LIBRARY

OF THE

UNIVERSITY OF CALIFORNIA.

RECEIVED BY EXCHANGE

Class

PRESENTED

'©trustees

OF

THE BRITISH MUSEUM.

-•--"' ''-':•• i

.

-'-'.'••

>..'-. ,••

M

NATIONAL ANTARCTIC EXPEDITION

1901-1904

NATURAL HISTORY

VOL. IV.
ZOOLOGY

(VARIOUS INVERTEBRATA)

LONDON

PRINTED BY ORDI.K or TNI. IRIMKES OF
I UK HKITISII MUSEUM

1908

(All Rights Reitrvat)

SOLD BY LONGMANS AND Co., 39 PATERNOSTER Row, E.G. ; BERNARD QUARITCH, 15 PICCADILLY, W.

DULAU AND Co., 37 SOHO SQUARE, W. ;
AND AT

THE BRITISH MUSEUM (NATURAL HISTORY), CROMWELL ROAD, LONDON, S.W.

PREFACE .

WHKS, in 1901, the Expedition of the 8.8. 'Discovery,' under Captain Scott, R.X.,
was sent to the Antarctic Regions, the Trustees of the British .Museum gave their
•ntstance to this national enterprise by allowing the cases containing the natural
history specimens which might lie obtained by the Expedition to be sent to the
Natural History Museum for unpacking and sorting. They further undertook to
publish a detailed report on the collections so obtained, under the superintendence
of the Director of the Natural History Departments.

8ome of the most important collections have l»een dealt with by naturalists who
were mem tars of the Expedition. Thus, the Mammals* and -Birds* arc dcscrilxxl by
Dr. Edward A. Wilson, tlie Isopoda and Pycnogonida* by Mr. T. V. Hodgson, and
the Rocks* (in relation to Field Geology) by Mr. H. T. Ferrar. Other groups have
been dealt with by members of the staff of the Natural History Dcjmrtments of
the British Museum : Mr. Boulenger describes the Fishes* ; Mr. E. A. Smith, the
Qtttropoda,* Lamcllibranchia,* and Brachiopoda* ; Mr. Jeffrey Bell, the Echinoderma* ;
Dr. Caiman, the Crustacea Decapoda,* and the Cumacea* ; Mr. Kirkpatrick, the non-
. al.iireous Sponges*; whilst Dr. G. T. Prior has prepared a petrographical description
of the Rock-specimens.*

It has been necessary to obtain the assistance of other specialists in order to deal
with the rest of the collections. So far as the latter group of contributors is
concerned, the following is a list of the subject-matters, together with the name of the
naturalist who has undertaken the work in each case :—

EMBRYOS OF SEALS . . .DR. MARRETT TIMS.

•ANATOMY OF EMPKROR PKNOUIN . . MR. W. P. PYCRAFT.

TUUICATA . . . PROF. HERDMAN.

•CBPHALODISCt-S . . DR. RlDE^VOOD.

•CEPHALOPODA DR. HOYLE.

•NUDIBRAUrHS AM> PTEROPODS .. SlR CHARLES ELIOT, K.C.M.O.

POLYZOA . . . MR. II. W. BiRROwa

•EGOS AND YOUNO OF AsTBRIA.H . PROF. \l \< I'.!'.! M

•AMPHIPODA . . . . MR. A. 0. WALKER.

•SCHIZOPODA MR. HOLT.

•NKBALI.K . .... DR. J. THIELE.

•OSTRACODA PROF. BRADY.

•COPBPODA .DR. WoLFESDW.

iv PREFACE.

*CIRRIPEDIA ...... PROF. GRUVEL.

*MYZOSTOMA ...... PROF. v. GRAFF.

*ACARI ....... DR. TROUESSART.

*COLLEMBOLA ...... PROF. CARPENTER.

POLYCH^ETA ...... PROF. EHLERS.

*GEPHYRIA ...... MR. A. E. SHIPLEY.

*CHy£TOGNATHA ...... DR. FOWLER.

NEMERTINES ...... PROF. HUBRECHT.

FREE PLATYHELMINTHES .... MR. F. F. LAIDLAW.

*CESTODA ....... MR. A. E. SHIPLEY.

*NEMATODA ...... DR. v. LINSTOW.

*ZOANTHARIA . . . . . .MR. CLUBB.

*ALCYONARIA AND PENNATULIDA . . . PROF. HICKSON.

HYDROMEDUS.E ...... MR. E. T. BROWNE.

*CALCAREOUS SPONGES ..... MR. FREWEN JENKIN.

RADIOLARIA ...... MR. LEWIS H. GOUGH.

* MOSSES ....... M. JULES CARDOT.

LICHENS ....... MR. DARBISHIRE.

(MARINE) ...... MRS. GEPP.

; (FRESH- WATER) ..... DR. FRITSCH.

(CALCAREOUS) . ... DR. FOSLIE.

PHYTOPLANKTON ...... DR. LEWIS H. GOUGH.

The work of securing the assistance of these specialists and of distributing the
collections has been performed by Mr. Jeffrey Bell, of the Zoological Department, who
has also acted as sub-editor of the Zoological and Botanical portions of the reports.
The Keeper of Minerals, Mr. Fletcher, has superintended the reports in the subjects
belonging to his department.

The Director desires to acknowledge the ability and energy which have been
brought to bear on the preparation of the Zoological reports by Mr. Jeffrey Bell.
Owing to his care, the reports have been got ready by the various contributors and
published within a reasonable time after the return of the ' Discovery ' from the
Antarctic Regions. Neither trouble nor expense has been spared in order to
render the illustration and presentation of the Natural History of the Expedition
worthy of the generous efforts both of Captain Scott and his fellow-explorers and
of those who provided the funds for that enterprise.

E. RAY LANKESTER.
October, 1906.

* Have been or are now published.— 1/3/1908.

( ONTENTS OF VOL. IV.

MOLLUSCA.
\II.-SoLEXooA8nuB. By DR. H. F. NIERSTRASZ . (1.3 pp., 2 Pis.)

ARTHROPODA.

(A) IN8ECTA.
APTKRA. By G. H. CARPENTER, B.Sc.,M.R.I.A. (5pp., 1 PI.)

(B) CRUSTACEA.
VII.— SCHIZOPODA. By W. M. TATTBRSALL, M.Sc. . . (42 pp., 8 Pis,)

VIII.— COPBPODA. By R NORRIS WOLFENDEN, M.I). . . (44 pp., 7 Pis.)

ECHINODERMA.

I. — ECHINODERMA. By F. JBFKREY BELL, M.A. . . (16 pp., 5 Pis.)

II.— ECHINODERM LARV^. By E. W. MACBRIDE, M.A., F.R&,

and J C. SIMPSON, RSc. . (9 pp., 1 PI.)

MYZOSTOMIDAE.
By DR. RUDOLF RITTER v. SITMMKR-TRACNFEL.H (26pp., 1 I'l )

BIPUNCDLOIDBA

By W. F. LANCHESTER. .M.A. . . (''• i>|- »

IV. A.TIXL*. By .1. A. CLUBB, M.Sc. . . (12 pp., » Pis.)

II Ti-rrK.\\..Mi.A. By R. KIRKI-ATRI. K . . . (56 pp., 19 Pis.)

Ill TALAUKA. By ( K. JENKIX, B.A. . (50 pp., 12 Pk)

MOLLUSCA.

VH. SOLENOGASTEES.

By l>r. 11. F. NIKRSTRASZ.
(2 Plates.)

THR * Discovery ' expedition brought back one specimen of Solenogastres from 77° 50'
30" 8., 165° 40' 5" E., 100 fathoms, of which Fig. 1, A gives a representation. The
length is 23 mm., the diameter 2*25 mm., the index therefore about 10. Proximally
the animal is blunt, slightly broadened, with a slit-like mouth (Fig. 1, B). The distal
part terminates in a dorsal prolongation. On the ventral side the rhomboid opening
of the cloaca is distinctly visible, as also the ventral groove, which runs OH far as that
opening (Fig. 1, C). The opening of the ciliated groove (" Flimmerhohle " of Win'-n)
is also visible. The animal has no lustre. When slightly enlarged, spicules are seen
crossing each other at obtuse angles. The colour of the animal in alcohol is a yellowish
brown.

After being decalcified in nitric acid (1 per cent in alcohol 90 per cent.) nn<l
being stained in iron carmalum for twenty-four hours, transverse sections won- ma«Ir
of the proximal and distal portions for the study of the internal organs. The .-]>i< ulrs
of the different parts of the body were isolated in Eau de Javelle. They are small,
luit strictly uniform. They belong to the true Proneomenia type : hollow, straight, or
more or less curved or 8-shaped calcareous tubes, which end either sharply or bluntly
(Fig. 2). In some cases the base is slightly broadened. I have not succeeded in
tilling spicules of different shapes along the ventral groove, nor at the proximal and
distal portions. At the proximal end, however, they are somewhat more curved.

The spicules are arranged in different layers upon each other. The whole of tin-
strong cuticle is pierced by numerous papillae. The greater part of these are oval, pear-
"i- dub-shaped, sometimes round, and they show some transparent cells (Fig. 4). They
are placed on strong multinuclear stalks, the cells of which are more or less fibrous.
Amongst the transparent cells granular supporting cells are found. It is worthy of
note that these papillae are often branched. The side-branch never reaches the length
of the stalk, though ending also in a papilla. In most cases these epidermal papillae
do not open on to the exterior, but remain closed. Thiele believes the papillae to be
sensitive, an opinion which I share (18, p. 280). In the stalk a thin thread may
occur, arising from the circular muscular layer. Whether this thread is of a nervous
character or not remains undecided. There is, however, no penetration into the < in-ular

, IT. B

H. F. NIERSTRASZ.

muscular layer by nervous threads. I am also inclined to ascribe to the epidermal
papillae, and more especially to the granular cells amongst the transparent ones, a
secretory function, if not an excretory function as well. The purpose of such a
secretion is doubtful. Perhaps it prevents the cuticle from damage by causing small
particles to adhere to its surface. In any case, it is remarkable that the cuticle of the
Solenogastres is nearly always covered with a thin layer of mud firmly sticking to it.

The spicules are formed in small cellular accumulations in the epidermis. Such
accumulations frequently occur, and consist of a few transparent cells (Fig. 3, A). In
one of the cells the spicule is formed. This cell often has two nuclei, a fact which may
give rise to the supposition that there are two cells that form the spicule. The
spicules are pushed to the periphery of the cuticle, and are connected with the
epidermis by short stalks. The older spicules lose their connection with these stalks,
but preserve the mother-cells, which surround their base like a cap (Fig. 3, B). In
the case of the oldest spicules, however, this cap has also disappeared. Nowhere did
I meet with a formation of the spicules such as Thiele described in Amphimenia
neapolitana, Thiele (13, p. 246). Thiele is also of the opinion that it is one cell only
that produces a spicule.

• The ventral groove is distinct and deep. In it three folds of the epidermis are
found, a large and broad median fold and two small lateral ones. All three run as far
as the opening of the cloaca. Close before the latter the groove and the median fold
broaden considerably. The folds are formed of a single layer of epithelial cells. All
along the median fold the ventral glands open between its cells. These glands (" hintere
Bauchdriise," of Wiren) are strongly developed, especially at the proximal end of the
body, where they entirely surround the digestive tract (Figs. 6-11, vg). They are
conspicuous by their more or less distinct vesicular structure and granular contents,
which stain intensely. At the proximal part of the body, amongst the strongly
developed ventral glands, voluminous anterior glands (" vordere Bauchdriise," of
Wire'n) are also found. They are of a more delicate structure, and stain much less
intensely (Fig. 6, ag). They open out between the epithelial cells which invest
the ciliated cavity (" Flimmerhohle," Fig. 6, /). The latter is large, and divided by
a median dorsal fold. The wall carries strong cilia.

There are two dorso-terminal sense organs situated on the distal lip behind the
cloaca. Fig. 5 shows three sections through the more proximal of these organs. The
epithelium (ep) shows an evagination (Fig. 5, C) into which the muscular layers and the
connective tissue continue (m). This evagination separates ventrally from the epidermis
and pierces the thick cuticle (cut. Fig. 5, B). It forms a pedicle (peel.), upon which a
round knob is resting, the sense organ visible at the surface of the cuticle (Fig. 5, A,
did). The external layer of this knob consists of high epithelial cells, amongst which
slender, fine, easily staining cells are found, which I believe to be sense-organs.

The mouth-slit is terminal and vertical. It leads into a spacious mouth-cavity,
provided with numerous ramified cirri, as is the case in other forms. A horseshoe-

.I:N

i] iped MM orgui it mating, tfcoogfc n-mii*, praortfaiollM gwoii Ib >i; •• tin
tract opens to the exterior at the aamo place an the mouth-cavity, with which it has,
however, no communication (ef. Dinomt-nin, 6, p. 12, PI. 2, Fig. 43). I share Thiele's
opinion, who does not regard the so-railed mouthA-avity as a part of the digestive
ira. t. liut considers it ait imlcpcn.1. 'lit c< to<lcrmal sense-cavity (17, pp. 308 and 314).
A proper mouth-cavity occurs in the proximal part of the digestive tract, which in
slightly dilated, and the wall of which presents numerous folds. The epithelial cellM
of which the wall of this part is composed are very transparent and have large oval
mn-lei. The wall is surrounded by a thin circular muscular layer. The folds, as it
appears to me, allow this portion of the digestive tract to extend considerably after
tin- taking of food. I do not believe them to lie sensitive (Figs. G and 7, />/<).
*(!radually this portion passes into the following part of the alimentary canal. Tin-
latter is much wider, and likewise provided with a strongly folded wall (Fig. 8, re*).
Its cells are taller than those of the first portion, and have elongated nuclei ; they arc
covered with a cuticle. Consequently these two part* of the digestive tract arc easily
• list inguished from one another. More distally, however, between these slender cells
there occur broader ones with more oval nuclei, the cuticle no longer being easily
discerned throughout Finally the spacious intestine unites with the proximal caecum
(Fig. 8, c).

In the second portion of the digestive tract the radula, which is of a very
peculiar structure, is found. In the digestive tract a proximally directed cone appears
with a strong ventral fold (Fig. 8, en). This cone is beset throughout with radula-
teeth (Fig. 12, r). For its further structure compare figures 12-14 and 8-11. An
epithelial layer appears (a), which covers the radula. The ventral wall of the
digestive tract (6) unites with the ventral wall of the cone (c) ; the epithelial layer
a with the side-walls of the cone. The two lateral walls of the proximal part of the
alimentary canal unite dorsal from the cone. By this complicated process the cone,
which on the outside is beset with small teeth, is divided into two closed pouches ; the
latter are on the inner side closely beset with teeth and are situated outside the
alimentary canal. The dorsal pouch is the radula-sac, the ventral contains adult
radula-teeth and, distally, ends blindly. Consequently the ventral pouch is but
a ventral caecum of the proximal part of the digestive tract, directed distally ;
teeth on the ventral wall are wanting, in the distal part teeth only occur <m
the dorsal wall The radular teeth are well developed (Fig. 16). They are
triangular, with a broad base, and more or less strongly curved, often recurved
at the point (Fig. 17, B). Frequently forms occur as represented in Fig. 17, A ;
:i other forms are met with, which, however, I consider fragments of whole
teeth. Fig. 17, C shows a tooth as seen from the ventral side : the base
appears to be broader than the curved part The teeth are placed in irregular
rows, and rest in crypts of the epithelium (Fig. 16). A basal membrane is
present This is remarkable; Thiele lays. stress on the alwencc of such a membrane

I |

4 H. F. NIERSTRASZ.

in different forms and doubts its existence (17, pp. 268 and 315). As has been
indicated by me elsewhere, the existence of a basal membrane has to be accepted for
several forms (9). The membrane is stained red with carmalum, consequently its
structure differs from that of the chitinous teeth. Probably it must be considered
as a cuticle, and so its origin is entirely different from that of the radula-teeth, which
are formed in the distal portion of the radula-sac by large odontoblasts. In this
respect Proneomenia discoveryi resembles those forms, already described by me
elsewhere (9). So we have to do here with a multiserial radula with basal membrane.
Of any double character of the radula, viz., of the existence of a right and a left
half, no trace is to be found. The distal portion of the radula-sac, however, is
divided into two parts, a right and a left one (Figs. 10, 20) ; in both parts radula-
teeth are formed by a thick cushion of odontoblasts. Here also the radula originates*
from the coalescence of a right and a left portion.

The wall of the cone and the dorsal wall of the ventral pouch consist of
epithelial cells with round nuclei, between which numerous extremely slender
supporting cells occur with elongated nuclei (Fig. 16). On the ventral wall of the
ventral pouch and on the dorsal wall of its distal end supporting cells are absent ;
this ventral wall is formed by cylindrical epithelium with oval nuclei, whilst the
constituent, cells of the dorsal wall of the distal portion are higher and more
cylindrical. One part of this dorsal wall more especially deserves notice ; the distal
and median part is invaginated and of an entirely different structure. Here very long
and slender cells are found, with pointed nuclei and granular contents. It leaves the
impression that these cells have pushed themselves wedgewise between the epithelial
cells; yet no sharp line can be drawn between these two elements (Fig. 18). Upon
this wedge-shaped part the proximal rows of the radula-teeth are resting ; the basal
membrane, however, fails here. As to the significance of this wedge I am at a loss.
There seems to be a connection between the wedge and the layer of muscular and
connective tissue surrounding the radula-sac, numerous fibrils pushing themselves
between the cells of the wedge. I am bound to mention its existence, having found
in a similar ventral pouch containing the proximal portion of the radula in Proneomenia
weberi, a regeneration either of the cuticle or of the radula (6, p. 4). But in this
new form I do not believe we have to do with a regeneration ; its significance remains
undecided.

Still, in another respect the ventral pouch is of great importance, as the
salivary glands open into it (Figs. 11 and 15 sg). These are tube-shaped, and extend
a long way distally ; they merge separately into the ventral wall of the ventral
pouch.

The connective tissue and the muscles, which are in connection with the radular
apparatus, are strongly developed. Ventrally and laterally from the radula-sac
numerous crystal-clear cells occur ; usually those are small, sometimes, however, of
large size (Fig. 10 cc). They serve as a _ support for the surrounding connective

r.\'". VSTIM-> 5

tissue anil muscle*. The course of these muscles is ditficult to trace ; they surround
the railula-sae and the distal portion of the ventral pouch ; a very broad transverse
muscle takes its course between the radula-sac and the ventral pouch (Fig. 10, m).
Around the salivary glands a thin muscular layer likewise occurs. I do not know
how to explain the mechanism of the radular movement No doubt, the diverse
muscles allow of movement in various directions.

• •

A Bubradular organ, as found by Heath (1) in Proneununia and Itlwpalomenia,
is wanting.

The spacious intestine is formed by the coalescence of the already described
portion of the digestive tract with the dorsal caecum (Fig. 8). I have avoided making
VW of the names pharynx and oesophagus, as there cannot IKS question here of any
sharp separation between the two, even though the proximal part is of a different
structure from the following portion. Thiele mentions a strong fold in which the
salivary glands open out, and which separates pharynx from oesophagus (18, p. 288).
A similar condition does not exist here. An oesophagus, in the sense Thiele ascribes to
it, fails entirely.

The spacious intestine shows strongly folded walls, with the exception of the
dorsal wall, which remains smooth, except in the median line, where a small fold
occurs. The dorsal wall carries cilia ; the ventral and lateral walls show glandular
epithelium. Lateral caeca are met with as usual ; the transverse sections do not teach
much in regard to their relative situation. The rectum has strong cilia (Fig. 21, r).

The cloaca with its strongly folded wall opens to the exterior through a wide
longitudinal slit. Gills are entirely wanting. The wall of the cloaca is very thick and
consists of several layers of small transparent cells with small nuclei, between which
slender ciliated supporting cells occur with narrow long nuclei, which stain well
(Fig. 19). As it seems to me, the wall of the cloaca does not perform any respiratory
function, it being too thick. A secretory function must not IKJ excluded. Oftentimes
in the cloaca and against its wall little feebly staining globules are found, which I
believe to be formed by the wall -cells of the cloaca. Yet we must take into
consideration that the cloaca is surrounded by extremely loose connective tissue with
large spaces, filled with blood-corpuscles. This arrangement more especially obtains
in that part of the cloaca which is contiguous with the dorsal lip ; this part is entirely
surrounded by blood-corpuscles, and, as it is only clothed by a single layer of epithelial
cells, a respiratory function may \te ascribed to it The lateral walls of the ventral
slit, by which the cloaca communicates with the exterior, are coated with ciliated
cylindrical epithelial cells (Fig. 19).

Properly speaking, the praecloacal organ does not open into the cloaca, but into
the ventral slit, as well as both cloacal spicules (Figs. 19, 20). The latter (Kloaken-
spicula of Thiele) are well developed, and consist of small bars of about * 75 mm. length.
These bars are probably calcareous. Around the liars a strong cuvclojM- i.s found,
which i- not disM ilvt-d in nitric acid, but stains easily with carmaiuni. Probably

6 H. F. NIERSTRASZ.

this envelope is of chitinous character (Fig. 24). The two extremities of the envelope
are tube-shaped (Fig. 24, B) ; it is open in the middle portion (Fig. 24, A), which may
have beon caused by damage. Around the proper spicule a cellular envelope is found,
consisting of transparent epithelial cells with round nuclei, between which there are
long supporting cells with oblong nuclei (Fig. 24). Between the cellular envelope and
the spicule an almost homogeneous mass occurs, staining pale-red with carmalum, which
I believe to be a secretion of the cells of the envelope, though perfect certainty cannot
be obtained. In this case the envelope might perform the duty of a poison gland, the
spicule having consequently the power of attack and defence. At the proximal
end of the spicules numerous small muscles are attached, which enable the animal
to move the spicules in various directions.

The nervous system does not furnish much worthy of note. The cerebral
ganglion is large (Fig. 6, eg) ; the latero-ventral ganglia are large also (Fig. 7, gr),
whereas the latero-dorsal ones are very small (Fig. 6, gd). As in other Proneomeniidce,
the cerebro-ventral and the cerebro-lateral connectives take their origin from the
cerebral ganglion separately. The buccal ganglia are distinct, as well as the buccal
commissure (sub-lingual commissure of Simroth), which takes its course dorsally from
the radula-sac (Fig. 10, be}. The posterior superior ganglia are large (Fig. 21, gps),
the commissure between them strong. The posterior inferior ganglia are much smaller ;
the connectives between the posterior superior and inferior ganglia are strong. From
the posterior superior ganglia some strong nerves take their course distally toward the
lip, situated behind the cloaca (Fig. 19, n).

