-51-0-57
0-06-0-07
10-12

Remarks. — There are in the British Museum four type specimens received from
Prof. Hutton in 1875. I select specimen Brit. Mus. Beg.

No. 75.1.5.57a as holotype. It is a dried tuft of stems
about one and a quarter inches long bearing very numerous
gonothecae. Hutton (1872) described the gonotheca as
having an acutely toothed crown. Probably the tooth
was the eccentric terminal tube. The “ Terra Nova ”
material agrees closely with the type ; and so do the small
dried types of S. capillaris, Allman. In these types of
Allman the diameter of the terminal tubes of the gono-
thecae is slightly larger than in S. delicatulus, being from
0-09-C-l mm. The shape and size of the terminal tubes
of the gonothecae probably differs in the two sexes, as in
those of related forms. Sexual dimorphism makes the
determination of the numerous closely related species a
matter of some difficulty at the moment, as the specific differences between the
trophosomes are not very marked. S. delicatulus and S. johnstoni appear to grow
alongside one another. The “ Terra Nova 55 specimens did, and specimens taken by
Farquhar in Ohiro Bay, New Zealand, appear to have done so too. The species
may be separated at once by their gonothecae, and by the presence or absence of
internal hydrothecal teeth.

scyphus delicatulus, Hutton.
Gonotheca, X 56.

184

“TERRA NOVA” EXPEDITION.

Symplectoscyphus spiritualis, sp. n. (Text-fig. 34.)

Material. — A tuft of intertwining stems up to one and a half inches in length
growing on a tubular sponge, from Station 144, N.Z., 35-40 fathoms. A doubly
pinnate stem and a tuft of stems from Station 134, off North Cape, N.Z., 11-20 fathoms.

Description. — Stems monosiphonic, straggling, pinnate, sometimes bipinnate,
interconnected by tendrils. Nodes scarcely or not at all marked. Hydrothecae

lateral, three-toothed, orifices lateral, free part of adcau-
line side equal in length to fused part, no internal teeth
or thickened margins. Gonothecae adnate except distal
end, laterally compressed, with from 16-18 annulations
covering entire length, distal two or three only with free
frilled margins ; distal ends not enlarged, aperture funnel-
shaped, slightly flaring, eccentric.

Measurements :

scyphus spiritualis, sp. n.
X 56.

Hydro theca,

maximum length
maximum width
Gonotheca,
length
width

diameter of aperture,
internal
external

number of annulations

mm.

0-37

0- 17

1- 27-1 -54
0-55

0-06

0-15

16-18

Three stems only from Station 144, and none from Station 134 bore gonothecae.
Remarks. — The large number of annulations on the gonothecae distinguishes this
species from forms previously described as varieties of S. divaricatus by Bale (1914).

A stem with gonothecae has been selected as holotype and mounted on a slide,
Brit. Mus. Reg. No. 29.10.10.11.

Symplectoscyphus confusus, sp. n. (PI. I, figs. 4, 6 ; text-fig. 35.)

Material. — Three tufts of stems about one and a half inches long, growing on
pieces of primnoid Alcyonarians, from Station 91, off Three Kings Islands, N.Z.,
300 fathoms.

Description. — Stems monosiphonic, pinnate or doubly pinnate. Nodes faintly
marked. Hydrothecae lateral, three-toothed, without internal teeth ; free part of
adcauline side equal to or a little less in length than fused part, orifices inclined upwards.
Gonothecae not compressed laterally, annulated except where they are adnate to pinna
for two-thirds of their length ; annulations from eleven to twelve in number well
marked and over entire length, margins not produced into free frills. Apertures on
cylindrical tubes, equal in diameter to length of tubes ; proximal aperture of tube a
little smaller than external one.

HYDROIDA — TOTTON.

185

Measurements :

Hydrotheca,

mm.

maximum length

0-37

maximum width

0-19

Gonotheca,

length

1-25-1-34

width

0-59

diameter of external aperture

0-17

,, internal ,,

0-13-0-14

number of annulations

11-12

Remarks. — I have not observed any dimorphism
of gonothecae ; but I believe the gonads are all of
one sex. The species appears to be nearest S.filiformis
(Allman). That species has dimorphic gonothecae,
the female having stouter terminal tubes which
were not described by Allman. As holotype of
S. confusus I select a tuft of stems, Brit. Mus.
Reg. No. 29.10.10.6, growing on a primnoid stem
43 mm. long, the whole measuring about 8x6x3
cm.

Text-fig. 35. — Symplectoscyphus
confusus, sp. n. Gonotheca from
Holotype, X 56.

Symplectoscyphus epizooticus, sp. n. (PI. I, figs. 5, 6 ; text-fig. 36, a-b.)

Material. — Epizootic on S. confusus, sp. n., from Station 91, off Three Kings
Islands, N.Z., 300 fathoms.

Description. — Stems simple or occasionally pinnate, about 5 mm. long, internodes
long and very slender. Hydrothecae three-toothed, apertures inclined upwards, with-
out internal teeth. Gonothecae small, flattened, smooth or irregularly undulating,
aperture central, on short tube.

Measurements :

Hydrotheca,

mm.

maximum length

0-32

maximum width

0-15

Gonotheca,

length

0-66

width

0-34

diameter of aperture

0-07

annulations

vestigial

Remarks. — Very few gonothecae were observed. Specimens (? juv.) generally
spring from a rounded foot plate about half a millimetre in diameter.

As holotype I have selected a specimen, Brit. Mus. Reg. No. 29.10.10.7, mounted
as a microscopical preparation. From a footplate arises one simple stem which con-
sists of six internodes. Another stem issues from the basal plate and remains attached
for a short distance to a specimen of S. confusus. It gives rise to a third simple stem,
consisting of ten internodes, and a fourth of seven. From below the distal hydrotheca

180

“TERRA NOVA” EXPEDITION.

on the fourth stem two short simple pi
a single gonotheca is developed.

sp. n. a, immature gonotheca ; b, hydro-
thecae. Both figures magnified X 55.

mae arise. From the penultimate hydrotheca

Symplectoscyphus tuba , sp. n. (Text-fig.

37, a-b.)

Material. — A small tuft, measuring one
inch square by half an inch, attached to
a polyzoan, from Station 90, off Three Kings
Islands, N.Z., 100 fathoms.

Description. — Stem monosiphonic, strag-
gling multipinnate. Internodes long and

Text-fig. 37. — Symplectoscyphus tuba, sp. n.
a, gonotheca, X 56 ; b, hydrotheca, X 45.

slender. Hydrothecae three-toothed, aperture inclined upwards, no internal teeth,
free part of adcauline wall equal to fused part. Gronothecae ovate with 11 or 12 well
marked annulations ; terminal tube central, flared, like the mouth-piece of a telephone.

Measurements :

Hydrotheca,

maximum length
maximum width
Gonotheca,
length

maximum diameter
diameter of external opening
,, internal ,,
number of annulations

min.

0-34

0-17

. M0
. 0-59

. 0-25
. 0-11
. 11-12

Remarks. — This species resembles Symplectoscyphus confusus, sp. n., but the
hydrothecae are a little smaller and the gonangia are shorter and have a different
terminal tube. The tuft itself is the holotype specimen.

HYDROIDA— TOTTON.

187

Symplectoscyphus vanhoeffeni, nom. n. (Text-fig. 38, a-d.)

Sertularella subdichotoma, Vanhoft’en, 1910, p. 326, fig. 41, a-e (not Sertularella subdichotoma,
Kirchenpauer, 1884, p. 46, pi. XVI, figs. 1-16) ; not Sertularia divaricata, Busk, 1852,
p. 388 ; not Sertularella cumberlandica, Jaderholm, 1905, p. 27, pi. X, figs. 8-11. The
synonymy of S. divaricata auct. (nec Busk) has not been set in order.

Material. — Tufts of specimens up to a length of 6 or 7 cm., with immature and
spent gonothecae, in association with those of S. glacialis on the main stems of Schizo-
tricha unifurcata, on polyzoa and on Terebellid worm tubes, from Station 220, off Cape
Adare, 45-50 fathoms ; some branched fragments 3 cm. in length, with immature
gonothecae from Station 339, McMurdo Sound, 140 fathoms ; a tangle of specimens
with immature gonothecae from Station 356, off Granite Harbour, 50 fathoms ; and a
tangle of specimens from Station 340, McMurdo Sound, 160 fathoms.

Description. — Stems non-fascicled, straggling, up to 6 or 7 cm. long, branching by
false dichotomy and anastomosing. Distal branches pinnate. Hydrocaulus tending

Text-fig. 38. — Symplectoscyphus vanhoeffeni, nom. n. a, hydrothecae, X 55 ; 6, terminal tubes of male
gonothecae, X 33 ; c, spent female gonotheca, X 33 ; d, distal end of spiral fold of gonotheca, X 32.

not to form distinct nodes. Hydrothecae usually a length apart, short, with about
half length free.

Gonothecae obovate, narrowed and tubular at distal end, with flared opening,
spirally folded walls forming a frill with about eight turns ; dimorphic, one, probably
the female, with larger mouth tube and deeper frills.

Measurements :

Hydrotheca,

length

diameter

Gonotheca,

length

diameter

tube,

length

diameter at margin
frill depth

0- 33
0-20

Male.

1- 63
0-75

0-15-0-18

0-15-0-16

0-13

Female.

mm.

0-30-0-37

0-25-0-28

0-18-0-23

v. 5.

8

188

“TERRA NOVA” EXPEDITION.

Remarks. — In a smaller variety from Granite Harbour, McMurdo Sound, the
liydrothecae are more distant.

Through the kindness of Dr. Thiel of the Hamburg Museum I have been able to
re-examine Kirchenpauer’s syntypes of S. subdichotoma. They have been previously
re-examined by Hartlaub and mentioned in his “ Revision der Sertularella-Arten.”
Hartlaub did not redescribe the original of Kirchenpauer’s figure of S. subdichotoma
which, in the absence of a selection by Kirchenpauer, I now select as the holotype ;
nor did he redescribe the syntypes. He gave a general description both of the syntypes
and of a quantity of material in the Hamburg Museum which he ascribed to Kirchen-
pauer’s species, at the same time expressing his unwillingness to decide whether the
Australian and S. American forms were the same. This is not the occasion to publish
a redescription of Kirchenpauer’s species, but it is to be recorded that I have since
identified by many fine details the roughly figured specimen from Australia, the holo-
type of the species, Hamburg Museum number C 4311. In S. subdichotoma the free
part of the hydrotheca is shorter than in S. vanhoeffeni, there are no free frills to the
gonotheca and the mouth of the gonotheca is on a very short and wide tube. The
species is distinct from the Antarctic one.

I have found at last the original specimen of Busk’s S. divaricata, of which sketches
are already known. I select this specimen, Brit. Mus. Reg. No. 99.7.1.6624, as the
holotype of the species. It is closely related to Coughtrey’s Thuiaria subarticulata,
but not to the Antarctic S. vanhoeffeni.

As to all the varieties ascribed by Bale and other authors to Busk’s species divari-
cata, until they have been classified and named, and a new description of S. divaricatus
is published, there will be much confusion. I do not think that any of them belongs to
the Antarctic species. Both Vanhoffen and I have noticed the association between
this species, S. vanhoeffeni, and S. glacialis.

It is remarkable that the collection contains no specimens with fascicled stems
that can be assigned to S. cumberlandica, Jaderholm, which is a closely related species.
I was able to make careful re-examination and sketches of the type slides of that species.

What we desire now is a careful comparison of the two species showing how to
differentiate between them. It is possible that they will turn out to be synonymous.

Symplectoscyphus glacialis (Jaderholm). (Pis. I, figs. 8, 9 ; II, fig. 7 ; text-fig. 39,
a-b .)

Sertularella glacialis, Jaderholm, 1904, p. 9 ; 1905, p. 26, pi. X, figs. 3-7 ; idem, Ritchie, 1913,
p. 29, fig. 10 ; idem, Billard, 1914, p. 23 ; idem, Jaderholm, 1917, p. IX, pi. I, fig. 9 ;
Symplectoscyphus glacialis, Stechow, 1922, p. 148 ; 1923, p. 174 ; Sertularella glacialis,
Jaderholm, 1926, p. 5 (not S. glacialis, Vanhoffen, 1910, p. 325 = S. plcctilis, Hickson and
Gravely).

Material. — Two small tufts up to 2 cm. in length from Station 316, McMurdo
Sound, 190-250 fathoms ; several tufts of multipinnate stems up to 6 cm. in length

HYDROIDA— TOTTON.

189

with gonothecae, growing on other hydroids, from Station 220, off Cape Adare, 45-50
fathoms ; and from Station 356, entrance to McMurdo Sound, 50 fathoms. A few
fragments with rather larger hydrothecae coming from Station 194 off Oates Land,
180-200 fathoms, may belong to this species.

Remarks. — It has been stated by Ritchie (1913) that there is little difference
between the trophosomes of Sertularella glaciaiis and of “ Sertularella subdichotoma ”
(= Symplectoscyphus vanhoeffeni, nom. n.). It is, indeed, often a matter of great diffi-
culty to determine specimens of these two species in bulk, especially when the two are
growing side by side. But while in the abcauline wall of the hydrothecae of S. van-
lioejjeni there is a pronounced re-entrant angle, which makes them gibbous below,

those of S. glaciaiis are generally without it, and in form are more nearly tubular.
Attention has already been drawn by Ritchie (1913) and Jaderholm (1905) to the
proportions of the “ free ” and “ adnate ” portions of the thecal walls. Half, or more
than half, of the hydrothecal wall in S. glaciaiis is said to be free. This is not constantly
so, for I find that the average length of the free part of 16 hydrothecae of a topotype
is distinctly less than that of the adnate part. Measurements of 41 hydrothecae on
different parts of a specimen from Station 220 give the same result. The average
measurements of 9 hvdrothecae on a fragment from Gauss Station give an inverse
proportion, however. I have tabulated measurements below. Table I gives, in
addition to new measurements, some of the figures previously published. Table II
gives an analysis of the measurements of one specimen. From this it will be seen that
the mean measurements of similar lengths from successive series in one specimen
decrease distalwards.

190

“TERRA NOVA” EXPEDITION.

Symjj lectoscyphu s glacialis , Jaderholm.

Table of Measurements I. Unit : 1 micron.

Nature of the material.

Intemode.

Hydrotheca.

Gonotheca.

Length.

, Diameter.

Length.

Diameter.

Length.

Diameter.

Diameter of
orifice.

Adnate

part.

Free

part.

Inter-

nal.

Exter-

nal.

1. Jaderholm’s type ( teste

minimum

440

290

770

370

Jaderholm).

maximum

600

370

1000

480

number

measured

mean

150

2. Small fragment (topo-

minimum

405

202

215

126

type) from Graham

maximum

886

265

265

151

Land. Swedish Ant-

number

arctic Exp., Brit. Mus.

measured

(19)

(16)

(16)

(.11)

Reg. No.

mean

607

151

254

239

144

481

430

(Immature)

3. Small specimen from

minimum

405

189

253

101

481

Gauss Station, German

maximum

734

240

278

506

Antarctic Exp., Brit.

number

Mus. Reg.No.11.12.8.6.

measured

(19)

(9)

(9)

(2)

mean

531

126

213

267

126

708

493

101

151

4. Specimens from Gauss

minimum

340

160

800

440

Station ( teste Van-

maximum

400

200

1000

620

hoffen).

number

measured

mean

500

5. Specimens from Cape

minimum

640

140

200

240

170

Royds, Br. Antarctic

maximum

1150

160

270

300

200

None

Exp., 1908 (teste

number

Ritchie).

measured

mean

895

150

235

270

185

6. Specimens from Graham

minimum

610

135

230

215

135

Land, “ Pourquoi Pas? ”

maximum

840

160

270

270

150

None

Exped., teste Billard.

7. Fragment, 9 mm. X

minimum

683

215

266

278

165

8 mm. Station 220,

maximum

405

158

177

190

139

“ Terra Nova ” Exp.,

number

Forma typica.

measured

(46)

(33)

(41)

(40)

(45)

mean

505

188

227

224

149

None

8. Fragment, Station 220,

minimum

329

1013

481

101

1 65

“ Terra Nova ” Exp.,

maximum

658

127

177

Forma typica.

number

measured

(32)

(2)

(1)

(2)

(2)

mean

484

Not measured

114

171

9. Fragment, Station 220,

minimum

633

203

228

“ Terra Nova ” Exp.,

maximum

1392

241

278

Forma elongata.

number

measured

(19)

(24)

(20)

mean

950

228

248

None

10. All recorded specimens

of “forma typica ” in-

eluding previously pub-

lished figures.

mean

525

155

256

243

150

797

472

108

161

11. “ Terra Nova ” and

“ Nimrod ” specimens

of “ forma elongata."

mean

972

150

231

259

185

HYDROIDA — TOTTON.

191

Symplectoscyphus glacialis, Jaderholm.

Table of Measurements II. Unit : 1 micron.

Internode.

Hydrotheca.

Specimen from Station 220 “Terra Nova” 1910
Expedition (9 mm. long).

Length.

Length.

Diameter.

Adnate

part.

Free part.

Diameter.

Main stem

minimum

430

176

203

203

146

maximum

683

203

266

278

164

number

15

7

11

11

12

mean

598

202

240

234

152

Branch No. 1,* penultimate series

minimum

506

176

228

203

152

maximum

506

203

241

215

152

number

2

3

2

2

3

mean

506

200

234

209

152

Branch No. 2, penultimate series (8 mm.

minimum

456

177

177

203

127

long).

maximum

633

215

241

253

177

number

13

7

12

12

14

mean

526

196

225

230

152

Branch No. 2a, ultimate series (3 mm.

minimum

456

177

177

190

127

long).

maximum

506

203

228

253

152

number

6

4

6

6

5

mean

468

190

213

222

149

Branch No. 2b, ultimate series (2 mm.

minimum

405

158

127

long).

maximum

456

184

—

—

152

number

2

3

0

0

3

mean

430

173

—

—

148

Branch No. 3, penultimate series (4 mm.

minimum

456

171

215

209

127

long).

maximum

532

203

241

253

152

number

7

7

8

7

8

mean

532

185

229

233

147

Branch No. 3a, ultimate series

minimum

506

164

203

203

127

maximum

506

177

241

228

152

number

1

2

2

2

2

mean

506

171

222

215

146

Main stem

mean

598

202

240

234

152

3 Branches penultimate series

mean

521

194

231

224

150

3 Branches ultimate series . .

mean

468

178

218

219

00

rH

Main stem and all branches

mean

529

191

230

226

150

* Branch No. 1 is proximal to Nos. 2 and 3. Secondary branch a is proximal to b.

192

‘TERRA NOVA” EXPEDITION.

Hy droid colonies appear to be able to adjust their growth-form to the environment
with rapidity, as more than one form may occur in different parts of one specimen.
The elongated form included in Table I appears to be the same as that recorded by
Ritchie (1913) from Cape Royds. The mean value * for the length of the internodes
is 0-972 mm. compared with 0-525 mm. for the typical form noted in Table I.

There has been some confusion in the past between S. plectilis, Hickson and
Gravely, and S. glacialis, Jaderholm. Having compared a specimen brought back by
the German Antarctic Expedition from Gauss Station, Brit. Mus. Reg. No. 11.12.8.6,
with a topotype of S. glacialis brought back by the Swedish Expedition from South
of Snow Hill, Graham Land, Brit. Mus. Reg. No. 19.12.10.1, I have no hesitation at
all in saying that although S. glacialis was collected at Gauss Station, the specimens
described by Vanhdffen (1910, p. 325) are not S. glacialis, but S. plectilis as suggested
by Ritchie (1913). f Vanhoffen’s suggestion (1910) that the type of S. plectilis is only
a specimen of S. glacialis is rejected, since re-examination of the type of S. plectilis
confirms all that has been said by Ritchie (1913) to show that it is a distinct species :
Vanhoffen appears to have confused the two. The gonothecae of the two, apart from
the hydrothecae, are quite distinct, as the text-figures will show. Vanhoffen’s figure 40
(1910) depicts a characteristic fragment of S. plectilis, but the orifice of the gonotheca
does not bear the distinctive marks of either S. glacialis or S. plectilis. It seems highly
improbable that the two can represent the two sexes of one and the same species.
Ritchie has even figured (1913) the gonothecae of the two sexes of S. plectilis, but with-
out histological evidence to support his statements. For the sake of convenience the
two described variations may be referred to in future as forma typica and forma elongata.
They do not appear to be very stable.

Symplectoscyphus curvatus (Jaderholm). (PL II, figs. 1,2,3; text-fig. 40, a-b.)

Sertularella curvata, Jaderholm, 1917, p. XI, pi. I, figs. 10-11.

Material. — A dozen pinnately branched specimens, up to 5-5 cm. in length, bearing
gonothecae on the branches ; hydrorhiza creeping up stems of Schizotricha unifurccita,
from Station 220, off Cape Adare, 45-50 fathoms.

Description. — There are three or four twists where the short bases of the stems
spring from the hydrorhiza. Main branches pinnate, in one plane, secondary branches
irregular. Gonothecae borne only on branches, large, elongate-ovate, with ten deep
independent circular frills not forming a spiral. Aperture borne on a short wide
cylindrical tube, hardly projecting beyond rim of distal frill.

Measurements :

mm.

Hydrorhiza, diameter . . . . . . . . . . . . . . 0-37

Base of stem,

length . . . . . . . . . . . . . . . . . . 0-5 -0-9

diameter . . . . . . . . . . . . . . . . . . 0-37

* This was obtained by combining figures for Ritchie’s material from Cape Royds and for “ Terra
Nova ” material.

I This is denied by Stechow (1923c).

HYDROIDA — TOTTON.

103

Internode,

length (over all)

diameter at base of hydrotheea

Hydrotheca,

length, adnatc . .

„ free

breadth
distance apart
Gonotheca,
length

breadth (over all)
mouth tube,
length
diameter

A serious discrepancy will be noticed
between the diameters of the gonothecae
of the Swedish and British specimens :
Jaderholm gives the diameter of gono-
thecae as 1-5 mm. The other measure-
ments correspond well enough.

Remarks. — This species was de-
scribed in a supplementary report on
the hydroicls of the Swedish Antarctic
Expedition published in the yearly
report of Norrkoping Higher School,
Sweden. Only two half-grown gono-
thecae were described. The only two
records of this species are from the
Graham Land region and the Ross
Sea.

mm.