The heart shows some very interesting peculiarities. As far as can be observed,
the structure generally corresponds with that of the other forms, of which I have
already given a detailed account (8). Fig. 21 represents the proximal offsets of the
cloaca, which gradually disappear (cl}. Ventrally from these the distal offsets of the
pericardium are found (p), into one of which — the left — the left cloacal duct opens (cd).
The medial walls of these pericardial offsets invaginate ; these invaginations together
fopm the atria. There exists also a median distal offset of the pericardium (Fig. 21, p).
By the union of these three offsets the spacious pericardial cavity is formed ; at the
dorsal wall the two atrial invaginations still exist, separated (Fig. 22, a) ; more
proximally, however, they unite and form together the atrium. The double character
of the atrium remains here plainly visible. The atrium is consequently open on the
distal side ; on the dorsal side it is more or less closed by connective tissue. The
ventricle originates similarly to the atrium by the coalescence of two invaginations,
viz., the median walls of the pericardial ducts. The unpaired portion of the ventricle
unites with- the dorsally situated atrium (Fig. 23, a, v). Concerning this we must
notice the following differences : in other forms the ventricle is situated dorsally from
the atrium ; it also arises from the union of two parts, but is continued as a dorsal
unpaired invagination at the dorsal pericardial wall, and communicates at its ventral
side with the atrium (8). In this form, however, the ventricle remains small and is not

•DI.KX'w:. \STMI 7

continued along the dorsal wall «>f the pericardium ; consequently it is situated ventrally
fn>m the atrium. The blood of the atrium in driven into the ventricle through one
large opening, there being no question of two atrio-ventricular openings. The wall of
both parts of the heart here also remains thin ; it is strengthened by muscular fibres
belonging to the surrounding tissue. The ciliated bands (" bourrclets cilies " of
Pruvot) do not fail, at least not in the distal part of the pericardium (Fig. 22, <•>>). The
irl.u i vi-ly very large blood-corpuscles are of a peculiar shape, oblong, more or less
[x.inted corpuscles of various forms. They stain pale-red with carmalum and usually
have a small round nucleus, which may occur at different points, and often obviously
protrudes (Fig. 26, A). Further, there occur in the blood large round cells with round
granular nucleus (Fig. 26, B).

The animal in mature ; the gonads extend to close behind the radii la. The eggs
are formed on the median walls, the spermatozoa on the dorsal, ventral, and lateral walls.
Both the gonads are densely filled with products, more especially the distal part, which is
entirely filled with eggs, or rather with generative epithelium. The latter seems to
become free in the gonads ; in the pericardium, compact masses of small round cells occur
(Fig. 25, A), which arc also found on the median walls of the gonads, and which prolwbly
belong to the generative epithelium. Amongst these large cells mature eggs arc
observed, with granular contents and of various sizes (Fig. 25, B) ; even very large eggs
are found, though they are of rare occurrence. Such eggs are often surrounded by the
small cells of the generative epithelium, enveloping them like a follicle. Further,
numerous spindle-shaped cells occur, the extremities of which are more transparent and
stain less obviously (Fig. 25, C). Whether there exists any connection between the
round eggs and the spindle-shaped cells cannot lie decided, any more than we can follow
the further development of the eggs. Full-grown eggs occur sporadically ; they arc
large and round, finely granular, with round nucleus. Amongst the ripening eggs
there are always found numerous extremely small globules, the significance of which is
perfectly unknown to me. Besides the female genital products, numerous spermatozoids
occur in the pericardium (Fig. 25, D).

The cloaca! ducts are not in any way remarkable (Figs. 21-23, or/). They leave
the pericardium as wide tubes with ciliated epithelium of irregular height, the inner
surface being consequently undulating. First they take a proximal course ; afterwanN
they bend and run distally, becoming gradually broader and having a more glandular
epithelium, which produces a granular secretion. The state of preservation is imt
MI. h as to enable us to decide whether the structure is similar to that <>f
other forms, vie., glandular cells, alternating with supporting onea The two ducts
unite into a broad praecloacal organ (Fig. 22, /*»). Each cloacal duct bears, at
the point where it bends proximally, a spacious pouch, placed on a snort stalk,
uhi'-h I believe to be a receptaculum seminis ; in both of them numerous
spermatozoids occur.

Now it remains to l»c derided to what form this new species, for \v lii.-li I propose

H. F. NIERSTKASZ.

the name Proneomenia discoveryi* is related. In the first place we must mention the
different forms, which are now considered as belonging to the genus Proneomenia. The
thick cuticle with the needle-shaped spicules, the polystichous radula, the two tube-
shaped salivary glands, and the absence of gills point towards a relation to that genus.
The genus Proneomenia includes at present no less than fourteen species, viz., sluiteri
Hubr., acuminata Wir., vagans Kow. a. Mar., desiderata Kow. a. M.&r.,gorgo)wphila Kow.
a. Mar., weberi Nierstr., longa Nierstr., thulensis Thiele, valdiviae Thiele, australis
Thiele, amboinensis Thiele, gerlachei Pels., liawaiensis Heath, and neapolitana Thiele.
As already mentioned by me (8), the expression " polystichous" radula has to be taken
in a wide sense, as very different types belong to it, which must lie kept sharply
separated. The salivary glands likewise show remarkable differences.

All the above-named forms have a thick cuticle beset with numerous needle-
shaped spicules. This characteristic, however, is not typical of Proneomenia, but also
holds good for other genera, e.g., Rhopalomenia. The length-index varies considerably,
from 6 (P. desiderata) to 50 (P. longa).

Epithelial papillae are of general occurrence ; however, they are said to be
wanting in Proneomenia sluiteri, though represented and described by Heuscher for
P. langi. On studying Hubrecht's sections of sluiteri more closely, I came to
the conclusion that epidermal papillae really are present in this form. In the
outer part of the cuticle many of these papillae are found, just as Heuscher observed
them in langi (3, p. 3). Owing to this fact another point of difference between the
two forms disappears, and we have the more reason to share Thiele's opinion, and to
consider both forms as belonging to one species (17, pp. 261).

All the forms of Proneomenia have one or more dorso-terminal sense-organs.

The structure of the coelomoducts (cloacal ducts) and their appendages varies
considerably. The latter always occur at the point where the proximal directed part
of the coelomoducts bends to take a distal direction. Proneomenia sluiteri shows
glandular, ramified tubes ; langi, winding coeca ; weberi, longa, acuminata and gerlachei,
one vesicular appendage ; thulensis, twenty, often lobate vesicles ; valdiviae, a tube ;
australis, thirteen stalked vesicles.

The abdominal and cloacal spicules (abdominale Hakenbiindel und Kloakcnspicula,
Thiele (13, p. 291) ) show also important points of difference : sluiteri, langi,
desiderata, gorgonophila, weberi and longa have the so-called abdominal spicules in
great numbers. Thiele found their homologon in vagans and australis (13, p. 261 ;
17, p. 256). In longa they even are numerous, though of smaller size. In
discoveryi two complicated copulation-spicules likewise are present, which show affinity
to those of vagans and to the penis-spicules of Neomenia.

Regarding the cloaca, we must observe that gills are always wanting, the gills of
the Solenogastres being merely folds of the cloacal epithelium. It is a very remarkable
fact that they are absent from all species of Proneomenia.

* Sit venia verbo ; we have a precedent in CJiallengeri.—'En.

9

Aa a coi).T,|ii«-nre of these facts, to which might }»• added many others concerning
the structure <>f the digestive tract, nervous system, etc., we must acknowledge that
there exist great differences lietween the various species of Proneomenia, and
consequently it would be rational not to class them in one genus. The difference

L'tWeeil ' ;. :i!i.| /•/-'.;/-. !',,]• m-r.ill. •-. 1- •_•!•••., ;,T li.u, i],.,; !..:\\,.;i W|j .,!.,:

Rhopalomenia indica or DJ/I.-//I. •///«/. The same holds good for Rliopalomenia itself.
This genus was created by Simroth on account of the presence of club-shaped
• •pi.lermal papillae (12, p. 229). It is with good reason that Thiele observes, that
the absence of a radula is more typical of Rhopalomenia than the presence of the said
jiapillar (13, p. 272). In this sense the genus Rhopalnrnenia was enlarged by me,
some years ago, to contain new species (6, p. 22).

As to the geographical distribution, nothing confirmative is given. Arctic
are sluiteri and tliulrnsu ; both forms are most certainly not closely related
(14, p. 115).

Antarctic forms : gerlachei and discoveryi. Between these two forms there exist
great differences (dimensions, radula, dorsal spicules, buccal gland of yrrltichei,
structure of the proximal portion of the digestive tract). In dimensions, gerlachsi
resembles sluiteri ; for the rest, according to Pelseneer (10, p. 34), these two forms
do not correspond in many respects.

Mediterranean forms: neapolitana, gorganophila, desiderata, vagans. No doubt
neapolitana is a distinct form (Amphimenia of Thiele, 18, p. 244). Vagans,
desiderata and gorgonophila agree in the structure of the pharynx and salivary glands
(5, pp. 59, 76); the index, however, varies from 6 to 25. Yet I am inclined to
consider these forms related to each other. Vagans differs importantly from sluiteri,
gerlachei and discoveryi, by the structure of the radula ; on the other hand, a relation
with thulensis is very well possible (radula, salivary glands, abdominal spiculcs),
tin nigh in the latter form cloacal spicules are wanting.

Indo-pacific forms : tceberi, longa, amlminenxis, australis, valdiviae and haicaiensi*.
Of these forms haicaiensis is insufficiently known ; only the nervous system has \>een
accurately studied by Heath (1). Weberi and longa, no doubt, are related ;
MH&0«MMM^ however, must be considered an isolated form, on account of the structure
of the radula and the coelomoducte. Australit has a bisehal radula, valdiviae a
multiscrial one (9, pp. 681, 672).

Atlantic forms: acuminata shows a multiserial radula and two salivary glands,
which open out separately. Abdominal and cloacal spicules are wanting.

Recapitulating these facts, we come to the following conclusion. The known

-;.. • :• - "'. /' • •• ' ' i I.P-M L'l.Ml -liti". -r.-n.-.., ; .-ill .,(' t':..-ni. i,..\\. \.T. h.iv.- .1 IM-I ;!.,.

a thick cutii-le with many layers of pointed, hollow spicules, numerous epidermal
sense-organs, and one or more dorso- terminal sense-organs ; gills are wanting. If no
other characteristics were considered, we might keep the genus Proneomenia. In my

opinion, IIOW.-V.T. \w ha\r good reason t<> divide the L'fiiu-* into a nuniltrr of sinalli-r

VOL. IT. C

10 H. F. NIERSTRASZ.

ones, considering in the first place the exact structure of the radula and salivary
glands. I propose the following division :—

FAMILY PRONEOMENIIDAE.

Index — at least— 6. Cuticle thick, with many layers of spicules, by far the greater part of which is
pointed and hollow. Epidermal papillae present. One or more dorso-terminal sense-organs. Radula and
salivary glands present. Gills are wanting.

A. Radula multiserial. Salivary glands tube- or bag-shaped.

Proneomenia Hubrecht.

a. Radula with basal membrane. Salivary glands tube-shaped, uniting before opening

out into the digestive tract. Receptacula seminis strongly folded, glandular. The
spicules remain connected with the epidermis by multinuclear stems. Abdominal
spicules present. Index 9-14.

Proneomenia sluiteri Hubr. (4.)
(= langi Heuscher),

•Barent's Sea, Spitzbergen.
70-160 fathoms.

b. Radula with basal membrane. Salivary glands tube-shaped, opening out separately.

Receptacula seminis vesicular or tube-shaped.

a. With abdominal 'and cloacal spicules. Receptacula seminis vesicular.
One dorso-terminal sense-organ. Index 22-50.

Proneomenia weberi Nierstr. (6, p. 2.)

East-Indian Archipelago, 22-1633 M.
Proneomenia longa Nierstr. (6, p. 7.)

Saleyer, East Indian Archipelago, 1158 M.

ft. Without abdominal, with cloacal spicules. Receptacula seminis vesicular.
Two dorso-terminal sense-organs. Index 10.

Proneomenia discoveryi Nierstr.

77° 50' 80" S., 165° 40' 5" E. 100 fathoms.

y. Without abdominal or cloacal spicules. One dorso-terminnl sense-organ.
Receptacula seminis vesicular or tube-shaped. Index 9-18 '5.
Proneomenia valdiviae Thiele (16.)
3° 7' S., 40° 45' 8" E. 748 M.
Proneomenia oxuminata Wircn (18, p. 68.)

West India. 300 fathoms.
Proneomenia hawaiemis Heath (1.)
(? insufficiently known.)

Hawaian Islands.

Proneomenia gerlachei Pels. (10, p. 16.)
70° S., 80° 48' W. 500 M.

c. Radula without basal membrane. Salivary glands tube- or bag-shaped, opening out

separately. Receptacula seminis vesicular.

a. With abdominal and cloacal spicules. Salivary glands tube-shaped.
1-3 dorso-terminal sense-organs. Index 6.

Proneomenia vagans Kow. a. Mar. (5, p. 29.)

Mediterranean.

/3. With abdominal, without cloacal spicnles. Salivary glands bag-shaped.
Receptacula seminis numerous, often lobate. One dorso-terminal sense-
organ. Index 11.

Proneomenia thulensis Thiele (14.)
80° 8' N., 16° 55' E. 480 M.

y. With "Mnmhft1. without ijoaral •picnic*. Salivary glands tnbc-thaped.
Index C

Pnnuomnua ftryomfMa Kow.a. Mar. (5, p. 75.)

Mediterranean.
/VMMMMNM tltudtrata Row. a. Mar. (6, p. 59.)

Mediterranean.

It. Itadnla biaerial Salivary glandt tnbi-«hap«l and opening oat separately. With
abdominal ipicnle*. RecepUcola •eniinu num. T..IM. One dono-toriniiial
Index 18.

Sjpimt*MHuutniit (TMM) (17, p. 255.)
N.W. coast of Australia. 60 fathoM.

C. Kadula bi-|«ctiniil. Salivary gland* consulting of more or IGM developed mvuinu-

hitions of glandular oclb, opening out through *alivury ductn, which remain M-purated.
Abdominal spicnle* prvaent. RecvpUcula M-tuinu numerous The coclouioducta remain
wparated. Index 24.

AnamntM Nientnv.

Anammia ambtuutuu (Thiele) (15.)
Ambon.

D. Badnla bi-pertinid (?). Salivary glanda lobod, conaiiting of aocnmnlations of

glandular wlU, and opening ont through salivary dncta, which remain Kparatai.
Without abdominal and cioacal tpicnlea. Index 20.

Amphiuifiiia Thiele.

AmpMimtttia mapolitana Thk-lu (13, p. 244.)

The same holds good for Rhttpalomenia. Up to now, of this genus five species
are known, viz., aglaopheniae Kow. a. Mar. ; euigi Thiele ; indica Nierstr. ; debili*

tr. ; and scanden» Heath. All these forms resemble the Proneomcniidae ; they
differ, however, in the absence of a radula. Traces of the latter are found in
the radula-sac. The structure of the salivary glands presents important points of
difference. R. aglaopheniae has two pairs of salivary glands, one pair vesicular and
the other lobate, just as they occur in Amphimenia. Undoubtedly, It. eisiyi in related
to aglaopheniae. R. indica shows numerous receptacula seminis ; cioacal spicules,
however, fail. True abdominal spicules are absent, though around the cioacal opening
spicules occur of special form. Moreover, there only occur two tube-shaped salivary
glands; relation with Proneomenia t/ntlensi* and Epimenia auttralis should, therefore,
not be excluded. Rhopahmenin </•/•///«, thoroughly known to UH, prol«bly is closely
related to Rhopalomeiria indica. Our knowledge of the structure of the internal
organs of Rhopalomenia scamlrn*, with the exception of the nervous system, is
insuth' ient. It appears to me that these forms are connected with different genera
of the Proneomeuiidae, and, therefore, the genus Hhopalomenia may be a hetero-
geneous one. A division into smaller genera is for the present not to be recom-
mended ; in still stronger degree the same holds good for Chaet<«l>rinn (7, p. 368).

c 2

12

H. F. NIERSTRASZ.

MEMOIRS REFERRED TO.

1 1904 HEATH, HAROLD .

2 1905 „ „

3 1892 HEUSCHER, J.

4 1881 HUBRECHT, A. A. W.

5 1887 KOWALEVSKY, A. 0.

6 1902 NIERSTRASZ, H. F.

7
8

14
15

17

1903
1903

9 1905 „ „ .

10 1903 PELSENEER, P. .

11 1891 PRUVOT, G.

12 1892-94 SIMROTH, H.

13 1894 THIELE, J. .

1900
1902

16 1902

1902

18 1892 WIR£N, A. .

The nervous system and subradular organ in two genera of

Solenogastres. Zool. Jahrb., Vol. 20, Anat.
The Morphology of a Solenogastre. Zool. Jahrb., Vol. 21, Anat.
Zur Anatomie und Histologie der Proneomenia sluiteri Hubrecht.

Jen. Zeitschr. Naturw., Vol. 27.

Proneomenia sluiteri, gen. et sp. n. Niederl. Arch. Zool.
Kowalevsky, A. 0., et Marion,

1'histoire des Solenogastres ou

Hist. Nat. Marseille. Zoologie.

Suppl.

A. F. Contributions a
Aplacophores. Ann. Mus.
Vol. 3.

The Solenogastres of the Siboga Expedition, in Sib. Exp.,

Monogr. 47.

Neue Solenogastren. Zool. Jahrb., Vol. 18, Anat.
Das Herz der Solenogastren. Verhand. Kon. Akad. Wet.

Amsterdam. Sect. 2, Vol. 10, No. 2.
Kruppomenia minima und die Radula der Solenogastren.

Zool. Jahrb., Vol. 21, Anat.
Mollusques (Amphinenres, Gastropodes et Lamellibrancb.es).

Result: Voyage Belgica, 1897-1899, Zoologie.
Sur 1'organisation de quelques Neomeniens des cotes de France.

Arch. Zool. Exper. Ser. 2, Vol. 9.
Aplacophora. Bronn's Klassen u. Ordnungen des Thierreichs.

Mollusca, pp. 128-226.
Beitrage znr vergleichenden Anatomie der Amphineuren.

Zeitschr. f. Wiss. Zool., Vol. 58.
Proneomenia thulensis, nov. spec. Fauna arctica (Romer u.

Schaudinn). Vol. 1, Lief. 1, pt. iii.
Proneomenia amboinensis, n. sp. Semon, Zool. Forschungs-

reisen in Austr. u. d. Mai. Arch., pp. 735-737.
Proneomenia valdiviae, n. sp. Wiss. Ergebn. deutsch. Tiefs.

Exp. Valdivia., Vol. 3, pp. 169-174.
Die systematische Stellung der Solenogastren nnd die Phylogenie

der Mollusken. Zeitschr. f. AViss. Zool., Vol. 72.
Studien iiber die Solenogastren. II. Kongl. Svenska. Vet.

Ak. Handl.,Vol. 25.

PLATES I. AND II.

FIG. 1. — A. Proneomenia discovery!. The animal (x 5), drawn from spirit specimen.

B. Anterior extremity. ( x 5.)

0. Posterior extremity. ( x 5.)
FIG. 2.— Spicula. (x 330.)
FIG. 3. — Spiculum with mother-cells, (x 730.)
Fio. 4.— Epidermal papilla. ( x 330.)

80LBNOGA8TRB8. I :«

Flo. .V— A. Done-terminal sense-organ at the surface of the cuticle.
l:
C. trf. M evagination of Uw epidermis, (x 124.) n£ catide ; «to. dcfso~tertuiii»l

••iso mniii ; «p. epidermis ; M. circular muscular Uy«T ; /«*»/. pedicle.
Pros). 6-1 1.— Transverse sections through the anterior extremity. ( x 53.)
of. ulterior mncotu gland.
be. boccal commissure.
*y. buccal nerve.

e. proximal caecum of the intestine.
ec. cartilaginous cells.
rg. cerebral ganglion.
t». cone with radnla-teeth. (S* text)
/. ciliated cavity.
i. intotine.

gr. latero-ventral ganglion.
ffil. Utero-donal ganglion.
In. lateral nan*.
M. moacle.
<*». oeaophagna.
pk. pharynx.
n. radnla iac.
•g. salivary gland.
ey. ventral gland,
rn. ventral nerve.
M. ventral sac of the radola.
Fios. 12-14. — Schematic transverse sections through radula, ventral sac of the radula, and

pharyngeal wall.

Flo. 15. — Salivary glands (•?.), opening into the ventral we of the radnla. (x 124.)
Flo. 16.— Radula-teeth, basal membrane and pharyngeal wall. ( x 380.)
Fio. 17.— Radnla-teeth. (x 780.)

Fio. IH.— Distal portion of the ventral sac of the radnla. ( x 330.) (Stt text.)
FIGS. 19-23. — Transverse sections through the distal extremity, (x 29.)
a. atrium.

et>. ciliated bands. (Stt text)
at. cloaca! dnct.
el. cloaca.
t*. cloacal-spicnle.
yp». ganglion poster! us snperiu* with commissure.

i. intestine.
In. lateral nerve,
n. distal nerves.
p. pericardium.
I*, praedoacal organ.
r. rectum.
r. ventricle,
r/i. ventral nerve.

Fio. 2 1.— Transverse section through a cloacal-«picale. (x 880.)
Fio. 25.— Genital products in the pericardium. (Stt text) (x 460.)
Fio. 26.— Blood-corpuscles. (Sw text) (x 780.)

•-.

•I.

5"

...

INSECTA APTERA.

By GEORGE H. CARPENTER, B.Sc. (LoND.), M.R.I. A.,
Professor of Zoology in the Royal College of Science, Dublin.

(1 Plate.)