0-73-1-10

044-0-56

0-30-0-46
0-56-0-66
0-29-0-34
M0-1 -59

1-83-1-96

0-98-0-10

0-075-0-10
0-15 -0-20

Text-fig. 40. — Symplectoscyphus curvatus (Jader-
holm). a, distal end of gonotheca, X 35 ; b,
hydrothecae, X 35.

Symplectoscy pirns plectilis (Hickson and Gravely). (PI. II, figs. 4, 5 ; text-fig. 41.)

Sertularella plectilis, Hickson and Gravely, 1907, p. 20, pi. Ill, fig. 21 ; Sertularella pleclilis,
Ritchie, 1913, p. 30, figs. 8-9 ; Sertularella glacialis, Vanhoffen, 1910, p. 325, fig. 40, a-c
(not S. glacialis, Jaderh., 1904, p. 9) ; Sertularella plectilis, Jaderholm, 1917, p. VIII,
pi. I, fig. 8 ; 1926, p. 5.

Material. — Inextricably tangled, non-fascicled, branched and anastomosing stems
with gonothecae from Station 316, McMurdo Sound, 190-250 fathoms ; and Stations
331, 339 and 356 at entrance to McMurdo Sound, 190-250, 140 and 50 fathoms. From
the last station come a few young stems.

Remarks. — Sound and critical notes on this species were given by Ritchie (1913).
A study of new material and of the types of this species and of S. glacialis enables me
to endorse Ritchie’s remarks. The adcauline marginal hydrothecal tooth in S. glacialis

194

“TERRA NOVA” EXPEDITION.

is generally recurved, whereas in S. plectilis it is only the extreme tip that is ever, if at
all, slightly recurved.

The hydrothecae of S. plectilis are, generally speaking, more distant than those
of S. glacialis, but specimens of each can be selected in which the reverse holds true.
The proportions of the neck and aperture of the gonothecae of the two species are
quite different, as pointed out by Ritchie.

The original description did not mention the few fully formed gonothecae present
on the types. Ritchie’s description and figures hold good for them.

The few specimens taken on Feb. 9 from Station 316 are light brown in colour

and bear a few fully formed gonothecae. Those
taken on Jan. 14 had numerous immature gono-
thecae. Ritchie recorded male gonothecae in
June, and female in July and August. Jader-
holm recorded male, but no female gonothecae on
Feb. 10.

Text-fig. 41. — Symplectoscyphus plec-
tilis (Hickson and Gravely). Hydro-
theca, X 160.

Text-fig. 42. — Symplectoscyphus Jiliformis (Allman).
a, male gonotheca ; b, female gonotheca. Both
magnified X 34.

Symplectoscyphus jiliformis (Allman). (PI. Ill, fig. 9 : text-fig. 42, a-b.)

Sertularia gracilis, Allman, 1888, p. 51, pi. XXIV, figs. 1-la (not S. gracilis, Hassal, 1884,
p. 2223) ; S. Jiliformis, Allman, 1888, Expl. pi. XXIV.

Material. — A tangled mass measuring 7 mm. x 7 mm. X 7 mm., with a few
gonothecae, from Station 38, S.W. of Falkland Islands, 125 fathoms.

Description. — The hydrothecae have one adaxial and two lateral marginal cusps
as in other species. Gonothecae sexually dimorphic, females longer, broader, usually
with one or two more annulations, with a longer and wider distal tube, and containing
two gonophores ; male containing one spermary.

HYDROIDA — TOTTOX.

195

Measurements (of types) :

Hydrothecae,

mm.

mm.

distance apart

*.

. 0-85-1-10

length of adasial wall,

adnate part

. 0-25-0-30

free part

. 0-17

Gonothecae,

Male.

Female.

length

1-40

1-98

(1-07-1-55)

(1-87-2-23)

diameter (maximum)

0-70

0-77

(0-62-0-76)

(0-69-0-84)

tube,

length

0-25

0-35

(0-17-0-25)

(0-32-0-37)

diameter at margin . .

0-11

0-24

(0-10-0-12)

(0-21-0-26)

Remarks. — The type specimens

(Challenger Expedition) were

taken on

1876, and at that date bore gonothecae containing ripe sperms and eggs, the discharge
of which had not then begun. In the specimen taken on April 13, 1912, the discharge
had taken place and most of the gonothecae had fallen off. A careful search revealed
a few empty ones of either sex.

Allman’s description of the male gonotheca is misleading. The “ short wide
neck ” needs no special mention, as it is merely the distal part of the gonotheca, not
differentiated in any way. The “ saucer-shaped summit ” is the last frilled annulation
which may be more or less marked. The narrow cylindrical tube does not broaden
out at its base as Allman represented it. The hydrothecae, too, were incorrectly
drawn.

Sertularella robusta, Coughtrey.

S. robusla, Coughtrey, 1876, p. 300 ; S. microgona, v. Lenclenfeld, 1881, p. 116, pi. VII, figs. 1-3 ;
S. angulosa, Bale, 1891, p. 102, pi. IV, fig. 6 ; S. robusta, Trebilcock, 1928, p. 16, pi. VI,
fig. 3 ; not Sertularia tcnella, Alder. 1856, p. 357, pi. XIII, figs. 3-6.

Material. — Fifteen small unbranched shoots, without gonothecae, from a stock
associated with Sertularia marginata, creeping over an alga from Station 134, near
North Cape, New Zealand, 11-20 fathoms.

Remarks. — The hydrothecal walls show distinct transverse undulations and
three internal denticles are present. None of the shoots bears more than five or six
hydrothecae, and in several cases one only is present.

This species, which was treated by Hartlaub (1 901) and Bedot (1912) as synonymous
with S. ienella (Alder), has been redescribed and figured with two new varieties by
Trebilcock (1928). It differs from S. tenella, as pointed out by Hartlaub (1901) and
Bale (1924), in having three internal hydrothecal denticles, and also in its smaller
size. Comparison with v. Lendenfeld’s types of S. microgona in the British Museum
shows that S. robusta and S. microgona are identical,
v. 5.

9

196

“TERRA NOVA” EXPEDITION.

According to Trebilcock (1928), who seems to have had a plentiful supply of
material for study, this species is a very variable one in size, degree of transverse
undulation of thecal walls, branching, shape of distal part of gonothecae and distance
apart of hydrothecae. In S. simplex too, with which it may be confused, he remarks
on the variation in size, branching, shape of distal part of gonothecae, distance apart
of hydrothecae and their position on hydrocaulus. The only character that seems to
separate the two species, as far as our present knowledge goes, is transverse undulation
of hydrothecal walls ; and even this is “ often rather faint.” So it still seems possible
that after all S. simplex and S. robusta are only varieties of one species. Further
studies no doubt will clear up this point.

Sertularella geodiae, sp. n. (PI. Ill, figs. 7, 8 ; text-fig. 43.)

Material. — Two specimens, the largest 5 inches in height, from Station 90, off

Three Kings Islands, N.Z., 100 fathoms ; and two
more from Station 96, off North Cape, N.Z., 70
fathoms, the largest of which is 6 inches in height and
has been chosen as the Holotype.

Description. — A species with the habitus of S. gayi,
Lx. Hooting masses large. Stems and branches
polysiphonic and well defined. Stems up to 5 mm.
in diameter. Hydrothecae similar to those of S. gayi,
but larger than British specimens. Gonothecae resem-
bling those of S. polyzonias except that they generally
have only three blunt apical teeth, occasional branches
bear gonothecae with four teeth.

Remarks. — This appears to be a form new to New
Zealand or Australia, and seems to be distinct from
S. polyzonias and S. gayi, combining some of the
characters of both. The specimens from Station 96
were evidently growing on a stratum of dissociated

Geodine sponge spicules, many of which are attached

Text-fig. 43. — Sertularella qeodiae, , , • mi , r x o i

tj , ., to the rooting masses. I he stems bear a few Scal-

sp. n. Hydrothecae, X 34. _ ®

pellids, Foraminifera, Ephyropsid scyphostomes —
“ Stephanoscyphus ” — and specimens of a species of Lafoea.

Measurements :

Hydrotheca,
length,

abaxial wall
adaxial wall, free part
,, „ fused part

diameter at orifice

mm.

. . 0-60-0-74
. . 0-47-0-55
. . 0-68-0-76
. . 0-38-0-40

HYDE OIDA— TOTTON.

197

Internode,

length,

branch

hydrocaulus

diameter,

branch

hydrocaulus

Gonotheca,

length

diameter (maximum)

mm.

1-22

1- 28-1 *39
0-31

0-33-0-39

2- 0 -2-04
0-89-1-03

Sertularella spiralis, Hickson and Gravely. (Pis. I, fig. 7 ; III, fig. 5 ; text-fig. 44,

b-d.)

Sertularella spiralis, Hickson and Gravely, 1907, p. 19, pi. Ill, figs. 19-20 ; S. bifurca, Billard,
1914, p. 22, fig. 13 ; IS. articulata, Allman, 1888, p. 61, pi. XXIX, figs. 3-3a ; ? S. elongata,
Jaderholm, 1904, p. X.

Material. — Unattached branching specimens up to nine centimetres in length and
four centimetres in diameter without gonothecae, attached to stems of Polyplumaria
antarctica (Jaderh.) from Station 220, off Cape Adare, 45-50 fathoms ; also broken
specimens from Station 338, McMurdo Sound, 207 fathoms. From Station 194 off
Oates Land, 180-200 fathoms, there are two characteristic fragments of stems.

Description. — The stems remain non-fascicled through life. The internodes of the
main stem and branches are progressively hypertrophied, at the base of a specimen
measuring 0-95 mm. x 0-21 mm., and at the distal end 4-23 mm. X 0-48 mm. The
stem forms a gentle spiral, one turn in a length of ten, twenty, thirty or even more
internodes. The paired branches may rebranch as many as six times, at every third
internode, alternately to right and left ; but occasionally all the branchlets form on
one side, giving rise to a scorpioid sympodium. The ends of branches freely stolonise.
In case of a break any branch may become hypertrophied and behave as the main
stem, with elongated internodes and paired branches below each newly formed hydro-
theca. The hydrothecae are generally inclined to one side of branches and pinnules,
and have triangular mouth openings with a well-marked intrathecal tooth in the
shallow embayment between each tooth.

The polyp has an oblique annular attachment which, viewed in optical section from
the side, runs from a point about or a little below the middle of the abaxial thecal wall,
where there is generally a slight re-entrant angle, to the inner angle of the floor of the
theca. In extension that part of the pol)q) within the theca is narrow and tubular,
spreading out below to its attachment, and with the curious ectodermal sheath noted
by Hickson. The cavity of the polyp appears not to be continuous with that of the
hydrocaulus, but there is a smal] excentric fibrillar connection from the polyp base
through the foramen at the outer side of the thecal floor to the stem apophysis. In
retraction the polyp folds in the form of the letter u S.” Hickson’s original figure 19
does not adequately represent the arrangement. There were no gonothecae on these

198

“TERRA NOVA” EXPEDITION.

specimens collected on January 3rd and 23rd respectively, but those of the £i Discovery ”
types taken on July 3rd bore acrocysts. The gonothecae are smooth-walled, elongate-
ovate in shape, with tapered base but no distinct pedicel, and rounded apical end.
There is a thickened margin to the aperture but no collar or projection. Acrocysts are
formed as described by Hickson.

Text-fig. 44. — a, Sertularella articulata, Allman. Schizoholotype (Brit. Mus. Reg. No. 88.11.13.49) ;
b, c, d, Sertularella spiralis, Hickson and Gravely, showing internal hydrothecal teeth. All figures
magnified X 40.

Measurements :

Hydrotheca,

length,

abaxial wall . .
adaxial wall, free part
,, „ fused part

diameter at orifice

Internode,

length,

branch

hydrocaulus

diameter,

branch

hydrocaulus

Gonotheca, (? $) of types,
length

diameter (maximum)
diameter of aperture

mm.

0-40-0-42

0-21-0-26

0-29

0-14-0-15

0-75

0-95-4-23

0-12-0-18

0- 21-0-48

1- 13

0-40-0-45

0-17

Remarks. — The original description does not mention the internal hydrothecal
teeth, which are present in the types and indicated in the original figures. They are
alluded to by Billard (1914) in his description of S. bifur ca which appears to be a
synonym.

HYDROIDA — TOTTOX.

199

This species appears to be related to S. neglecta. Thompson, and its associated
group of species.

The " Terra Nova ” specimens were taken farther north in the Ross Sea than the
" Discovery ” types, and oh Oates Land. The £' Pourquoi Pas ? ” specimens came
from oh the coast of Graham Land. Its further distribution is for the present doubtful.

Jaderholm’s S. elongata , 1904, is very close to if not identical with this species.
The habitus of the two is the same, and the only known differential character of
elongata is that it has no internal liydrothecal teeth. In 1905 Jaderholm sank
S. elongata in the synonymy of S. articulata, Allman, 1888. Allman’s fragmentary
specimen (text-fig. 44, a) is probably a small branch ; and until the species has been
taken again at Kerguelen and described it would be best not to use the name. It was
badly figured and described.

Jaderholm did not say whether the hydrothecae of S. elongata were bilateral as
in articulata or inclined to one side as in spiralis. From “ Terra Nova ” Station 339,
McMurdo Sound, 140 fathoms, come some pieces of stem with branches that corre-
spond with Jaderholm's description. The hydrothecae are bilateral and have no
internal teeth, and the stem internodes are longer than in typical S. spiralis. How-
ever, there are specimens that may be considered intermediate between those that
I identify with Jaderholm’s S. elongata and S. spiralis. They come from Station 338,
off the entrance to McMurdo Sound, 207 fathoms. The hvdrothecae are bilateral, but
internal teeth are present.

It is possible, therefore, that S. spiralis sometimes grows without internal hydro-
thecal teeth. We already know of anomalies concerning these teeth in other species.

Sertularella elongata, Jaderholm.

Sertularella elongata, Jaderholm, 1904, p. X ; S. articulata, Vanhoft’en, 1910, p. 328, fig. 42, a-d ;

? Sertularia articulata, Allman, 1888, p. 61, pi. XXIX, figs. 3-3« ; l Sertularella spiralis,
Hickson, 1907, p. 19, pi. Ill, figs. 19-20.

Material. — Fragments of branched stems without gonothecae, from Station 339,
McMurdo Sound, 140 fathoms.

Remarks. — This species has been dealt with under S. spiralis, Hickson, above.
If it can be shown that the two species are really one, the name elongata will have
priority. This may eventually prove to be a synonym of articulata, Allman, but fresh
material from Kerguelen must first be studied. Vanhoffen’s specimens had hydrothecae
which were entirely free, a characteristic of S. biformis, Jaderh., but in that species the
hydrothecae are unilateral.

Sertularella biformis, Jaderholm. (PI. II, fig. 8 ; text-fig. 45.)

Sertularella biformis, Jaderholm, 1905, p. 28, pi. XI, figs. 1-3.

Material. — The apical three centimetres of two stems, without gonothecae, with
close-set spreading branches giving the specimens a diameter of 2-5 cm., from Station
220, off Cape Adare, 45-50 fathoms.

200

“TERRA NOVA” EXPEDITION.

Description. — Hydrocaulus a non-fascicled sympodium of short, stout internodes
with alternate hydrothecae. From the base of each hydrotheca arises a pair of

branches which bifurcate three or four times. The stem is thus
surrounded by whorls of branchlets. Hydrothecae unilateral,
to a large extent if not entirely free from internodes.
Measurements :

Stem internode,
length
diameter
Hydrotheca,
length

diameter (maximum)

mm.

2-45

0-98-1-0

0-76-0-83

0-27

rella biformis, Jader-
holm. Hydrothecae,
X 34.

Remarks. — There seems little doubt that these specimens
would correspond with the apical part missing from Jaderholm’s
type from Graham Land. If so this species shows an interesting
series of growth-stages, progressing in complexity from base to
apex. In the type the paired lateral buds arise from points
slightly distal to the stem hydrothecae, but further growth makes the original hydro-
thecae appear to lie between the paired branches. The mode of growth in many ways
resembles that of S. elongata and S. spiralis, which form part of the “ neglecta ” group
of Sertularella species.

Sertularella edentula, Bale. (PI. Ill, fig. 6 ; text-fig. 4G.)
Sertularella edentula. Bale, 1924, p. 237, fig. 6.

Text-fig. 46. — Sertularella edentula,
Bale. Hydrotheca showing lateral
folds round hydropore, X 54.

Material. — A small unbranched fragment 8 mm.
in length without gonothecae ; from Station 90, off
Three Kings Islands, N. Zealand, 100 fathoms.

Description. — Hydrothecae projecting slightly
from hydrocaulus, inoperculate, margins entire.
Floor of theca rising slightly from “ arris 55 to
abcauline side, with two lateral dependent folds
flanking the hydropore.

Remarks. — This fragment is of interest because
although it is without doubt to be referred to Bale’s
S. edentula it varies from the original in having
slightly projecting hydrothecae.

The dependent folds flanking the hydropore
are indicated in Bale’s figure, but received no
mention. They form a readily recognisable specific
character.

HYDROIDA — TOTTOX.

201

Genus DICTYOCLADIUM, Allman.

Genotype. — D. dickotomum, Allman, 1888, p. 77, pi. XXXYI, figs. 2, 2a.

Dictyocladium, Allman, 1888, p. 76 ; Dictyocladium, Stecliovr, 1923, p. 169 ; not Selaginopsis,
Allman, 1876, p. 272.

Billard and Bale both assign species of Dictyocladium to the genus Selaginopsis.
The hydrothecae of the holotype of the genotype of Selaginopsis, S. allmani , Norman,
do not have the tripartite operculum to be found in Dictyocladium and Symplectoscyphus.
In S. allmani there are two large adcauline lateral flaps and possibly a very narrow
abcauline flap. The habitus of this species is not like that of species of Dictyocladium,
and there seems no reason for uniting the genera. In both genera the hydrothecae,
which are really on opposite sides of the hydrocaulus, are alternately staggered to
£i front ” and “ back/’ a fact which has led some observers to say that the hydrothecae
are in several rows. The original arrangement is best seen near the growing points of
the branches. As growth proceeds this arrangement of hydrothecae is masked.

Bale (1915) rejects the genus Dictyocladium because he considers that it is not
sufficiently characterised. Although its species have affinities with Symplectosciypkus
spp. I think that they are so distinct that it is convenient to retain the generic name
Dictyocladium for them. The distinctive features in this genus are the shape and
situation of the gonothecae, and the twist of the axis between successive pairs of hydro-
thecae alternately to right and left, so that only those of every other pair are in align-
ment and there appear to be four longitudinal rows. There does not appear to be any
subsequent interpolation of new rows of hydrothecae. Parallelisms occur in other genera.

The genotype is D. dickotomum, Allman. S. reticulata, Krpr., 1884, is obviously
referable to the genus, and may be conspecific.

Nutting’s D. flabellum should be re-examined. It is said to have a four-flapped
operculum, but otherwise would be included in the genus. As Bale has pointed out
D.fuscum, Hickson and Gravely, must be excluded : it should be assigned to Staurotkeca.

For the New Zealand species there have been used the names monilifera, Hutton,
1873, and cerastium (Alim., 1876). Bale (1924) maintains that these are synonyms
of dichotomum, Alim., but further study of the forms is required before this view can
be accepted. The gonothecae of both Australian and New Zealand specimens are of
two lands, and the old problem arises whether we are dealing with two species or with
a single sexually dimorphic one. In one kind, which is typical of dickotomum, the
spiral fold of the wall makes about eight turns, is deep, and nearly conceals the terminal
cylindrical tube, the penultimate coil making a relatively wide turn ; whereas in the
other kind, typical of cerastium, the spiral fold is shallow and makes from 17-21 turns,
the last two being of small diameter and not concealing the flared terminal tube. The
first kind is generally more robust and of greater proportions.

From the Australian region the British Museum collection contains, besides the
types of dickotomum, and another specimen of this species from off South Cape,

202

“TERRA NOVA” EXPEDITION.

Tasmania, presented by Bale, some fine dried specimens of the other ( cerastium ) form
collected by Flower in 1850. From New Zealand, beside the types of cerastium from
North Island and other cerastium forms from Palliser Bay, Wairarapa, collected by Miss
Rye in 1890, it contains specimens of the dichotomum form from Wellington. The two
forms do not appear to have been taken together, but very little material has been collected.

I shall follow Bale in regarding cerastium, Alim., as a synonym of moniliferum,
Hutton, but I think that it is premature to state that dichotomum, Allman, is only one
sex of moniliferum. It is to this latter form that the “ Terra Nova ” specimens belong.

Dictyocladium moniliferum, Hutton. (Text-fig. 47, a-b .)

Dictyocladium moniliferum, Hutton, 1873, p. 257 ; Thuiaria cerastium, Allman, 1874, p. 179 ;
idem, Allman, 1876, p. 271, pi. XVIII, figs. 3-4 ; Dictyocladium dichotomum, Allman, 1888,
p. 77, pi. XXXVI, fig. 2 ; Selaginopsis dichotoma, Bale, 1915, p. 266 ; S. monilifera,
Bale, 1924, p. 237 ; idem, Trebilcock, 1928, p. 9, pi. IV, fig. 4.

Material. — Four small fertile fragments, each about one square inch in size, from
Station 134, off North Cape, New Zealand, 11-20 fathoms, and a sterile fragment from

Station 90, off Three Kings Islands,
100 fathoms.

Description. — Specimens are said
to branch in a truly dichotomous
fashion, but it appears to be a false
dichotomy, the branch being given
off at an angle rather more acute
than that at which the main axis
diverges at that point. It is the
branch that bears at its base on the
inner side a rounded tubercle from
which later the gonotheca develops.
On occasions a free hydrotheca de-
velops from this tubercle.

In specimens from Palliser Bay
the gonothecae are sometimes de-
veloped in addition from hydro-
thecae. On younger parts of branches, where reduplication of hydrothecal margins
has not yet taken place, the original shape of the hydrotheca margin can be seen
well. It is just like that found in the genus Symplectoscyphus.

Measurements :

Hydrothecae, mm.

distance apart . . . . . • • • • • • • • • • • 0-75-0-8

length, adnate part . . . . • • • • • • • • • • 0-37

„ free part . . . . . . • . • • • • • • • • 042

,, (with reduplications) . . . . . . . . . . . . 0-25-0-3

diameter . . . . . . . • • • • • • • • • • • 0-12

Text-fig. 47. — Dictyocladium moniliferum, Hutton, a,
specimen from Station 90, X 53 ; b, specimen from
Station 134, showing origin of branch and seat of gono-
theca, g, X 39.