TDK exploration* of the National Antarctic K\|»-diiii>n have established the
of a wingless insect of exceptional interest, far south in the Continent of Victoria
bind. From Granite Harbour, 77° S. lat ami 162° E. long., on the south-treinling
i-i'iitincntal coast-line, almost opposite Ross Island, on which stand Mounts Erebus
and Terror, and about 100 miles N.W. of the Winter Quarters of the ' Discovery,' a
jar-full of moss believed to contain Collembola was secured. Examination of this
material has resulted in the detection of half-a-dozen very imperfect specimens of a small
dark-blue springtail. Unfortunately, these insects are in a poor state of preservation ;
cither the spirit in which the moss was placed was too strong, or the insects had died
and shrivelled before the moss was collected. The result, however, is. that no really
good example of the species can IK? obtained, and the following descriptions with the
accompanying figures have l>een made from various fragments. Under these
circumstances the descriptions are necessarily imperfect It is hoped that the account
will not require correction by the discovery of further and more satisfactory specimens
by some future expedition, but it will assuredly need supplementing. The remoteness
of the locality, and the difficulty with which the insects must have l«een obtained,
render it a duty to make the l>est possible use even of such unpromising material.

At first I spent many hours pulling the leaves of the moss plants asunder with
the help of a dissecting microscope, and this labour was rewarded by the discovery of
one or two shrivelled specimens, almost useless for study. Afterwards it was found
that the bases of the shoots afforded a better hunting-ground, and several more days'
work resulted in the unearthing of a few specimens in a less shrivelled state. From
these, when cleared in potash and mounted in glycerine jelly, it was possible to make
out various details of structure, and the systematic position of the insect hat been
gradually traced. All the specimens seen are referable to the same species.
Unfortunately, the general shape of the body can only l»e surmised, and the
segmentation of thorax and abdomen cannot be clearly distinguished in any HJN-. m,.-n.
On the other hand, the delicate sense-organs on the feelers and head, and some details
of the jaws, can be plainly seen in one or two examples. A large quantity c.f the
moss still remains unexamined, so that further light may yet be thrown on the species

2 G. H. CARPENTER.

by the collection already in hand ; but I think it unlikely that any specimens more
satisfactory than those here described and figured will be found in it.

Tliis is the second species of Collembola known from the Antarctic Continent
of South Victoria Land. The first, described five years ago (Carpenter, 1902) was
brought home by the ' Southern Cross ' Expedition from Geikie Land, on the shores of
Robertson Bay near Cape Adare, some 380 miles north of Granite Harbour. That
species was an Isotoma, a member of the family Entomobryidae. The present insect
belongs to the family Poduridae, so that the two largest families of Collembola are both
represented on the Antarctic Continent. A summary of the distribution of the known
species of Antarctic and sub-Antarctic Collembola will be found in Wahlgren's recent
paper (1906) on the insects of this order collected by the Swedish Expedition. To the
species that he enumerates have now to be added the insect here described and the two
new species brought from the South Orkney Islands by the ' Scotia ' Expedition
(Carpenter, 1906). The last-named paper contains arguments, from the distribution of
Antarctic springtails, in favour of a former wide extension of the Antarctic Continent.
The present species, having no near allies, does not throw any additional light on such
geographical problems.

COLLEMBOLA.

FAMILY PODURIDAE.
GOMPHIOCEPHALUS, gen. nov.*

Cuticle very finely granulate. Abdomen with two anal spines. Spring and
catch vestigial. Empodium of foot without appendix (" inferior claw "). Feeler
four-segmented ; sensory setae on the third segment ; a sub-apical, depressed sense-
organ, and a protrusible apical sense-organ on the fourth. On each side of head a
simple post-antennal organ surrounded by a single ovate or cordiform rim ; and a stout
sensory tooth-like spine behind the post-autennal organ. Mandible with molar
surface close to base. Maxillula with apical teeth and a feeble but distinct molar
area.

This genus will probably require a new sub-family (Gomphiocephalinae) to express
its peculiar affinities. On the whole it comes nearest to the Hypogastrurinae in the
recent classification of Bb'rner (1906). But the dentiform sense-organ on the head, the
shortened mandibles, and the maudibuliform maxillulae appear to be absolutely
distinctive characters ; while the feeble granulation of the cuticle, and especially the
simple post-antennal organ, show an approach to the Anurophorini, a tribe of the
Isotominae which belong to the Entomobryidae. Indeed the genus GomphiocepJialus

* From folios, a back-tooth, and « </mX>';. I have to thank the Editor, Prof. F. J. Bell, for kindly proposing
this name to replace my suggestion Odontorrphalus, which is pre-occupicd. The allusion is to the peculiar
dentiform sense-organ on the head (fig. I,/).

I\M:« i \ \ITI:I:\ 8

seems • l'.»lund. with .ithnities to the Kut<>m..lirvnlm\ just as tin- Anurophorinac
—which include tliiK remarkable Antarctic genus OrjfptofyytU (Willem, 1902) — are
Knt.itniiliryids with .illiiuti. - t<> tin- I'oduridM. The presence of such ancient

cuiiiie. tin:; links i m tin- Antarctic Continent mid Mauds might reasonably ha\e l-.cn

' \|>ected.

I have pleasure in naming the type-species of Gomfthiocephnlttx after Mr.

T. V ll..d^..i....f the 'Discovery.'

GOMI'HIIK Kl'HALf.S HOtXJSONI.

Length, 1 mm. Colour dark blue- violet Fooler* rather shorter than head ;
pn>[Hirtional length of segments 8 : 11 : 12 : 15. Foot with three long, slendor
ten. -lit haire ; claw simple without teeth; cmpodium vestigial. Anal spines nearly
as long as foot-daw, on short contiguous papillae.

Habitat, In moss (Bryttm <tl</?im, Cardot *), Granite Harlx>ur, South Victoria bind.
ri'\ |M«S in British Museum.)

The whole insect is of a very deep blue-violet colour. The cuticle is finely
rugose, and bears numerous bristles, many of which, especially on the terminal
alxluinmal segments, arc elongate, but none are feathered. The general aspect of the
species and the form of the foot are suggestive <»f a Xmi/lla.

Feelers. The feelers (figs. 1, 2) have characteristic sense-organs both on the
third and fourth segments. On the dorsal aspect of the extremity of the third
segment is a group of four short, sharp spines arising from large circular depressions in
the cuticle, and surrounded by a sinuous ridge (fig. 1, d). The two outer spine* of
this organ are stronger than the two central. At the tip of the fourth segment is ft
long tapering sensory bristle (fig. 1, b), and near it is a 8ul>-apical sense-organ (fig. 1, a)
consisting of an oval depression within which is placed a slender bristle. Close to this
are two small rounded papillae with a bristle at the tip of each. At the apex of the
fourth segment and directed ventral wards, a bladder-like protrusible organ (fig. 2, c)
can be seen in two specimens.

Cephalic sense-organs. The most careful search has failed to discover any
trace of ocelli, and I believe that they are absent. In every specimen, however, the
poet-antennal organ (fig. 1, e) is prominent a simple chitinous ridge of oval or heart-
shaped form, varying in the details of its outline in different specimens <•' •"). Ik-hind
this is a structure, which appears to lie a sense-organ, hitherto undescribed, as I !.<•];, -\ .-.
in any s|iringt«il. It is a stout curved tooth-like outgrowth (fig. 1, /), probably

•ruling a tactile function in connection with the shelters beneath which those
insect- Inc.

Jaws. The front region of the head is distinctly |»n>duced towards the-mouth.
producing the " pmgnathou^ " aspect which characterises this group of (

-<* tbew) Report*, vol. iii. ; MUMU. p. 6.
•».

4 G. H. CARPENTER.

The labrum (fig. 8) is narrow and rounded in front, and bears a proximal row of four
and a distal row of three long bristles.

The jaws are in many respects highly remarkable. The mandibles (figs. 3, 4) have
four strong teeth at the apex, and a very well-developed molar area, the teeth at its
distal region being especially strong. At the proximal end of the molar area are two
prominent recurved hooks directed dorsalwards. On the outer edge of the mandible is
a strong " shoulder," and just proximal to this a rounded " boss." The large basal
extension of the mandible so characteristic of the Collembola generally is entirely
wanting in the present genus. At first I could not believe this to be so, supposing
that the basal region had somehow got destroyed in the specimens that I was
examining. But on finding the mandibles in all the specimens in exactly the same
condition, I can only conclude that this insect shows a modification of the jaws unique
in the Collembola. As the muscles are usually attached to this missing basal part of
the mandible, fresh specimens of the present genus would be especially welcome
subjects for anatomical research.

The maxillulae (fig. 5) are also very remarkable on account of their likeness to the
mandibles. I have not succeeded in isolating these appendages, but one of the heads is
so transparent that their shape can be easily seen. At its apex the maxillula has three
delicate pointed teeth, while the inner edge of the base carries, not, as is usual, one row
of spines, but four rows of short denticles, forming a molar area analogous to that of
the mandible. I have not succeeded in seeing the tongue (hypopharynx) so that I
cannot state whether the relation of the maxillulae to that structure is of the usual
nature or not. In this genus they are more unmistakably a pair of jaws than in any
Collembolan that I have examined.

Each maxilla (fig. 6) has a " head " of the usual Collembolan form with a two-
toothed galea (</) and three internal lamellae (I) ; on account of their delicate texture
these latter structures are somewhat shrivelled, and the shape given to them in the
figure may be inaccurate. The base of the maxilla (fig. 6, st] can only be traced a
short distance into the head capsule, but I cannot satisfy myself that the cardo,
with its usual associated sclerites, is absent. There is a blunt palp (fig. 6, />) carrying
three or four bristles at its tip.

The second maxillae combine, as is usual in the Collembola, to form a transverse
labium (fig. 7) with alternating narrower and broader lobes, carrying long bristles.

Legs. The legs (fig. 9) are of the usual Collembolan form. The tibio-tarsus
carries a simpfe pointed claw without trace of tunica, and a small rounded empodium
without appendix. There are three elongate tenent hairs, without terminal club,
springing from large circular depressions of the cuticle.

Ventral Tube. The ventral tube (fig. 10) on the first abdominal segment is
abrupt t>n its front and rounded on its hinder face, beset with numerous bristles like
those of the cuticle generally. It appears to be in the undeveloped state characteristic
of the section to Avhich this genus belongs.

INSBCTA AITI:I:\

Spring and Catch. Most of ti. m.-ns r\.iminr.i *li..w no Iran- «>f thm«

structures, luit in two the appearance drawn in tig. 1 1 mn In- plainly Keen. The spun-
(fig. 1 1, ») in clenrlv in a vestigial ••••n.litii.n, its manuhrium being merged in the ventral
regiuii <>t tin- fourth alxlniuitml segment, whih- the combined den* ami mucro terminate
in ii liltuit, li.ini'lr.l kiiuli.

Tli.- <-at i-li (ivtinarulum) is represented by a pair of small rugose prominences

1 1 /•) <>n the tliinl •bdoBUMl segment.

Anal Spines. The anal spines are nearly as long as the foot-claw, slightly and
• •\nilv nirved (figs. 11, 12). They are situated on short, contiguous papillae. In all
tin- .-|M-i-inu-iis examined the spines arc slightly asymmetrical (fig. 12), hut thin is
l>r.>luil.lv due to the position of one having lieen slightly shifted.

REFERENCES.

1906. C. B"fl"1T" DM Sjstem dcr C'ollembolen nebst Itesrhreibun;; neuer Collembuk-n de* Hamburger

naturhistoruciicn Miun-uniK. 2 Beihcft turn Jahrb. titr Hamburg, tritttntth. Aimfnltni, zxiii.,

I'JiKJ, pp. 147-1K*.
1902. 0. II. CARPKXTKR. " InwcU Aptera," in Report on Collation*, \'oyagt tf Soutkern Crott,

pp. 221-3, pi. xlvii. London (British Mtucnm), 1902.
: Keottuh National Antarctic Expedition. '.Scotia' Collection*. Collembola

from the South Orknej Inlands. I'ror. Itoyal Sor. Edinl>., xxvi., HKiC, pp. i ;:;- i-::. pb. 1-2.
1906. F. WAHUIKRN. AnUrktischc und ralNuitarktuKhc Collembollen f^aammelt von dc>r nchwcdiiielien

SOdpoUn-xpfdition. Wt*>+n»ch. Ergtbn. dtr Sfkmti. SwIpolartxptdHioit, I'.Hil-S, vol. v .

1906, pp. 1-22, pk. 1-2.
1902. V. WILLKM. Collembolt* : Ritultatt du Voyagt •!• & Y. ' Brifica ' en 1H97, 189M, IM-.I-.I.

Sritntifaur*. Anren, 19O2.

EXPLANATION OF PLATE.

hodgtoni.

Ki-i. 1.— Left feeler with part of head, donal aspect; a, saWpicai, depreaed •enn-drpnn : ^ teatorj
brittle ; (I, Benwry »i>ine« on third anU-mml segment ; r, poit-antcnnal organ ; /, dentiform
cephalic KMe-organ ; 1, f, variations in form of pott-anU'iitml organ from other specimen*.
Ki<.. 2.— Right filler, lateral aspect of extremity ; a and d as in fig. 1 : e, apical, protrtwible iente-organ.
:t. — I*ft mandible, seen from inner aspect.
\.~ I A ft iiiuii.iil.li-. donal aspect.
5. — Right maxillnla, shoving teeth and molar area.
. FID. 6. — Left maxilla ; «, stipes ; y. galea ; /, lamellae ; />, |*lp.
Ki<i. 7. — .Second pair of maxillae forming labium.
l'i'.. 8. — Labrum.

Kit;. 9. — Right hind leg, showing daw, vestigial empodiam, and tenent hairs on foot.
Fie. lc. — Ventral lube on first abdominal segment, seen from right side.

11.— Part of hind abdominal segment*, showing anal spiuea, and vestigial spring (*) and caU I

lateral view.
Kn; 12.— Tip of abdomen, with papilhe and anal spines ; postcro-ventnl view.

All figures an magnified GOO diameters.

'•'JHJJ

•&

Antarctic ( l)i«cov*ry)*E\p.

Gomphtoceptulus hodgsoni.

CRUSTACEA.

VII.-SCHIZOPODA.

By W. M. TATTKRSALL, M.Sc.
(8 Plates.)

THE collection of ' Discovery ' Schizopoda is a large one from the point of view of the
number of specimens it contains, considerably over ten thousand, including larva',
having been collected. By far the majority of these specimens, nearly ten thousand,
however, belong to one species, and the total numlwr of species reaches thirteen. The
collection is in a generally excellent state of preservation, though many specimens have
undoubtedly suffered from being frozen after coming out of the water and from the
sulwequent thawing before preservation.

In the preliminary notice of this collection (Holt and Tattersall, 1906 (1)*) ten
species were noted, of which five were absolutely new, two were only known under
manuscript names given to them by Dr. Hanson from the collections of the ' Belgica '
Antarctic Expedition, while the remaining three were already descrilied forms.

Since the publication of that notice further material has come to hand, collected
on the homeward journey of the ' Discovery.' It contained two species of Euphauxia
not included in the earlier material, one, an immature form which was too young t<>
be specifically identified ; the secDnd, a species very close to E. lucens, Hansen
(= E. splendens, G. O. Sara), but which Dr. Hansen has kindly informed me is distinct,
and will be descril>ed by him in a forthcoming paper. A re-examination of the two
specimens which in the preliminary note were referred to Mysis maxima,
Hansen (MS.), has shown that they represent in reality two very closely allied
species, the second of which will also be descril>ed by Dr. Hansen in a future work.

Previous to the recent activity in South Polar exploration only three expeditions
to the Antarctic had brought back zoological material from which Schizopoda were
recorded. Dana (1852) in his great work on Crustacea records two species from
Antarctic waters (i.e., south of lat 60° S.) viz., Euphausia superba and Eucopia australis.

H.M.S. 'Challenger' in 1874 collected, in the same region, Euphausia superba,
E. murrayi, E. antarctica, Thysanoessa macrura and a species of Pxnulommn,
doubtfully referred to P. sarsi by Prof. Sars, who described the Schizopoda of that
expedition. The second and third species in this list are, however, only synonymous
with the first, so that the real total of 'Challenger' Antarctic species in three.

• The*e dates refer to the list of Memoirs on p. 88.

W. M. TATTERS ALL.

Mr. Hodgson (1902) described two species as iiew, Euphau&id glacialis and E. mistral!*,
collected by the ' Southern Cross ' South Polar Expedition. Both species are
synonymous with E. superba, Dana, so that previous to 1903 only four Antarctic
species of Schizopoda were known, viz., Eucopia australis, Dana, Euphausia superba,
Dana, Thysanoessa macrura, G. 0. Sars, and Pseudomma sarsi, Will.-Suhm.

Of the recent expeditions to the South Pole, which number seven, the results of
the Schizopoda of the French Antarctic Expedition have been published in full, and of
the 'Valdivia' Expedition in part only. M. Coutiere (1906) notes from the French
Antarctic collections, Euphausia superba, Dana, E. similis, G. 0. Sars, Thysanoessa
macrura, G. 0. Sars, and Antarctomysis maxima (Hansen, MS.), a species also recorded
in the preliminary note on the present collection.

Preliminary descriptions have been published of two Antarctic Mysidse collected
by the 'Valdivia' (lllig, 1906), Dactylerythrops arcuata and Eckinomysis chuni. The
first of these is synonymous with a species Dactylamblyops hodasoni, described below.

This completes the bibliography as regards purely Antarctic Schizopoda, though a
number of sub-Antarctic species are known^

Ten of the thirteen- species collected by the >' Discovery ' were taken in Antarctic
waters, and when, as we have seen above, the total previously recorded species number
seven, it will be recognised that the present collection has added considerably to our
knowledge of South Polar species of this group.

The most abundant species in the collection is a small Euphausia, E. crystal-
lorophias, H. and T., which evidently has its head-quarters under the ice, since all the
specimens were collected from ice-holes at Winter Quarters, and none were met with in
the open sea. On the other hand, the dominant species of the collections from open
Antarctic waters are Euphausia superba, Dana, and Thysanoessa macrura, G. 0. Sars.

The abundant material of the first-named species has afforded opportunity for
some observations on the sexual characters and life history of the species, with the
result that four other species, hitherto regarded as distinct from E. superba, must now
be allocated to its synonymy, having been founded either on characters which are
sexual and not specific, or else from immature specimens.

The scarcity of fully grown males of E. superba seems worthy of note. This fact
seems to be established by the results of the various collections of which we have
knowledge, since, so far as I am aware, the only one recorded is Sars' ' Challenger '
specimen. M. Coutiere (1906), it is true, notes that male specimens were more
numerous than females in the collection he examined, but it is equally clear that none
were fully grown, since the largest measured only 4?> mm., and was in the stage pre-
vious to the last moult into completely grown specimens. There are no fully-grown
males in the ' Discovery ' collections, but to judge from the development of the copulatory
apparatus on the first pleopods, many of them must be sexually mature at any rate.

Thysanoessa macrura, G. 0. Sars, too, seems never to have been previously met
with in full-grown condition, and but very few of the ' Discovery ' specimens can be

SCHIZOI'ODA. 3

said to have mched that state. Still, they afford material for a rc-«lc.s. ripti<>n and
figures of the s{x-< i<-> in the adult state, with some notes on the growth change*.

Tin- M\-i'lacea consist of eleven specimens, referable to six species. The most
interesting of these species is Ifanscnomysis anttirctioi, an Antarctic representative of
• genus hitherto known from hut three specimens from Arc-tit- and Itoreal waters.

The chief interest of the collection lies in the evidence which it may afford as to
the similarity or dissimilarity of the fauna at the two poles. There are no species
common to the fauna of both polar regions in the collection ; hut, on the other hand,
all the genera save one, Antarclomysi*, are represented in northern waters hv species
which arc quite distinct from their southern allies.

Exploration of the bottom fauna of the deep waters of the globe, especially in
tropical and sub-tropical regions, is as yet only in its infancy, and it is therefore
extremely probable that what are now known to In- bipolar genera and species will in
future be found to IK? cosmopolitan in their distribution. The Schixo|»oda were lonj.'
thought to have in Lopkogatttr typicw* a stock instance of a bipolar form, but the
gaps in ita distribution have \tccn almost completely filled up as a result of recent dcep-
sca work, and, with the exception of the tropical Atlantic, it* range is known to be
complete from Norway to the Cape.

Two bipolar species of Mysidse are known, however — Boreomyisi* Hcyplmjix, G. O.
Sarn, from Arctic waters, and Lat. 50° 8., near the Crozct Islands, and J /;</»///•</<*
crozetti, from the seas of Greenland and Ian Mayen in the north and the Crozet Islands
in the south. They are not known from the intermediate waters.

Of the genera of Antarctic Schi/opoda, Eupluiw'ui, T/iymtiwessti, and Euctip'm are
world-wide in range, but the northern and .southern species are quite distinct, even the
hitherto supposedly cosmopolitan Ewwpiti aufitralis, Hansen having shown to contain
at least two species, probably three.

P*fiuloinmnt Hansenomysis, DactylanMyttps, and Mytidetf* are, as at present
known, bipolar genera, but Psetulomma, at least, ranges far from both poles, and
further exploration will prol>ably extend the known range of the other genera also.

The most interesting case is presented by the genus Anturclninyxitt. It is rlo.-«-ly
related in structure to the northern species Michtheimy*i* mijcta (Lilljeliorg), a species
inhabiting chiefly the colder waters of the northern hemisphere. The two genera arc
separated only in the characters of the male pleopods, which are more primitive in the
Antarctic form, and neither genus is likely to he found to have a distribution which
extends very far from the poles they frequent

In the preparation of this report I have received much valuable help from many
sources. The authorities of the British Museum kindly allowed me to examine and
dissect two specimens from the 'Challenger' collections in their charge. To Dr.
Caiman, of the British Museum, I have been much indebted for information on many
points connected with the 'Challenger' material, and he has. also, at my request,
furnished me with drawings of various species. Mr. E. \V. L Holt kindly examined

4 W. M. TATTERSALL.

the British Museum material, and provided me with valuable notes as a result, which
I have freely used in this report. The Rev. T. R. R. Stebbing, F.R.S., placed two of
the type specimens of Euphausia vallentini at my disposal, with full permission to
dissect them if found to be necessary, while Prof. D'Arcy W. Thompson, C.B.,
allowed me to examine and retain for some time a small collection of Antarctic
Schizopoda from the collections of the Museum at University College, Dundee. I am
specially indebted to Dr. Hansen, of Copenhagen, for the most generous and valuable
assistance. Not only did he kindly confirm or modify my determinations of the
more subtle species of Euphausia, but he placed at my disposal his notes and drawings
of the male copulatory apparatus of the species of this genus, of which he has made
a special study. Without his help I should have failed to recognise that the two
specimens of Antarctomysis belonged to two distinct species, while I should have
recorded as E. lucens the new Antarctic species which Dr. Hansen will himself describe.
For all this assistance I wish to express my best thanks.

ORDER EUPHAUSIACEA.

FAMILY EUPHAUSIID^.

SUB-FAMILY EUPHAUSIN^:, Holt and Tattersall.

GENUS EUPHAUSIA, Dana.

EUPHAUSIA SUPERBA.

(Plate L, Figs. 1-12.)

Euphausia superba, Dana, 1852.
Euphausia superba, G. 0. Sars, 1888 and 1885.
Euphausia murrai/i, G. 0. Sars, 1883 and 1885.
Euphausia antarctica, G. 0. Sars, 1883 and 1885.
Euphausia ylacutlis, Hodgson, 1902.
Eupliausia australis, Hodgson, 1902.
Euphausia superba, Holt and Tattersall, 1906 (1).
Eupftausia superba, Coutiere, 1!)00.

Localities of captures :—

Lat. 61° 46' S., long. 141° 12' E., 16. 11. 01, five specimens, 14-20 mm.

Off Scott Island, 26. 12. 01, two specimens, 15-19 mm.

Lat. 66° 52' 9" S., long. 178° 8' 15" E., 3. 1. 02, eight specimens, 12-16 mm.

From stomach of Lobodon carcinophaga, 3. 1. 02, eleven specimens, 43-47 mm.

Lat. 72° 5' S., long. 172° 23' E., 10. 1. 02, eleven specimens, 15-47 mm.

Lat. 72° 10' 33" S., long. 172° 26' 2" E., 11. 1. 02, ninety specimens, 35-48 mm.

From River Koettlitz, 2. 1. 03, sixty-nine specimens, 13-27 mm.

Lat. 70° 29' 27" S., long. 168° 51' 46" E., 26. 2. 04, five specimens, 45-47 mm.

None were taken at Winter Quarters from the holes dug in the ice.

BOHBOPODA.

Tin- .-ynoiiyniy given here, whereby five supposed <listinc( species of En
arc merged into E. supcrba, Dana, is the ramlt of n careful examination of the
abundant ami valuable 'Discovery' material, aided by a coinjtarison with the

• »'li;tll« -II.T,. ,- ' ;,ii«l ' Southern Cross' types (for which I am greatly indebted to
Mr. K \V. I. II. .It), and the small collection from the zoological museum of University
' 'iillc^e. hundee.

bason* foi !..,-. riem mr brief)] -•.••• • • < pralimiiMi notia I I

• •••lie. tion, and the opinions as to the synonymy of this species, as far as they concern
/•.'»//'/»'/«*/./ nntarctica and E. murrayi, have recently l>een confirmed and adopted by
routii'-re (190G) as a result of his examination of the collection of the French
Antarctic Expedition. A fuller justification for these opinions is given Mow, together
with some notes on the growth changes, and sexual differences.

Female. — This sex has been very well descriU-d under the names E. mwntyi by
G. 0. Sara (1885), and E. auttrali* by Hodgson (1902).

The ' Discovery ' specimens present some slight differences from Sars' description,
but the examination of his type specimens proves them to be due to errors on San*
jwrt. They may be noted under their separate heads as follows :—

(1) I'l-rttnal fpine. — Sars states that this spine is wanting in /.'. murrayi, but the
type specimen shows it to l»e distinct, well-developed and simple, but not visible from
the side from which Sars took his drawing All the ' Discovery ' sjH-cimciis show a
well -developed simple preanal spine.

(2) Small />l n n f ftpinf on lite outer ilixtttl Cttrner qftktjlnt jtnnt of the itnteimulur
petlunde. — This spine is not shown in Sars' figures nor mentioned in his description.
It is, however, clearly visible in the type in lateral view, but in dorsal view is quite
olwcured by the numerous seta' arming the lm.sal joint of the peduncle, which are well
preserved and very opaque. The ' Discovery ' material conforms to the type in
possessing this spine well-developed.

(3) Terminal spine on the outer man/in of the nntennal xctil?.— Sars mentions this
spine in his description as very small, but docs not figure it It is, in fact, not visible
in his type from the dorsal aspect, owing to its being slightly vcutrally dcflexed, and
the specimen is so well preserved and rigid, that the pressure necessary to place it
dorsal surface uppermost for drawing is not sufficient to straighten out the spine anil
render it visible in dorsal view.

(4) Shape of the epimtral plate of the penultimate segment of the pleon, — Sars Ix.lh
dMcribes and figures the penultimate epimeral plate as acute and triangular, but in the
t\[«- and the 'Discovery' material, whereas these plates have substantially the same
shape as depicted by Sars, the apex in all is bluntly rounded instead of acutely pointed

(5) Spinutes on the dorsal surface of the telsoii. Sars figures and describes three
pairs in E. murrayi, but, as a matter of fact, the number is subject to variation, nil
additional |>air anterior to the three shown by Sars being frequently noticed. In all

VOL. IV.

6 AV. M. TATTERSALL.

other respects the ' Discovery ' material arid the ' Challenger ' types are in perfect
agreement, and the facts noted above establish the identity of the females here
referred to E. superba, with the species described by Sars as E. murrayi. It now remains
to show that the differences between E. murrayi and E. superba are only sexual.

Male. — Under the name E. superba Sars has described and figured this sex
adequately. The only point in which his description is deficient is the structure and
armature of the telson. He figures no dorsal spinules on the telson, and both describes
and figures the apex as slightly produced and obtusely pointed. Examination of Sars'
type shows that the apex of the telson is clearly broken, so that Sars' figure is in this
respect entirely imaginary. In the present material the apex of the telson is much
produced and acutely pointed, and the number of dorsal spinules is usually three pairs,
but may be four or two, placed as in Sars' figure of the telson of E. murrayi. One pair
of spinules still remains in Sars' type of E. superba, but the others had probably been
broken off (or obsolete ?).

The most conspicuous difference between E. superba and E. murrayi, as described
by Sars, is the presence in the latter and absence in the former of a lateral denticle on
the carapace. But both Sars' E. murrayi were females, and his single specimen of
E. superba a male. In all the females in the present collection, the largest of which is
47 mm. in length, the spine on the lateral margin is large and prominent, and even in a
female, 50 mm. in length, in the collection from University College, Dundee, the spine
is equally well-developed. I have figured the spine of the latter specimen on PI. I.,
Fig. 10. In male specimens, on the other hand, only those which are less than 42 mm.
in length have the spine well-developed (cf. PL I., Fig. 12, taken from a male, 39 mm.
in length). In males above 42 mm. up to 47 mm. in length the lateral spine on the
carapace is nearly obsolete and persists only as a blunt protuberance (cf. PI. I.,
Fig. 11, taken from a specimen 45 mm. long, and also Coutiere (1906), PI. II., Fig. 22,
taken from a male of the same size). The ' Discovery ' collection contains no male
specimens exceeding 47 mm. in length, but the ' Challenger ' type measures 48 mm.
It is well preserved and shows no trace of the lateral spine at all. Obviously, then, the
absence of a spine is a sexual character confined to absolutely full-grown males only.
The remaining differences between E. superba and E. murrayi given by Sars are
as follows :—

(l) E. superba has the antennules considerably more robust than in E. murrayi
and the lobe from the second joint almost obsolete. This difference is, I think, a
purely sexual one, affording a parallel instance to that seen in the northern species,
Nyctiphanes couchi. PL L, Figs. 1 and 2 are taken respectively from male and
female specimens of the same size, viz., 45 mm., and from the same bottle. They
indicate, clearly, the difference in relative stoutness in the two sexes, and that of the
male shows the lobe from the second antennular joint in an intermediate stage of
reduction between that of the female and that shown by Sars in his figure of the male
E. superba, 48 mm. in length.

(-') 1" '•' tin- rostrum is shorter and blunter than in K. mumn/i, and

tin- margins lc-~ .li.|.|\ .-..n.ave. This, a^ain, is clearly shown to l»e a s-
• litlcrcinT in I'l I., Fiji's. I ami •_'. The rostrum of the male figured (Fig. 2) IB shorter
than that of the female. I. in i> still rather more acute than in Saw" figure of K. *i*/»rrfa.
Redn, ti-.M is probably not complete till a size of at least 48 mm. is attained.

A further ditlcren.-e 1,,-tween the two sexes is brought out by the figures here
given, namely, the reduction in the male of the spine on the outer distal corner of the
basal joint of the aiiteiuiular peduncle. It is not visible in dorsal view, Iwing hidden
l.y the slightly j.mj.rtmg anterior margin of the joint, but it still persists as a small
1-liint protuberance. In the female, on the contrary, it is well-developed, distinctly
\ isilih- in dorsal view, and acutely pointed throughout life.

A fourth distinction shown in the figures, the absence in the female of the curved
setae on the dorsal surface of the basal joint of the antcnuules, is due to the accident
that in the female from which the figure was taken, these setae had Iwcome broken off.
They are, in reality, present, and equally developed in lx>th sexes.

The above detailed description proves, I think, clearly, that E. su/terhi and
E. murrayi are the adult male and female, respectively, of one species which must l»ear
the name E. superba Dana.

I also give (Plate I., Figs. 5-9), figures of the mouth organs and cndopods of the
first two thoracic limbs, to show two characters in which E. superba differs from all
other Euphausia yet described. The first of these points is the narrow and elongate
form of the terminal joint of the mandibular palp, with its peculiar armature of four or
five terminal strong plumose set®. In all the other species of the genus (with the
exception of E. antarctica, Sars, and E. fflticialis, Hodgson), the terminal joint of the
mandibular palp Is much shorter and stouter. In the two exceptions just mentioned
the maudibular palp is figured by Sars and Hodgson respectively, almost exactly as
here given for E. superba. This fact first suggested to me that these two species were
only developmental stages of E. superba, a suggestion fully l»orne out by the evidence
derived from a study of the present collection. The second distinctive character of the
appendages is found in the great length of the seta; arming the joints of the thoracic
limlis. They are very much longer than in any other species of the genus, and with
the character of the mandibular palp serve for recognition of E. superba at any stage
in its development.

Euphausia superba is the giant of the genus, and the only one of Dana's original
four species which is now retained by Hansen (1905 (2) ), the other three having been
cancelled by that author as unrecognisable.

Some Notes on the development of E. superba.

These notes were made chiefly with a view to confirming the suspicion, aroused
by the similarity in mouth organs, that Euptuiuttia antnrclica and E. glacialis were
merely developmental stages of E. xuperba. The changes which accompany growth to

E 2

W. M. TATTERSALL.

maturity concern chiefly the rostrum and the antennules, juul these notes refer to these
organs more particularly.

The smallest recognisable specimen of E. superba measured 12 '5 mm. in length.
The anterior end is represented in Plate I., Fig. 4. The rostrum is a bluntly rounded
triangular plate. The spine on the outer distal corner of the basal joint of the
antennule is still larval in character, being much longer than in larger specimens.
There is no trace of the lobe from the basal joint of the antennule, but the one from
the second joint already shows as a slight membranous projection of the anterior
margin. The antennal scale also shows larval characters in that the outer margin
is shorter than the inner. Finally, the telson has assumed adult form, but the dorsal
spinules immediately anterior to the sub-apical spines are still long and plumose. The
lateral spine of the carapace is present, but small. Between 12*5 mm. and 15 mm.
the spine on the basal joint of the antennule gradually shortens up and assumes the
characters seen throughout adult life. The antennal scale also assumes adult form,
and the spinules immediately anterior to the sub-apical spines on the telson lose their
plumose character and shorten to adult size. The rostrum, however, still remains
obtusely rounded.

The next stage is that described by Sars as E. antarctica, and measures 17 mm.
The rostrum has now become a broad, acutely pointed triangular plate, while the lobe
from the basal joint of the antennules first becomes evident as a slight inflation of the
anterior margin. This is shown by Sars in his 'Challenger' Eeport (Plate XV.,
Fig. 2). He has, however, overlooked the lobe on the second joint of the antennules,
which is now considerably forward in development. The spine on the lateral edge of
the carapace is now quite conspicuous.

Sars describes E. antarctica as being without lateral denticles. Examination of
his type specimen, however, shows that, while the side from which he took his
figure is rather damaged and the spine not visible, on the other side the
spine is quite conspicuous and perfect. This removes the only serious difference
which existed between the young E. superba here noted and Sars' description of
E. antarctica.

The transition from E. antarctica at 17 mm. to E. glacialis, Hodgson, is simple
and obvious. I figure (Plate I., Fig. 3) the anterior end of a typical glacialis stage
from a specimen 26 mm. in length. The only differences to. be noted from the
antarctica stage are the better development of the antennular lobes and the shortening
and broadening of the rostrum, which is still, however, pointed at the apex. The stage
figured agrees well with Hodgson's figures and description of E. glacialis.

After a length of about 27 mm. the sides of the rostrum gradually become more
and more concave till at about 30 mm. the completely adult form is reached. Very
little change takes place in either the form of the rostrum or the antennular lobes after
a length of 35 mm. has been attained, except, of course, in the changes accompanying
the last two or three moults in the male, already noted above. Examination of the

0
mouth organ* at various stages confirms the iil«-ntitiuation of the specimens with E.

This brief rtswmi of the development, 1 think, justifies the view that E.
(intarcHca and E. •jlacialis represent stages in the development of E. mtjtfrba, and must
therefore be regarded as synonymous with that special.

It should be- mentioned that Hodgson's types of E. australi* differ in no way from
E. superlxi (females), except in being considerably damaged.

EUPHAUSIA CRYSTALLOROPHIAS.

(Plate II., Figs. 1-10 ; Plate IV., Fig. 10.)
Euiikaiuia rryttallorophuu, Holt and TiUU-nwll, 1900 (1).
Localities of Captures :—
Winter Quarters.

26. 1. 02-8. 3. 02, 1 specimen, 24 mm.

,, ,, 216 specimens, larval.

No. 3 Hole, 52 specimens, larval to 25 mm.
No. 4 Hole, 4572 specimens, larval to 32 mm.
No. 6 Hole, 13 specimens, larval to 25 mm.
No. 8 Hole, 4642 specimens, larval to 32 mm.
No. 13 Hole, 50 specimens, larval.

No specimens were captured either on the outward or homeward journey.

Form, moderately robust

Carapace (Plate II., Figs. 1 and 2), with a prominent, rather long and acute
spine on its lateral margins, a little anterior to the middle, and just alx>ve the
insertion of the second thoracic limb ; nritero-lateral angles terminating in an
acute spine ; anterior margins inflated al»ove the eyestalks and produced into a long
acute rostrum extending to the visual part of the eye and about half-way along the
l«sal joint of the antennular peduncle ; there is a faint gastro-hepatic groove and
a distinct keel runs forward medio-dorsally from the latter into the rostrum.

Pli-im (Plate II., Fig. 1) without ridges or dorsal spines; none of the epimerel
plates much produced ; sixth segment about one and a half times as long as the
fifth ; preanal spine well developed and usually simple, but in large examples bifid.

Eyes (Plate II., Fig. 1) globose and rather large; greatest diameter of the
cornea exceeding half the length of the last pleon segment ; pigment black.

Antennulai- pedunck (Plate II., Fig. 2), with the basal joint as long as the second
and third joints combined and much wider ; no lota or lappet ; a row of altout
twelve long curved plumose setie set en a ridge on the distal part of its length ;
a short stout spine on the outer distal corner, which is more or less concealed by
the numerous seta- which arm the outer half of the anterior margin and the distal

10

W. M. TATTERSALL.

half of the exterior margin ; a bunch of coupling setae on the inner distal corner ;
second joint slightly longer than the third and without a lobe, its anterior margin
a little oblique.

Antennal peduncle about equal in length to the basal two joints of the antennular
peduncle, the third joint only very slightly shorter than the second.

Antennal scale reaching the centre of the third joint of the antennular peduncle,
about three times as long as broad, outer margin entire and terminating in a spine,
apex broadly rounded ; spine on the outer corner of the basal joint long and slender,
extending one-third of the way along the scale, plumose at least on the
proximal part.

The mouth parts (Plate II., Figs. 3, 4, 5) are figured for comparison with those
of other species. They do not appear to present any striking peculiarities.

First thoracic limit (Plate II., Fig. 6), has the penultimate joint of the endopod
longer than either the preceding or ultimate joints ; the latter has the lower margin
armed with a row of short fine setae in addition to the longer ones at the apex.

Second thoracic limb (Plate II., Fig. 7), with the terminal joint armed with a row of
three (sometimes four) short, rather stout and curved spines on the inner face.

The remaining thoracic limbs have the penultimate joint in all cases longer than
the ultimate and slightly longer than the antepenultimate. The following table
gives the lengths of the joints of the first six thoracic limbs in millimetres and the
total length of the limbs from a specimen 27 mm. long.

Thoracic
limb.

Lengths of the joints in mm.

Total length
of limb in mm.

1

2

3

4

5

6

1

•50

1-11

1-50

"77

1-00

•00

5-54

2

•5d

1-22

1-77

1-11

1-22

•50

0-32

3

•00

1-50

1-88

1'22

1-28

•72

7-2C

4

•00

1-88

2-05

1-28

1-88

•83

7-98

5

•fil

2-00

2-11

TOO

1-05

•00

7-43

8

• 55

2-00

2' 00

•72

•83

•44

C-54

First pleopod of the male (Plate IV., Fig. 10) with both movable processes on the
inner plat* of the endopod shorter than the plate itself ; distal process feebly curved,
bifid at the tip ; proximal process expanded at the tip into two lobes not in the
same plane, the outer lobe the larger, and wider than long, the inner lobe but little
expanded ; uncinus of the inner plate of the endopod without secondary spiuule.

Telson about one and a half times as long as the last segment of the pleon ;
apex acutely pointed ; sub-apical spines extending for half their length beyond the

8CH 1 7.0PO D.V 1 1

.•i|H'\ df the tclsoti and l»earing a few minute spinuK>s un their inner margins : <l-.r-M»l
denticles usually in two pairs, the first nltout half-way towards, the second at the
base of, the sub-apical spines.

Uropod* reaching to the level of the insertion of the sub-apical spines, the outer
very slightly longer than the inner, with a prominent denticle at its outer extremity.

Length of the largest adult specimens of both sexes, 32 mm.

Euphaiuia crystallorophins approaches most nearly among the species of the
genus to E. similif, G. 0. Sars, but differs ( 1 ) in the different shape of the rostral
projection, (2) in the shape of the cpimeral platen of the fourth and fifth segment*
of the plcon, (3) in the absence of antennular lobes and lappet*.

From K. uplenden*, G. O. Sars (E. lucent, Hanson) the present specie* is dis-
tinguished by the greater length of the rostrum and by the a!>sence of antennular
loin-* and leaflets, the types of E. splendent, G. 0. Sars, In-ing possessed of a small
but distinct nntennular lobe. E. crystallorophiit* is an enormously abundant species
under the ice, some ten thousand specimens having l>een taken. None, however,
were met with in any other locality except Winter Quarters.

Larvae of E. CRYSTALLOKOPHIAM.

The collection contains individuals in all stages of development from the
Mttiniiin/iliuM to the adult condition.

The Calyptopis Inrnr (Plate II., Fig. 8) first appear at the l>cginning of January
and continue in the tow-nettings till nearly the end of February. The IH..H! of the
carapace is very obtusely pointed in front and has the margins quite smooth. There
is no posterior median spine on the carapace, while the telson has the apical margin
lightly emarginate. I can see the Itegiuuing of the lateral spine of the carapace at
this stage. The largest Calyptopi* larva measures 3 -9 mm.

The Furcilia stages (Plate II., Fig. 9) first appear during the last week of February
and are abundant all through March. They cease after the first week in April.
The emargination of the apex of the telson is most marked during this Stage and
serves readily to connect it with the early Calyptopi* larva;. The spine on the
lateral margin of .the carapace is now well developed. The size of the FurcUia larva;
is from 4 • 5 mm. to 8 mm.

The Cyrtopia larvae (Plate II., Fig. 10) first occur about the last week in March,
and late post-larval stages are still to be had at the beginning of August. The size of
this stage is from 8 to 1 1 mm. The rostral projection is now an acutely pointed
triangular plate, but the sides are still but little concave. The final shape of the
rostrum is not assumed till the animal is in all other respects like the adult.

At a size of 1 1 mm. the telson assumes its adult shape, but the pair of spines
immediately anterior to the sub-apical spines arc still long and plumose. They
finally become reduced to adult size when a length of 13 mm. is reached. At this

12 W. M. TATTERSALL.

latter size the species has all the adult characters, except perhaps the rostrum, which
has the margins hardly as concave as fully adult specimens. Examples of 13 mm.
in length are to be met with in January, and so were presumedly larvge of the
preceding season, from which it would appear that the species takes at least one
year, and very probably longer, to reach the final adult size of 32 mm.

EUPHAUSIA TRIACANTHA.

(Plate IV., Figs. 1-3.)
Euphausia triacantha. Holt and Tattersall, 1906 (1).

Locality of capture :— Lat. 66° 52' 09" S., long. 178° 08' 15" E., 2030 fathoms;
one specimen, immature male, 23 mm.

Carapace (Plate IV., Fig. 1), with a single lateral denticle posterior to the centre
of the lower margin of the carapace ; antero-lateral margins somewhat inflated over
the eyestalks, and then produced into a long and very acute rostrum, which extends
beyond the eyes and almost to the distal end of the basal joint of the antennular
peduncle ; a faint keel is present on the carapace behind the rostrum.

Pleon (Plate IV., Fig. 1) with the posterior dorsal margin of the terga
of the third, fourth, and fifth segments produced into rather long, slender, very acute
and slightly curved median spines ; sixth segment rather long, nearly twice as long
as the fifth segment without the spine.

Eyes somewhat damaged in the single specimen, but apparently rather small,
pyriform in shape.

Antennular peduncle (Plate IV., Fig. 2) bearing on the inner distal corner of
the basal joint a well-developed bifid leaflet, the lappets of the leaflet of about equal
size ; outer corner of the basal joint rounded and adorned with numerous rather long
plumose setae ; a row of six curved plumose setae on the dorsal surface of the basal
joint ; second joint with a simple acutely spiniform lappet arising from the median
anterior margin ; third joint slightly narrower and shorter than the second.

Antennal peduncle shorter than the scale, the third joint a little shorter than

the second.

Antennal scale reaching very slightly beyond the distal extremity of the second
joint of antennular peduncle, broadly " oval in shape, about three times as long as
broad, apex broadly and obtusely rounded, spine at the distal end of the outer margin
small but distinct ; spine on the outer distal corner of the basal joint long, slender
and smooth.