HYDR01DA — TOTTOX.

203

Gonotheca,

mm.

length

2-0-2-65

diameter

1-1-1-45

No. of turns of spiral fold

17-21

height of fold

0-07

terminal tube.

length

0-91

diameter, pros, end . .

0-2

„ distal end . .

. . 0-35

Sertularia unguicuktta, Busk. (Text-fig. 48, a.)

Sertulana unguiculata, Busk, 1852, p. 394 ; \ S. divergens, Busk, 1852, p. 392 ( =Tridentata
xantha, Stechow, 1923, p. 12) ; not S. divergens, Lx., 1816 ( —S . bicuspidata, Lk., 1816, II,
p. 121) ; Dynamena australis, Kirchenpauer, 1864, p. 11 ; Thuiaria ambigua, Thompson,
1879, p. Ill, pi. XIX, figs. 2-2a ; ? Sertularia australis, Thompson, 1879, p. 105, pi. XVII,
figs. 4— 4a (not S. australis, Bale, 1887, p. 93 = Desmoscyphus inflatus, Versluys, 1899,
p. 42) ; Desmoscyphus unguiculala, Allman, 1885, p. 144, pi. XVII, figs. 5-7 ; ? Thuiaria
heteromorpha, Allman, 1885, p. 147, pi. XX, figs. 1-5 ; Desmoscyphus pectinatus, Allman,
1888, p. 71, pi. XXXIV, figs. 1, la-6 (in part) ; Sertularia challenged, Nutting, 1904, p. 54,
pi. II, figs. 1-2 ; Sertularia challengeri, Billard, 1910, p. 19, fig. 6 ; Amphisbetia ungui-
culala, Stechow, 1921, p. 258 ; ? Tridcntala xantha, Stechow, 1923, p. 12.

Material. — A group of fifteen stems about 7 cm. in height growing on a sponge,
from Station 144, off Cape Maria van Diemen, N.Z., 35-40 fathoms ; and a group of
a dozen stems 3-5 cm. in height, from Station 134, off North Cape, New Zealand,
11-20 fathoms. Gonothecae absent.

Remarks. — The specimens from Station 144 closely resemble Busk’s type material
in every way. Others from Station 134 and 144 are more slender in form, have more
distant pinnae and more widely separated hydrothecae, in which the adnate portion
is very much shorter than the free part. In typical specimens the adnate and free
parts of the hydrothecae are equal in length. An examination of a fertile topotype
of D. australis named by Kirchenpauer, Brit. Mus. Reg. No. 90.8.27.15, enables me to
confirm Bale’s opinion that that species is synonymous. The holotype of Thuiaria
heteromorpha, Allman, Brit. Mus. Reg. No. 27.1.1.7, is a typical S. unguiculata, and
cannot be distinguished by any peculiarities of the hydrotheca. Allman’s enlarged
figures of this species are misleading.

Tridentata xantha, Stechow (=S. divergens, Busk, non S. divergens, Lx.), appears to
be a smaller form of the same species.

D’Arcy Thompson (1879) described several forms closely related to S. unguiculata,
namely S. australis, S. Jlosculus, S. pulchella and S. flexilis. S. Jlosculus is almost
certainly a synonym of S. marginata (Krpr.), and S. australis a synonym of
S. unguiculata. Further study of this most variable of species is desirable before its
synonymy can be finally settled.

V. 5.

10

204

“TERRA NOVA” EXPEDITION.

Sertularia marginata (Kirclienpauer.) (Text-fig. 48, b.)

Dynamena marginata , Kirclienpauer, 1864, p. 13, fig. 8 ; Sertularia floscuius, Thompson, 1879,
p. 104, pi. XVII, fig. 2 ; Desmoscyphus gracilis, Allman, 1888, p. 71, pi. XXXIV, figs 2,
2a-c ; Desmoscyphus pectinalus, Allman, 1888, p. 71 (part) ; Sertularia versluysi, Nutting,
1904, p. 53 (part) ; Sertularia pluma, Hartlaub, 1905, p. 661 ; Sertularia versluysi, Ritchie,
1907, p. 505, fig. 144; Sertularia marginata, Bale, 1913 (part), p. 125, pi. XII, fig. 9;
pi. XXIV, figs. 2-6 ; Amphisbelia marginata, Stechow, 1921, p. 258 ; not Sertularia
marginata, Allman, 1877, p. 23 (= S. striata, nom. nov.) ; not Sertularia amplectens,
Allman, 1885, p. 141, pi. XVI, figs. 3-4 ; not Desmoscyphus inflatus, Versluys. 1889, p. 42,
figs. 11-13.

M aterial — Numerous pinnate and simple specimens up to 2-5 cm. in height on a
shell and organisms growing thereon, also two dozen small unbranched specimens
from 7-8 mm. in height without gonothecae, associated with Sertularella robusta, on
an alga, all from Station 134, off North Cape, Newr Zealand, 11-20 fathoms.

Text-fig. 48. — a, Sertularia unguiculata, Busk, X 54 ; b, Sertularia marginata, Kirchenpauer, x 33.

Description. — Between the hinge- joint at the base of the stem and the lowermost
branch there are from one to five pairs of hydrothecae. They are opposite below, but
become alternate above. At the distal end of specimens, beyond the last pinna, the
hydrothecae are arranged in pairs again, as on the pinnae. Intermediately the arrange-
ment is a pinna and three hydrothecae to each internode. On the simple stems there
is no displacement of members of a pair of hydrothecae.

Remarks. — An analysis of the hydrothecal arrangement indicates that the nodes
are determined by the position of the pinnae. A subopposite or alternate pair of
hydrothecae, from the bases of which no pinna arises, alternates with two pairs
of hydrothecae from whose bases pinnae do arise, one from each pair alternately

HYDROIDA — TOTTON.

205

left and right. It thus happens that the nodes separate the members of every third
pair of hyclrothecae, which have become displaced and alternate.

This hypothesis enables us to link up forms whose stem articles bear three liydro-
thecae with the more simple and primitive forms like Sertularia desmoides, Torrey, in
which no displacement of members of pairs of stem hydrothecae takes place, and in
which the branching arrangement of pinnae has not become regular.

The specimens are scarcely distinguishable from the types of D. gracilis, Allman,
from Bermuda, and closely resemble Kirchenpauer's figured holotype of the species,
Hamburg Museum, C. 3910 ; but differ from the types of S. amplectens, Allman, which
Bale (1913) cites as a synonym. In S. amplectens the proximal part of the stem up to
the centre of the hinge-joint is from 0-45-0-7 mm. in length, whilst in D. gracilis
and the “ Terra Nova ” specimens it is from 2-5-3-G mm. In S. amplectens it is
rare to find any hydrothecae on the stem below the proximal pinna, and then only one
pair. In S. marginata it is usual to find many pairs. In size S. amplectens is smaller
in every way, stems often being less than 1 cm. in height, although Nutting (1904)
assigns to it specimens as tall as 2 inches taken on Gulf Weed by “Albatross.” It
seems doubtful whether S. amplectens should be regarded as a synonym of S. marginata.
The holotype of I). inflatus, Versluys, is very like S. amplectens, except that the basal
part of the stem is stated to be 4 mm. long. In S. versluysi, Bitchie (1907), from Cape
Verde Islands this part of the stem was from 3 mm. to 10 mm. long. S. versluysi,
Nutting (1904) taken on Gulf Weed by “Albatross,” must certainly be the same species
as S. amplectens, Allman, but it is doubtful whether it is conspecific with D. gracilis,
Allman.

The syntypes of Desmoscyphus pectinatus, Allman, 1888, belong to two species.
Those from Bahia are to be assigned to S. marginata, Kirchenpauer. They bear gono-
thecae like those described by Ritchie (1907) for S. versluysi, and, like them, are male.
They differ in having no stalk. There is an additional ridge nearer the base, and the
ridges are generally produced into slightly upstanding free membranous edges. There
is no trace of the toothed slit-like aperture. There are two gonophores, one larger
than the other, consisting of central column and spermary, but no blastostyle remains.
The state of preservation is not good. Kirchenpauer’s MS. name pluma has found its
way into literature (Hartlaub, 1905). Kirchenpauer later changed the word to margi-
nata. The specimen Brit. Mus. Reg. No. 90.8.27.17 bearing Kirchenpauer’s MS. name
pluma is part of the holotype of his D. marginata.

Nomenclature. — Synonymies for the many allied forms, both simple and pinnate,
have been suggested by Billard (1910) and amended by Bale (1913), who recognises as
distinct two simple forms 8. turbinata, Lx. (including S. loculosa , Busk), and a species to
which he gives the homonym S. loculosa, Bale. Although S. loculosa, Busk, for the
present may be regarded as a synonym of S. turbinata, Lx., it may be necessary to
separate the two later on. For this reason the homonym S. loculosa, Bale, 1913, cannot
be retained. I give to Bale’s species the new name Sertularia restricta (nom. n,=

206

“TERRA NOVA” EXPEDITION.

Sertularia loculosa, Bale, 1913, nec Busk, 1852). For the pinnate forms of this group
Bale uses Kirchenpauer’s specific name marginata. This necessitates the institution
of a new name for the little known Sertularia marginata, Allman, 1877, which I rename
Sertularia striata , nom. n. ( =Sertularia marginata, Allman, 1877, nec Kirchenpauer,
1864). .

Crateriiheca zelandica (Gray). (PI. Ill, figs. 1, 2, 3 ; text-fig. 49.)

Thuiaria zelandica, Gray, 1843, p. 294 ; T. dolichocarpa, Allman, 1876, p. 270, pi. XIX, figs. 3,
4-4a ; T. zelandica, var. valida, Quelcli, 1883, p. 249 ; T. hippisleyana, Allman, 1885, p. 146,
pi. XIX, figs. 1-3 ; T. zelandica, Farquliar, 1896, p. 464 ; T. zelandica, Billard, 1910, p. 15 ;
Stereotheca zelandica, Bale, 1924, p. 251.

Material. — A specimen 13-5 cm. in height, torn away from its root stock, bearing
ten gonothecae of the female type, from Station 90, off Three Kings Islands, New
Zealand, 100 fathoms ; and two sterile specimens from Station 144, oh Cape Maria
van Diemen, N. Zealand, 35-40 fathoms. Of these two one is 12 cm. in height, com-
plete with small footplate ; it bears on its lower half, which is free from pinnae and
stem hydrothecae, three incomplete branches. The other is 7 cm. in height and
incomplete. The spread of the pinnae may be as much as 4 cm.

Text-fig. 49. — Crateriiheca zelandica (Gray). Distal end of hydrotheca, showing internal folds,/,

X 155.

Description. — The margins of the stem hydrothecae and of the proximal thecae
of the branches are entire. Passing from the base to the tip of a branch one sees
a gradual change from an entire margin, through a sinuous form to the definitely
toothed one.

The gonothecae are from 7-8 mm. in length, tubular, rather less than 1 mm.
in diameter in the middle and tapering gradually towards each end, with a flattened,
truncated slightly recurved orifice about 1 mm. wide. One side is notched, and has
below the notch on the inside a projection. The notch and projection resemble the
adcauline sinus and projection of the hydrotheca. The single gonotheca of the “ Terra

HYDROIDA — TOTTON.

207

Nova ' material contains six eggs, and three more appear to have been extruded.
Some dry specimens in the Museum collection from Palliser Bay bear many gonothecae
of a different type, presumably male. They are of the same diameter as the female
ones, about 3*5 mm. in length and with obovate terminal ends, in each of which there
is a small circular pore flush with the surface.

Remarks. — There is evidence to show that Gray's, Allman’s and Quelch's types,
which are now in the British Museum, were all collected by Dr. Andrew Sinclair in
North Island, Hew Zealand. In my opinion they should not be separated, but should
all bear Gray’s specific name. Mrs. Eden {nee Gatty) recently presented the holotype
of T. hippisleyana, Allman, to Sheffield Museum, whence it has been transferred to
the British Museum, together with a number of other type specimens of Hydroida
described by Allman at the same time. This species of Allman does not differ appreci-
ably from T. zelandica, Gray. The early figures of Allman are not reliable.

Bale has used for this form the generic name Stereotheca , Stechow (genotype Sertu-
laria elongata, Lx.). In my opinion these two species of Lamouroux and Gray have
no affinities. There is no significance in the fact that the margins of the hydrothecae
in both are toothed, for the arrangement of teeth in the one is entirely different from
that in the other. It is true that in both forms the fully developed hydrothecae are
inoperculate.

Neither Gray ’s nor Quelch’s types nor Allman’s of T. liippisleyana bear gonothecae,
but the holotype of T. dolichocarpa does, as also do some small specimens in the British
Museum Collection from Palliser Bay, Wairarapa, New Zealand, and the “ Terra
Nova ” material. Allman appears to have been misled by irregular growth marks on
some of the gonothecae into thinking that they had narrowed tubular distal ends. In
his figure he has wrongly reconstructed the dried terminal part, which he represents
as having a circular aperture. Re-examination shows that the terminal parts of these
same gonothecae agree with those of the “ Terra Nova ” material.

I have examined the holotype of Pericladium novae-zelandiae, Thompson, 1879
(genotype of Crateritheca, Stechow, 1921), and consider it to be congeneric with this
species of Gray. Its hydrothecae are smaller and arranged in many rows on stem
and branches.

Gray’s species in habitus resembles D. pinaster, and in its longitudinal ridges
recalls I). scalar ifor mis , Kirkpatrick. Crateritheca and Diphasia appear to have diverged
from a common ancestral stock.

208 “TERRA NOVA” EXPEDITION.

PLUMULARIIDAE.

Osivaldella bifurca (Hartlaub). (Text-fig. 50.)

Schizotricha bifurca, Hartlaub, 1904, p. 16, pi. Ill, figs. 4-8 ; Osivaldella bifurca, Stechow, 1919,
p. 40 (32).

Material. — Three rooted tufts of stems with hydrocladia, bearing mature male
gonothecae, from Station 340, entrance to McMurdo Sound, 160 fathoms ; also a

rooted tuft of six stems, with gonothecae, 24 cm. in
length, from Station 339, McMurdo Sound, 140
fathoms.

Description. — Stem nodes may not be marked at
all. Hydrocladia homomerous, either unbranched or
with a single branch which is usually of same size
as hydroclade : small intermediate articles present
following fractures. Hydrothecae of hydroclade and
branch not strictly opposed, but facing slightly
upwards.

Two axial nematoplioral foramina present on upper
surface of base of apophysis. A large median foramen
behind margin of each hydrotheca, and a small, shallow,
scoop-shaped infrathecal nematotheca on a process of
each hydrothecate article except the basal one : here the process is not at all or only
very slightly marked.

Gonothecae (male) without nematothecae, compressed, bean-shaped, without
pedicel ; vertical adaxial side becoming concave distally, where there is a subterminal,
oval, thin area or operculum ; abaxial wall first horizontal, then turning up parallel
with other wall ; distal end broadly rounded ; origin in median line, from large area
below infra-thecal nematotheca.

Measurements :

Stem, mm.

diameter . . . . . . . . . . . . . . . . . . 0-69-0-83

Branch internode,
length *
breadth
Hydrotheca,

length (abaxial wall;
breadth at margin,
dorso-ventral
latero-lateral
breadth at base,
dorso-ventral
latero-lateral

* The length of proximal internodes is commonly between T25 mm. and 1-30 mm. Distally they
often measure only 1-05 mm. in length.

0-13-0-18

0-15-0-18

0-20

0-22

0-10

0-20

l-J

n

Text-fig. 50. — Osivaldella bifurca
(Hartlaub). x 23.

HYDE OEDA— TOTTON.

209

Gonotheca, male,
length
breadth
thickness . .
opercular area

mm.

. . 1-54-1-83

. . 0-51-0-59

. . 0-30-0-37

. . 0-30 x 0-32

Remarks. — This species, the genotype of Oswaldella, Stechow (1919), has been
recorded only once before. It was taken by “ Belgica ” on Nov. 29, 1898, in 459
metres at 70° 20' S., 83° 23' W. The gonothecae were at that time spent.

Some important modifications of and additions to Hartlanb’s original description
are given above. It is distinguished at once from the related species O. antarctica
(Jaderholm) by the once-branched hydrocladia, absence of infrathecal nematotheca
and process from proximal hydrothecate article of each hydroclade ; and in many
other ways.

Osualdella antarctica (Jaderholm). (Text-fig. 51.)

Schizotricha antarctica, Jaderholm, 1901, p. 12 ; 1905, p. 35, pi. XIV, figs. 6-8 ; Schizolricha
antarctica, Billard, 1906, p. 11 ; Schizotricha antarctica, Vankoft’en, 1910, p. 336, fig. 18 ;
Polyplumaria antarctica, Billard, 1911, p. 28, fig. 17 ; Oswaldella antarctica, Stechow,
1919, p. 10 (32) ; Polyplumaria antarctica, Jaderholm, 1926, p. 6.

Material. — Two tufts of from one to two dozen stems up to 350 mm. in height
with immature gonothecae, from Station 220, off Cape Aclare, Ross Sea, 45-50 fathoms.

Text-fig. 51. — Oswaldella antarctica { Jaderholm). x 39.

Description. — From the bases of some of the older stems other younger stems
arise, with the result that the stems form tufts. Both the hydroclade and its first
branch, which diverge from one another, rebranch unilaterally two or three times ;
branches face each other and arise from lower sides of hydrothecae of primary hydro-
clade and its first branch, the hydrothecae facing each other. The main stem is thus
not hidden when viewed in a plane at right angles to the branching plane. The infra-

210

“TERRA NOVA” EXPEDITION.

thecal nematotheca and process of the internode are present on the proximal internode
of each hydroclade as well as on all the others. A pair of lateral “ mamelons ” present
on the upper surface of each stem apophysis. There is a characteristic step down on
the dorsal side from the distal end of one internode to the proximal end of the next.
It is much more marked in this species than in 0. bifurca (Hartl.). There is a slightly
thickened margin to the hydrotheca which, in these specimens, is not narrowed as
illustrated by Billard (1914). The gonothecae were immature on January 3rd. They
spring from the middle line of the basal hydrothecate articles of the main branches,
and are distal to the infrathecal nematophores. Mature gonothecae have not yet
been recorded. They have been found only between the end of November and the
middle of February.

Measurements :

Stem,

diameter

Branch internode,
length
breadth

Hydrotheca,

length (abaxial wall)
breadth at margin,
clorso-ventral
lateral

breadth at base,
dorso-ventral
lateral

mm.

MO-1-22

0-63-0-75

0-10-0-22

0-20-0-26

0-17-0-20

0-22-0-23

0-17

0-17

Remarks. — Comparison with the figure 17 of Billard (1914) shows that if the
specimens taken by the “ Terra Nova ” and “ Pourquoi Pas ? ” belong to one and
the same species there is a large amount of variation in the shape of the hydrotheca.
The “ Terra Nova ” specimens appear to resemble closely figures of Jaderholm’s types.

Billard (1914) refers to the two lateral foramina mentioned by Vanhoffen (1910)
in the “ Belgica ” report and says that they are absent from the “ Pourquoi Pas ? ”
specimens. They are not the paired supracalycine nematothecae, but are the dupli-
cated median supracalycine foramina to be found in Oswaldella species at the forking
of the primary hydroclade and the first secondary branch. Elsewhere there is a single
median supracalycine foramen (dactylotreme).

Genus ANTENNELLA, Allman.

Antennella , Allman, 1877, p. 38.

It is still convenient to retain this name for a group of unbranched or little-
branched Plumularians, although a study of their minute structure does not afford
grounds for considering them to form a monophyletic group.

HYDROIDA— TOTTON.

211

The characters used for differentiation of species are :

1. Fusion with or freedom from hydroclade of adaxial wall of theca.

2. Presence or absence of second pair of pleurohydrothecal nematothecae.

3. Whether or not the nematothecae, including infrathecal one, are paired.

4. Degree of obliteration of nodes.

5. Presence or absence of median nematotheca on distal end of hydrothecate

article, and the length of this part of the article.

6. Relative length of intermediate article and number of nematothecae borne

on it.

The gonothecae are in some, probably all, cases sexually dimorphic, the male being
smaller and having a smaller distal aperture. Both kinds are found on a single
hydroclade.

Antennella ritchiei, sp. n. (Text-fig. 52, a-b.)

Material. — A specimen, with immature gonothecae, consisting of 70 stems, each
about 1 cm. long, which spring from a stem-like bundle of thirty hydrorhizal tubes

7 cm. in length ; and a fragment, consisting of a bundle of 7 tubes measuring 0-4 mm.
in diameter and 6 mm. in length, bearing a dozen hydroclades ; from Station 91, off
Three Kings Islands, 300 fathoms.

Description.— A heteromerous species of Antennella with two pairs of pleurohydro-
thecal nematothecae and with most of the nodes obliterated, the intermediate inter-
v. 5. 11

212

“TERRA NOVA” EXPEDITION.

nodes being short and bearing two nematothecae. A well-marked oblique hinge-joint
below the first two hydrothecae ; elsewhere oblique joints seldom to be seen. The
second pair of pleurohydrothecal nematothecae very short. Hydrothecae close
together with flared margins projecting well out from axis, slightly compressed dorso-
ventrally, nearly cylindrical when viewed laterally. Early stages in the development
of gonothecae are present on some stems, each having a pair of nematothecae near the
base.

Measurements :

Thecate internode,
length *

Intermediate internode,
length
breadth
Hydrotheca,
depth

diameter at margin
Length of basal part of hydroclade,
in Holotype
in Paratype

mm.

0-4-0-45

. 0-52
. 0-10

. 0-25

. 0-22-0-25 X 0-23-0-28

. 1-5-2-75

. 1-25-1-5

The species is close to A. quadriaurita, Ritchie, but the thecae are closer together
and project further ; the intermediate internodes are shorter and bear only two
nematothecae ; and the supernumerary pleurohydrothecal nematotheca are shorter.
The measurements, except those of internode lengths, exceed those of A. quadriaurita.
The larger specimen, Brit. Mus. Reg. No. 29.10.10.13, is selected as holotype.

Antennella serrata, sp. n. (Text-fig. 53.)