First pleopods of the male (Plate IV., Fig. 3) obviously not fully metamorphosed,
since both the proximal and distal movable processes on the endopod are small
and simple, and the uncinus on the middle lobe is without a secondary spinule.

Tdson with the portion between and posterior to the sub-apical spines acutely
produced and smooth ; sub-apical spines extending beyond the apex of the telson,

SCHIZOl'ODA. 13

smooth ; dorsal denticle* in two pairs, the first situated at aboat two-thirds of the

u •«• from the base of the telson to the insertion of the sub-apical spines, second
almvr ill-' spines.

/ /**/* sub-equal in length, rather slender, extending to the level of the
insertion of the sub-apical spines of the telson.

Preaiutl ttpine small and simple.

A fuller description of this species is not possible, since the single specimen is
in bad condition and dissection was not desirable.

The species belongs to that group of the genus with a posterior median dorsal
spine on the third segment of the pleon, and is distinguished from the other meml>ers
of the group by having .an equally developed spine on the fourth and fifth segments
of the pleon in addition. It presents no very near kinship with any dcMcribrd
species of the genus, and from the depth at which it was captured is prolmbly a deep-
water form.

K I I' 1 1 A I > I A V A I . I . K N T I N I .

(Plate IV., Figa. 4-6.)

ni itpltitdtn* (part), 0. 0. Btn, 1881.
ralltntini, Su-bhing, 1MB.
valltntou. Holt and TaUenoll, 1906(1).

Localities of captures : — Lat. 56° 54' S., long. 170° 28' E, two specimens, male
and female, 19 mm.

I have carefully compared these two specimens with two of the types from the
Falkland Islands which the Rev. T. Stebbing kindly sent me, giving me at the
same time full permission to dissect them if necessary. The ' Discovery ' specimens
are in perfect agreement with the types, and I have nothing to add to Stebbing's
description except a note on the copulatory organs on the first pleopod of the male.

In the course of working out this collection the authorities of the British
Museum kindly allowed me to examine and dissect two of the ' Challenger ' specimens
labelled Euphausia, splendent by Sars. They were from the second of the localities
given by Sara on p. 82 of his " • Challenger ' Report," viz., " October 21, 1875, South
Pacific." It became at once apparent on examination that one of these specimens
did not agree with Sara' description, since the antennule was furnished with a large
evenly rounded lappet on the basal joint, very conspicuous in lateral view.* Further
examination showed that it probably, indeed almost certainly, belongs to the present
species. It is true that I could not see the spine on the third pleon segment, but
the specimen is in very poor condition, and if, as I suspect to be the case, the
spine has been broken off, the scar would be difficult to detect

I give (Plate IV., Fig. 5) an outline sketch of the rostrum and the basal joint

• Examination of the 'Challenger' type* of E. ipltndtn* shown that thU cpecie* poiieMM a mall
antennnlar lobe, bat it in nothing like no well developed and conxpicooai a* in K. ratlmtini (*•< Hunwn
(IMS 9) ), Holt and Tattenall (1906 (1). and below, p. 14.)

vou IT. K

14 \V. M. TATTERSALL.

of the antennule of the 'Challenger' specimen and (Plate IV., Fig. 4) a sketch of
the antennule of one of the ' Discovery ' examples for comparison with those given
by Stebbing (1900). This shows clearly, in my opinion, that all three specimens
belong to one species, and that the absence of the spine on the third pleon segment
of the ' Challenger ' example is the result of accident or possibly an abnormality.
The peculiar shape of the antennular lobe is practically the same in the ' Discovery '
and ' Challenger ' individuals, and only differs from Stebbing's types in degree, a
result of more complete growth.

The rostrum of E. vallentini is very like that of E. splendens, G. 0. Sars, but is
slightly longer, and the angle formed by its margins a little more acute. Sars may
have been misled by the resemblance between the rostra of the two species, which
caused him to overlook the marked differences which exist in the antennulse. Dr.
Hansen has seen the 'Challenger' specimen referred to, and agrees with my
interpretation of its specific identity.

One of the ' Discovery ' E. vallentini is a male, but unfortunately the copulatory
apparatus on the first pleopods is considerably damaged, so that I am obliged to refer
to the ' Challenger ' example, which is likewise an adult male, for a description and
figure of this apparatus (Plate IV., Fig. 6). The figure represents the inner lobe of
the endopodite of the first pleopod of the male. This inner lobe bears internally two
movable processes, the inner and more distal of which is feebly curved, slightly over-
reaching the inner lobe and bifid at the tip. The external and more proximal of the
two processes has the distal extremity greatly expanded, the expansion very much
broader than long, oblique, and divided into two lobes, the more distal of which is the
larger. On the under side of the expansion of the proximal process as viewed in the
figure there is a small spine-like process. The inner lobe of the endopodite itself bears
a strongly curved uncinus with a small secondary spine near the tip.

Distribution. — Southern Pacific, between New Zealand and Chili (' Challenger ') ;
Falkland Islands (Stebbing).

EUPHAUSIA, sp.

Locality of capture. — Lat. 57° 25' 30" S., long. 151° 43' E., nineteen specimens,
10-18 mm.

On first looking over these specimens I identified them with Euphausia splendens,
G. 0. Sars (1885), a species which Hansen (1905 (2)) considers to be different from
E. splendens, Dana, and which he has re-named E. lucens. Hansen, in the same
paper, notes that E. splendens, G. 0. Sars, has the first joint of the antennular
peduncle without a leaflet, but distally produced above. Holt and Tattersall (1906
(1)) have confirmed this statement by an examination of Sars' type specimens of E.
splendens, in which they found that in the female type the lobe is quite conspicuous
both in lateral and dorsal view ; while in the male type, which is considerably smaller
than the female, the lobe is less developed, but still easily seen in lateral view. Sars

SCHIZOPODA. 15

was therefore in error when he <lc8cril>e<l the antennular* peduncle of his E. splendent
M "more particularly distinguished by the total absence of any dorsal leaflet or lobe."
Moreover, it is apparent from what has already been written above in dealing with
Euphausia valient ini that Sars confused at least two distinct species under the name
E. splendent. It was subsequent to the publication of the preliminary notice of the
' Discovery ' collection that the present specimens came to hand. I therefore appealed
to Dr. Caiman for further information with regard to the ' Challenger ' species, and he
very kindly sent me a sketch of the dorsal aspect of the anterior end of both types.
From these sketches and Sars' description in the ' Challenger ' report I identified the
' Discovery' specimens as Euphausia splendent, G. 0. Sara = E. lucens, H. .1. IlaiiHcn.
Wishing, however, to have confirmation of my identification, I submitted the
specimens to Dr. Hansen, who at first was inclined to agree with me that they
belonged to E. splendent, G. 0. Sara. I may, perhaps, IKJ allowed to quote Dr.
Hansen 's remarks. They read as follows : " E. lucens (splendent). — I have specimens
from the southern Atlantic and the southern Pacific, and in all these the leaflet from
first antennular joint is easily seen, triangular, but not acuminate, with the end often
a little obtuse. In the material from the Swedish Antarctic expedition I have a large
number of specimens which differ only from the Copenhagen specimens in the feature
that the antennular leaflet is extremely small (visible as a very small triangular
distally obtuse plate when seen from in front . . . .)or rudimentary, but I find it
necessary to consider this difference only as a variation " (Hansen, in lilt.). Then,
after some remarks in which he noted that my specimens agree with the latter
condition, he concludes by saying that he considers them to l>elong to the more
Antarctic variety of E. lucens. In a later communication Dr. Hansen kindly informed
me that, after an elaborate study of the copulatory organs on the first pleopodn of the
males of the genus Euphausia, he had found that these two varieties were readily
distinguishable in the characters of the male pleopods, and that he proposed to
consider them as two species. At the same time he was good enough to send me
sketches of the first pleopods of both species for comparison with my own specimens.

The largest ' Discovery ' specimen is a male 18 mm. in length, and as far as I can
judge, it appears to be quite adult The copulatory apparatus on the first plcopods
agrees exactly with the sketch which Dr. Hansen sent me of the same apparatus in liis
Antarctic form. It would therefore appear that these specimens belong to Han.sen's
new Antarctic species. I have not attempted to give a detailed description with
figures of this form, since it is quite evident that an accurate djagljoan can only be
drawn up from a close study of this species and the true E. lucens side by side, and
a careful comparison, character by character. There are no specimens of the true
E. lucens in the ' Discovery ' collection, so I leave the descriptions of the two species
to Dr. Hansen, who has abundant material for the purpose.

I may mention here that some specimens of an Euphausia (labelled E\ splendens,
G. 0. Sars) in the small collection of Antarctic Schizopods kindly lent me by Prof.

p 2

16 W. M. TATTERS ALL.

D'Arcy W. Thompson, from the collections of the University College, Dundee, appear
to belong to this species. They were collected in the Antarctic Ocean, the exact
locality being uncertain, but it is believed to be in the neighbourhood of the South
Shetland Islands.

E u P H A u s I A, sp. ?, juv.
(PL IV., Figs. 7-9.)

Localities of captures:— Lat 49° 40' S., long. 172° 18' 30" W., five specimens,
immature, 8-9 mm.

Lat. 58° 49' 45" S., long. 154° 48' W., three specimens, immature, 10 mm.

The specimens from the above two localities all belong to the same species. They
were submitted to Dr. H. J. Hansen of Copenhagen, who agreed with my suggestion
that they were too young for absolute specific determination. I give here only a brief
description, pointing out a few of the characteristic features.

Carapace with a prominent slender denticle on lateral margins just over the base
of the third thoracic limbs ; antero-lateral margins slightly undulate, only partially
concealing the eyestalks and produced into a long, narrowly acute rostrum (Fig. 7)
extending almost to the anterior end of the eye and about half-way along the basal
joint of the antennules.

Pleon having the third segment provided dorsally on the median posterior margin
of the tergum with a slender spine (Fig. 9) ; sixth segment long and slender, about
twice as long as the fifth.

Antennular peduncle (Figs. 7 and 8), with a minute bluntly pointed simple lobe
on the inner distal corner of the basal joint ; a thin oblique lamella-like ridge running
across the third joint from the inner proximal to the outer distal corner and partly
continued down the inner side of the second joint.

Antennal scale reaching to about half-way along the terminal joint of the
antennular peduncle.

Telson having the portion beyond the sub-apical spines produced into an acute
apex with smooth margins ; two pairs of spinules present.

Uropods reaching to the level of the insertion of the sub-apical spines.

This species belongs to that section of the genus provided with a spine on the
dorsal surface of the third segment of the pleon. Among members of this section it
approaches most nearly to E. gibboides, Ortmann (1893), but Dr. Hansen has kindly
pointed out to me that it differs from that species in the much greater length of the
rostrum and its different shape.

The smallest of the specimens, i.e. all under 9 mm. in length, have the spine on
the third pleon segment still undeveloped, only the two largest ones, 9*5 mm. and
10 mm. in length, showing it fully formed. We have here slight evidence as to the
stage in development at which this spine appears. None of the specimens present any
larval characters in the form of the telson or antennules. It would appear, then, that

BOBHOPODA. 17

the spine, at any rate in this species, developes late in life, only after the final adult
form is reached.

The species was taken on the homeward voyage of the 'Discovery,' in the
extreme southern part of the Pacific Ocean between New Zealand and Cape Horn.

SUB-FAMILY NKMATOSCKLIN.K, Holt and Tattcrwll.

GENUS THYSANOESSA, Brandt

THYSANOKSSA MACRURA.

(PI. III., Figs. 1-12).

Tkyttmtitta mamtra, U. 0. San, 18*8; id. (1885); Orttmimi, 1M98 ; Htebbing, 1900; Holland
TaUenall, 190C (1) ; Conttare, 190C.

Localities of captures :-?-

Winter Quarters.

No. 4 Hole, 47 specimens, 7-18 mm.
No. 8 Hole, 40 „ 8-21 mm.

No. 12 Hole, 2 „ 7 and 20 mm.

From River Koettlitz, 2. 1. 03, G specimens, 9-14 mm.

Outward or Homeward Journey.

Lat 61° 46' S., long. 141° 12' E., 16. 11. 01, 18 specimens, 14-20 mm.

Lat 57° 25' 30" S., long. 151° 43' E, 20. 11. 01, 35 specimens, 12-22 mm.

Lat 54° 1' 15" S., long. 170° 49' E., 27. 12. 01, 1 specimen, 6 mm.

Lat 61° 13' 30" S., long. 173° 33' E., 31. 12. 01, 30 specimens, 12-19 mm.

Lat 66° 52' 9" S., long. 178° 8' 15" E, 3. 1. 02, 3 specimens, 8-9 mm.

Lat 70° 29' 27" S., long. 168° 51' 46" E., 26. 2. 04, 1 specimen, 28 mm.

Lat 49° 40' S., long. 172° 18' 30" W., 12. 6. 04, 2 specimens, 8-12 mm.

Lat 58° 49' 45" S., long. 154° 48' W., 24. 6. 04, 4 specimens, 10 mm.

Lat 59° 34' 30" S., long. 106° 28' 12" W., 28. 6. 04, 3 specimens, 7 mm.

Lat. 55° 44' S., long. 95° 43' 30" W., 1. 7. 04, 3 specimens, 6-8 mm.

•

Form (Fig. 1) of the body rather slender.

Carapace (Fig. 1) with a single rather long slender spine on the lower lateral
margin posterior to the middle, just above the origin of the sixth thoracic limb ;
antero-lateral corners acute and somewhat produced ; anterior margins very concave
and produced forwards into a long, slender, acute rostrum which reaches beyond the
eyes and far beyond the middle of the basal joint of the anteunules ; there is a low
keel on the anterior part of the carapace behind the rostrum, and a very faint
gastro-hepatic groove.

Pleon (Fig. 1) rather elongate, narrow and attenuate; segments unarmed; sixth
segment equal to or slightly less than the combined lengths of the preceding two;

18 W. M. TATTERSALL.

preanal spine well developed, provided with an external strong tooth and a comb-like
row of finer teeth up to twelve in number.

Antennular peduncle (Fig. 2) rather long and slender, considerably longer than half
the carapace ; basal joint rather flattened, considerably wider than the remaining two
joints, bearing on its outer distal corner a rather long and slender spine, its anterior
margin somewhat overlapping the basal part of the second joint, fringed with
numerous setae and exhibiting near the inner side a shallow fissure : terminal two
joints very long and slender, subequal in length, their combined lengths slightly
exceeding that of the basal joint.

Antennal peduncle (Fig. 3) very long and slender, nearly as long as the scale,
third joint less than half as long as the second.

Antennal scale (Fig. 3) extending to about midway along the terminal joint of
the autennular peduncle, almost five times as long as broad, outer margin terminating
in a well-marked spine, inner margin sloping away obliquely towards the apex ; spine
on the basal joint moderately long, slender and smooth.

Mouth parts (Figs. 4-6) do not offer any marked differences from those of
Thysanoessa gregaria as figured by Sars (1885, PL XXII).

Second thoracic limbs (Fig. 8), with the endopod remarkably long and well
developed, in full-grown specimens equalling nearly three-quarters of the total length of
the body from the eyes to the telson ; meral joint the longest, extending beyond the
tip of the antennular peduncle ; carpal joint a little shorter than the meral and more
slender, nearly four times the length of the propodal joint, armed with four slender
spiniform setae on the outer margin at the distal extremity, and five similar setae on
the inner distal margin ; propodal joint with five long spiniform setae on the outer
and six on the inner margin ; terminal joint small, tipped with six spiniform setae.

The remaining thoracic limbs (figs. 7, 9, 10) not differing greatly from the same
limbs in T. gregaria.

Copulatory apparatus (Fig. 11) on the first pleopod of the male exhibits a
structure very similar to that figured by Sars for T, gregaria, except that the distal
extremities of the two movable processes on the inner lobe do not seem to be serrate.

Telson slender, apex acutely produced and smooth ; sub-apical spines smooth ;
dorsal denticles two pairs, the first pair just anterior to the centre of the telson, the
second pair a little anterior to the insertion of the sub-apical spines.

Uropods exceedingly slender, inner one reaching the apex of the telson, outer
slightly shorter.

Length of the largest specimen, 28 mm.

I have thought it well to describe and re-figure this species, since Sars' original
description was taken from admittedly young forms, and I cannot find that mature
specimens have ever been described and figured. The changes that take place
during growth affect principally the rostrum, the second thoracic (or elongate) limb
and the preanal spine.

SCniZOPODA. 19

The rostrum is figured correctly by Sars from a specimen, 13 mm. in length.
At that stage in development the rostrum is an acutely triangular projection, tin-
sides of the triangle very nearly straight and but little concave. As growth prnimxli,
however, the sides of th6 triangular plate become deeply concave and the rostrum
assumes the form of a long narrow acutely spiniform projection, as shown in Fig. 2.
This gives it something of the form as seen in T. gregaria, but it is much longer and
more slender than in that species.

Bars describes and figures the preaual spine in his 13 mm. s]>ccinicn as armed
with only two teeth. This is correctly stated for a specimen that size, but increase
in size is also accompanied by an increase in the number of teeth in the prcanal spine,
till in a specimen 28 mm. in length I found the teeth to number twelve in addition
to the strong external one. The teeth are, moreover, fine and comb-like, und the
whole spine closely resembles that figured by Sars for T. gregaria, I have noticed
all intermediate stages in the present material. This clearly undermines the value of
the preanal spine as a specific character, since the number of teeth is dependent first of
all upon age, while individual variation must also be taken considerably into account.

The most considerable changes due to growth are, however, exhibited by the
second thoracic limbs. Sars describes them as much smaller than in T. gregaria,
with the nil-nil joint scarcely reaching beyond the middle of the antennal scale.
But the above description and accompanying figures show that the limb is
quite as well developed as in T. gregaria, and that the meral joint in full-grown
individuals actually extends beyond the antennular peduncle. Stebbing (1900)
has already called attention to the fact that in specimens of this species from
the Falkland Islands the elongate limbs were proportionately longer than in Sars'
drawing, but he gives no figures. I find that this statement applies generally to the
material in the present collection, even in specimens of the same size as Sars'. Only
two exceptions were noted, and in these two specimens the elongate limbs were
proportionally as in Sars' figures. I will deal with these two exceptions and Sara'
specimen below, but will proceed first to briefly note the changes which take place in
these limbs during growth. In two specimens, 8 and 9 mm. in length, the second
thoracic limbs are developed about as fully as in Sars' figures. They are considerably
less than one-half of the total body-length of the specimens, the meral joint extends
very little beyond the centre of the antennal scale, and is alxmt equal to the terminal
three joints combined, while the carpal joint is only about twice as long as the
propodal.

In specimens from 13 mm. to 18 mm. in length, these limbs are about equal in
length to the half of the length of the body, the meral joint extends to the tip of the
antennular peduncle, and is a little shorter than the terminal three joints combined,
while the carpal joint is now three times as long as the propodal. This is the
condition noted by Stebbing in his Falkland Islands specimens. In the full-grown
specimen, 28 mm., the elongate limlw measure 21 mm. in length, or three-quarters of

20 AV. 11. TATTERSALL.

the total body-length. The meral joint extends beyond the tip of the antennular
peduncle, and is about one-tenth shorter than the combined lengths of the terminal
three joints, while the carpal joint is four times as long as the propodal. It will be
seen, therefore, that these changes during growth are quite considerable, and most
evident in the elongation of the carpus.

It was noted above that Bars' specimen, 13 mm. in length, had the elongate
limbs only as well developed as in a specimen 8 mm. long. I have found in the
present collection two specimens, 16 and 17 mm. long, in which these limbs were at
the same stage of development. They were accompanied by specimens with more
fully developed elongate limbs, and I would suggest that the comparative shortness
of these legs in the above-mentioned two specimens and in Sars' type is due to their
having at some earlier period been broken off and grown again. This is a common
occurrence in Decapods, and I should think is by no means rare in Euphausians,
especially those with elongate second legs, which must be specially liable to be
detached. The re-developed limb is usually shorter than the one it replaces.
I believe this explanation to be the correct one in the present instance, though it may
be that we have here a case of retarded development. Further slight changes during
growth are seen in the proportional length of the last pleon segment, which in very
young specimens is slightly longer than the combined length of the preceding two,
while in full-grown specimens the reverse is seen, though the differences either way
are not great.

The antennular peduncle is a little longer proportionally in full-grown specimens,
while the outer uropod likewise approaches more nearly to the length of the inner
ones, though even in full-grown specimens it remains slightly shorter than the latter.
This account of the growth changes in T. macrura reduces considerably the points of
distinction formerly supposed to exist between it and T. gregaria. The most
conspicuous difference is in the proportional length of the last pleon segment, and this
distinction would seem to hold throughout life. T. macrura is, besides, a much more
slender and graceful form than T. gregaria, with its parts proportionally attenuated.

Distribution. — T. macrura was recorded from several localities in the Southern
Ocean, South Atlantic, and Antarctic Ocean by the ' Challenger.' It has since been
noted by Stebbing from the neighbourhood of the Falkland Islands, and by Coutiere
from the collections of the French Antarctic Expedition. These records, together with
the list of localities for specimens in the present collection, indicate that the species
has a wide range in the waters of the southern temperate and Antarctic regions, but
its northern -limit would seem to be the 40th parallel of south latitude.

EUPHAUSIAN LARVAE.

The larvae listed below could not be referred to any species, and are merely noted
here for completeness.

SCHIZOl'ODA. 21

Lat 57° 25' 30" a, long. 151° 45' E, 21. 11. 01, numerous Euphausian
\fftanauj>lii and Ctilyptoj>i« larvae, and two Furcilias.

The Mttanauplii and Calyptopis larvae arc characterised by the presence of a
short blunt posterior median spine on the carapace. The front part of the latter, which
forma the hood over the eyes, appears to have ita margin quite smooth. In the
majority of the Calyptopis larvte a small spine on the lateral margins of the carapace
is clearly present The apex of the telson is very slightly emargiuate. The largest
Calyptopis larva measures 4 mm. in length.

Lat 56° 31' &, long. 156° 19' 30" E, 22. 11. 01, seventeen Calyptopu and fifty
Furcilia larvae.

The Calyptopis larvae belong to the same species as in the preceding lot

The Furcilia larva) measure from 2'5 to 4 mm. They present no features of
note, but from their size they proliably belong to a smaller species of adult than the
Calyptopis larvie which accompany them. All present a lateral spine on the carapace,
and the rostrum is in the form of an acutely produced triangular plate.

Lat 54° 1' S., long. 170° 49' E, 27. 12. 01, numerous Metanaujtlii, Calyptopi*,
and Furcilia larvae.

All these are apparently stages of one species. The largest Calyptopis larva
measures 3 mm. and is without a posterior median spine on the carapace. The
anterior margin of the hood is smooth ; there is a lateral spine on the carapace, and the
apex of the telson is very lightly convex. The largest Furcilia measures 4 • 5 mm.

Lat 61° 13' 30" S., long. 173° 33' E., 31. 12. 01, one Pwrdlin larva, 5 mm.,
apparently the same species as the following larvae.

Lat 63° 4' 24" S., long. 175° 47' 57" E, 1. 1. 02 ; numerous Furcilia larva) from
4 mm. to 5*25 mm. in length, and apparently the same species as the preceding larva.

Wood Bay, 22. 2. 04, numerous small Metanauplii and Calyptopis larvae, the
largest of which measured 2 mm.

Lat, 70° 29' 27" S., long. 168° 51' 46" E, 26. 2. 04, one late Cyrtopia larva,
9 mm. in length, which, from its size, is probably Euphausia superla.

Lat 49° 40' S., long. 112° 18' 30" W., 12. 6. 04; forty-eight larva-, from
Calyptopis stage to post-larval form, measuring 5 mm., and probably the larva- of
T/iystimnissa macrura.

ORDER MYSIDACEA.
FAMILY PETALOPHTHALMID^E, Czerniavsky.

Pttalophthalmidm, Crerniavaky, 1882.
PftalopUhalmidte, Holt and Tattenall, 1906 (2).

When defining this family Holt and Tattersall, 1906 (2), overlooked the fact that
it had been established and defined, albeit rather incompletely, by Czerniavsky (1882)
a quarter of a century previously. Czerniavsky, however, had no specimens of any of

VOL. IT. O

22 W. M. TATTEESALL.

the genera of the family before him, but drew up his definition from the descriptions
and figures of Willemoes-Suhm (1875). He includes in the family the single type
genus, Petalophthalmus, Will.-Suhm, with two species — P. armiger, Will.-Suhm, and
P. willemoesii, a new species which he founds for the reception of the female ascribed
by Suhm to P. armiger.

Faxon and Hansen, however, have since pointed out that the female specimen
described by Will.-Suhm is in reality a Boreomysis, probably B. scyphops, G. 0. Sars,
so that if this latter view of its identity be adopted, P. willemoesii becomes a synonym
of B. scyphops.

Czerniavsky's definition of the family is inadequate, inasmuch as no reference is
made therein to the remarkable characters of the carapace, and the first and second
thoracic limbs, while the importance which is given to the supposed characters of the
exopods of the thoracic limbs is exaggerated, the difference in development as
compared with those of the Mysidse being very slight.

The definition given by Holt and Tattersall, 1906 (2), may, therefore, be adopted
with some slight alteration in the characters ascribed to the eyes, rendered necessary
by recent discoveries.

Examination of British specimens of Hanseiwmysis fyllse (Hansen, 1887) has
revealed the fact that eyes are, in reality, present in this species. They resemble those
described below for the Antarctic species, except that the lappets are much reduced
and almost obsolete.

Further, in a new species of Petalophthalmus, P. oculatus, recently defined by
Illig (1906), the eyes are described as well developed, with the cornea bright brown in
colour and distinctly facetted.

In the amended definition of the family, therefore, the description of the eyes
would read : — Eyes (first cephalic appendages) small, either imperfectly developed as
lamellar or spiniform organs, without visual elements, or furnished with a distinct
cornea in which visual elements are clearly defined and functional.

GENUS HANSENOMYSIS, Stebbing.

Arctomysis, Hansen, 1887 (non Czerniavsky, 1883).

Hansenomysis, Stebbing, 1893.

Hansenomysis, Holt and Tattersall, 1906 (1 and 2) ; Tattersall, 1907.

The name Arctomysis, given to this genus by Hansen (1887), having been already
used by Czerniavsky (1883) for an entirely different form, was changed to
Hansenomysis by Stebbing (1893). Arctomysis Czerniavsky is itself a synonym of
Boreomysis G. 0. Sars.

Of the other three genera belonging to the Petalophthalmidse — Pttolophthalmus,
Ceratomysis and Scolophtlialmiis, Hansenomysis comes nearest to the last. Both
agree in having the first thoracic limbs devoid of exopods and lacking the internal
lamelliform moral lobe, and in the presence of well-developed exopods to the second

scnizoponA.

thoracic limbs. Whereas, however, in Scolophthalmw the rostrum is prominent and
the eyes are modified into sharp spiniform organs, Hansenomysi* baa the rostrum
obsolete and the eyes more or less leaflike.

Males of this genus have not yet been noted, but specimens of that sex of the
northern species, H. fyllee (Hansen, 1887), have come into my hands. Detailed
examination and description are reserved for a future occasion, but it may lie
mentioned here that, besides having the pleopods biramous, males also have the Iwwal
portion of the inner flagellum of the antennule considerably thickened and adorned
with rings of setae.

HANSKNOMYSIS ANTARCTICA.

(PI. V., Figs. 1-19.)
ffanMnoiqysis antarctic*, Holt and Tatteraall, 190C (1).

Locality of capture. — Off Coalman Island, 100 fathoms, two specimens, females,
20 mm.

Form (Fig. 1) compact, rather slender, tapering considerably towards the
posterior end.

Carapace (Figs. 1 and 3) short, sub-membranous, leaving the last two thoracic
segments completely exposed, and part of a third visible behind its posterior
emargination ; anterior border produced, but very slightly, into a broadly and evenly-
rounded but somewhat strongly upturned rostrum ; antero-lateral angles evenly
rounded and extending forwards as much as the rostrum ; cervical sulcus well marked
and rather deep, the posterior margin bounded by a conspicuous and rather sharp ridge
formed by the carapace. Behind the cervical sulcus is a shield-shaped dorsal area,
indicating the attachment of the carapace to the thorax, behind, and on either side of
which the wings of the carapace are free. A slight ridge runs from the antcro-lateral
angles, first downwards and then posteriorly, to meet the cervical sulcus, while a
shallow groove runs forward on each side from the dorsal shield-shaped area, thus
marking off a hepatic area, on which is a prominent forwardly -directed spine with a
broad base. A shallow depression follows the base of the rostrum, and merges on
either side into the groove formed by the ridge from the a ntcro- lateral angles. A
small blunt and rounded spine is present on the gastric area.

/'/•«/• (Fig. 1)9 mm. in length, a little longer than the thorax, which measures 8 mm.
from the eyes to the posterior margin of the last free segment ; segments cylindrical,
postero-lateral inferior margins not at all produced as epimera ; first segment arcuate
in dorsal contour, its anterior margin slightly raised above the level of that of the last
thoracic segment, its posterior margin broadly produced, so as to partly cover the
second segment, the whole forming a sort of " cap " over the junction of the thorax and
pleon ; second to fifth segments sub-equal in length and succeediogly narrower ; sixth
segment narrower than any of the preceding ones, and nearly twice as long.

a I

24 W. M. TATTERSALL.

Eyes (Figs. 1, 2 and 3) small, united at their base into a thick flattened pad, from
the anterior part of which proceed two thin, short, sub-triangular, slightly-diverging
lappets, which do not reach the middle of the basal joint of the antennular peduncle ;
visual elements entirely absent.

Antennular pedumle (Fig. 2) short and stout, its three joints sub-equal in length
and quadrangular in outline ; basal joint, with a single long seta, on its internal distal
corner, and a more or less continuous submarginal row of setae across the anterior
dorsal region ; second joint with about seven long stout plumose setae on its inner
margin and two or three long setae on the outer distal corner ; third joint with about
eleven long stout plumose setae on the inner margin. On the dorsal surface of the
basal joint, partly concealed by the eye in dorsal view, is an organ of rather
problematical function (Figs. 2, 4 and 5). It appears to consist of a shallow depression
bounded by a raised ridge marked with pigment, and overhung by a membranous flap,
which apparently rises from its posterior border. The flap only imperfectly covers the
depression. In the preliminary notice of this collection it was suggested that this
organ might prove to be auditory in function, but under moderately high powers of
the microscope no otoliths or even sensory hairs could be distinguished in the
shallow pit.

Antennal peduncle longer than the antennular, and considerably more slender ;
distal joint shorter than the preceding.

Antennal scale (Fig. 2) lanceolate in shape, about three-and-a-half times as long as
broad, apex evenly rounded, the whole of the inner margin and distal third of the
outer margin setose ; proximal two-thirds of the outer margin devoid of setae, but
armed with eleven strong spines, the proximal one of which is the shortest, and is
situated at the end of the proximal quarter of the outer margin, the spines increasing
in size distally ; spine on the outer distal corner of the basal joint short, but
prominent.

Mandibles (Fig. 6) with the cutting edge prominent and molar process well-
developed and rather long ; between the cutting edge and the molar process is a
single spine-like seta, and in the left mandible a lacinia mobilis in addition ; palp
(Fig. 7) rather long and powerfully developed, terminal joint shorter than the
penultimate, both joints armed on both outer and inner margin with numerous long
and rather stout setae.

First maxilla (Fig. 8) having the outer lobe much larger than the inner, and
armed at its apex with about thirteen spines, behind which is a row of five plumose
setae ; inner lobe armed at its tip with four long plumose setae.

Second maxilla (Fig. 9) consisting of the usual three lobes, a two-jointed palp
and outer setiferous plate, the setae arming the appendage being numerous and
rather strong.

First thoracic limb (Fig. 10) short and stout, devoid of exopod, but with well-
developed epipod ; third joint small ; fourth joint with numerous setae and a row of

arnrzopoDA. 25

six short stout spines on its inner margin ; fifth joint with three, sixth joint two, and
seventh joint four rather long, strong plumose spines on their inner margins as well
as numerous setae.

Second thtiracic limb* (Fig. 11) with the endopods longer and rather more slender
than the first, ezopods well developed ; fourth or meral joint produced internally into
a large setiferous lamelliform lobe nearly as long as the fifth joint ; the latter, the
longest joint of the limb, longer than the combined length of the sixth and seventh
joints, its outer margin armed with a single seta, the distal half of the inner margin
slightly excavate with a row of eight short, closely-set plumose spines and a single
long plumose seta on the cmarginate portion, and a few long simple setiu, set wiildv
apart, on the proximal portion of the inner margin ; sixth joint longer than the
seventh, it« outer margin armed with a few long seta), the proximal portion of the
inner margin bearing a row of altout nine short closely-set plumose spine* and a sin^l.-
long plumose seta, the distal portion of the inner margin with a few long simple sctaj ;
seventh joint small, armed with numerous long and rather stout simple setaj.

Third to fifth thoracic liiiJm (Fig. 12) with the endopods feeble, long and slender;
sixth joint slightly longer and more slender than the fifth ; seventh joint very small,
forming with two strong spines a very minute chela, densely clothed with short fine
setae ; the rest of the endopod armed with a few short scattered seta}.

Sixth to eighth tlutracic limb* (Fig. 13) with the endopods slightly longer and
stouter than those of the three preceding pairs ; sixth joint shorter than the fifth ;
seventh joint small and bearing a long slightly-curved nail, the junction between the
nail and seventh joint being indicated by a seta on the inner margin ; rest of the
endopod feebly armed with short seta}.

Etopods of the second to eighth thoracic limbs well-developed ; basal joint long
and rather narrow, the outer distal corner rounded ; flagelliform part composed of from
ten to thirteen joints.

Incubatory lamellte, seven pairs, situated on the second to eighth thoracic limbs.

Pleopods (Figs. 14 to 18) in the female uniramous, the first pair small, succeeding
pairs increasing in size to the fifth pair, which are slightly longer than the sixth
segment of the pleon ; first four pairs one-jointed ; fifth pair two-jointed, the second
joint longer than the first ; all the pleopods bearing long seta} at the apex.

Tekon (Fig. 19) rather massive, longer and a little wider than the hut segment
of the pleon, dorsally grooved, oblong in shape, slightly wider at the apex than at the
base, its margins lightly arcuate ; apex truncate or very lightly emarginate, bearing
a single median spine with six or seven long spines on cither side ; lateral margins
armed with from twenty-five to thirty fairly long spines arranged more or less
in series.

Inner uropods broken in both specimens.

Outer uropods (Fig. 19) nearly twice as long as the sixth segment of the pleon,
two-jointed, the terminal joint about one-seventh as long as the basal ; outer margin

26 W. M. TATTERSALL.

of the basal joint without setae, but- armed with twenty-one stout spines increasing in
size posteriorly.

Length of adult and ovigerous female, 20 mm. from the eyes to the tip of the
telson.

Colour of preserved specimens light brown, with a broad band of dark brown
pigment across the dorsal surface of the first segment of the pleon and scattered
patches of dark pigment on the lateral parts of the carapace, basal joints of the
antennules and antennae and the basal membranous pad of the eyes.

One of the specimens has young, considerably advanced in development, in the
incubatory lamellae.

In both specimens the telson is considerably damaged, and the description and
figures have been drawn up from both specimens and fragments of the telsons found
along with them. This fact must be borne in mind in dealing with specimens of this
species which may be found by future expeditions. It was a matter of considerable
surprise and no little interest to find in this collection two specimens belonging to
a genus hitherto known only from a single specimen from Greenland and two taken
off the cost of Ireland.

There can be no doubt that //. antarctica is co-generic with //. fyllse (Hansen,
1887). All the distinctive characters of generic importance in the mouth parts and
thoracic limbs of the latter are reproduced in //. antarctica down to the minutest
detail. The points of difference between the two species are, however, sufficiently
well-marked and numerous enough to justify specific separation. They may be pointed
out as follows :—

H. antarctica is in general build a more robust and less fragile species than
H. fyllse.

Antennae. — In H. fyllse the terminal joint of the peduncle is longer than the
penultimate, whereas in H. antarctica the reverse obtains.

Antennal scale. — In H. fyllse the outer margin bears only five spines, between
which are numerous setae. In H. antarctica, on the other hand, there are eleven spines
on the outer margin and no setae between them.

Telson. — The telson in Hansen's type-specimen was broken, but so much of
it as remained showed that the armature consisted of both spines and setae. In
H. antarctica the telson is armed with spines only, which are probably more numerous
than in //. fyllse. The shape of the telson in both species is also somewhat divergent.
That of //. antarctica recalls rather markedly the telson of Petalophthalmus armiger as
figured by Sars in the ' Challenger ' Report.

Outer uropods. — //. fyllce has the outer uropods armed with both spines and
setae, whereas in //. antarctica there are spines only present. The spines in H. fyllce
number six, while in H. antarctica there are twenty-one.

Pleopods. — The type specimen of H. fyllce had only one pleopod remaining. This
was one of the third pair, and is described by Hansen as biarticulate. Presumably,

SCH120PODA. 27

therefore, fhe fourth and fifth pleopods will likewise be found to IMJ biarticulatc when
perfect specimens are examined. In H. antarctica only the fifth pleopods are
Inarticulate, the remaining pairs consisting of a single joint only.

The eyes in the genus are described for the first time. They arc remarkable
chiefly for their small size and degenerate structure, for the complete absence of visual
elements, and the subservience, either entirely or in great part, of ophthalmic functions
to those of prolwbly a tactile nature.

The cap-like form of the tergum of the first segment of the pleou recalls the
somewhat similar form of the third pleon segment in many Carida, and suggest* that
the posterior part of the body is capable of great ventral flexure. The ' Discovery '
expedition is to be congratulated on the finding of this species, by far the most
interesting Schizopod in the collection.

FAMILY MYSID^E.

•

SUB-FAMILY LEPTOMYSIN.E, Norman.
GENUS PSEUDOMMA, G. 0. Bars.

PsEUDOMMA BEUJNXB.

(Plate VI., Figs. 1-8.)
Ptntdomma btlgifm, Holt and Tatteraall, 1906 (1).

Locality of capture.— Lat 78° 25' 40" 8., long. 185° 39' 6" E., 300 fathoms, one
specimen, immature female, 23 mm.

Form (Fig. 1) compact and moderately stoutly built.

Carapace (Fig. 1) large, less than half the total length of the body, covering
laterally all the segments of the thorax, but dorsally exposing the last one behind its
posterior emargination ; its anterior margin very slightly produced into a blunt, very
broadly rounded rostrum ; antero-lateral angles rounded ; cervical sulcus well marked.

Pleon (Fig. 1), excluding the telson, about half the total length of the body from
the eye to the tip of the telson ; first four segments subequal in length and slightly
longer than the fifth ; sixth segment twice as long as the fifth.

Eye plates (Fig. 1) contiguous, exhibiting only a very slight anterior median
cleft ; each plate subquadrangular or rhomboidal in shape, nearly twice as broad as
long, antero-lateral angles rounded, anterior margin nearly straight, no serrations or
armature of any kind ; no pigment present in preserved specimens ; corneal leans
absent, but the ramifications of the optic nerve arc clearly visible in dorsal view.

Ant. a n til- 1 r peduncle (Fig. l) short and stout, not extending beyond half the
length of the antennal scale ; basal joint almost entirely covered by the ocular lamina-,
a few plumose seta; on each anterior corner ; second joint very short, more than twice
as broad as long, a few short setae on the outer distal corner, inner margin with a few
longer plumose setae ; third joint longer than either of the other two and slightly

28 W. M. TATTERSALL.

narrower, rectangular in shape, outer margin unarmed, inner margin beset with a few
plumose setae.

Antennal peduncle (Fig. l) equal in length to the antennular, but more slender;
last two joints subequal in length.

Antennal scale (Fig. 1) slightly longer than the last segment of the pleon and
twice as long as the antennular peduncle, about three and a half times as long as
broad, outer margin entire and terminating in a very strong spine, beyond which the
apex of the scale is but slightly produced ; spine on the basal joint short and acute.

Mouth parts (Figs. 2, 3, 4, and 5) exhibit no salient points of difference from
those of the type species of the genus P. roseum, G. 0. Sars.

First and second thoracic limbs (Figs. 6 and 7) agreeing in the main with those
figured by Sars for P. roseum, but a dactylus is distinctly visible among the dense
mass of plumose setae arming the terminal joints, though it is smaller and shorter
than the terminal joint ; exopods having the outer distal corner of the basal joint
round, though produced, and the flagelliform part composed of ten to twelve joints.

Telson (Fig. 8) slightly shorter than the last segment of the pleon, tapering to
an evenly rounded apex which is one-third as wide as the base ; apex armed with a
median pair of plumose setae and four pairs of strong smooth spines, the innermost
and largest of which equals one-sixth of the telson in length ; distal half of the lateral
margins armed with five shorter spines.

Inner uropods half as long again as the telson, armed with a single long spine
in the region of the inner posterior corner of the otocyst.
Outer uropods about twice as long as the telson.

Length of an immature female, 23 mm. By an error the specimen was described
in the preliminary report as adult. This is scarcely correct, since the incubatory
lamellae are still only about half developed, so that the adult female probably reaches
to nearly 30 mm. The specimen is badly mutilated, the third to the eighth thoracic
limbs being entirely absent.

P. beljicce is far and away the largest species of the genus yet described, none
of the other ten known species exceeding 15 mm., whereas adult specimens of this
species must reach to nearly 30 mm. It is most nearly related to P. sarsi Will.-
Suhm, described by Sars (1885) from the ' Challenger ' collections for specimens taken
at Kerguelen Island. Besides the great difference in size (P. sarsi measures only
14 mm., adult specimens), the only other conspicuous difference is in the ocular
laminae. In P. sarsi the antero-lateral angles of the eye-plates are serrate, whereas
in P. belf/ic(e they are quite smooth. Minor differences in the shape of the antennal
scale and telson may also be noted. The antennal scale in P. sarsi has the spine
terminating the outer margin less strong than in P. belaicce, while the apex of the
scale is more produced. The telson in P. sarsi has the apex more truncate than
P. Iielgirrn and the lateral margins, according to Sars, bear eight short spines.
Mr. Holt, however, who has kindly examined the types of P. sarsi in the British

snilZOI'ODA. 2U

Mil-, ii 111. iii (onus me that only five spines tire present on the lateral margin*, MI that
the armature of the telson of P. wirai approximates closely to that of P. Myica.

The only other species of Pseudoinma having Hmooth ocular lamina it
P. australe, G. 0. San (1885), from Bass Straits, Australia. The vastly different
form of the antennal scale in the latter, however, abundantly distinguishes it from
P. Mgicat.

Besides the single ' Discovery ' specimen, this species is also known from the
' Bclgica' collections, and has been described by Hansen in MS. under the name which
is here used. It is possible that the mutilated specimen noted by Sara (1885, p. 191)
from 1675 fathoms in the Antarctic Ocean may have belonged to this species rather
than to P. sarsi Sars notes that it was much larger than the latter.

Genus DACTYLAMBLYOPS, Holt and Tattersall.

Duetylamblyopt, Holt and Tatteraall, I'Jm; (1).
DartyltrythrofM, 11%, 1 !>"•;, iion Holt and TaUcreall, 1II05.
? Amblt/opi (par*), Ohlin, 1901.
Daetylamblyopt, TatU-reall, 1907.

This genus was established for the reception of the single rather mutilated specimen
<>t° />. /(.•» /•/.•.-••/// in the present collection. Since the publication of the preliminary
notice of the 'Discovery' Schizopoda, however, two closely allied species have been
discovered off the west coast of Ireland (Tattersall, 1907). A clearer idea of the
exact relationships of the genus has thus been gained, and while the species referred
thereto appear, in the present state of our knowledge, to form a natural group, it w
undeniably very nearly allied to Dactylerythrops, Holt and Tattersall (1905), to
which genus, indeed, the present species was referred by Illig (1906).

The definition of the genus given by Holt and Tattersall, 1906 (1), may therefore
l>e amended as follows :—

DACTYLAMBLYOPS, Holt and Tattcrsall.

Characters generally as in AnMynps, G. 0. Sars, except : —

Eyes placed close together, but not contiguous, more or less pyriform in shape,
furnished with distinct and definite peduncles ; visual element* imperfectly developed,
numerous, reaching to the surface of the eye, and probably directly functional as
organs of sight ; outer distal corner rounded, and not produced into a digitiform
process ; a short blunt process always present on the inner and upper surface.

Second thoracic limbs with the endopods not noticeably short, but well developed,
and considerably longer than the endopods of the first thoracic limbs.

Tebnii not very long, triangular in shape, the distal part* of its margins armed
with more or fewer spines ; median sefoe absent.

Type species, D. hodgsoni, Holt and Tattersall.