Material. — Sixty shoots 10-12 mm. in length, springing from a plexus of hydro-
rhizal tubes which cover a sponge, from Station 90, Three Kings Islands, N. Zealand,
100 fathoms ; and three dozen shoots growing on sponges at the base of a group of
stems of S. unguiculata, from Station 144, off Cape Maria van Diemen, N.Z., 35-40
fathoms.

Description. — A heteromerous species of Antennella with two pairs of pleuro-
hydrothecal nematothecae, intermediate articles bearing a single nematotheca or
occasionally two. Rhizal plexus bears numerous scattered, narrow, funnel-shaped
nematothecae. Shoots divided into two parts. The proximal millimetre or two
without hydrothecae, but with one or two transverse nodes, and often some nemato-
thecae ; these are either normal ones, or of the type found on the rhizo-plexus. The
distal hydrothecate part of the shoot is separated from the proximal part by an oblique
node or hinge- joint. Another oblique node regularly appears below the second hydro-
theca. There are oblique nodes between every two or three successive hydrothecate
articles ; and the distal part of each is generally separated off by inconspicuous trans-

* I have made measurements similar to those Ritchie (1909a) made of A. quadriaurita.

H YDR OIDA— TOTTON.

213

verse nodes as an intermediate article bearing one or, in some specimens, two nemato-
thecae. There may be as many as two dozen hydrothecae on a shoot. Hydrothecae
with abaxial and adaxial sides parallel, free part of adaxial wall slightly longer than
diameter at margin. Floor of theca curved up on abaxial side so as to be at right
angles to axis of hydroclade. Besides the usual three nematothecae round each hydro-
theca there is a pair of small nematothecae springing
from the bases of the lateral nematothecal apophyses,
as in A. quadriaurita, Ritchie. The nematothecae on the
hydrocladia have the adaxial side shorter than the
abaxial.

Measurements in millimetres :

Station 96. Station 144.

Tkecate article, length . . 0-33-043 0-38-0-48

Intermediate article, length . . 0-35-0-43 0-30-0-58

,, ,, breadth.. 0-08 0-08

Hydrotheca, depth . . . . 0-20 0-18-0-23

,, diameter at margin 0-16-0-20 0-19-0-20x0-23

Remarks. — Broch (1914) has recorded specimens,
which he names A. quadriaurita, Ritchie, forma africana,
that have only one nematotheca on each intermediate
article. But as this is the only precise information
which he gives about the specimens it has seemed to
me preferable to regard this form, raised to specific rank
by Stechow (1923) and figured by him (1925), as in-
completely described, and to give the “ Terra Nova ”
specimens a new specific name. Subsequent examina-
tion will show whether or not the W. African and
N. Zealand forms belong to the same species. I think
that the close approximation of the hydrothecae is
sufficient to distinguish a separate species, especially as
Ritchie’s is an Atlantic one.

The specimens from the deeper water station,
number 90, constantly bear single nematothecae on the sp n Schizoholotype,

intermediate articles, but those from the shallower x 155.
water station, number 144, bear sometimes two. In

these latter specimens, too, there is rather more difference between the distances
apart of hydrothecae at the proximal and distal ends of a shoot. At the base
the distance is 0-7 mm. and at the tip 0*43 mm. In the specimens from Station
90 the respective distances are 0-58 mm. and 0-45 mm. The group of hydro-
clades from Station 90, Brit. Mus. Reg. No. 29.10.10.8, have been chosen as the
holotypes.

214

“TERRA NOVA” EXPEDITION.

Genus PYCNOTHECA, Stechow.

Genotype. — Diplocheilus mirabilis, Allman, 1883, p. 49, pi. VIII, figs. 4-7.

Pycnotheca, Stechow, 1919, p. 110 ; Diplocheiius , Allman, 1883, p. 48 (preoccupied by v.

Hasselt, 1823).

Pycnotheca, as the genotype of which I select the species mirabilis, Allman, appears
to be closely related to the genus Halicornopsis, Bale, 1882, genotype elegans, and only
distantly to Kirchenpaueria, Jickeli, 1883, as the genotype of which I accept Sertularia
pinnata, 1758, of Linnaeus. It will be as well to recognise all three genera as distinct.

In Pycnotheca and Halicornopsis there is an intrathecal fold on the fore-side of
the hydrotheca, or in other words, there is a secondary theca external to the true
opercular opening. In Halicornopsis the true thecal cavity sends diverticula into the

two lateral and the single median
marginal teeth (text-fig. 54, a) ; but
it should be noted that in the
basal theca of each hydroclade the
marginal teeth and the diverticula
contained in them are much less
pronounced than elsewhere, giving
the theca more the appearance of
those of Pycnotheca, which is pro-
bably the more primitive form.
In Pycnotheca I have observed the
operculum in situ in immature
thecae ; and in suitably stained
preparations the scar of the line of
attachment can be seen at all stages
after the loss of the operculum.
In these two genera, again, there
is a structure in the axils of the
branches which is absent in Kir-
chenpaueria. This is a tubular
“ mamelon ” (text-figs. 54, b, 55, b) resembling that found in Nemertesia and other
genera. The growth-form and arrangement of branching and nodes are comparable
in the two genera, but differ from those in Kirchenpaueria. By carefully cleaning,
staining and dissecting the perisarc of K. pinnata it can be seen that there is no
“ mamelon ” in the axil ; but that there is a large foramen in the side of the stem
for the exit of a long protrusible nematophore, and two smaller ones on the hydro-
cladial apophysis (figs. 54, c-d). At present there is no evidence that a nematophore
protrudes from the tubular “ mamelon ” in Pycnotheca and Halicornopsis, though the
“ mamelon ” has a large pore.

Text-fig. 54.— a, Halicornopsis elegans (Brit. Mus. Reg.
No. 88.11.13.97), showing diverticula, d, of thecal cavity ;

b, the same, showing tubular mamelon in axil of branch ;

c, Kirchenpaueria pinnata, Linn., axils of three branches,
showing the three foramina ; d, the same, lateral view of
axil of branch, showing two foramina of apophysis, af,
in optical section, and the foramen in hydrocaulus, hf.
All figures magnified X 38.

HYDROIDA— TOTTON.

215

At the present time two Australasian species are recognised, P. producta (Bale)
and P. mirabilis (Allman). The chief differences between them are that in P. mira-
bilis (1) the theca is much deeper from centre of margin to floor, being 0-30 mm.
compared with 0-25 mm. ; (2) the free back of the theca is longer — 0-10 mm : 0-05 mm. ;

Text-fig. 55. — a-d, Pycnotheca mirabilis (Allman), a, subspecies waneni, nov. (Brit. Mus. Reg.
No. 22.3.6.140), X 166 ; b, axils of branches of specimen from Station 134, X 39, to show mamelon
and cauline nematothecae ; c, the same, lateral view of hydrotheca, X 166 ; d, the same, frontal view,
X 55 ; e, Pycnotheca producta (Bale) from Botany Bay (Brit. Mus. Reg. No. 99.5.1.308b), X 160.

(3) the theca proper has a lesser maximum depth below the ridge from front to back —
0-15 mm. : 0-20 mm. ; (4) the portion on which the infrathecal nematotheca is seated
is less prominent, the opening or foramen in the wall being in a vertical plane, whilst
in producta it is in a horizontal one, and seated on a distinct prominence : in the last

216

“TERRA NOVA” EXPEDITION.

respect the S. African form is intermediate ; (5) the marginal opening is circular rather
than elliptical ; (6) the fore edge of the intrathecal ridge seen from the side is
relatively deeper than in P. producta. P. mirabilis reaches a height of three inches,
whilst P. producta is a dwarf, with a maximum height of about one-third of an inch.
Bale (1894) has alluded to differences in the gonothecae.

Whether we are dealing with varieties of one species, producta, or whether we
have two species is not clear. Two atypical forms have been described, one by Ritchie
(1911) which resembles mirabilis but has the ridge of producta, and the other by Inaba
(1890, 1892) and Stechow (1909, 1913) which resembles producta but has a longer free
dorsal wall to the theca. Torrey ’s P. allmani seems to resemble the Japanese form
except in size. It is smaller even than the S. African form. This latter, studied by
Warren (1908), appears to me to be a distinct geographical race and may be known as
P. mirabilis (Allman) subspecies warreni nom. nov.* It is a pigmy form and its
hydrothecae have a characteristic shape (text-fig. 55, a).

Warren states that the anterior flexure is not completely filled in with solid
perisarc. His figure “ C ” may not represent a sagittal section, for the filling in can be
seen clearly in whole mounts.

To sum up it appears probable that we have several geographical races of these
two species. Though we still need further data there is some evidence to justify us in
naming a S. African race, P. mirabilis warreni nom. nov., two N. Pacific races, P. pro-
ducta allmani (Torrey) (for Diplocheilus allmani, Torrey, 1904) and P. producta inabai
nom. nov. (for Plumularia producta, Inaba 1890, 1892).

The “ Terra Nova ” specimens closely resemble Allman’s Challenger ones.

Pycnotheca mirabilis (Allman). (Text-fig. 55, a-e.)

Diplocheilus mirabilis, Allman, 1883, p. 49, pi. VIII, figs. 4-7.

Material. — Fifty shoots up to five centimetres in length, growing on an eroded
shell from Station 134, near North Cape, New Zealand, 11-20 fathoms; three small
shoots on a shell from Station 90, off Three Kings Islands, New Zealand, 100 fathoms ;
a single stem 3 cm. long from Station 144, off Cape Maria van Diemen, 35-40 fathoms.
Gonosome absent.

Description. — There is little to add to what has been published. In the holotype
an occasional branch springs from the mid- ventral line just below an apophysis. There
are two short, bare internodes at its base. The normal arrangement of cauline nemato-
thecae is that there is a simple scoop- or cowl-shaped nematotheca near the distal end
of each internode, even if the node is not marked, i.e. there is one to each hydroclade.
There is another on the antero-superior side of the apophysis, and a tubular mamelon
in the axil on the postero-superior side of the apophysis (text-fig. 55, b).

* Stechow (1925) has published a name robusta for a new variety of P. mirabilis from South-West
Australia, and has referred to the Natal race by this varietal name.

HYDROIDA— TOTTON.

217

Genus HALOPTERIS, Allman.

Genotype. — Halopteris carinata, Allman, 1877, p. 33, pi. XIX, figs. 3-7.

Halopteris, Allman, 1877, p. 32 ; Thecocaulus, Bale, 1915.

I select Plumularia catharina, Jolmston, as the kolotvpe of Thecocaulus, Bale, 1915.
I regard the words of Bale (1915, p. 294) as a characterisation and establishment of a
genus Thecocaulus separated or excluded from Plumularia sensu stricto ; and ascribe
the genus to Bale rather than to Bedot (1921), who excluded from it both P. catharina
and P. campanula which Bale considered typical. Both P. catharina and Halopteris
carinata have cauline thecae and a second pair of pleurohydrothecal nematothecae. I
consider that they are congeneric, and therefore sink Thecocaulus, Bale, in the synonymy
of Halopteris, Allman.

Halopteris constricta, sp. n. (Text-fig. 56, a.)

Material. — Half a dozen branched stems up to 9 mm. in height, without gonosome,
from Station 144, off Cape Maria van Diemen, N. Zealand, 35-40 fathoms.

Description. — Basal millimetre of stem without hydrothecae, but with a few
median nematothecae and a distal oblique hinge- joint. Upper part of stem and
branches heteromerous, longer hydrothecate articles with basal oblique and distal
transverse nodes alternating with articles bearing a single median nematotheca. An
apophysis without nematothecae at origins of branches which have from one to four
hydrothecae, usually three. First intermediate article of branch longer than rest and
with nematotheca near proximal end. Supracalycine nematothecae raised on finger-
like processes, and reaching margin of hydrothecae. Intrathecal nematothecae with
little or no free adcauline wall. All nematothecae scoop-shaped. Hydrothecae free
in distal half, nearly cylindrical, slightly constricted just below top of adcauline and
lateral walls.

Remarks. — The species resembles Nutting’s P. alternitheca, but its hydrotliecae
differ in shape and the nematothecae of constricta are larger in proportion and
differently shaped.

Halopteris heterogona (Bale). (Text-fig. 56, b-d.)

Thecocaulus heterogona, Bale, 1924, p. 255, fig. 13.

Material. — Two large tufts of about sixty stems, and one or two small ones, from
Station 144, off Cape Maria van Diemen, N. Zealand, 35-40 fathoms ; three tufts of
from two to six stems up to 12 cm. in length, from Station 90, off Three Kings Islands,
100 fathoms. Gonothecae present.

Description. — There are some important points to add to or to amend in Bale’s
description. On the main stem the regular supracalycine pair of nematophores is
said to be wanting ; but they are present in the distal region. The lower part of the
stem, which bears neither hydrothecae nor hydrocladia, has two or three transverse

218

“TERRA NOVA” EXPEDITION.

nodes. At a distance of about 1-5 cm. from the base there are two very oblique nodes
close together. These act as hinges to allow the distal part of the stem a certain amount
of rotary motion on a nearly vertical axis. Similar structures are common elsewhere,
especially in the Plumulariidae, and have been referred to by Nutting (1900), but
their function appears not to have been understood. Nutting refers to them as
“ pinched places.” In its most highly developed form a structure of this sort consists
of an upper and a lower articulation of the point-and-socket type with two opposed
oblique knife-edges between them. There is often a complete interruption of the

l

Text-fig. 56. — a, Halopteris constricta, sp. n. Base of a pinna from holotype, X 156 ; b, Hahpteris
heterogona (Bale). Lateral view of lrydrothecae, showing minute median supracalycine nematothecae,
X 55 ; c , the same, frontal view of axil of branch, showing nematothecae, X 32 ; d, the same, male and
female gonothecae, X 24.

coenosarc at such points. I refer to them as hinge-joints. Between the two hinge-
joints is the first cauline hydrotheca. As in T. simplex (Warren) and in T . polymorpha
(Billard) the first cauline hydrotheca is the point of origin of a pair of hydrociadia.
In T . simplex, however, the first cauline hydrotheca does not lie between the two
hinge-joints but above them. In T. heterogona there is a series of about five mesial
nematophores below the proximal hinge- joint, and two or three between the first
cauline hydrotheca and the uppermost hinge-joint. Distally to this joint each inter-
node at first bears the usual three nematophores round a hydrotheca, and an apophysis

HYDROIDA — TOTTON.

219

from which springs a branch alternately to left and right. The supracalycine nemato-
phore in the axil of the branch is much smaller than its opposite number, and later on
one or both of them may fall off. In addition there is a minute median supracalycine
nematotheca, with a second and occasionally a third pair of supracalycine ones at a
short distance from the theca on the stem as Bale has described. Bale does not appear
to have seen the median nematophore above each hydrotheca of the hydrocladia, at
the junction between theca and hyclroclade. It is necessary to free the chitinous
parts from organic matter and to stain before all the nematothecae and pores can be
seen satisfactorily. As in many other hydroids there is a progressive change in shape
of the articles, hydrothecae and nematothecae of the hydrocladia when followed from
base to tip. At the distal ends the supracalycine nematothecae are larger and closer
together than in the proximal region. Bale says that in the condition of the anterior
(median infrathecal) sarcothecae the species is a marked exception to the general
ride in Thecocaulus. Either the species that I am describing is not the same one as
Bale’s, which I doubt, or else the species is not an exception, for in the “ Terra Nova ”
specimens, which have been carefully cleared and stained, these nematothecae are
quite plainly not funnel-shaped with truncate top as drawn by Bale, but short and
scoop-shaped, the adaxial side of the distal chamber, which is flattened, being cut
away much more than in the other nematothecae. The type specimens should be
re-examined to determine this point.

The female gonotheca is flattened and somewhat truncated. The operculum is
not circular, as described by Bale, but elliptical. The nematophores at its base are
arranged in a transverse arc. There are usually three or four, but occasionally five.

Specimens are found growing in tufts, formed by the connection of the lower
parts of neighbouring stems by lateral anastomosing hydrorhizal tubes.

Remarks. — This striking species has been described, since the return of the Expe-
dition, from Chilton’s specimens at Canterbury College, N.Z. They were taken at a
spot some eight miles a little West of North from the locality from which the “ Terra
Nova ” specimens came.

Halopteris campanula (Busk), var. zelandica, var. nov.

Plumularia campanula , Busk, 1852, p. 401.

Material. — A fine group of specimens growing on a mollusc shell from Station 134,
near North Cape, N. Zealand, 11-20 fathoms; and a young specimen growing on
P. tenuissima from Station 91, off Three Kings Islands, N. Zealand, 300 fathoms.

Description. — The “ Terra Nova ” specimens differ from the type in that (1) the
abcauline wall of the theca and its continuation down to the mesial nematophore is
distinctly thickened ; (2) the free part of the adcauline wall of the theca is longer and
diverges more from the hydroclade ; (3) the distal ends of all the nematothecae are

rounded off and scoop-shaped, and not truncated with inrolled sides as in the type,
v. 5. 12

220

“TERRA NOVA” EXPEDITION.

Lastly each stem article regularly bears two suprathecal mesial nematothecae, whereas
this is only rarely seen in the type. On account of these differences it is advisable
to establish a new variety, H. campanula (Busk), var. zelandica var. n., characterised
above.

Remarks. — The differences between these specimens and the types are not suffi-
cient to warrant specific separation, although Busk had long ago received specimens
from N. Zealand from a Capt. Bolton and given them the MS. name neozelanica.

There are still in existence at the British Museum several of Busk’s type specimens
and sketches of this species, as well as specimens labelled by him as coming from other

than the type locality. More
than one species was labelled with
this name by Busk, one of them
being a Monostaechas. I have
selected as holotype a dried speci-
men from the type locality, Brit.
Mus. Reg. No. 99.7.1.6626.

The specimens bear no gono-
thecae, the breeding season pro-
bably being over at the time of
capture. We are without data
concerning the gonothecae of this
species, so that it may be useful
to give some for a specimen, Brit.
Mus. Reg. No. 88.4.16.41, coming
from Port Phillip, Australia. They
are dimorphic and both sexes are
borne on the same shoot. The
female is larger with a truncated end, a large operculum, and two nematophores
at the base, whilst the male is much smaller, oblong, with a minute operculum
and only a single nematophore. There is a single embryo in the female gonotheca.
It seems to undergo a development similar to that described by Warren (1914)
for Thecocaulus simplex (Warren), a very minute species from Pondoland. At
any rate the specimens from Port Phillip show a stage in which there is a large
single embryo with lateral walls closely appressed to the gonothecal walls, and
partly surrounded by the two horns of the blastostyle. Warren did not observe later
stages than this. Here there is another stage with what appears to be vestigial or
reduced blastostylic horns and a small free embryo : the operculum is still attached
at this stage. It is desirable to get fresh material with gonophores so that the final
stages of development and liberation can be followed. The material in the Museum
does not enable me to seriate the stages of development.

Text-fig. 57. — Halopteris campanula (Busk), a, var.

zelandica, var. n., lateral view of hydrotheca, x 54 ; b, male
and female gonothecae of specimen (Brit. Mus. Reg.
No. 88.4.16.41) from Port Phillip, Australia, X 40.

HYDROIDA — TOTTON.

221

Plumularia pulchella, Bale. (Text-fig. 58, a-d.)

Plumularia pulchella, Bale, 1882, p. 42 (30), pi. XY, fig. 6 ; Plumularia flexuosa, Bale, 1894,
p. 115, pi. 5, figs. 6-10 ; P. pulchella, Trebilcock, 1928, p. 24.

Material. — A dense growth of stems about half an inch long covering a gasteropod
shell ; and a small tuft on a fragment of an alga, from Station 134, ofi North Cape,
New Zealand, 1 1-20 fathoms.

Description. — These specimens correspond in all details with specimens from Port
Phillip, Australia. There are no gonothecae, but occasional branches are present ;
and one hydroclade which was
broken off has been replaced by a
branch with three hydrocladia.

There is little to add to pre-
vious descriptions. The distal end
of the hydroclade bears a thickened
chitinous cap which is kidney-
shaped in optical section when
viewed from the posterior side.

The specimens are twice the size
of those described by Bale, some
being half an inch in height and
bearing 48 hydrocladia. There is
a “ mamelon ” on one side of the
upper face of the stem apophysis
and a nematotheca on the other,
with its opening directed towards
the hydrotheca. There is a second
nematotheca on the stem article
on a level with the “ mamelon ”
and on the same side. It is curved
so that its opening is facing up
the stem. There is a third in the
middle of the article, in line with
the opposite row of apophyses.

Some of the stems in one

Text-fig. 58. — Plumularia pulchella, Bale. Hydrothecae
seen from different view points ; a, h, c, X 155 ; d, X 320.

clump are finer and flexuous, while others are stouter and straight. The base of a
stem may be from 0-25-0-6 mm. in length, and consists of two or more irregular
articles separated by transverse nodes. They appear to be devoid of nematothecae.
The hydrorhizal plexus bears irregularly disposed nematothecae not arranged in
pairs.

The coenosarc is not preserved, but a closely allied and undescribed species that
occurs on the Atlantic Gulf Weed has polyps with a circle of twelve tentacles and an

222

“TERRA NOVA” EXPEDITION.

adcauline blind-sac. This Gulf Weed species bas been confused with P. obliqua
(Jobnston), 1847, but is distinct.

Remarks. — Nutting (1900) raised Kirchenpauer’s subgenus Monopyxis, 1876
(name preoccupied ; genotype obliqua, Jobnston) to generic rank and renamed it Mono-
theca because tbe bydrocladia of tbis and other species, wbicb be considered to be
congeneric, bore only a single bydrotbeca eacb. Now Motz-Kossowska (1903) bas
shown that at Banyuls P. obliqua, one of these monotbecate forms, grows on plants
in rough water, and that where most exposed tbe perisarc is thinnest. It is known
also that these monotbecate forms are very flexible and have weak nodes. It seems
probable, therefore, that tbis habitus bas been adopted by more than one race in response
to environmental stimuli, and that it does not necessarily indicate very close relation-
ship. Motz-Kossowska (1907) bas shown that P. obliqua bas a single free-swimming,
though reduced, medusoid gonopbore of an interesting type. We are without informa-
tion on tbis point for tbe other monotbecate forms of Plumularia, and it will be as
well to retain them for tbe present in tbis polypbyletic group. It is probable, also,
that many of these forms, like some specimens in tbe British Museum from Tasmania,
are parasitic. These Tasmanian specimens bear numerous short stems from 3-4 mm.
in length wbicb arise from hydrorhiza embedded in an alga, as in Thecocarpus
parasiticus (Warren), but to what extent tbis habit can be correlated with reduction
of bydrocladia we cannot say.

P. flexuosa, Bale, 1894, whose tropbosome is smaller, though very much like that
of P. pulchella, and whose gonosome is slightly different, appears to fall well within tbe
range of variation of tbis species. For tbis reason tbe older specific name bas been
used, in spite of tbe absence of tbe gonosome.

Plumularia diploptera, sp. n. (Text-fig. 59, a-b.)

Material. — Three small groups of stems up to 3-5 cm. long, growing on a small
nodule of shell and sand, another on a fragmentary bydroid stem, and a third on a
polyzoan, tbe groups consisting of three, seven and two stems respectively, from
Station 144, off Cape Maria van Diemen, N. Zealand, 35-40 fathoms ; also numbers
of stems 2-5 cm. in length, with gonotbecae, from Station 134, off North Cape,
N. Zealand, 11-20 fathoms.

Description. — Stem non-fasciculate, cymose. Tbe stem apophyses, wbicb bear
eacb a small “ mamelon ” on tbe “ dorsal ” or upper surface, midway between stem
and apophysial ring, have a pair of nematotbecae flanking tbe axil, and a sub-terminal
thickened ring. They give rise to delicate heteromerous bydrocladia of about six
bydrotbecate articles. Tbe first of tbe intermediate articles is shorter than tbe rest,
and bas a thickened ring in tbe middle but no nematotbecae. Tbe other intermediate
articles have a sub-basal and sub-terminal ring with a nematotheca in tbe middle.
Tbe bydrotbecate articles have tbe theca and its usual complement of nematotbecae
in the middle, tbe supracalycines not reaching tbe distal end of tbe article. None of

HYDROIDA — TOTTON.

223

the nematothecae are borne on protuberances. These articles have sub-basal, sub-
terminal and sub-thecal thickened rings, and one at the level of the thecal floor. Each

Text-fig. 59. — Plumularia diploptera, sp. n., paratype from Station 134. a, gonothecae, X 33 ; b, stem

and branches, X 54.

stem internode bears a single lateral nematotheca on the side opposite the apophysis
and a little below the middle region.

Measurements :

Hydrocaulns,
diameter
internode length
apophysis length
Hydrocladia,

internode diameter
internode length,
non-thecate
hydrothecate
Hydrotheca,
length

width (lateral view),
at margin . .
at base
Nematotheca
length

diameter (maximum)
Gonotheca,
female,
length

breadth (maximum)
aperture, diameter
male,

length

breadth (maximum)
aperture, diameter

mm.

0-088-0-120

0-37-0-46

0-08

0-25

0-17-0-19

0-37

0-08

0-11

0-05

0-070-0-078

0-029-0-032

1-12-1-3

0-25-0-39

0-08-0-10

0-58-0-96

0-15-0-25

0-05-0-06

224

“TERRA NOVA” EXPEDITION.

This species is closely allied to P. setacea (Ellis) but is distinct in having con-
stantly a second lateral nematotheca on the “ mamelon ” side of the apophysis.
Examination of specimens of P. setacea from different localities round Britain shows
that there is regularly only one axillary nematotheca. Bedot (1921) whilst admitting
that this is generally so yet states that very often there are two, so that he is unable
to accept this point as a specific character. There are, of course, other species of this
group that constantly have two axillary nematothecae, among them one collected by
Lendenfeld at Lyttelton, New Zealand, and labelled P. tripartita, Brit. Mus. Reg.
No. 86.6.8.73. No part of this material, which belongs to two distinct species, can be
the type of P. tripartita since the material comes from Lyttelton instead of Timaru.

It is desirable to have further information on the following points in the morpho-
logy of P. setacea: — (1) Are the branches ever on opposite sides of stem, or do they
always come off to one side ? (2) Is there always a ring at the basal end of the stem

internodes ? (3) Is a sub-hydrothecal ring always present on the first and second

hydrothecate articles but absent from the distal ones ? (4) Is there only one lateral

axial nematotheca ?

Plumularia spirocladia, sp. n. (Text-fig. 60.)

Material. — Four complete shoots, the largest 5-5 cm. in length, with five or six
smaller shoots that have settled down upon it, all bearing
female gonothecae, and fragments ; from Station 134, near
North Cape, New Zealand, 11-20 fathoms. Also a fragment
from Station 144, off Cape Maria van Diemen, N. Zealand,
35-40 fathoms.

Description. — Stem simple, articles two or three times as
long as broad, each bearing at distal end one or sometimes
two apophyses, and single nematotheca facing hydroclade at
a point one-third of distance from base. Apophyses at base
alternate : distally, arranged in an open right-hand * spiral,
five apophyses in two turns, the sixth being in line with the
first. Each bears a small “ mamelon ” in the centre of the
upper and inner surface, and a pair of lateral nematothecae at
the base. Hydrocladia heteromerous ; articles as in P. setacea,
but there is a second ordinary intermediate internode between
the short basal internode and the first hydrothecate one of
each hydroclade. In hydrothecate internode a distinct pro-
tuberance between the basal and the sub-hydrothecal thicken-
ings of the wall, superior nematothecae nearly reaching end of article. Hydrothecal
margin not level but dipping down to meet hydrocladia! wall. Gonothecae as in
P. setacea, arising from sides of apophyses.

* Ascending to right, looking from inside.

Text-fig. 60. — Plumularia
spirocladia, sp. n. Part of
hydroclade of paratype
from Station 134, X 160.

HYDROIDA — TOTTON.

225

Measurements :

Hydrocaulus,
diameter
internode length
apophysis length
Hydrocladia,

internode diameter,
proximal
distal

internode length,
non-thecate,
basal
proximal
distal

hydrothecate,

proximal

distal

Hydrotheca,

length

width (lateral view;,
at margin
at base
Nematotheca,
length

diameter (maximum) . .
Gonotheca,
female,
length

breadth (maximum)
aperture, diameter
(?) male,
length

breadth (maximum)
aperture, diameter

mm.

. . 0-175-0-28
. . 0-57
. . 0-12

.. 0-10

. . 0-06

. . 0-10
. . 0-27
.. 0-19

. . 0-52
. . 0 32

. . 0-10

.. 0-11
. . 0-06

. . 0-067-0-070

. . 0-032-0-035

1-24-1-40

0-32-0-37

0-10-0-11

0-98-1-20

0-30

0-05

Plumulana triangulata, sp. n. (Text-fig. 61.)

Material. — Half a dozen short stems, some unbranched, up to 7 mm. in height,
one bearing two gonothecae, arising from a hydrorhiza that creeps over the stem of
Plumularia tenuissima , from Station 91, off Three Kings Islands, New Zealand,
300 fathoms.

Description. — Stem apophyses with “ mamelon ” and one (occasionally two)
nematotheca near distal end, bearing homomerous hydrocladia without basal article.
All nematothecae monothalamic, straight on inner, convex on outer side, with oblique
orifice ; one near each end of stem article, one intrathecal, one median supracalycine.
Distal parts of articles very long. Hydrotheca with rim very slightly everted, diameter
greater than depth. Gonothecae triangular in section, two and a half times as long as
stem articles, inserted on apophyses on side opposite “ mamelon,” and lying parallel
to stem.

Remarks. — This species most nearly resembles Billard’s P. ventruosa, but differs

226

TERRA NOVA” EXPEDITION.

in having no basal article to the hydrocladia, and nematothecae with straight inner
side and oblique orifice. One gonotheca was immature, and the other was somewhat
damaged at the distal end, so that a full description cannot be given.

Plumularia brachiata, sp. n. (Text-fig. 62, a-b.)

Material. — The apical part of a large specimen, 30 cm. in length, and smaller
detached branches, with gonothecae, from Station 96, off North Cape, N. Zealand,
70 fathoms.

Description. — Stem a fasciculate irregular sympodium, irregularly branched in
one plane, branches ascending close alongside the stem, often 15 cm. long and them-
selves with very few branchlets. The original stem is surrounded by innumerable
secondary tubules, which arise from the axils of the hydrocladia, and, higher up, them-
selves may be transformed into branches which repeat the characters of the main stem.
The secondary tubules have on either side a row of nematothecae. Hydrocladia
alternate, on opposite sides, about 5 mm. long, very delicate, borne on stem apophyses,
the abaxial sides of which become greatly thickened. The secondary tubules for the
most part run along the two sides of the stem which do not bear the hydrocladia, a
section of the stem being in the form of an ellipse. The stem is irregularly marked by
nodes, internodes bearing from one to four apophyses. There is often a lateral nemato-
phore just above the origin of the hydroclade. There is a pair of nematothecae on the
stem flanking the upper part of each apophysis, on which, at some little distance from
the stem, is an upwardly directed open tubular “ mamelon.” Distal to the “ mamelon ”
is a pair of nematothecae. The apophysis has a subterminal thickened ring. Gonangia
borne singly on sides of stem apophyses, lying close alongside stem, short, cornuate,
obliquely truncate, with wide circular operculum. Hydrocladia heteromerous. The

Text-fig. 61. — Plumularia tnangu-
lata, sp. n. Part of liolotype,
showing median supra- and infra -
thecal nematothecae, X 54.

Text-fig. 62. — Plumularia bracliiata,
sp. n. a, stem and branches ;
b, gonotheca, both figures magni-

fied X 54.

HYDROIDA — TOTTON.

227

intermediate articles have a sub-basal and a sub-terminal thickened band with a
median nematotheca on a small projection just above the sub-basal one. The first
(proximal) intermediate article is shorter than the rest and may be duplicated. The
hydrothecate articles, too, have a sub-basal and a sub-terminal thickening with a median
basal infra thecal nematotheca. The theca is in the middle of the article, and is
accompanied by a pair of longer supracalycine nematothecae which reach the distal
end of the article. Between the theca and the basal nematophore is a thickened ring.
There is a second on a level with the hydropore, and another slightly above the hydro-
thecal margin. The hydrotheca is adnate, deeper at the margin than at the base, and
with the abcauline side rather longer than the adcauline one, the margin being slightly
everted, and dipping down to meet the internode.

Measurements :

Stem diameters

Hydrocaulus,
diameter
internode length,

1 hydroclade
4 hydroclades

apophysis and basal article, length . .

Hydrocladia,

diameter,

proximal part
distal part
internode length,
non-thecate,

proximal part
distal part
hydrothecate,

proximal part
distal part

Hydrotheca,

length

breadth (lateral view),
at margin
base

Nematotheca,

length,

snpracalycine
intermediate
breadth (maximum),
supracalycine
intermediate

Gonotheca,

length

diameter (maximum)

mm.

3x4

0-18-0-20

0- 82-0-99

1- 88

0-20-0-22

0-06-0-07

0-05

0-11-0-16

0-11-0-16

0-30-0-32
0-20- 0-22

0-070-0-073

0-077-0-080

0-055

0-080

0-110

0-040

0-050

0-77-0-89
0 32-0-37

V. 5.

13

228

“TERRA NOVA” EXPEDITION.

Plumularia tenuissima, sp. n. (Text-fig. 63, a-b.)

Material. — A fragment, 1 cm. in length, of a fascicled stem 0-75 mm. in diameter,
branched and rebranched in one plane, with vestiges of gonothecae ; from Station 91 ,

off Three Kings Islands, N. Zealand,
300 fathoms.

Description. — Resembling P. bra-
chiata, but with finer branches and
hydrocladia, smaller thecae and more
elongated internodes. Hydrocaulus
unjointed and without nematothecae
except laterals on apophysis. Heteio-
merous hydrocladia alternate at distal
ends of branches, in places becoming
sub-opposite and decussate. The basal
intermediate article bears a nemato-
theca on a basal boss, but is not
divided from the stem apophysis.
This bears a prominent tubular
C£ mamelon,” but no distal nemato-
thecae. Accessory tubules, bearing
nematothecae, arise from sides of
apophyses. The arrangement of
nematothecae in the axils of hydro-
cladia cannot be ascertained with
certainty, but in some axils one nematotheca can be seen, and in one a pair. The
normal arrangement is probably a pair. Gonothecae incomplete, without nematothecae.
straight, compressed, distal end truncated, with wide aperture.

Text-fig. 63. — Plumularia tenuissima, sp. n. Holotype.
a, part of hydroclade, x 155 ; b, axil of hydroclade,
X 155.

Measurements :

Hydrocaulus, diameter
Hydrocladia
Iuternode, length,
hydrothecate
intermediate
1st intermediate
apophysis . .
Hydrotheca,
length

width (lateral view)
Nematotheca,
length

diameter, maximum
Gonotheca,
length

breadth, maximum
at aperture

mm.

0-11

0-04

0-37

0-19

0-11

0-15

0-05

0-06

0-07

0-45

0-33

0-18

0-15

HYDROIDA— TOTTON.

229

Nemertesia elongcita, sp. n. (Text-fig. 64.)

Material. — Two complete specimens with rooting masses, 27-5 cm. and 59 cm.
in length respectively, from Station 90, off Three Kings Islands, N. Zealand, 100
fathoms.

Description. — A fasciculate cymose stem, rising from a
mass of fine hydrorhizal tubes, gives rise irregularly, at
intervals of from one to two centimetres, to simple fascicu-
lated branches, which may reach a length of 9 cm. and very
occasionally rebranch. The branches in the lower region are
shed as growth proceeds. The hydrocladia on the stem and
branches are arranged in alternating whorls of four, and reach
a length of about 6 mm. On the branches they are so
arranged that a transverse section of the space that they
occupy is ovoid. In the hydrocladia there is a node at the
base of each hydrothecate article, but generally none to
mark off the intermediate articles. The stem apophysis,
which probably includes a fused intermediate article bearing-
no nematotheca, has a flanking pair of nematotkecae in the
axil, a pronounced distally directed tubular “ mamelon,”
characteristic of the genus, in the centre of the upper surface,
with a pair of supracalycine nematothecae just above it.

The apophysis is followed by the first of the hydrothecate
articles of the hydroclade, to each of which is fused the
intermediate article with its single nematotheca. The gono-
thecae are borne on the branch apophyses and slightly exceed
in length the distance between successive whorls. They are
slightly curved, narrow at the base and obliquely truncate
above, with wide circular aperture, the rim of which dips
somewhat on the adcauline side.

The most distinctive character is the general growth form.

Remarks. — I select the larger specimen, Brit. Mus. Reg. No. 29.10.10.9, as the
holotype.

Text-fig. 64. — Nemertesia
elongata, sp. n. Lateral
view of hydrotheca from
apical part of a hydro-
clade, x 255.

Genus SCHIZOTRICHA, Allman.

Genotype. — S. unifurcata, Allman, 1883, p. 28, pi. VII, figs. 1-3.

Schizotricha, Allman, 1883, p. 28.

Diagnosis. — Eleutlieroplea with homomerous, often branched, hydrothecate stem
and accessory supporting stem. This is formed by nematotheca-bearing tubules,
which arise from out of the older proximal stem hydrothecae. Each stem article bears
a long apophysis growing out alternately left and right from below the hydrothecae.

230

“TERRA NOVA” EXPEDITION.

The supracalycine pairs of nematothecae of the stem hydrothecae are thus found in
the axils of the apophyses. From the apophyses grow out homomerous hydrocladia
which branch unilaterally and fanwise, forming scorpioid sympodia. These primary
hydrocladia may rebranch as many as fifteen times. The main axis of a branch is
formed usually not by the primary hydroclade but by a series of successive apophyses
together with the proximal parts of the basal articles of the successive hydrocladia of
higher orders.

Hydrothecae deep, tapering below with abaxial edge of margin projecting beyond
level of adaxial edge. Dioecious, gonangia of heteromedusoid type, gonothecae cornu-
copia-shaped, borne immediately below hydrothecae on branches, with basal nemato-
thecae and below them a small basal article.

Remarks. — The following species should be assigned to the genus : —

Sertularia frutescens, Ellis and Solander, 1786 ; Schizotricha unifurcata, Allman,
1883 ; S. multifurcata, Allman, 1883 ; Plumularia variabilis, Bonnevie, 1899 ; P. 'pro-
funda, Nutting, 1900 ; Schizotricha dichotoma, Nutting, 1900 ; S. anderssoni, Jaderholm,
1904 ; S. turqueti, Billard, 1906 ; and Plumularia glacialis, Hickson and Gravely, 1907.

Diplopteron grande, Nutting, 1900, and D. longipinna, Nutting, 1900, should
probably be included too. In a fragment of D. grande, Brit. Mus. Reg. No. 02.6.6.16,
there appear to be no cauline hydrothecae, although Nutting says there is a hydrotheca
at base of each hydroclade ; but in other respects it resembles typical Schizotricha
species. The stem hydrothecae have probably been overgrown by accessory tubules.
In S. grande (Nutting) the reduced secondary hydrocladia are tending to become
specialised as phylactogonia and are losing their hydrothecae. One remains in the
axil of each successive forking. The main branch axis is formed by the primary
hydroclade.

To be excluded from the genus are the following species which have from time to
time been assigned to it: Plumularia sulcata , Lamarck, 1816 (genotype of Heterotheca,
Stechow, 1921) ; P. catharina, Johnston, 1833 (genotype of Thecocaulus, Bale, 1915) ;
P. campanula, Busk, 1852 ; P. diaphana, Heller, 1868 ; P. gracillima, Sars, 1873 ;
P. tenella, Verrill, 1873; P. buslci, Bale, 1884; P. liechtensterni, Marktanner, 1890;
Diplopteron quadricorne, Nutting, 1900; Plumularia parvula, Nutting, 1900; Schizo-
richa bifwrca, Hartlaub, 1904 ; Schizotricha antarctica, Jaderholm, 1904 ; Schizotricha
simplex, Warren, 1914 ; and Plumularia zygocladia, Bale, 1914.

Bifurcation of hydrocladia is found in several widely separated forms which
Bedot (1921) has grouped together under Schizotricha; but a number of common
features characterises the group of species which includes the genotype of Schizotricha.
Only this group should bear that generic name.

There is an evident connection between this genus and Nuditheca, Nutting, 1900.

Distribution. — There are three Antarctic and four N. Atlantic species, but none
in the Arctic area. There are three isolated records of the N. Atlantic species
S. frutescens from the northern part of the Adriatic, another isolated one from the

H YDR OIDA — TOTTON.

231

Bay of Biscay, and one by Ivrauss from Algoa Bay. The only temperature records
appear to be 3-5° C. for the N. Atlantic species S. variabilis and 2-4° C. for S. profunda
in Davis Strait.

Schizotricha unifurcata, Allman. (PL III, fig. 4 ; text-fig. 65, a-f.)

Schizotricha unifurcata, Allman, 1883, p. 28, pi. VII, figs. 1-3 ; S. multifurcata, Allman, 1883,
p. 29, pi. VII, figs. 4-5; S. turqueti, Billard, 1906, p. 15, fig. 5 ; ? Plumularia glacialis,
Hickson and Gravely, 1907, p. 22, pi. Ill, figs. 23-24.

Material. — Many broken stems up to 21 cm. in length bearing male and female
gonothecae ; and root masses measuring up to 3 cm. in diameter, from Station 220,
off Cape Adare, Boss Sea, 45-50 fathoms ; also the distal 5 cm. of two branched stems
with immature male gonothecae from Station 316, off Glacier Tongue, McMurdo
Sound, 190-250 fathoms.

Text-fig. 65 .—Schizotricha unifurcata, Allman, a-d, from “ Terra Nova ” Station 220 ; e, f, parts of
holotype (Brit. Mus. Reg. No. 88.11.13.78), from off Kerguelen Island, Challenger Expedition. All
figures magnified x 40.

The specimens were broken up by the Agassiz trawl. A reconstructed tuft would
be about 30 cm. in height and would consist of a dozen or more tapering stems rising
from a compact root-mass.

Description. — The primary hydrocladium arises below and to one side of the stem
hydrotheca, and each successive branchlet arises from a similar point below the hydro-
theca of the basal article of the preceding branchlet. The sagittal plane of the hydro-
thecae of the branchlets is rotated gradually clockwise through about 90 degrees in a
series of 10 branchlets, while at the same time the direction of the branchlets changes
from “ latero-dorsal ” to “ dorso- ventral.” All branchlets are elevated some 45
degrees from the horizontal. The nematothecae are horn-shaped with an emargination

232

“TERRA NOVA” EXPEDITION.

on the adaxial side. In their arrangement there is much variation. Three nemato-
phores are usually present on the basal article of each branchlet, two infrathecals, a
proximal ventro-lateral and a lateral, homologous with the two median infrathecals,
described by Allman for ordinary articles in S. unifurcatci. and a third on the proximal
part of the apophysis ; all three being on the upper side. In the proximal and older
regions of a branch one or two additional nematophores may arise near any of these
three characteristic ones, it being quite usual to find a second one on the distal part
of the apophysis of the basal articles.

The proximal parts of two neighbouring branches are nearly at right angles to
one another ; but the axes gradually curve round towards each other on the ventral
side like ribs. The branchlets are arranged tangentially on the axes which consist of
the series of successive apophyses and proximal parts of basal articles of the branchlets
which grow from these apophyses.

Mature male gonangia have been obtained, but only immature female ones. The
species appears to be dioecious. Gonangia of both sexes are of the heteromedusoid

type, and similar to that which is found in many species of
Halecium and in Sertularia argentea. The single egg in the
female and the spermatic mass in the male is borne on a
process of the blastostyle, and is covered by a mantle of
ectoderm. It has not been possible to determine whether there
are other layers of ectoderm inside the mantle. In the earlier
stages of development at any rate the distal end of the blasto-
style is produced into a flattened endplate which lies under the
blastogenic layer of ectoderm which secretes the chitin. At a
later period all these tissues seem to be absorbed, leaving only
the egg or sperms inside the perisarc. This latter is of the well-known pear-shape
in the mature gonotheca, and has a circular terminal or sub-terminal operculum
and generally two proximal nematophores : one or two additional ones may be
present, but they are not always perfectly formed.

The gonangia grow out from a point between the infra-hydrothecal nematophores
and the hydrothecae of the ordinary articles of the branches, or from just below the
hydrothecae of the basal articles. Sometimes additional gonothecae are present
in these regions. There is always a node at a short distance from the point of
origin.

Text-fig. 66. — Schizo-
tricha turqueti, Billard.
Nemato theca from
holotype, X 75.