VOL. IV. H

30 W. M. TATTEKSALL.

The absence of median setae from the apex of the telson is not necessarily
of generic importance, since the genera Pseudomma and Dactylerythrops both contain
species in some of which these setae are present, and others in which they arc
wanting. As, however, all three species at present referred to this genus are without
median apical setae, it is convenient to retain this character in the generic definition.

DACTYLAMBLYOPS HODGSONI.
(PL VI, Figs. 9-16.)

Dactylamblyops hodgsoni, Holt and Tattersall, 3906 (1).
Dactylerythrops arcwta, Illig, 1906.

Locality of capture :— Lat. 66° 52' 09" S., long. 178° 08' 15" E., 2030 fathoms,
one specimen, male, 13 mm.

The single specimen in the collection is considerably damaged, the antennules,
antennal scales, and the third to the eighth thoracic limbs being missing. A complete
description is therefore not possible, but it is hoped that as many of the characters as can
be made out with certainty will suffice for future recognition of the species in collections.

Carapace submembranaceous, covering all the thoracic segments except the last
one, anterior margin produced into a blunt, broadly but evenly rounded rostrum
projecting between the eyes ; cervical sulcus well marked ; antero-lateral angles rounded.

Pleon slightly longer than the carapace ; first five segments subequal in length ;
sixth nearly twice as long as the fifth.

Eyes (Fig. 9) small, placed on definite peduncles, not in any way contiguous,
pyriform in shape, external angle evenly rounded, a short digitate process arising from
the inner dorsal face ; visual elements imperfectly developed, apparently represented
by numerous minute granular bodies with a refractive centre ; a large opaque ganglionic
mass, probably the optic nerve, visible in the peduncle, from which a nerve fibre
proceeds to the cornea.

Antennal peduncle short, composed of three subequal quadrangular joints.

Antennal scale broken on both sides, but there does not appear to be a spine on
the outer corner of the basal joint.

Mouth parts (Figs. 10-13) not differing markedly from those figured by Sars for
Amblyops abbreviata (1870-79).

First thoracic limbs (Fig. 14) with the endopod substantially of the same form as
in A. abbreviata.

Second thoracic lim})S (Fig. 15) of essentially the same structure as in A. abbreviata,
but with the endopod apparently much longer, being nearly twice as long as the
endopod of the first thoracic limbs.

Genital appendix to the last thoracic limbs of the male terminating in two lobes,
the larger of which bears six long setae, the smaller one being devoid of setae, but
apparently having a covering of very fine hairs.

SOIIIZOPODA. 31

PUopodf in the male agreeing in nil points with those of males of the genus
Amblyop.i.

Tfl»m (Fig. 1G) not quite as long as the last segment of the pleon, triangular in
slta|ie, tapering evenly to a narrowly rounded apex, nenrly twice as long an broad at
its l»ase ; distal half of each lateral margin armed with nineteen spines, increasing in
length towards the apex, the terminal ones alxmt one-tenth of the total length of the
telson ; median sct«e absent.

Uropoda broken on both sides, but the inner one poaMMM a single strong spin*-
on the ventral surface at the inner posterior angle of the otocyst

Length of the single specimen, an apparently adult male, 13 mm.

There can be little doubt, 1 think, that Dactylerythropa arcwita, Illig (190G), is
the same species as the present one. Minor differences, it is true, are to !•<• noticed.
For instance, the visual elements of the eye in Illig's species are represented as larger
and less numerous than in D. hodgsoni. It may l»e that in the former preservation has
caused the visual elements to mass in groups of more or fewer lenses, since in the
'Discovery' specimen, as the figure (Fig. 9) shows, the visual elements are not
regularly arranged, but more numerous in some places than others, an effect pmlmbly
due to the mode of preservation.

Moreover, in D. arcvuiia, the telson is not quite so long compared with the
breadth at its base as in D. hodgsoni, and is furnished with only eighteen spines on
each lateral margin. But these differences are insignificant when compared with the
general agreement between the two species in most points. One or two characters
may be added to the above diagnosis from Illig's description and figures.

Antennular peduncle long and rather stout, last joint the largest and nearly equal
to the combined length of the other two.

Antennal scale slightly over-reaching the antenuular peduncle, about four times
as long as broad, outer margin entire, and terminating in a short spine beyond which
tin- apex of the scale is only slightly produced.

Four species of this genus are now known, D. sarsi (Ohlin, 1901), D. thawnatvp*
and D. goniops, Tattersall (1907), and the present species.

From the other three species, D. h^gsoni is at once distinguished by the shajte
of the eye. In D. sarsi the eye is acutely pointed in front ; in D. ttuiumati>ps it
is of quite peculiar form, with an outer equatorial membranous ridge, while in
D. goniops the eye is quadrangular rather than pyriform in shape. Otherwise the
four species are rather closely allied and form quite a distinct generic group to
themselves.

Z>. hodgsoni is at present only known from very deep water in the Antarctic
Ocean. Illig's specimens were collected over a depth of 4000-5000 metres, while the
present specimen was dredged in 2030 fathoms (ca. 3700 metres).

H 2

W. M. TATTERSALL,.

SUB-FAMILY MYSIDETIN.E, Holt and Tattersall.

GENUS MYSIDETES, Holt and Tattersall.

? Mi/sidopsis, G. 0. Sars, 1883 and 1885, non G. 0. Sara, 1864.
Mysideis (pars), Holt and Tattersall, 1905, non G. 0. Sars, 1869.
Mysidetes, Holt and Tattereall, 1906 (1) and (2).
Metamysidella, Illig, 1906.

A full diagnosis of this genus and a statement of its possible affinities have
already appeared, Holt and Tattersall (1906 (2)). While agreeing in most characters
with the normal genera of the Leptomy since, the rudimentary nature of the pleopods
of the male offers a feature of sharp distinction from members of that sub-family
and has necessitated the establishment of a new sub-family for its reception. In
the form of the telson and armature of the inner uropods it approaches rather closely
to the genera Heteromysis and Mysidella, but whereas in the former the third, and
in the latter the first, thoracic limbs are peculiarly modified and strongly armed, in
Mysidetes both these limbs are of normal structure. The external resemblance of
females of the present genus to those of Mysidopsis and Mysideis has already been
noted, and a comparative table of their respective characters was given (Holt and
Tattersall 1906 (2)). It will suffice here to mention that Mysidetes differs from
Mysidopsis, (1) in having a well-developed molar tubercle to the mandible; (2) in
the presence of a setiferous expansion of the inner margin of the basal part of the
second maxilla, and (3) in having the endopods of the first thoracic limbs seven-
jointed instead of six.

From Mysideis it is distinguished by having the endopods of the first and
second thoracic limbs of normal stoutness and usual armature, instead of being
unusually massive and strongly armed ; while from both genera it differs, (l) in
having the cleft of the telson armed with spines ; (2) in the uropods having a row
of spines in their inner edges extending well over half-way down their length, and
finally, (3) in having the pleopods of the male rudimentary.

Mysidopsis indsa, G. 0. Sars (1885), probably belongs to this genus. It was
described in the 'Challenger' Report from a specimen taken off Australia. This
specimen is a female much mutilated, and diasection was neither practicable nor
desirable. The telson and inner uropods, however, conform to the type found in
Mysidetes.

The genus Metamysidella (Illig, 1906) is undoubtedly synonymous with this
genus, though no mention is made in the diagnosis of the character of the pleopods of
the male. In all other features the two genera agree absolutely.

33

MYSIDKTRS POSTHON.
(PL VII., Figs. 1-13.)
MytidtUt pttOun, Holt and Tattcmll, 1906 (1).

Localities of captures:— Off Coulman Island, 100 fathoms, one specimen, male.
85 mm.

Winter Quarters, 5. 6. 02. 56 fathoms, one specimen, female, 23 mm.

Winter Quarters, 29. 8. 03. No. 12 Hole, 25-30 fathoms, three specimens, two
females and one male, 21 mm.

General form (Fig. 1) compact and moderately robust

Carapace (Fig. 1) leaving the last segment of the thorax exposed posteriorly;
antero-lateral margins produced into a short obtuse rostrum not extending l>eyond
the eyestalks ; antero-lateral corners rounded, cervical sulcus well marked.

Pleon (Fig. l) longer than the carapace ; first five segments sub-equal in length ;
sixth segment aliout once and a half to twice as long as the fifth.

Eyes (Fig. 1 ), large, globose ; pigment brown.

Antennular peduncle (Fig. 2) much shorter than the nntennal scale ; luisal joint
longer than the terminal joint, and having ite outer corner produced Insyond the
distal extremity of the second joint, the produced part tipped with four or five long
setae ; second joint small ; third joint almost square in shape ; antennular brush in
the male rather small and feebly hirsute.

Antennal peduncle (Fig. 3) slightly shorter and more slender than the antennular,
and little more than half as long as the scale ; third joint shorter than the second.

Antennal scale (Fig. 3), lanceolate in shape ; between four and five times M
long as broad ; setose all round ; a minute second joint at the apex ; a spine on the
outer distal corner of the l>asal joint.

Mandibles (Fig. 4) with a well-developed molar process and cutting edge ; palp
(Fig. 5) with the second joint somewhat expanded and armed with long setae on
both margins ; third joint not much expanded, a row of strong plumose seta? on
the lower edge, and two very strong simple spine-like seta: at the tip.

Second maxillae (Fig. 7) with the setiferous expansion of the basal joint well
developed.

Endopods of the first thoracic limbs (Fig. 8) of about the same build as in the
genus Mysidopsis, but seven-jointed ; masticatory lobe well developed ; inner margins
of the proximal four joints armed with numerous plumose setae ; sixth joint bearing
a well-developed nail and beset with numerous plumose sette.

Endopods of the second thoracic limbs (fig: 9) very similar to those of Mysidvpti* ;
longer than the first ; sixth joint armed with numerous plumose seta;, but in the
specimen dissected I was unable to detect a nail, though it may have been
broken off.

34 W. M. TATTERSALL.

Endopoda of the third thoracic limbs (Fig. 10) with the merus longer than the
tarsus ; latter composed of six joints ; nail well developed and longer than the last
joint of the tarsus.

Endopods of the remaining thoracic limbs become successively longer and more
slender from the fourth to the eighth ; the increase in length takes place chiefly in
the ischial joint ; the number of joints in the tarsus of the endopods also increases
in the more posterior limbs ; in one specimen there were six joints in the tarsus of
the third limb, six in the tarsus of the fourth, ten in the tarsus of the seventh, and
twelve in the tarsus of the last limb.

Genital appendix (Fig. 11) on the last thoracic limb of the male exceedingly
long and slender, equal in length to the first three joints of the limb to which it is
attached.

Pleopods (Fig. 12) similar in both sexes, consisting of a single ramus bearing
proximally and externally a rather large process tipped with setae.

Telson (Fig. 13) a little longer than the last segment of the pleon, and more
than twice as long as broad at its base ; cleft at the apex for nearly a quarter of its
length, cleft rather wide, its margins armed with about eighteen teeth on each side ;
the apex of each lobe of the cleft armed with a pair of spines, the inner one the
shorter ; lateral margins armed throughout their whole length with about seventy
spines, which become arranged in series towards the apex.

Inner uropods slightly longer than the telson, with a row of moderately slender
and long spines on its inner ventral margin, varying in number from twenty-six to
twenty-eight, and extending from the otocyst to near the apex ; spines not arranged in
series, but increasing in size distally. In some specimens the spines extend further
down the uropod than in others.

Outer uropods about half as long again as the inner.

Length of an adult female with embryos in the brood pouch, 21 mm. ; of an
apparently adult male, 25 mm. A second female with embryos in the brood pouch
measured 23 mm.

It is not a little interesting that this genus should have been discovered almost
simultaneously in the northern and southern hemispheres, M. farrani, Holt and
Tattersall (1906 (2)) having just been described when the 'Discovery' collections
came to hand.

M. posthon is a more stoutly built form than M. farrani, and is further
distinguished from the latter in the following characters : —

Antennules. — M. farrani has not the outer corner of the basal joint of the
peduncle produced nearly as much as in M. posthon.

Thoracic limbs. — The tarsus of the endopods in M. posthon is composed of six to
twelve joints, while in M. farrani there are only four.

Genital appendix to the last thoracic limb of the male is much longer and more
slender in M. posthon than in M. farrani.

35

Pbopod*. — The lateral lobe is less developed in M. farrani than in M. jxmthtm.

Telson, — In M. /arrant the cleft in armed with only al>out thirteen spine*,
whereas in M. posthon there are about thirty-six. In the former, moreover, the lateral
margins of the telson are armed with not more than twenty-six spines, not arranged in
aeries, and situated only on the distal two-thirds of the margin. In M. posthon the
lateral margins are armed throughout the entire length with about seventy spines,
arranged, at any rate, distally in series. The whole telson in M. farrani is more
slender than in M. posthon.

The spines on the inner uropods of M. posthon appear to be somewhat longer than
in M. farrani. I have already expressed the opinion that the genus M<-tamysidtlla of
Illig is synonymous with Mysidetes. The type species of the former, M. kerguelensis,
Illig (1906), is, however, a much smaller species than M. posthon, measuring only
10 mm. in length. It is otherwise closely allied to the latter, but differs in having
the antennular peduncle almost equal in length to the antennal scale and in the details
of the armature of the telson.

If Mysidopsis incisa, G. 0. Sars (1885), should in future be found referable to the
genus Mysidetes, as seems probable, it differs from the present species in size, in
having fewer joints in the tarsus of the thoracic Mini-, and in the details of the
armature of the telson.

I should mention here that I do not attach too great an importance to the
difference in size between M. kerguelensis and Mysidopsis incisa as compared with
M. posthon as a specific character, for I have found l>oth males and females of
M. farrani quite sexually mature at 15 mm. (judging from the characters of the
antennular brush in the male and the incubatory lamella: in the female), while the
species, fully grown, reaches to 28 mm. in total length.

SUB-FAMILY

GENUS ANTARCTOMYSIS, Coutiere.

JTfMf, Holt and Tattcreall, I'.HU; (1).
Antarttomytu, Coutiere, 1906.

This genus has been recently established by Coutiere for the reception of the
species briefly noted as My*i* maxima, Hansen (MS.), in the preliminary notice of
this collection.

There can be little doubt as to the correctness of the reasons which have led to it«
formation, since the biramous character of the fifth pair of pleopods in the male offers a
character of undoubted generic value, as distinguishing Mysis maxima from the genus
Mysis (sens, stricto). The genus Ilemimysis has the fifth pair of pleopods in the male
biramous and natatory, but the third pair are only imperfectly biramous, the outer
ramus being very minute and single-jointed, whereas in Antarctomysi* the third pair
resemble the fifth in having both rami multiarticulate and setose.

36 W. M. TATTERSALL.

I became aware, only after the plates illustrating this report had been printed,
that the two specimens of Antarctomysis in the ' Discovery ' collections, referred in the
preliminary note to one species A. maxima, in reality belong to two distinct but very
closely allied species. On my appealing to Dr. Hansen, he very kindly sent me some
notes and sketches of A. maxima, and a second species of the genus discovered by him
in a collection which he is engaged in working out. These notes and drawings placed
the matter beyond doubt, the larger of my two specimens clearly belonging to Hansen's
second species. I note the species here, and give the points of distinction, but leave a
full description and name to Dr. Hansen. The drawing on PL VIII. , Fig. 1, was taken
from the real A. maxima, but the remaining figures on the plate represent the
appendages of the second species, which, at the time, I took to be A. maxima also.
They will probably be of use, however, in illustrating how closely allied the two species
are when they are compared with the figures given by Coutiere (1906) of the true
A. maxima.

ANTARCTOMYSIS MAXIMA.

(PL VIIL, Fig. 1.)

Mysis maxima (j>ars), Holt and Tattersall, 1906 (1).
Antarctomysis maxima, Coutiere, 1006.

Locality of capture.— Winter quarters, 5. 6. 02, D-net hole, 56 fathoms, one
specimen, immature male, 33 mm.

Coutiere (1906) has recently described this species in great detail from mature
examples collected by the French Antarctic Expedition. I have practically nothing to
add to his description, but since no figure of the entire animal was given by him, my
drawing on PL VIII., Fig. 1 may be useful.

Coutiere does not mention the spines arming the inner ventral edge of the inner
uropod. They extend from the posterior inner corner of the otocyst to the extreme
tip of the uropod, and posteriorly, at least, are arranged in series of twos, threes and
occasionally fours. The species would appear to be circumpolar in distribution, since,
besides the single specimen in the ' Discovery ' collection, it has been taken by the
French, Swedish and Belgian Antarctic expeditions ; by the two former, in considerable
numbers.

ANTARCTOMYSIS sp.
(PL VIIL, Figs. 2-12.)
Mysis maxima (pars), Holt and Tattersall, 1906 (1).

Locality of capture :— Lat. 78° 25' 40" S., long. 185° 39' 6" K, 300 fathoms, one
specimen, immature male, 40 mm.

This species is so closely allied to A. maxima that I only became aware that it
was distinct when too late to properly illustrate it. The figures 2-12 on Plate VIIL
were taken from the appendages of this specimen.

HCIIIZUPODA. .17

The species will be fully descril>ed and named by Dr. Hanson, HO here I will
merely note the point* of distinction between it and . I. maxima.

(1) Eyt. — In .1. maxima (PI. VIII., tig. 1) the eye is large and the visual rlniicnt-
occupy a large part of the outer side of the eye-stalk, so that in dorsal view the inner
eye-stalk proper is mut-h longer than the outer, and in external lateral view very
little of the latter is visible. In the present form the eye is smaller and narrower than
in .1. maxima, the visual elements occupy the terminal part of the eye-stalk only, so
that the inner and outer margins of the latter arc sul»cqiial in length, and in external
lateral view practically the whole of the eye-stalk is visible.

(2) Rostrum. — In A. maxima the angle contained by the antero-latcral margins
of the carapace which form the rostrum is equal to or slightly greater than a right
angle, so that in lateral view the antero-lateral margins are not very oblique. The
tip of the rostrum is produced into a very small spine.

In the new species the angle of the rostrum is considerably less than a right
angle, so that the antero-lateral margins of the carapace in lateral view are very
oblique. The apex of the rostrum is bluntly rounded.

(3) Antenna. — In A. maxima the basal joint of the antenna, from which the
anteunal scale and peduncle arise, bears two spines ventrally, one at each of the outer
and inner distal corners. In the new form, only the one on the outer distal corner is
present, the inner corner l>eing rounded.

(4) In .1. maxima the tarsus of the third to the eighth thoracic limbs is seven
to eight-jointed (excluding the nail); in the present species the tarsus is six to
seven-jointed, so that the two distal joints before the nail are proportionately longer
than in .4. maxima (cf. PI. VIII., Fig. 8, with Coutiere (1906), PI. I., Fig. 11).

In other characters the two species are practically identical.

VOL. IV.

38 W. II. TATTERSALL.

LIST OF AUTHORITIES QUOTED.

COUTIEBE, H., 1906.— Expedition Charcot Crustacea Schizopodes et Decapodes, Paris.

CZERNIAVSKY, W., 1882-8. — Monographia Mysidarum imprimis Impcrii Rossici, fasc. i.-iii., St. Petersburg.

DANA, J. D., 1852. — United States Exploring Expedition, Crustacea.

HANSEN, H. J., 1887. — " Malacostraca marina Groenlandiae occidentalis." Vid. Medd. naturh. Foren.

Kjobenavn.
HANSEN, H. J., 1905 (1).—" Prelim. Report Schizopoda 'Princess Alice,' 1904." Bull. Mus. Ocean.

Monaco, no. 80.

HANSEN, H. J., 1905 (2). — "Further notes on the Schizopoda." Bull. Mus. Ocean. Monaco, no. 42.
HODGSON, T. V., 1902. — Schizopoda in " Report collections Natural History ' Southern Cross.' " London.
HOLT, E. W. L., and TATTERSALL, \V. M., 1905.—" Schizopod. Crust. N. E. Atlantic Slope." Report

Sea and Inland Fisheries, Ireland, 1902-3, Pt. ii., Scientific Investigations, Appendix no. iv.
HOLT, E. W. L., and TATTERSALL, W. M., 1906 (1).— "Prelim, notice Schizopoda 'Discovery.'" Ann.

and Mag. Nat. Hist. ser. 7, vol. xvii.

HOLT, E. W. L., and TATTERSALL, W. M., 1906 (2).— "Schizopod. Crust. N. E. Atlantic Slope. Supple-
ment." Fisheries, Ireland, Sci. Invest., 1904, v.
ILLIO, G., 1906. — "Bericht ii. die neuen Schizopoden-Gattung und ArtenderDeutschciiTiefsec-Expedition,

1898-1899." Zoologischer Anzeiger, Bd. xxx., no. 7.
NORMAN, A. M., 1902. — "Notes on the Natural History of East Finmark." Ann. and Mag. Nat. Hist.,

ser. 7, vol. x.

OHLIN, A., 1901.— "Arctic Crustacea." Bihang Kongl. Sven. Vet.-Akad. Handl., Bd. 27, Afd. iv.
ORTMANN, A. E., 1893. — Decapoden und Schizopoden der Plankton-Expedition. Ergeb. Plankton Erf. der

Humboldt-Stiftung, Bd. ii. G, b.

SARS, G. 0., 1864. — Beretning om en i Sommeren 18C3 foretagen Zoologisk Reise i Christiania Stift.
SARS, G. 0., 1869. — Undersogelser over Christianiafjordens Dybvandsfanna. Christiania.
SARS, G. 0., 1870-79. — Carcin. Bidrag til Norges Fauna. I. Monog. Norges Mysider.
SARS, G. 0., 1 88;!—" Prelim, notices Schizopoda ' Challenger.' " Forhandl. Vidensk. Selsk. Christiania,

no. 7.
SARS, G. 0., 1885.— Report on the Schizopoda collected by H.M.S. 'Challenger.' Zool. 'Challenger'

Exped., Pt. xxxvii., vol. xiii.

STEBBINO, T. R. R., 1893.— History of Crustacea. London.

STEBBINO, T. R. R., 1900. — " On some Crustaceans from the Falkland Islands." Proc. Zool. Soc., London.
TATTERSALL, W. M., 1907. — " Prelim, diagnoses of six new Mysidae from the West Coast of Ireland."

Ann. and Mag. Nat. Hist., ser. 7, vol. xix.
WILLEMOES-SUHM, R., 1875.— " Some Atlantic Crustacea from the ' Challenger ' expedition." Trans.

Linn. Soc., London, ser. ii., vol. i.