The mature male gonothecae appear to be indistinguishable from those of
S. glacialis (Hickson and Gravely).

Measurements. — The average length of 126 branch internodes of a specimen from
off Glacier Tongue, McMurdo Bay, is 0-765 mm. Below is a table of some of the chief
measurements in microns compared with those of two other Antarctic species.

HYDROIDA — TOTTON.

233

Species.

Ordinary

Branch internode.

Hydrotheca.

Length.

Breadth.

Length.

Breadth.

S. unifurcata, Alim. “ Terra Nova,” 1910-13

544-965

227-297

260-300

190-210

S. unifurcata, Alim. “ Challenger,” 1873-6. .

506-1013

135-150

285-310

200-215

S. multifurcata , Alim. “ Challenger,” 1873-6

660-780

135 160

230-310

160-200

S. unifurcata, Alim., var. turqueti , Billard . .

770-860

160-175

280-315

175-190

S. anderssoni, Jaderh. “ Antarctica,” 1901-3

730-1100

129-148

280-350

150-180

S. glacialis, H. & G. “ Discovery,” 1901-3. .

—

198

334

210

Range of Variation. — Every article on three separate branches of Allman’s type of
S. unifurcata was measured, with startling results ; for in one branch the range of
variation in length of the ordinary hydrothecate articles was from 0-506-0-727 mm.
and the maximum length for a basal article was 0-776 mm. ; while in another branch
the range was from 0-810-1-013 mm. and the maximum length of a basal article was
no less than 1-443 mm. The figures given for the length of the ordinary branch inter-
node of Schizotricha unifurcata are based on a series of 105 articles from three branches.

Remarks. — It is not possible to say what difference there is between Allman’s two
species E. unifurcata and S. multifur cata. In both the primary hydroclade may
rebranch several times. The types of unifurcata show as many as six branchlets.
The arrangement of nematophores in one does not seem to differ characteristically
from that in the other.

The ways in which the “ Terra Nova ” specimens vary from the types are : in
having no node to separate off what Billard (1910) calls the “ article basal ” from the
apophysis of either stem article or first hydrothecate article of a branchlet ; in that
the ordinary branch articles (i.e. excluding basal articles) with one exception have but
one infrathecal nematophore ; and in having rather broader branch internodes.

Hickson and Gravelv’s types of P. glacialis, Brit. Mus. Beg. No. 07.8.20.37, include
male and female specimens. The long branches generally carry each a long secondary
branch, but more material is needed before we can decide how many species of Schizo-
tricha there are in the Antarctic area.

Aglaophenia laxa, Allman. (Text-figs. 67, 68.)

Aglaophenia laxa, Allman, 1876, p. 275, pi. XXI, figs. 5-7 ; A. laxa, Bale, 1924, p. 260, fig. 15 ;
idem, Trebilcock, 1928, p. 25, pi. V, fig. 5 ; not Thecocarpus laxus, Billard, 1913, p. 98.

Material. — Six tufted colonies from two to three inches in height and fragments,
all bearing mature corbulae, growing on a shelly bottom, from Station 134, near North

234

“TERRA NOVA” EXPEDITION.

Cape, N. Zealand, 11-20 fathoms ; also a small group of specimens with corbulae, and
a branched specimen without corbulae, 4-5 cm. in height, anchored by rooting processes
to a mass of shells and sand, from Station 144, off Cape Maria van Diemen, N. Zealand,
40 fathoms.

Only one new record, Trebilcock (1928), has been made of the capture of this New
Zealand species during the period of fifty years that has elapsed since Allman described
it. Specimens have been taken from time to time and the British Museum collection
contains a large fertile specimen from Wellington, N.Z., which formed part of the
Hincks collection, a fertile specimen from Lyall Bay, near Wellington, N.Z., sent to
the Museum in 1898 by Farquhar, and another specimen taken by Haddon in Torres
Straits.

Description. — In the largest “ Terra Nova ” specimen, seven stems up to 6 cm. in
length rise from the centre, and several other smaller ones from the margin of the

tangled mass of hydrorhiza. This is 4 cm. in dia-
meter, and is attached to shell fragments. The
only branching sensu stricto that takes place is in
the formation of hydrocladia and the gonosome.
In addition to this, however, hydrorhiza! tubes grow
up the stems, occasionally proceeding from the stems
themselves, and from them at intervals spring other
secondary stems. There is intercommunication be-
tween stem and hydrorhizal tubes by means of ostia
in the adjacent walls. This system of false branch-
ing seems to be confined to the Hydroida.

For the first two or three millimetres the
primary stems are smooth ; the next portion, of
variable length, bears a row of nematothecae, doubtless homologous with the mesial
intrathecal nematothecae. After this the distal part is regularly segmented, each
internode having a mesial and an external lateral nematotheca of normal size.
The hydrocladia spring from fronto-lateral apophyses which do not develop a
“ mamelon.” One centimetre’s length of the stem bears on either side about two and
twenty hydrocladia, which have a length of from 4 mm. to 6 mm. For the proximal
2 mm. or 3 mm. the secondary stems are more or less regularly segmented, each
internode bearing a single median nematotheca.

The gonosome consists of a number of modified, branched hydrocladia, termed
gonocladia. When fully formed each gonocladium is 6 mm. long, and generally turned
up parallel with the stem. Each bears seventeen branches or costae on either side,
and is of the type described by Allman for A. acantliocarpa.

The term “ corbula ” may be applied to each of these branching gonocladia which
bear and enclose the gonangia. These structures are not strictly comparable with the
somewhat similar phylactocarps of Lytocarpus ramosus, Fewkes, where we find the

HYDROIDA— TOTTON. 235

hydrocladia on the proximal parts of branches modified to form what at first sight
look like the costae of a corbula. In A. laxa the peduncle of the corbula bears a single
hydrotheca, and each of the short, broad internodes of the rachis is provided with a
median infrathecal and an inner lateral nematotheca like the stem. There is no outer
lateral nematotheca and no “ mamelon.” There is a rounded fronto-lateral eminence
on each internode of the rachis, and from the outer side thereof arises a vestigial branch,
which we may term the gonohydrocladium. In species of Thecocarjms it is well

Text-fig. 68. — Aglaophenia laxa , Allman. Apical end of immature corbula, showing gonohydrocladia,

q.hy., and costae, c, X 168.

developed, but here it is nothing more than a short process bearing on its distal end
a pair of generally asymmetrical nematothecae, a small median infrathecal nemato-
theca and a lateral rounded eminence, from which the final order of branch — the costa
— arises. Each costa is segmented. Every internode except the first, which bears
only one, is provided with a pair of opposite nematothecae. The gonangia arise from
the rachis in the axils of the branches. They are shaped like doubly convex lenses.

I am unable to find any reliable point of difference between the corbulae of
A. acanthocarjpa , of which I have examined the types, and A. laxa. The corbulae of
v. 5. 14

236

“TERRA NOVA” EXPEDITION.

the types of acanthocarpa have as many as five and twenty pairs of costae, whereas
those of laxa have not more than seventeen in specimens that I have examined. In
acanthocarjpa corbulae there is usually only one of the pair of lateral suprathecal
nematothecae on the gonohydrocladium, whereas in laxa both of them are nearly
always present. But little stress can be laid on the first of these points unless we know
whether the corbulae are fully grown or not. The presence of ripe gonangia is no
criterion because these are present at an early stage. As to the second point, there is
much variation from corbula to corbula.

I have noted an interesting branched corbula, a thing not unknown in other
species. One costa has become a secondary rachis, and one or two of the neighbouring
costae are abnormal. In one of these the gonohydrocladium bearing it and its
proximal internode are normal, but the second internode, instead of bearing a pair of
opposite nematothecae, bears a vestigial hydrotheca with its median infrathecal and
paired lateral suprathecal nematothecae. The infrathecal nematotheca is similar in
all respects to the other costal nematothecae. In Thecocarpus laxus, Billard, I have
seen a corbula in which the proximal gonohydrocladium consists of a succession of
normal hydrothecae instead of a single one, and bears no costa. The very next gono-
hydrocladium of this same corbula bears a costa which is transformed into a secondary
normal corbula similar to the one I have described above.

Gonangia bearing spermatic masses are in corbulae separate from those bearing
the gonangia in which the eggs are produced, six or seven in each gonangium. Corbulae
of the two kinds appear not to be borne on the same stem. There is no appreciable
difference between the two kinds of corbulae.

From Station 134 comes a specimen which is growing in the midst of a tuft of
Sertularella sp. A careful examination of the latter revealed some very interesting
developmental stages of A. laxa, presumably springing from recently settled planulae.
Each has a rounded footplate, and from it springs a stem bearing hydrothecae. There
is a bare proximal article 0-27 mm. long over all, followed by a still shorter one
015 mm. over all, which has a single median nematotheca. The nodes at each
extremity of this short article are of the oblique hinge type. Following this we find
from three to four normal hydrothecae, hydrocladia being developed at an early stage
from the second or third and successive articles. The point of interest is that after
the third or fourth hydrothecate stem articles the hydrothecae are suppressed, leaving
only the infrathecal and the outer supracalycinal nematothecae. This primitive
arrangement of hydrothecae is only to be seen at the bases of stems arising directly
from planulae, and not in those springing from subsequently formed hydrorhizal tubes.
A similar arrangement has been described by Bedot (1919) for A. kirchenpaueri and
A. pluma, but in both forms the hydrocladia were only produced from articles formed
subsequently to the hydrothecate ones. There is no evidence that these hydrothecae
disappear later on ; but old stems arising directly from the larval rosette or footplate
are seldom, met with or recorded.

HYDROIDA — TOTTON.

237

Remarks. — Nothing was added to Allman’s description till 1924, when Bale figured
and described some type material that I had sent him together with some notes for this
purpose. In 1921 I was fortunate enough to be able to identify the figured holotype
and paratypes of the species in a packet of New Zealand hydroids in Busk’s collection,
which had been presented to the British Museum in 1899. The specimens bore no
specific name. The gonosome was overlooked by Allman. The figured holotype bears
eleven incompletely developed gonocladia of the type predicted by Bale.

Thecocarpus ctenatus, sp. n. (Text-fig. 69, a.)

Material. — A small complete colony, three centimetres in length with branched
rooting fibres but no gonosome, from unknown locality ; and the lower five and a half
centimetres of the stem and hydrocladia of a specimen without rooting mass, bearing
an open corbula without gonangia, from Station 91, off Three Kings Islands, N. Zealand,
300 fathoms.

Description. — Stem polysiphonic, unbranched, the accessory tubes being con-
stricted diagonally at intervals to allow the stem to twist. Hydrocladia very long,
up to 3-5 cm. Full complement of three cauline nematophores, all of which are simple
and gutter-shaped. Hvdrothecae of an aberrant type with intrathecal ridges on both
abcauline and adcauline sides ; margin with three pairs of inconspicuous teeth and a
small anterior one. Median nematothecae adnate up to abcauline ridge, with septum
at base and thickening on abcauline side near apex, single gutter-like opening, and not
communicating with theca. Gonosome an open corbula of seventeen pairs of short
gonohydroclades without theca. Gonohydroclades give rise to longer costae, and have
three dorsal nematothecae and one median below origins of costae. Costae with five
pairs of nematothecae and one apical. Gonoclade with two nematothecae to each
article representing infrathecal and interior supracalycine, and single dorsal one.

Thecocarpus rostratus (Bale). (Text-fig. 69, b.)

Halicornaria rostrata, Bale, 1924, p. 264, fig. 18; Thecocarpus formosus (Busk), var. inarmatus,
Trebilcock, 1928, p. 26, pi. V, fig. 6 ; not Plumularia formosa, Busk, 1851, p. 118.

Material. — Thirty stems, up to 11*5 cm. in length, growing in tufts, their bases
encrusted with polyzoa and other organisms, from Station 90, off Three Kings Islands,
N. Zealand, 100 fathoms ; and a number of stems up to 7 cm. in length, growing on
sponges and other organisms, from Station 144, off Cape Maria van Diemen, N. Zealand,
35-40 fathoms.

Description. — Stems up to 10 cm. long, unbranched except for hydrocladia and
occasional abnormal regenerations. Basal five or six millimetres devoid of hydrocladia.
A short, smooth portion is followed by five or six internodes each bearing a double
mesial nematotheca. Thereafter about twelve hydrocladia a side in each centimetre
of stem, the hydrocladia being about 1 cm. long. The stem internodes have the full

238

“TERRA NOVA” EXPEDITION.

complement of nematothecae and a “ mamelon.” Hydrothecae in middle of hydro-
cladia with prominent median and four pairs of lateral pointed teeth, a pointed
prominence at the base of the median one. First and third lateral teeth prominent
and everted, fourth reduced, especially at distal ends of hydrocladia, where the third
lateral teeth are particularly strongly developed. First and second laterals single and
not double as in T. formosus. Median prominence hollow and in open communication
with hydrotheca. Median and lateral nematothecae with gutter-shaped openings.
The paired supracalycine nematothecae become progressively larger and more prominent
distally. The proximal hydrotheca has a small median prominence, and its mesial
nematotheca, instead of projecting, lies close to the theca. Corbulae not present.

Remarks. — The trophosome of this species differs from that of T. formosus (Busk)
in that its hydrocladia are longer, more distant from one another, and do not terminate
in spines. The hollow prominence at the base of the median hydrothecal tooth is
markedly shorter than in that species. The arrangement of the marginal teeth, too,
is different. As lectotype of T. formosus (Busk) I have taken two stems 4-5 cm. long
arising from a small rooting mass which is provided vTith short processes on its under
surface, indicating a parasitic habit. The stems bear about twenty-seven hydrocladia
on each side to every centimetre. The hydrocladia are about 4 mm. long. The speci-
men, Brit. Mus. Reg. No. 99.7.1.6115, which bears closed corbulae varying from
2-5 mm. to 5 mm. in length, comes from Algoa Bay, S. Africa. I agree with Mark-
tanner and differ from Billard in recognising four pairs of lateral teeth in T. formosus.
The projections, well figured by Billard (1907) on the ventral edges of the first and
second lateral teeth of the thecae of T. formosus, and making these teeth appear
to be divided into two, are absent in Thecocarpus rostratus, Bale.

Because of its simple lateral marginal teeth, and lack of terminal spines I regard
Trebilcock’s T. formosus var. inarmatus, 1928. as a variety of T. rostratus, Bale, rather
than of T. formosus, Busk. The distribution of the two species makes this seem
probable.

Thecocarpus spiralis, sp. n. (Text-fig. 69, c-d.)

Material. — Two colonies 16 cm. over all, complete with rooting masses and
branches, one with a single corbula, from Station 90, off Three Kings Islands,
N. Zealand, 100 fathoms.

Description.—' Colonies forming dextral scorpioid cymes. Stems from 2 mm. to
3 mm. in diameter at base, fascicled, lower 8 cm. or 9 cm. without branches, remainder
giving off pinnate branches from 3 cm. to 4 cm. long in a loose spiral, six branches to
each revolution. The branches all originate from the stem and not from the accessory
tubes, which grow up over them. There are secondary communications between the
stem and accessory tubes. Origin of accessory tubes undetermined. Hydrocladia
1-5 mm. apart, up to 14 mm. in length, with twelve hydrothecae to a centimetre.
Hydrothecae of the elongated cylindrical type, the mesial nematotheca arising at an

HYDROIDA- — TOTTON. 239

angle of 43° from the internode, and projecting only a short way beyond the thecal
wall, paired supracalvcine nematothecae not reaching margin of theca. All three

Text-fig. 69. — a, Thecocarpus ctenatus, sp. n., X 160 ; b, Thecocarpus roslratus (Bale), X 250 ; c, d,
Tliecocarpus spiralis, sp. n. hydrotheca, X 54 ; d, the same, proximal part of gonoclade, X 35 ; e,
Thecocarpus chiltoni, Bale. Lateral view of hydrotheca, X 150, showing variations in profile of median
hydrothecal tooth further enlarged.

nematothecae openly canaliculate. A pronounced median, and seven pairs of small
pointed lateral marginal teeth. Two or three internodal thickening beside one at the

240

“TERRA NOVA” EXPEDITION.

level of the paired nematothecae, and another at the level of the arris* Full comple-
ment of cauline nematothecae and a “mamelon.” There is an additional median
nematotheca below the usual one. The corbulae are of an interesting type, showing
transition from a normal hydroclade in the proximal region, to a stem with triply
branched hydrocladia in which the thecae are suppressed and the branches form a
closed corbula, in the distal part. In the intermediate region they have the form of a
simply branched stem, with cauline thecae suppressed, and with progressively reduced
thecae one on each of four secondary hydrocladia, the distal two of which each have
a tertiary branch. The supernumerary mesial nematotheca found on the main stem
is present on some articles of the corbula rachis. Each gonohydroclade or branch of
the corbula rachis, apart from four proximal ones, which do not form part of the
corbula, consists of a flat expansion — the “lateral spur.” This represents the
growing point of a hydroclade, and generally bears on its distal edge three nemato-
thecae, as well as from one to five on the other. On each lateral spur may be seen the
three nematothecae which always accompany a theca, but the thecae themselves are
suppressed.

From each “ lateral spur ” (gonohydroclade) arises a branch to form the broad
corbula “ leaflet ” (costa) provided with nematothecae on its distal edge. Close to
the point of origin of each secondary branch and on its distal edge, arises a small
tertiary branch (costal apophysis) surrounded by three thecal nematothecae. Each
tertiary branch fuses with the proximal edge of the corbula “ leaflet ” in front, and a
secondary communication is established in this way between the interiors of the
leaflets. Fusion between the distal ends of leaflets takes place in such a way that
small free expansions are left.

Remarks.— This species is closely allied to Aglaophenia tenuissima, Bale, 1914.
Colonies of both species form cymes, the hydrothecae are of the same type, although
differing in size and in number of marginal teeth, and the corbulae are similar.
A. tenuissima still retains reduced hydrothecae on the gonohydrocladia, whereas in
the distal region of the corbula of T. spiralis they are suppressed. The arrangement
of cauline nematothecae is the same in both, save for the presence of a supernumerary
mesial one in the “ Terra Nova ” species. These facts show what an unnatural group
is formed by giving the generic name Thecocarpus to all forms which have corbular
thecae.

Thecocarpus chiltoni, Bale. (Text-fig. 69, e.)

Thecocarpus chiltoni, Bale, 1924, p. 261, fig. 16.

Material. — Eight specimens bearing corbulae, some complete with rooting masses
and up to nine inches in height, from Station 90, off Three Kings Islands, N. Zealand,
100 fathoms ; and fragments bearing corbulae, from Station 144, off Cape Maria van
Diemen, N. Zealand, 40 fathoms. A fragment also from Station 91.

* See footnote, p. 163.

HYDROIDA— TOTTON.

241

Description. — From a vertically elongated rooting mass of hydrorhiza springs a
stout fascicled cymous stem, as much as 3 mm. in diameter at the base, branched and
rebranched pinnately in one plane as many as four times. The branches all arise from
the stem and not from the supporting tubes. One main supporting tube grows out
on the dorsal surface of each branch at an early stage, and later is accompanied by
subsidiary ones. Branches about 1 cm. apart. Stem and branches provided with the
full complement of nematothecae and “ mamelon.” Proximal parts of branches with
five normal hydrothecate internodes, after which the arrangement is as described for
stem and branches. Corbula peduncle with two normal hydrothecate internodes,
arrangement on rachis being similar to that on stem and branches. Corbulae closed,
about 5 mm. long inclusive of peduncle, consisting of about ten pairs of leaflets. The
rachis bears three orders of branches, the gonohydroclade, the costa and the costal
apophysis, the base of each being surrounded by the usual three nematothecae. The
internodes of the rachis are not all alike in this respect. The first three have the normal
complement of nematothecae, but distally all except the median inf rath ecal one are
suppressed. The gonohydroclade bears one hydrotheca with its three nematothecae
and a distal median supracalycine nematotheca. The costa or corbula-leaflet grows
out from between the hydrotheca and the mesial infrathecal nematotheca, and its
distal edge bears seven or eight normal hydrothecae. The costal apophysis is small
and triangular. It grows out from the base of the distal edge of the costa and is
surrounded by the usual three nematothecae. It unites with the hinder edge of the
base of the costa distal to it, filling up a space in this region. Hydrothecae with a
pronounced median and four lateral pairs of teeth, none of them everted. The median
tooth bears a short finger-like hollow process. First tooth small, second most pro-
nounced. Mesial nematotheca short, openings of all of them gutter-like. Infrathecal
ridge small with lateral uprising sinuous processes. Two internodal thickenings.
Opening from internode into hydrotheca small and oval.

Halicornaria regalis, sp. n. (Text-fig. 70.)

Material.— One detached branch two centimetres in length from Station 90, off
Three Kings Islands, N. Zealand, 100 fathoms.

Description. — Full complement of three simple gutter or scoop-shaped cauline
thecal nematothecae present. Calicles not closely approximated ; margin with a very
small anterior, three pairs of lateral, short, simple, broadly pointed teeth, the second
everted, and a rounded posterior notch. No intrathecal ridge. Supracalycine
nematothecae small, simple, gutter-shaped. Mesial nematotheca nearly straight,
aclnate right up to anterior tooth, with terminal pore and larger, though rather small,
basal opening ; length of free part about two-thirds of the distance from the margin
to the hydropore in the base of the theca. Gonothecae pedicellate, sub-cylindrical,
longer than broad, arising singly from anterior surface of branch apophyses, each
containing a single basal sporosac enclosed in a fine chitinous membrane.

242

“TERRA NOVA” EXPEDITION.

Measurements :

Stem, mm.

diameter . . . . . . . . . . . . . . . . . . 0-25

internode, length . . . . . . . . . . . . . . . . 0-43

Pinna,

length (15 thecae) . . . . . . . . . . . . . . 6-3

diameter . . . . . . . . . . . . . . . . . . 0-07-0T0

internode length . . . . . . . . . . . . . . . . 0-40

Hydrotheca,

length . . . . . . . . . . . . . . . . . . 0-22

diameter at mouth,

frontal view . . . . . . . . . . . . . . . . 0-25

lateral view . . . . . . . . . . . . . . . . 0T7

mesial sarcotheca,

free part, length . . . . . . • . . . . . . . . 0-20

Gonotheca,

diameter . . . . . . . . . . . . . . . . . . 0-32

Remarks. — This species is allied to H. Jiumilis , Bale, and to H. prolifera , Bale.
From the former it differs in that its mesial nematothecae are adnate as far as the
anterior tooth, and in the smaller size of this tooth. From H. prolifera it differs in

Text-pig. 70. — Ilalicornarid regalis, sp. n. Schizoholotype, X 160.

its smaller supracalycine nematothecae, its finer and straighter mesial nematothecae,
less strongly everted lateral marginal teeth and smaller anterior tooth, as well as by
having more distant hydrothecae. The holotype is the single specimen, Brit. Mus.
Keg. No. 29.10.10.14.

HYDROIDA— TOTTON.

243

V. -REFERENCE LIST OF WORKS.

The abbreviations of titles of periodicals are based on the World List ( Pollard and Smith , 1927).

Agassiz.. A. 1865. North American Acalephae. 111. Cat. Mus. Comp. Zool. Harvard. II, NIV -j- 234 pp.,
360 text-figs.

Agassiz, L. 1862. Contributions to the Natural History of the United States of America. Monogr. II,
parts 3-5, 383 pp., 16 pis., 48 text-figs.

Alder, J. 1856. A notice of some new genera and species of British Hydroid Zoophytes. Ann. Mag.
Nat. Hist., (2j XVIII, pp. 353-362, pis. 12-14.

1860. Descriptions of a Zoophyte and two species of Echinodermata new to Britain. Ann. Mag.

Nat. Hist., (3) V, pp. 73-75, pi. 5.

Allman, G. J. 1864. On the construction and limitation of genera among the Hydroida. Ann. Mag.
Nat. Hist., (3) XIII, pp. 345-380.

1872. A monograph of the Gymnoblastic or Tubularian Hydroids, London, Kay Soc., 1871-2.

Conclusion of Part I and Part II, pp. XXIII-XXIV -f- 155-450, pis. 13-23.

1874. Report on the Hydroida collected during the Expeditions of H.M.S. “ Porcupine.” Trans.

Zool. Soc. London, VIII, pp. 469-481, pis. 65-68.

1876. Diagnoses of new genera and species of Hydroida. J. Linn. Soc. Zool. London, XII,

pp. 251-284, pis. 9-23.

1877. Report on the Hydroida . . . Exploration of the Gulf Stream by L. E. de Pourtales. . . .

Mem. Mus. Comp. Zool. Cambridge Mass., V. 2, 66 pp., 34 pis.

1883. Report on the Hydroida . . . Part I. — Plumularidae. Report Sci. Res. H.M.S. “ Chal-
lenger ”... Zoology XX, 55 pp., 20 pis.

1885. Description of Australian, Cape, and other Hydroida, mostly new, from the collection of

Miss H. Gatty. J. Linn. Soc. Zool. London, XIX, pp. 132-161, pis. 7-26.

1888. Report on the Hydroida . . . Part II. — The Tubularinae, Corymorphinae, Campanularinae,

Sertularinae, and Thalamophora. Report Sci. Res. H.M.S. “ Challenger ”... Zoology
XXIII, LXIX + 90 pp., 39 pis.

Bale, W. M. 1882. On the Hydroida of South-eastern Australia, with descriptions of supposed new
species, and notes on the genus Aglaophenia. J. Micr. Soc. Viet., II, pp. 15-48, pis. 12-15.

1884. Catalogue of the Australian Hydroid Zoophytes. Australian Museum, Sydney, 8°, 198 pp.,

19 pis.

1887. The genera of the Plumulariidae, with observations on various Australian Hydroids.

Trans. Proc. Roy. Soc. Viet., XXIII, pp. 73-110.

1888. On some new and rare Hydroida in the Australian Museum collection. Proc. Linn. Soc.

N.S.W., (2) III, pp. 745-799, pis. 12-21.

1894. Further notes on Australian Hydroids, with descriptions of some new species. Proc. Roy.

Soc. Viet., (n.s.) VI, pp. 93-117, pis. 3-6.

1913. Further notes on Australian Hydroids. II. Proc. Roy. Soc. Viet., (n.s.) XXVI, pp. 114—

147, pis. 12-13.

1914. Report on the Hydroida collected in the Great Australian Bight and other localities. Biol.

Res. “ Endeavour,” 1909-14, II, 1, pp. 1-62, pis. 1-7.

1915. Report on the Hydroida collected in the Great Australian Bight and other localities. III.

Biol. Res. “ Endeavour,” 1909-14, III, 5, No. 6, pp. 241-336, pis. 46-47.

1924. Report on some Hydroids from the New Zealand coast, with notes on New Zealand Hydroida

generally, supplementing Farquhar’s list. Trans. Proc. N.Z. Inst., LV, pp. 225 -268, 18 text-figs.
Bedot, M. 1912. Materiaux pour servir a l’histoire des Hvdroides. 4me periode (1872-1880). Rev.
suisse. Zool. Geneve, XX, pp. 213-469.

1919. Le developpement des colonies W Aglaophenia . C. R. Soc. Phys. Hist. nat. Geneve, XXXVI,
pp. 50-57, 4 text-figs.

1921. Notes systematiques sur les Plumularides. Rev. suisse Zool. Geneve, XXIX, I, pp. 1-40.

v. 5. 15

244

“TERRA NOVA” EXPEDITION.

Billard, A. 1906. Hydroid.es. Exped. Antarct. Fran^aise (1903-1905). Paris, 20 pp., 5 text-figs.

1907. Hydroides de Madagascar et du Sud-Est de PAfrique. Arch. Zool. exp. gen. Paris, (4)

VII, pp. 335-396, pis. 25-26.

1909. Revision des especes types d’HydroIdes de la collection Lamouroux. Ann. Sci. nat. Paris,

(9) IX, pp. 307-336, 10 text- figs.

1910. Revision d’une partie des Hydroides du British Museum. Ann. sci. nat. Paris, Zool. XI,

65 pp., 24 text-figs.

1913. Les Hydroides de l’expedition du Siboga . I. Plumulariidae. Res. expl. zool. bot. . . .

Siboga VIIa, 115 pp., 6 pis., 96 text-figs.

1914. Hydroides. Deuzieme Exped. Antarct. Frangaise 1908-1910. Paris, 34 pp., 17 text-figs.

1918. Notes sur quelques especes d’hydroides de Texpedition du “ Siboga.” Arch. Zool. exp.

gen. Paris, LVII, Notes et Rev., pp. 21-27, 5 text-figs.

1924. Note critique sur divers genres et especes d'Hydroides avec la description de trois especes

nouvelles. Rev. suisse Zool. Geneve, XXXI, pp. 53-74, 3 text-figs.

1925. Notes sur quelques especes la plupart nouvelles de Synthecides et de Sertularides du

“ Siboga.” Bull. Soc. zool. Fr., XLIX, pp. 646-652, 2 text-figs.

Bonnevie, K. 1899. Hydroida. Den Norske Nordhavs-Expedition 1S76-1878. Christiania, Zool.
XXVI, 104 pp., 8 pis., 1 map.

Briggs, E. A. 1914. Hydrozoa from one hundred fathoms, seven miles East of Cape Pillar, Tasmania.
Rec. Aust. Mus. Sydney, X, pp. 285-302, pi. 25-26.

Broch, Hj. 1909. Hydroidenuntersuchungen II. Zur Kenntnis der Gattungen Bonneviella und
Lictorella. Nyt. Mag. Naturv. Kristiania (Oslo), XLVII, pp. 195-206, 6 text-figs.

1914. Hydrozoa benthonica. In : Michaelsen, W., Beitrage zur Kenntnis d. Meeresfauna West-

afrikas, Hamburg, pp. 19-50, 1 pi.

— — - 1918. Hydroida. (Part II.) Danish Ingolf-Expedition, Copenhagen, V, 7, 205 pp., 1 pi., 95
text-figs.

Busk, G. 1851. A list of Sertularian Zoophytes and Polyzoa from Port Natal, Algoa Bay, and Table
Bay, in South Africa ; with remarks on their geographical distribution, and observations on the
genera Plumularia and Catenicella. Rep. Brit. Ass. Adv. Sci. for 1850, London 8°. Notices
and Abstracts . . . pp. 118-120.

1852. An account of the Polyzoa and Sertularian Zoophytes . . . voyage of the “ Rattlesnake ” . . .

Sertularian Zoophytes. MacGillivray, J. Narrative of the voyage of H.M.S. “ Rattlesnake ”
. . . London, I, Appendix IV, pp. 385-402.

1857. Zoophytology. Quart. J. Micr. Sci. London, V, pp. 172-174, pi. 16.

1858. Zoophytology. Quart. J. Micr. Sci. London, VI, pp. 124-130, pis. 18-19.

Clarke, S. F. 1879. Report on the Hydroida . . . Exploration of the Gulf Stream, by Alexander
Agassiz 1877-78. Bull. Mus. Comp. Zool. Harvard, V, pp. 239-252, 5 pis.

Coughtrey, M. 1876. Critical notes on the New Zealand Hydroida. Trans. Proc. N.Z. Inst., VIII,
pp. 298-302.

Ellis, J., and Solander, D. 1786. The natural history of many curious and uncommon zoophytes
. . . London, 4° xii + 208 pp., 63 pis.

Farquhar, H. 1896. List of New Zealand Hydroida. Trans. Proc. N.Z. Inst., XXVIII, pp. 459-468.

Fleming, J. 1820. Observations on the natural history of the Sertularia gelatinosa of Pallas. Edin-
burgh Phil. Journ., II, pp. 82-89.

Fraser, C. M. 1911. The hydroids of the west coast of North America, with special reference to those
of Vancouver Island region. Bull. Lab. Nat. Hist. Iowa Univ., VI, No. I, pp. 3-91, pis. 1-8.

Gray, J. E. 1843. Additional radiated animals and annelides. In : Dieffenbacb, E. Travels in New
Zealand. Vol. II, Fauna of New Zealand. London, 8°, pp. 292-295.

Hadzi, J. 1913. Poredbena hidzoidska istrazivanja. I. Hebella parasitica (Ciamician) . . . Rad.
Jug. Akad. Znan. Umj. Zagreb., CXCVIII, pp. 105-210, 35 text-figs.

Hartlaub, C. 1901. Hydroiden aus dem Stillen Ocean. . . . Zool. Jahrb. Jena, Syst. XIV, pp. 349-
378, pis. 21-22.

1904. Hydroiden. Exped. Antarct. Beige, Res. Sci. Voy. Belgica, Zool. VII, Anvers, 19 pp.,

4 pis.

1905. Die Hydroiden der magalhaenischen Region und chilenischen Kuste. Zool. Jahrb. Jena,

Suppl. Bd. VI, Fauna Chilensis 3, pp. 497-714, 142 text-figs.

HYDROIDA — TOTTON.

245

Hassal, A. H. 1848. Definitions of three new British Zoophytes. The Zoologist, London, VI, p. 2223.

Hasselt, J. C. von. 1823. Algem. Konst en Letter Bode, XXXV, p. 132.

Hellek, C. 1868. Die Zoophyten und Echinodermen des Adriatischen Meeres. Verh. zool.-bot. Ges.
Wien, XVIII, Beilage, 88 pp. 3 pis.

Hickson, S. J., and Gravely, F. H. 1907. Coelenterata II. Hydroid Zoophytes. Nat. Antarct.
Exped., 1901-4, Nat. Hist. Ill, London, Brit. Mus., 34 pp., 4 pis.

Hincks, T. 1868. A history of the British Hydroid Zoophytes. 2 vols., London, 8°, pp. LXVIII + 338,
67 pis.

Hutton, F. W. 1873. On the New Zealand Sertularians. Trans. Proc. N.Z. Inst., V, pp. 256-259.

Inaba, M. 1890-1892. Hydroida obtained in Misalri, Minra and Soshu. Zool. Mag. Tokio, II-IV,
Nos. 17-42, text-figs.

1892. Additions to the Hydroids of Misaki. Zool. Mag. Tokio, IV, No. 49.

Jaderholm, E. 1904. Mitteilungen ueber einige von der sehwedischen Antarctic-Expedition 1901-03,
eingesamlete Hydroiden. Arch. Zool. exp. gen. Paris (4), III, Notes et rev. pp. I-XIV.

1905. Hydroiden aus antarktischen und subantarktischen Meeren, . . . Wiss. Ergeb. der

Sehwedischen Sudpolar-Expedition 1901-1903, Stockholm, V, 8, 41 pp., 14 pis.

1917. Hydroids from the South Seas. Redogorelse for Norrkopings hogre allmanna laroverk

under lasaret 1916-1917. 23 pp., 2 pis.

1919. Zur Kenntnis der Hydroidenfauna Japans. Ark. Zool. Stockholm, XII, 9, 34 pp., 6 pis.

1926. L’ber einige antarktische und subantarktische Hydroiden. Ark. Zool. Stockholm.

XVIII A, No. 14, 7 pp., 3 text-figs.

Jarvis, F. E. 1922. The Hydroids from the Chagos, Seychelles and other islands and from the coasts
of British East Africa and Zanzibar. Trans. Linn. Soc. London, XVIII, pp. 331-360, pis. 24-26.

Jickeli, C. F. 1883. Der Bau der Hvdroidpolypen. Gegenbaurs morph. Jahrb. Leipzig, VIII, pp. 373-
416, 580-680, pis. 16-18, 25-28.

Johnston, G. 1833. Illustrations in British Zoology. Mag. Nat. Hist. London, VI, pp. 320-4, 497-9,
text-figs. 61-62.

Kirchexpauer, G. H. 1864. Neue Sertulariden aus verschiedenen Hamburgischen Sammlungen,
nebst allgemeinen Bemerkkungen fiber Lamouroux’s Gattung Dynamena. Verb, k Leopold. -
Carol, deuts. Akad. Naturf. Dresden, XXXI, Abh. Ill, 16 pp., 1 pi.

1876. Ueber die Hydroidenfamilie Plumularidae, einzelne Gruppen derselben und ilire Frucht-

behalter. (II. Plumularia und Nemertesia.) Ahb. Naturw. Hamburg, VI, 2, pp. 1-59, 8 pis.

1884. Nordische Gattungen und Arten von Sertulariden. Abh. Naturw. Hamburg, VIII, Heft

III, 54 pp. pis. 11-16.

Kirkpatrick, R. 1890. Report upon the Hydrozoa and Polyzoa collected by P. W. Basset-Smith.
. . . Survey of the Tizard and Macclesfield Banks. . . . Ann. Mag. Nat. Hist., (6) V, pp. 11-24,
pis. 3-5.

Kramp, P. L. 1913. Hydroids collected by the “ Tjalfe ” Expedition to the west coast of Greenland
in 1908 and 1909. Vidensk. Medd. naturh. Foren. Kjob., LXVI, 36 pp., 2 text-figs.

Lamarck (de), J. 1816. Histoire naturelle des animaux sans vertebres, Vol. II, Paris, 8°, 568 pp.

Lamouroux, J. V. F. 1816. Histoire des polypiers coralligenes flexibles, vulgairement nommes
zoophytes. Caen. 560 pp., 19 pis.

Lendenfeld, R. von. 1884. The Australian Hydromedusae. Proc. Linn. Soc. N.S.W., IX for 1884,
pp. 206-241, 345-353, 401-420, 467-492, 581-634, pis. 6-8, 12-17, 20-29.

Levinsen, G. M. R. 1893. Meduser, Ctenophorer og Hydroider fra Grpnlands Vestkyst . . . Vidensk
Medd. naturh. Foren. Kjob., for 1892, 70 pp. 143-220, pis. 5-8.

1913. Systematic studies on the Sertulariidae. Vidensk. Medd. naturh. Foren. Kjob., LXIV,

pp. 249-323, pis. 4-5.

Marktanner-Turneretscher, G. 1890. Die Hydroiden des k.k. naturhistorischen Hofmuseums.
Ann. NatHist. Hofmus. Wien, V, pp. 195-286, pis. III-VII.

Motz-Kossowska, S. 1903. Sur Taction morphog&ne de Teau en mouvement sur les Hydraires. C. R.
Acad. Sci. Paris, CXXXVII, pp. 863-5.

1907. Sur les gonophores des Plumularia obliqua Saunders et Sertularia operculata L. Arch.

Zool. exp. gen. Paris (4) VII, notes et revue, pp. CXIV-CXVIIT, 3 text-figs.

Norman, A. M. 1875. In : Jeffreys, J. G., and Norman, A. M., Submarine-cable fauna. Part II.
Crustacea, etc. Ann. Mag. Nat. Hist., (4) XV, pp. 170-176, pi. 12.

246

“TERRA NOVA” EXPEDITION.

Nutting, C. C. 1900. American Hydroids. Part I. The Plumularidae. Smithson. Inst. U.S. Nat.
Mus. Special Bull., IV, 285 pp., 34 pis.

1901. The Hydroids. Papers from the Harriman Alaska Expedition. XXI. Proc. Washington

Ac. Ill, pp. 157-216, pi. 14-26.

1904. American Hydroids. Part II. The Sertularidae. Smithson. Inst. U.S. Nat. Mus. Special

Bull., IV, 325 pp., 41 pis., 139 text-figs.

1905. Hydroids of the Hawaiian Islands collected by the Steamer “ Albatross ” in 1902. Bull.

U.S. Fish Comm., XXIII, pp. 931-959, pis. 1-13.

Pictet, C. 1893. Etude sur les Hydraires de la Baie d’Amboine. Rev. suisse Zool. Geneve, I, pp. 1-64,
pis. 1-3.

Pieper, F. W. 1884. Erganzungen zu “ Heller’s Zoophyten, etc., des adriatischen Meeres.” Zool. Anz.
Leipzig, VII, pp. 164-169.

Pollard, A. W., and Smith, W. A. 1927. A world list of Scientific Periodicals published in the years
1900-1921. Vol. II. Abbreviated titles and locations of sets. Oxford Univ. Press, London.
Quelch, J. J. 1883. On Thuiaria zelandica, Gray. Ann. Mag. Nat. Hist., (5) XI, pp. 247-249.

1885. On some deep-sea and shallow-water Hydrozoa. Ann. Mag. Nat. Hist., (5) XVI, pp. 1-20

pis. 1-2.

Ritchie, J. 1907. The Hydroids of the Scottish National Antarctic Expedition. Trans. R. Soc.
Edinburgh, XLV, pp. 519-545, 3 pis.

1909. New species and varieties of Hydroida Tliecata from the Andaman Islands. Ann. Mag.

Nat. Hist., (8) III, pp. 524-528.

— - — 1909a. Supplementary report on the hydroids of the Scottish National Antarctic Expedition.

Trans. Roy. Soc. Edinb., XL VII, pp. 65-101, 11 text-figs.

— — - 1911. Hydrozoa (Hydroid Zoophytes and Stylasterina) of the “ Thetis ” Expedition. Mem.
Aust. Mus. Sydney, IV, pp. 807-869, pis. 84-89.

— - — 1913. The Hydroid Zoophytes collected by the British Antarctic Expedition of Sir Ernest
Shackleton, 1908. Proc. R. Soc. Edinburgh, XXXIII, pt. I, pp. 9-34, 11 text-figs.
Saemundsson, B. 1911. Bidrag til Kundskaben om de islandske Hydroider. II. Vidensk. Medd.
naturh. Foren. Kjub., LXIII, pp. 67-107, 6 text-figs.

Sars, G. O. 1873. Bidrag til Kundskaben om Dyrelivet paa vore Havbanker. Fork. VidenskSelsk.
Christiania for 1872, pp. 73-119.

1874. Bidrag til Kindskaben om Norges Hydroider. Forh. VidenskSelsk. Christiania for 1873,

pp. 91-150, pis. 2-5.

1851. Beretning om en i Sommeren 1849 foretagen zoologisk Reise i Lofoten og Finmarken. Nyt

Mag. Naturv. Oslo, VI, pp. 121-211.

Splettstosser, W. 1929. Beitrage zur Kenntnis der Sertulariiden. Thyroscyphus Alim., Cnido-
scyphus nov. gen., Parascyphus Ritchie. Zool. Jahrb. Jena, Syst., LVIII, pp. 1-134, 94 text-
figs.

Stechow, E. 1909. Hydroidpolypen der japanischen Ostkuste. I Teil : Athecata und Plumularidae.

In : Doflein, F. Beitrage zur Naturgeschichte Ostasiens. Abh. bayer. Akad. Wiss., Suppl.-Bd.
I, Abh. 6, 111 pp., 7 pis., 8 text-figs.

1913. Neue Genera thecater Hydroiden aus der Familie der Lafoeiden und neue species von

Thecaten aus Japan. Zool. Anz. Leipzig, XLIII, pp. 137-144.

— — 19136. Hydroidpolypen der japanischen Ostkuste. II. Teil : Campanularidae, Halecidae,

Lafoeidae, Campanulinidae und Sertularidae, nebst Erganzungen zu den Athecata und Plumu-
laridae. Abh. bayer. Akad. Wiss., Suppl.-Bd. Ill, Abh. 2, 162 pp., 135 text-figs.

1914. Zur Kenntnis neuer oder seltener Hydroidpolypen, meist Campanulariden, aus Amerika

und Norwegen. Zool. Anz. Leipzig, XLV, pp. 120-136, 9 text-figs.

1919. Neue Ergebnisse auf dem Gebiete der Hydroidenforschung. SitzBer. Ges. Morph. Munchen,

XXXI, pp. 9-45.

1919a. Zur Kenntnis der Hydroidenfauna des Mittelmeeres, Amerikas und anderer Gebiete.

Zool. Jahrb. Jena, Syst. XLII, Heft 2, 172 pp., 56 text-figs.

1920. Neue Ergebnisse auf dem Gebiete der Hydroidenforschung. SitzBer. Ges. Morph. Physiol.

Munchen, XXXI, pp. 9-45, 10 text-figs.

1921. Neue Genera und Species von Hydrozoen und anderen Evertebraten. Arch. Naturgesch.

Berlin, LXXXVII, A3, pp. 248-265.

HYDROIDA — TOTTON.

247

Stechow, E. 1922. Zur Systematik der Hydrozoen, Stroma toporen, Siphonophoren, Antliozoen und
Ctenophoren. Arch. Naturgesch. Berlin, LXXXYIII A Heft III, pp. 141-155.

1922a, Zur Erage des bipolaren Yorkommens mariner Tiere. Yerh. deuts. Zool. Ges., XXVII,

pp. 95-96

1923. Xeue Hydroiden der Deutschen Tiefsee-Expedition, nebst Bemerkungen uber einige andre

Eormen. Zool. Anz. Leipzig, LYI, pp. 1-20.

19236. Uber Hydroiden der Deutschen Tiefsee-Expedition, nebst Bemerkungen uber einige andre

Formen, Zool. Anz. Leipzig, LYI, PP- 97-119.

1923c. Zur Kenntnis der Hydroidenfauna des Mittelmeeres, Amerikas und anderer Gebiete.

II. Teil. Zool. Jahbr. Jena, Syst. XL VII, pp. 29-270, 38 text-figs.

1925. Hydroiden von West-und Sudwestaustralien. . . . Zool. Jahrb. Jena, L., pp. 191-269,

17 text-figs.

Stechow, E., and Muller, H. 0. 1923. Hydroiden von den Aru-Inseln. Abb. senckenb. naturf. Ges.,

XXXV, 4, pp. 459-478, pi. 27.

Stimpsox, AY. 1853. Synopsis of the marine Invertebrata of Grand Manan. . . . Smithson. Contr.
Knowl., AH, 5, 68 pp., 3 pis.

Thompson, D’Arcy AAt. 1879. On some new and rare Hydroid Zoophytes (Sertulariidae and Thuiariidae)
from Australia and New Zealand. Ann. Mag. Nat. Hist., (5) III, pp. 97-114, pis. 16-19.
Thomson, C. AVyville. 1853. Notes on some British Zoophytes. Ann. Mag Nat Hist., (2) XI, pp. 442-
444, pi. 16a.

Torrey, H. B. 1904. The hydroids of the San Diego region. Univ. Calif. Pub. Zool., II, pp. 1-43,
23 text-figs.

Trebilcock, R. E. 1928. Notes on New Zealand Hydroida. Proc. R. Soc. ATctoria, (n.s.) XLI, pp. 1-
31, pis. 1-7.

VanhOfeen, E. 1910. Die Hydroiden der Deutschen Sudpolar-Expedition 1901-1903. Deutsche
Sudpolar-Expedition 1901 -1903, Berlin, Zool. Ill Heft. IV, pp. 269-340, 49 text-figs.

Verrill, A. E. 1873. Results of recent dredging expeditions on the coast of New England. Amer.
J. Sci. New Haven, (3) Y, pp. 1-16, 98-106.

Versluys, J. 1899. Hydraires calyptoblastes recueillis dans la mer des Antilles. . . . Mem. Soc. Zool.
Fr., XII, pp. 29-58, 24 text-figs.

AVarren, E. 1908. On a collection of Hydroids, mostly from the Natal coast. Ann. Natal Mus.
Pmburg, I, pp. 269-355, pis. 45-48.

— 1914. On the development of the planula in a certain species of Plumularian Hydroid. Ann.

Natal Mus. Pmburg, III, pp. 83-102, pi. 6.

INDEX.

The more important references are indicated by black type.

Abietixella, 133, 166.
acanthocarpa, Aglaophenia, 23-1, 235.
Acryptolaria, 161, 162.
affinis, Selaginopsis, 178.

,, Staurotheca, 175, 176.
africana, Zygophylax, 165.

Aglaophenia, 233.
allmani, Diplocheilus, 216.

,, Pycno theca, 216.

,, Pycnotheca producta, 216.

,, Selaginopsis, 201.
alternitheca, Plumularia, 217.
ambigua, Thuiaria, 203.

Amphisbetia, 203.
amplectens, Sertularia, 204, 205.
andersoni, Acryptolaria, 161, 162.
anderssoni, Schizotricha, 230, 233.
angulosa, Sertularella, 195.
angusta, Hydractinia, 138
angustum, Hydronema, 138.
antarctica, Atractylis, 139.

,, Campanularia, 148.

,, Campanularia volubilis, var., 148, 150.
,, Gravelya, 139.

,, Lafoea, 160.

,, Oswaldella, 144, 209.

,, Polyplumaria, 209.

,, Reticularia, 160.

,, Schizotricha, 209, 230.

,, Staurotheca, 175, 176, 177.
antarcticum, Eudendrium, 140.

,, Filellum, 160.

,, Halecium, 144.

Antennella, 210, 211.
antipathes, Lictorella, 165, 166.

,, Sertularia, 165.
apophyses, Costal, 133.
arborea, Ophiodissa, 142.
arboreum, Halecium, 142.
arboreus, Ophiodes, 142.
arenosa, Atractylis, 139.
argentea, Sertularia, 232.
armata, Ophiodissa, 142.

,, Zygophylax, 165.

Arris, 163.

articulata, Sertularella, 197.

,, Sertularia, 199.

Atractylis, 133, 139.

australis, Acryptolaria conferta, var., 163.

australis, Cryptolaria conferta, var., 163.

,, Dynamena, 203.

,, Sertularella, 181

,, Sertularia, 203.

belgicae, Campanulina, 152.
biarmata, Zygophylax, 165.
bicuspidata, Sertularia, 203.
biformis, Sertularella, 199.
bifurca, Oswaldella, 208, 210.

,, Schizotricha, 208, 230.

,, Sertularella, 197.

Billardia, 150.

Bougainvillia, 139.
brachiata, Plumularia, 226, 228.
brevi theca, Zygophylax, 165.
brownei, Zygophylax, 165.

Brucella, 166.
buski, Plumularia, 230.

calcarata, Hebella, 155.

Calycella, 139.
campanula, Halopteris, 219.

,, Plumularia, 217, 219, 230.
Campanularia, 148.

Campanulina, 152.

campylocarpum, Synthecium, 169, 170.
capillare, Eudendrium, 140.
capillaris, Sertularella, 183.
carinata, Halopteris, 217.
carinatum, Synthecium, 169, 171.

Carolina, Zygophylax, 165.
catharina, Plumularia, 217, 230.
cerastium, Thuiaria, 201, 202.
cervicornis, Lictorella, 164, 165.
challengeri, Sertularia, 203.
chiltoni, Thecocarpus, 240.

Clytia, 146.

columnaria, Sertularella, 180.
columnarius, Symplectoscyphus, 180.
compressa, Clytia, 146.
concinna, Lictorella, 164, 165, 167.
conferta, Cryptolaria, 163.

,, var. australis, Acryptolaria, 163.

,, ,, ,, Cryptolaria, 163.

confusus, Symplectoscyphus, 184, 185, 186.
constricta, Halopteris, 217.
constrictus, Symplectoscyphus, 181.
contorta, Hebella, 155.

250

INDEX.

contorta, Hebellopsis, 155.
convallaria, Lafoea, 164, 165.

Corbulae, 132.

Costae, 133.

Costal apophyses, 133.
crassicaulis, Oswaldaria, 162.

Crateritheca, 206.
crateroides, Hebella, 155.

Crites has ales, 133.

Cryptolarella, 161.

Cryptolaria, 153, 161, 166.
ctenatus, Thecocarpus, 237.
cumberlandica, Sertularella, 187, 188.
curvatus, Symplectoscyphus, 192.
curvitheca, Zygopbylax, 165, 168.
cyatbifera, Lictorella, 165.
cylindrica, Hebella, 155.
cymiforme, Halecium, 144, 145.

delicatula, Sertularella, 183.

,, Sertularia, 181.
delicatulus, Symplectoscyphus, 183.
dentigerum, Synthecium, 172.
desmoides, Sertularia, 205.

Desmoscyphus, 203.
diaphana, Plumularia, 230.
dichotoma, Selaginopsis, 202
,, Schizotricha, 230.

,, Sertularella, 189.

,, Staurotheca, 175, 176, 177, 178, 179.

dichotomum, Dictyocladium, 201, 202.
Dictyocladium, 201.
dimorphism, Sexual, 132.

Diphasia, 207.

Diplocheilus, 214.
diploptera, Plumularia, 222.

Diplopteron, 230.

divaricata, Sertularia, 187, 188.

divaricatus, Symplectoscyphus, 182, 184, 188.

divergens, Sertularia, 203.

diverticulata, Tulpa, 145.

dolichocarpa, Thuiaria, 206.

dumosa, Lafoea, 158.

„ Sertularia, 158.

Dynamena, 203, 204.

edentula, Sertularella, 200.
elegans, Halicornopsis, 214.

,, Synthecium, 168, 172.
elongata, Clytia, 148.

,, Nemertesia, 229.

,, Sertularella, 197.

,, Sertularia, 207.

epizooticus, Symplectoscyphus, 185.

Eudendrium, 140.
exserta, Acryptolaria, 161.

„ Cryptolaria, 161.

,, Perisiphonia, 162.

False branching, 234.
fastigiata, Campanularia, 155.
fastigiatum, Stegopoma, 155.

Filellum, 160.
filiformis, Sertularia, 194.

,, Symplectoscyphus, 185, 194.

flabellum, Dictyocladium, 201.
flexile, Halecium, 144, 145.
flexilis, Sertularia, 203.
flexuosa, Plumularia, 221, 222.
flosculus, Sertularia, 203, 204.
formosus, Thecocarpus, 238.

,, var. inarmatus, Thecocarpus, 237.

frutescens, Schizotricha, 230.

,, Sertularia, 230.

fruticosa, Lafoea, 157.
fusca, Staurotheca, 201.

fuscum, Dictyocladium, 175, 176, 177, 178, 201.

gaussica, Lafoea, 159.
gayi, Sertularella, 196.
geniculata, Cryptolaria, 153.

,, Stegolaria, 154.

Genotypes, 133.

geodiae, Sertularella, 196.

glacialis, Plumularia, 230, 231, 233.

,, Schizotricha, 232, 233.

,, Sertularella, 188.

,, Symplectoscyphus, 188, 193.

Gonocladium, 132, 234.

Gonohydrocladium, 133, 235.
gracile, Halecium, 145.
gracilis, Acryptolaria, 162.

„ Desmoscyphus, 204.

,, Sertularia, 194.
gracillima, Campanularia, 158.

,, Lafoea, 158.

,, Plumularia, 230.

Grammaria, 133, 161.
grande, Diplopteron, 230.

,, Schizotricha, 230.
grandis, Campanularia verticillata, var., 153.

,, Stegella, 153.

,, Zygophylax, 166.

Gravelya, 139.

halecioides, Lictorella, 166.

Halecium, 143, 145, 232.

Halicornaria, 241.

Halicornopsis, 214.

Halopteris, 217.

Hebella, 155.

Hebellopsis, 155.
heterogona, Halopteris, 217.

„ Thecocaulus, 217.
heteromorpha, Thuiaria, 203.

Heterotheca, 230.

INDEX.

251

hicksoni, Campanularia, 148.

Hincksella, 133, 151.

Hinge-joint, 218.
hippisleyana, Thuiaria, 206, 207.
humilis, Acryptolaria, 161.

,, Halicornaria, 242.

Hydronema, 138.

inabai, Pycnotheca producta, 216.
inarraatus, Thecocarpus formosus, var., 237.
inflatus, Desmoscyphus, 203, 204.
insigne, Eudendrium, 140, 141.
irregularis, Stegolaria, 154.

johnstoni, Clytia, 146, 147, 148.

,, Sertularella, 181, 183.

,, Symplectoscyphus, 181, 183.

kirchenpaueri, Aglaophenia, 236.
Kirchenpaueria, 214.

laevis, Campanularia, 148, 149.

Lafoea, 157.

Laomedea, 179.
laxa, Aglaophenia, 233.
laxus, Thecocarpus, 233, 236.
levinseni, Zygophylax, 165.

Lictorella, 133, 165.
liechtensterni, Plumularia, 230.
loculosa, Sertularia, 205.
longipinna, Diplopteron, 230.
longithecum, Synthecium, 174.

Lytocarpus, 234.

macrocarpa, Sertularella, 181.

Mamelon, 214.

rnarginata, Amphisbetia, 204.

,, Dynamena, 204.

,, Sertularia, 133, 204.

Measurements, 132.
meridionalis, Sertularella, 180.
microgona, Sertularella, 195.
minima, Acryptolaria, 162.
mirabilis, Diplocheilus, 214, 216.

,, Pycnotheca, 214, 216.
monilifera, Selaginopsis, 202.
moniliferum, Dictyocladium, 202.

Monopyxis, 222.

Monostaechas, 220.

Mono theca, 222.

multifurcata, Schizotricha, 230, 231.
muscoides, Perigonimus, 139.

neglecta, Sertularella, 199.

Nemertesia, 214, 229.

Nodes, 133.

novae-zealandiae, Billardia, 150.

V. 5.

novae-zelandiae, Eudendrium, 141.
Nuditheca, 230.

obliqua, Plumularia, 222.
operculata, Zygophylax, 166.

Ophiodes, 142.

Ophiodissa, 142.
orthogonium, Synthecium, 168.
Oswaldaria, 161, 162.

Oswaldella, 132, 209, 210.
ovatum, Halecium, 143, 144.

pachyclada, Selaginopsis, 132, 176.
pacifica, Zygophylax, 165.

Parascyphus, 179.
parasiticus, Thecocarpus, 222.
Parathecium, 161.
parvula, Plumularia, 230.
patulum, Synthecium, 168.
pectinata, Perisiphonia, 166.
pectinatus, Desmoscyphus, 203, 204, 205.
pedunculata, Sertularella, 181.
Pericladium, 207.

Perigonimus, 139.

Perisiphonia, 133, 161, 162, 166.
pinaster, Diphasia, 207.
pinnata, Kirchenpaueria, 214.

,, Lafoea, 165.

,, Zygophylax, 165.

plana, Hebella, 156.

,, ,, striata, var., 156.

plectilis, Symplectoscyphus, 188, 192, 193.
plicatile, Stegopoma, 154.
pluina, Aglaophenia, 236.

,, Sertularia, 204, 205.

Plumularia, 221.

Plumulariidae, 132.
polymorpha, Thecocaulus, 218.
Polyplumaria, 209.
polyzonias, Sertularella, 196.
prima, Cryptolaria, 161, 166.
producta, Pycnotheca, 214.
profunda, Plumularia, 230.

,, Schizotricha, 230.

,, Zygophylax, 164, 165.
prolifera, Halicornaria, 242.
pulchella, Plumularia, 221.

,, Sertularia, 203.

Puncta, 145.

Pycnotheca, 133, 214.

quadriaurita, Antennella, 212, 213.
quadricorne, Diplopteron, 230.
quadriseriata, Perisiphonia, 161.

rameum, Eudendrium, 140.
ramosum, Bougainvillia, 139.

1G

INDEX.

252

ramosum, Eudendrium, 139, 140, 141
,, Synthecium, 168, 170.
ramosus, Lytocarpus, 234.
recta, Zygophylax, 165.
regalis, Halicornaria, 241.
repens, Eudendrium, 139.

,, Perigonimus, 139.
restricta, Sertularia, 205.

Reticularia, 160.
reticulata, Sertularia, 201.

,, Stauro theca, 177. 178.

ritchiei, Antennella, 211.
robusta, Grammaria, 161.

,, Pycnotheca mirabilis, var., 214, 216.
,, Sertularella, 195.
robustum, Halecium, 142.

,, Synthecium, 173.

rostrata, Halicornaria, 237.
rostratus, Thecocarpus, 237.
rufa, Lictorella, 165.

samauense, Synthecium, 173.
scalariformis, Diphasia, 207.
scandens, Hebella, 155. •

Scapus, 142, 161, 162.

Schizotricha, 132, 229.

Selaginopsis, 201.
serpens, Campanularia, 160.
serrata, Antennella, 212.

Sertularella, 195.

Sertularia, 133, 203.
sessile, Eudendrium, 139.
sessilis, Perigonimus, 139.
setacea, Plumularia, 224.

Sexual dimorphism, 132.
sibogae, Zygophylax, 165, 167.

,, Hincksella, 151.

simplex, Laomedea, 179.

,, Parascyphus, 179.

,, Schizotricha, 230.

,, Sertularella, 196.

,, Thecoeaulus, 218, 220.

singulare, Synthecium, 172, 173.

Spherules, 146.
spiralis, Sertularella, 197.

,, Thecocarpus, 238.
spiritualis, Symplectoscyphus, 184.
spirocladia, Plumularia, 224.

Staurotheca, 132, 175.

Stegella, 153.

Stegolaria, 153, 161.

Stegopoma, 133, 153, 154.

Stereotheca, 207^
striata, Hebella, 155, 156.

,, Sertularia, 204, 206.

striata, var. plana, Hebella, 156.
subarticulata, Thuiaria, 188.
subarticulatus, Symplectoscyphus, 182.
subdichotoma, Sertularella, 187, 188.
subrufa, Billardia, 151.

,, Campanularia, 150, 151.
sulcata, Plumularia, 230.

Symplectoscyphus, 132, 180, 201, 202.
Synthecium, 168.

tenella, Plumularia, 230.

,, Sertularia, 195.
tenuissima, Aglaophenia, 240.

,, Plumularia, 228.

Thecocarpus, 235, 240.

Thecoeaulus, 133, 217, 230.

Thuiaria, 203.

tizardensis, Zygophylax, 165, 168.
triangulata, Plumularia, 225.

Tridentata, 203.
tripartita, Plumularia, 224.
tuba, Symplectoscyphus, 186.

Tulpa, 145.

turbinata, Sertularia, 205.
turqueti, Schizotricha, 230, 231.

unguiculata, Amphisbetia, 203.

,, Desmoscyphus, 203.

,, Sertularia, 203.

unifurcata, Schizotricha, 229, 230, 231.
unilateralis, Zygophylax, 167.

valdiviae, Zygophylax, 165.

valida, Thuiaria zelandica, var., 206.

vanhoeffeni, Symplectoscyphus, 187, 189.

variabilis, Plumularia, 230.

ventruosa, Plumularia, 225.

versluysi, Sertularia, 204, 205.

verticillata, var. grandis, Campanularia, 153.

volubilis, Campanularia, var. antarctica, 148, 150.

warreni, Pycnotheca mirabilis, 216.

Wrightia, 139.

xantha, Tridentata, 203.

zelandica, Crateritheca, 206.

,, Halopteris campanula, var., 219, 220.

,, Stereotheca, 206.

,, Thuiaria, 206.

,, var. valida, Thuiaria, 206.

zygocladia, Plumularia, 230.

Zygophylacinae, 162.

Zygophylax, 164.

PRINTED IN GREAT BRITAIN BY WILLIAM CLOWES AND SONS, LIMITED, LONDON AND BECCLF.S.

Coelenterata, Part V. — Hydroida, PI. I.

PLATE I.

Fig. 1. — Zygophylax unilateralis, sp. n., x 5 ; stem and branches.

Fig. 2. — Zygophylax unilateralis, sp. n., x 12-4 ; branch.

Fig. 3. — Symplectoscyphus constrictus, sp. n., x 6*2.

Fig. 4. — Symplectoscyphus confusus, sp. n., x 3-9.

Fig. 5. — Symplectoscyphus epizooticus, sp. n., X 13-8.

Fig. 6. — Symplectoscyphus epizooticus, sp. n., growing upon S. confusus, sp. n., x 6-6.

Fig. 7. — Sertularella spiralis, Hickson and Gravely, Schizoholotype, McMurdo Bay, “ Discovery ” Expedi
tion 1901-4, x 16-3.

Fig. 8. — Symplectoscyphus glacialis, Jaderh., forma elongata, X 12-7.

Fig. 9. — Symplectoscyphus glacialis, Jaderh., showing gonothecae, X 12*7
Fig. 10. — Symplectoscyphus columnarius (Briggs), X 6-3.

Brit. Antarctic (Terra Nova) Exped. 1910.

Brit. Mus. (Nat. Hist.) Zoology, Vol. V, No. 5. Coelenterata, Part V, PI. T.

■

PLATE II.

Fig. 1. — Sympledoscyphus curvatus (Jaderh.), x 10-4.

Fig. 2. — Sympledoscyphus curvatus (Jaderh.), X 13-5 ; showing gonothecae.

Fig. 3. — Sympledoscyphus curvatus (Jaderh.), X 9.

Fig. 4. — Sympledoscyphus pledilis , Hickson and Gravely, X 13 ; “ Discovery ” Expedition, 1901-4 ;
to show form of gonothecae.

Fig. 5. — Sympledoscyphus pledilis, Hickson and Gravely, x 11-9.

Fig. 6. — Staurotheca antardica, Hartlaub, x 11*6 ; to show form of female gonotheca.

Fig. 7. — Sympledoscyphus f/lacialis ( Jaderh.), x 16*9.

Fig. 8. — Sympledoscyphus biformis ( Jaderh.), X 11.

Fig. 9. — Staurotheca dichtoorna, Allman, X 7-8 ; Marion Island, “ Challenger ” Expedition ; to show
form of male gonothecae.

ahL Um f /

my

M

# J

m

Jam

v \-&nh9I^ / jrf

Mis

yOTBjr

AjT-L

A' ^05428^

Bjy

I'TOj / /

•<A J ,4 /

A' ’l»0k

UsSL

Brit. Mus. (Nat. Hist.)

Brit. Antarctic (Terra Nova) Exped. 1910.
Zoology, Vol. V, No. 5.

Coelenterata, Part V, PI. II.

v. 5.

17

Coelenterata, Part Y. — Hydroida, PI. III.

PLATE III.

Fig. 1. — Crateritheca zelandica (Gray), x 15-5.

Fig. 2. — Crateritheca zelandica (Gray), x 6-3 ; stem and hydrocladia.

Fig. 3. — Crateritheca zelandica (Gray), X 10-2 ; with a single gonotheca.

Fig. 4. — Schizotricha unifurcata , Allman, x 8-5.

Fig. 5.- — Sertularella spiralis, Hickson and Gravely, x 9.

Fig. 6.- — Sertularella edenlula, Bale, x 9*1 ; the single fragment from Station 90.
Fig. 7. — Sertularella geodiae, sp. n., X 9-1 ; to show form of gonothecae.

Fig. 8.- — Sertularella geodiae, sp. n., X 6-2.

Fig. 9. — Symplectoscyphus filiformis, Allman ; to show form of female gonothecae.

Brit. Antarctic (Terra Nova) Exped. 1910.

Zoology, Vol. V, No. 5 . Coelenterata, Part V, PI. III.

Brit. Mus. (Nat. Hist.)

mmm

'~y £ la

2,

Z_QJ

7 /• , ■ ■ ..■■■■■■.

- . ' / ' .a . - ■ ■ , .

■ , '

“ 7 . • - . - . ... * • . ' .

W.

■-> ’ 7 > 7-'?'

/ v

■ ■'

-

'£ ■?!& i-'~- 1