ZIMMER, C., 1904. — Arktische Schizopoden. Fauna Arctica, ii.
ZIMMER, C., 1905. — " Biologische Notizen iiber Schizopoden." Verhandl. d. Deutsch. Zool. Gesellschaft.

II/.OPODA.

EXPLANATION OF THE PLATES.
PLATE I.

Fio. 1. — Male, -45 mm., dorsal view of anterior end x 10.

FKI. 2. — Female, 4."» mm., dorsal view of anterior end x 10.

Fio. 3. — " 0/dfMi/M '* stage, dorsal view of anterior end x 15.

Fin. 4.— I Ate Qyrtopia stage, dorsal view of anterior end x 30.

Flo. ft. — Mandilmlar palp x 13.

Fio. 6. — Second maxilla x 13.

Fio. 7.— Pint maxilla x 18.

FIG. X. — Pint thoracic limb, endopod x 9.

Flo. 9. — Second thoracic limb, endopod x 9.

Fio. 10. — Lateral spine on carapace of female, 50 mm. x 30.

Fio. 11. — Lateral spine on carapace of male, 45 mm. x 30.

FIG. 12. — Lateral spine on carapace of male, 89 mm. x 30.

PLATE II.

Eitphauxia rry*talloro/>huu.

Fio. 1. — Male, lntor.il view of entire animal x C.

Fio. 2. — Female, dorsal view of anterior end x 15.

Fio. 3.— Mandibnlar palp x SO.

Fio. 4.— First maxilla x 80.

Fio. 5.— Second maxilla x 30.

Fio. 6.— First thoracic limb X 20.

Fio. 7. — Second thoracic limb, endopod x 20.

Fio. 8.— Calyptopis larva x 40.

Fio. 9.— Furcilia larva x 20.

Fio. 10.— Cyrtopia larva x 20.

PLATE III.
Thytanvesta maerura.

Fio. 1. — Female, 28 mm., lateral view x 6.

Fio. 2. — Female, dorsal view of anterior end x 10.

Fio. 3. — Antenna! scale and peduncle X 15.

Fio. 4.— Mandibular palp x 30.

Fio. 5.— First maxilla x 80.

Fio. 6.— Second maxilla x 80.

Fio. 7.— First thoracic limb, endopod x 80.

Fio. 8. — Second thoracic limb, endopod x 15.

Fio. 9. — Seventh thoracic limb, endopod x 80.

Fio. 10.— Rudimentary eighth thoracic limb X 60.

Fio. 11.— Endopod of the first pleopod of the male X 60.

Fio. 12.— Endopod of the second pleopod of the male x 60.

I '2

40 W. M. TATTERSALL.

PLATE IV.

Euphausia, triacantha.
FIG. 1. — Male, lateral view x (J.
FIG. 2. — Male, dorsal view of anterior end x 20.
FIG. 3.— Inner lobe of the endopod of the first pleopods of the male to show copulatory apparatus x 80.

Euphausia vallentini.

FIG. 4. — Outline of antennular peduncle of ' Discovery ' specimen x 20.

FIG. 5. — Outline of rostrum and basal joint of antennular peduncle of ' Challenger ' specimen x 30.
FIG. 6. — Inner lobe of the endopod of the first pleopods of the male in ' Challenger ' specimen, to show
copnlatory apparatus x 80.

Euphausia sp, juv.

FIG. 7. — Dorsal view of anterior end of specimen, 10 mm. x 60.

FIG. 8. — Lateral view of antennular peduncle of the same specimen x 60.

FIG. 9. — Spine on the third segment of the pleon of the same specimen x 20.

Euphausia crystallorophias.
FIG. 10. — Inner lobe of the endopod of the first pleopods of the male, to show copulatory apparatus x 80.

PLATE V.

Hansenomysis antarctica.

FIG. 1. — Female, dorsal view X 10.
FIG. 2. — Enlarged view of anterior end x 23.
FIG. 3. — Side view of anterior end x 28.
FIG. 4. — Dorsal view of peculiar antennular organ x 70.

FIG. 5. — Side view of same x 70. ,

FIG. 6.— Mandible x 20.
FIG. 7.— Mandibnlar palp x 20.
FIG. 8. — First maxilla x 20.
FIG. 9.— Second maxilla x 20.
FIG. 10. — First thoracic limb x 20.
FIG. 11. — Second thoracic limb, endopod x 10.
FIG. 12. — Third thoracic limb, endopod x 10.
FIG. 13. — Sixth thoracic limb, endopod x 10.
, FIG. 14.— First pleopod x 20.

FIG. 15. — Second pleopod x 20.
FIG. 16.— Third pleopod x 20.
FIG. 17. — Fourth pleopod x 20.
FIG. 18.— Fifth pleopod x 20.
FIG. 19.— Telson and uropods X 20.

PLATE VI.

Pseudomma lelyicae.

FIG. 1.— Female, dorsal view x 10.

FIG. 2. — Mandible x 30.

FIG. 3. — Mandibular palp x 30.

FIG. 4.— First maxilla x 30.

FIG. 5. — Second maxilla x 80.

FIG. 6. — First thoracic limb, endopod X 13.

FIG. 7.- — Second thoracic limb, endopod x 13.

FIG. 8.— Telson x 20.

srill/.ol'ODA. 41

Dactylamblyofi* hotiytont.

Fio. 9.— Eye, external lateral view x 20.

Fio. 10.- Mandible X 40.

Fio. 11.— liandibnlar palp x 40.

Fio. 12.— First maxilla x 40.

Fio. 13.— Second maxilla x 40.

Fio. 14.— First thoracic limb, endopod x 27.

Fio. 15.— Second thoracic limb, endopod x 27.

Fio. 16.— Tekon x 20.

PLATE VII.

Mytulftef fwfthon.

Fio. 1.— Female, dorsal view x 13.

Fio. 2. — Antennnlar peduncle x 18.

Fio. 3. — Antennal peduncle with antenna! scale x I*.

Fio. 4.— Mandible x 18.

Fio. :>.— Mandibular palp x 18.

Fio. 6.*— First maxilla x 18.

Fio. 7.— Second maxilla x 18.

Fio. 8. — First thoracic limb, endopod x 18.

Fio. 9. — Second thoracic limb, endopod x 1*.

Fin. 10. — Third thoracic limb, endopod x 18.

Fio. 11. Genital appendix of male x 18.

Fio. 12.— First pleopod x 24.

Fio. 13.— Telaon x 18.

PLATE VIII.
Aniarrtomy*i« maxima.
Fio. 1. — Immature male, dorsal view x 6.

Antarctomijftt up.

Fio. 2.— Mandible x 20.

Fio. ».— Mandibular palp X 20.

Fio. 4.— First maxilla x 20.

Fio. :».— Second maxilla x 20.

Fio. 6.— First thoracic limb, endopod x 8.

Fio. 7. — Second thoracic limb, endopod x 8.

Fio. 8.— Third thoracic limb x 8.

Fio. 9. — Second pleopod of the male (immature) x 15.

Fio. 10. — Third pleopod of the male (immature) x 15.

Fin. 11. — Fourth pleopod of the male (immature) x l.v

Fio. 12.— Fifth pleopod of the male (immature) x l.v

INDEX OF GENEEA AND SPECIES.

Amblyops, 29.
Amblyops abbreviata, 30.
Amblyops crozetti, 3.
Antaretomysit, 3, 85.
Antarctomysis maxima, 2, 36.
Arctomysis, 22.
Boreomysis, 22.
Boreomysis scyphops, 3, 22.
Ceratomysis, 22.
Dactylamblyops, 3, 29.
Dactylamblyops goniops, 31.
Dactylamblyops hodgsoni, 2, 30.
Dactylamblyops sarsi,-Sl.
Dactylamblyops tluiumatops, 31.
Dactylerythrops, 29.
Dactylerythrops arcuata, 2, 30.
Echinomysis chtmi, 2.
Eucopia, 3.

Eucopia aus/ralis, 1, 2, 3.
Euphausia, 4.
Euphausia sp., 14, 1C.
Euphausia antarctica, 1, 4.
Euphausia australis, 1, 4, 5.
Euphausia crystallorophias, 2, 9.
Euphausia gibboides, 16.
Euphausia glacialis, 1, 4, 8.
Euphausia lucens, 1, 11, 14.
Euphausia murrayi, 1, 4, 6.
Euphausia similis, 2, 11.
Euphausia splendens, 1, 11, 13, 14.
Euphausia superba, 1, 2, 4.
Euphausia triacantha, 12.
Euphausia vaUentini, 13, 14.

Hansenomysis, 3, 22.
Hansmomysu antarctica, 3, 23.
Hansenomysis fyllse, 22, 26.
Hemimysis, 35.
Heteromysis, 32.
LopJi off aster typicus, 3.
Metamysidella, 32.
Metamysidella kerguelensis, 35.
Michtheimysis mizta, 3.
Mysideis, 32.
Mysidella, 32.
Mysidetes, 3, 32.
Mysidetes far rani, 31.
Mysidetes posthon, 33.
Mysidopsis, 32.
Mysidopsis incisa, 32, 35.
Mi/sis, 35.

Mysis maxima, 1, 35, 8C.
Nyctiphanes couchi, 0.
Petalophthalmus, 22.
Pelalophthalmus armiffer, 22.
Petalophthalmus oculatus, 22.
Petalophthalmus wilkmoesii, 22.
Pseudomma, 3, 27, 30.
Pseudomma australe, 29.
Pseudomma belgicse, 27.
Pseudomma roseum, 28.
Pseudomma sarsi, 1, 2, 28.
Scolophihalmus, 22.
Thysanoessa, 3, 17.
Thysanoessa gregaria, 18.
Thysanoessa macrura, 1, 2, 17.

X

a

Q
u

*••' OF THE

UNIVERSITY

OF

_

r;

T-rT-t;

'

5

i

<v

UNIVERSITY

OF
NXcfAL!FOR^>

00

y^TTm^l

CM

It UNIVERSITY

I

j

5 i.

'Z 5

•SJ

II

II
t!

&.*

i *

•2-3
|1

ff

OF THE

UNIVERSITY

OF

I

X

a

1

5

OF THE
UNIVERSITY

V

3.

a.

X

UJ

r>

I

1

5,

ko

•O

;

rt
c

00

» :

or THE
UNIVERSITY

OF

5

/ \

CVJ

1
5

5

2

c

CKUSTACEA.

VIII.-COPEPODA.*

BY R NORRIS WOLFENDEN, M.D. (Cantab.), F.L.S., F.Z.S.

(7 Plates.)
PREFATORY NOTR

THE collection handed to me for examination and report thereon was contained in 163
bottles, the contents of a few of which were in such bad preservation that they were
practically worthless for the purpose of identification. The collection of individual
species is not a large one, though there were great quantities of the more common
species. But few absolutely new forms were found ; these comprised a new genus
(Parahibidocera) and seven new species (Euch&ta ftimili*, Stephus antarcticum, Xantho-
ctdanus antarcticus and X. nuignus, Haloptilus ocelhttux, Faroella antarcticti, and Gnetanux
antarcticus). As such of these copepods as may be considered Antarctic were
collected within a small area, I have not deemed it necessary to occupy space by the
repetition of individual captures, which would be monotonous and of no particular
interest The drawings have been made by Miss Marion taes.

The signs used in the following pages are ft 1 and ft 2 for first and second basal* ;
Ri and Re for endopodite and exopodite ; Si for inner marginal and S? for outer
marginal spine (or bristle) ; Li for inner and Le for outer lobe ; Th for thoracic
somite. As they were first used in Giesbrecht's great work, and have been subse-
quently often employed by others as abbreviations, the author has thought no
excuse necessary for their use here, in order to avoid the constant repetition of the
words " exopodite " and " cndopodite," etc.

I.

UNTIL the expedition of the ' Belgica ' there existed no records of the collection of
Copepoda south of Kerguelen, except those of Dr. Brady, which referred to the
'Challenger' collections made from the south of Kerguelen to the pack-ice at 66° 29' S.
The collections made by the ' Discovery,' the ' Belgica,' and the ' (iauss ' form a
most important contribution to the planktonology of this southern region, and the

* Owing to the author's absence from England he wai unable to sec the " revise " of this Report.— Eft.

2 R. NORRIS WOLFENDEN.

results of any one expedition cannot properly be appreciated without reference to the
others.

The 'Belgica' collections were made S. and S.E. of Peter I. Island, between 69° 48'
and 71° 18' S., and 81° 19' and 92° 22' W., between April 21st and December 6th, 1898,
by means of nets lowered through holes in the pack-ice to a depth of 0-500
metres.

The ' Discovery ' collections were made by lowering and raising a vertically
actuated net through holes cut in the ice, while the ship was in Winter Quarters.*

The ' Gauss ' collections were made from the South of Kcrguelen to the winter
station in Gauss Bay, Kaiser Wilhelm II. Land, and were of very extensive character,
and as the collections were further made throughout the Atlantic traverse of the ship,
they afford an opportunity for the comparison of the purely Antarctic fauna with that
of the Southern Ocean.

In considering the question of the distribution of the Copepoda of the southern-
most area of the Atlantic (the Antarctic region) it is convenient to consider the results
of these expeditions together, since any conclusions drawn from the results of the
' Discovery ' alone would be incomplete and even misleading. The ' Belgica ' collections
have been reported upon by Dr. Giesbrecht (" Resultats du Voyage du S.Y. 'Belgica'
eu 1897-1898-1899"; Rapports Scientifiques, 1902), and the ' Gauss ' collections are still
under examination, and I only now refer to the results of my examination of that
collection in so far as they assist the elucidation of the ' Discovery ' results.

From the results of the three expeditions (' North American,' ' Challenger ' and
' Vettor Pisani ') which, previously to the ' Belgica,' had collected in the Southern
Ocean as far south as the pack-ice, Giesbrecht accepts seventeen species as
correct, after rejecting a number of species as "ungeniigend beschriebenen und nicht
zuverlassig genug bestimmten "),f viz., Aetidius armatus (50° S.), Calanus jinmarchichus
(52°), Calanus patagoniensis (47°), C. propinquus (64° 37'), simillimus (52°), Centropages
%rachiatus (52°), Clausocalanus arcuicornis (53°), Clytemnestra scutMata (46°),
Drepanopus forcipatus (53°), D. pectinatus (49° 16'), Metridia boeckii (45°),

* Mr. Hodgson has supplied me with the following. — ED.

" TOW-NETTING IN WINTER QUARTERS.

" After the surface of the sea was frozen over there was no means of dragging a tow-net through the
water, and as the current seemed sufficiently strong to hold the net out, it was attached to a line about a fathom
above a heavy sinker, 28 Ibs., and lowered to a depth of ten fathoms, except in special instances or during the
summer. This depth was decided on, it having been found that the formation of ice crystals on the nets could be
avoided. These crystals formed on the lines down to 5-8 fathoms, according to the season. The nets remained
down for twenty-four hours, sometimes longer if the holes could not be visited on account of the weather, or
opened on account of some difficulty with the ice. The mouth of the net was always in an approximately vertical
position, this was secured by the attachment of the line direct to the ring of the net and the sinker to the other
side of the ring."

•f The rejected species are, Acartia denticornis (52°), Candacia curia (50°) and truncata (64° 37'), Eucalanus
at/enuatus (47° 25'), Eucha-ta marina (47° 25'), Haloptilus aculeatus (46° 46'), Hfterorrhabdus spinifrons (50P)>
Lucieutia flavicornis (47° 25'), Pleuromamma abdominalc (65" 42'), Copilia stylifera (66° 29'). — Giesbrecht,
' Belgica ' report, p. 5.

COPEPODA.

grandis (49°), Oit/wmt similut (52°), Paracalanus parvw (52°),
wuutu* (52°), It. giga* (65° 42'), ScoUcithrix minor (46° 46').

This list contains a striking numlior of forms which are usually associated with
more temperate regions, and, as Dr. Giesbrecht remarks, the failure in agreement with
the pelagic species of the ' Belgica ' in very striking, for only two species an common
to all collections. Comparing it with the results of the ' Discovery ' the same
extraordinary differences are manifest, only four species (C. prvpiiiquw, C. simillimu*,
Clausixxtlamis arcuicorni*, Otthona ximilix) being common to both collections.

In the ' Gauss ' collections, in the area between Kerguelen and the Winter Station,
appear a great number of species in execs* of those either of the ' Belgira ' or
' Discovery.' Whereas in the ' Bclgira ' collection occur thirty specie*, of which nine-
teen only are pelagic, in the ' Discovery ' collection are twenty-four species of pelagic
Copepoda ; but in the ' Gauss ' c-ollection this nuiul>cr in more than doubled, and a
Dumber of species occur even in the collections made round al»out the Winter Station
which are not entirely Antarctic, but extend a long way northwards through the
deeper waters of the Atlantic Ocean, and have been brought there prolwbly by southern
currents. The species determined, however, show but little agreement with the list
enumerated alx>ve.

The very extensive number of species captured by the ' Gauss ' naturalists is
probably due to the fact that the tow-nets were used at much greater depths than in the
case of either the ' Belgica ' or ' Discovery.' In the former, 500 metres appears to have-
been the limit, whereas in the latter the collections may be considered to l»e practically
surface collections. If the tow-net had been used at the depths it was employed on
the ' Gauss,' viz., to 3,000 metres, the agreement between the respective captures
might certainly have been greater, and the number of species taken greatly
increased.

In the ' Gauss ' collections appear only six species which agree with any of
the species referred to above (viz., Aetideus armatwt, Caianua propinquu* and C.
simillimus, Clattsocalanus arcuicornix, Oithona sitnili*, Lucicutia fltivicorni*), and when
it is remembered that in the ' Belgica ' collection there are only two species, and in the
'Discovery' only four species, of the twenty-seven species enumerated by (Jiesbrecht
which are in agreement, the conclusion is inevitable either that the captures made by
the expeditions mentioned were unusual, or that the identification of species has in some
instances been erroneous. That unusual species do appear in these areas, even close to
the ice, is shown by the occurrence in the ' Gauss ' Antarctic collections of Corycoeu*
(tpecionUf Sapphirina inetallina, Aetideti* artnatu*, Labidocera aeutifroiis, Undtuchaeta
major, Arictellu* setofus, and others ; and in the ' Discovery ' collection in Lat 56°
31' S., Long. 156° 19' 30" occurred Eucalanu* subtenui*, and in Lat 49° 40' 8., and
Long. 172° 18' 30" W., Pleuromamnui gracilix, several young Gmdace, Euchceta
a HI r inn, and Cejitropagef violnottx, which belong undoubtedly to a subtropical or
warm temperate area, and arc to l>c regarded as accidental.

YOU 1%'.

R. NORRIS WOLFENDEN.

While the number of species captured was in each case comparatively small, the
number of individuals in any one haul in the case of the ' Discovery ' was very groat.
The rule which appears to hold good for tow-netting in the north part of the North
Atlantic, viz., that the further north we go the smaller the number of species, but the
immensely increased preponderance of individuals of certain species, certainly holds
good as regards the South Polar regions. Immense numbers of the small copepod
Ctenocalanus vanus appear in some of the hauls, to the almost entire exclusion of
any other species, and in other cases the larger copepod Euchceta antarctica appears
in great preponderance. Calanus acutus and, to a lesser extent, Calanus propinquus
also preponderate largely. Similarly, Metridia gerlachei appears in most of the
captures.

The collections of the ' Gauss ' provide information which is not given by those of
the ' Discovery ' or of the ' Belgica,' namely, that several species which appear
in the Southern Polar Sea also occur in the deeper water of the Atlantic Ocean
to the northwards of the Antarctic area. But as this properly belongs to the report
of the ' Gauss ' collections which I have in hand, I forbear its discussion in
this place.

Two questions are suggested by Dr. Giesbrecht in his ' Belgica' report, viz., (1)
Does the Antarctic area possess a peculiar fauna ? (2) Is the small agreement of the
Antarctic copepod fauna with that of the nearest seas due to defective research, or is it
that the area of the pack-ice has its own peculiar fauna ? and the further questions as
to whether the admixture of Polar and Antarctic fauna occurs in the deep ocean, or
whether there are physical and biogenetic conditions in the Polar regions which differ
from those in the warm seas and prevent such exchange of species, receive some
elucidation from the collections of the ' Gauss.'

With regard to the first question, viz., Does the Antarctic area possess its own
peculiar fauna ? it must be remarked that from the results of the three collections
named the typical copepod fauna (pelagic) of this region consists in the following :—

Calanus acutus
„ simillimus
„ propinquus

Rhincalanus grandis

Euchceta antarctica
„ austrina
„ similis

Ctenocalanus vanus

Heterorrhabdus austrinus

Euchirella magna

Spinocalanus antarcticus

Metridia gerlachei

Oncea curvata, similis, frigida, notopus,

conifera

Scolecithrix glacialis
Oithona similis

„ frigida
Gaetanus antarcticus
Haloptilus ocellatus
Paralabidocera hodgsoni
Stephus longipes

,, antarcticum
Ectinosoma antarcticum
Microcalanus pusillus

1. Of the Au tan- tic Copepoda the following are new specie* and genera :—

ParaLtbtdocfra . hodgtoni Euchirellu magna

ll<il,'j,tilus ocrllatua Faroflla antarcticn

* *

Stephos antarcticus Gaetamtx (infarction

'ioeta tiyiili* Xanthoctikinus antarcticu*

Calanus nmillimus (mentioned briefly by Giesbrecht, loc. cit.).

2. The following are species newly described by Dr. Gicsbrecht (' Bcljjica ' report),
occurring also in the ' Discovery ' collection : —

Euchceta antarctica

SttpMu longipes

Metridia gerUichei (nearly related to M. Ixtecki and M. lucens).

Oneta curvata „ „ 0. subtili*, Gieab.

lihincalamu grandis „ „ /P. gigas, Brady.

Harpacticus furcifer „ „ H.flexus, Brady.

3. Of species which occur in the North Polar regions there are only the following
'Discovery' •pecJM, which bear such slight modification as to be practically identical :
Microcalaniui ptisillus ( = Psewlocalanus pygnuetu) ; Oitlumu similis.

4. The ' Discovery ' collection would therefore lead us to suppose that, so far as
copepod fauna is concerned, there is little resemblance between the characteristic fauna
of both Polar regions. In the ' Belgica ' collection, Oncea conifera and notopus ; and in
the ' Gauss ' collection, Oncea conifera, Gaidius tcnuispinu.1 and brevispinus, and
Amallvphora maijiia, that is, seven species of a total of 55-60 species occurring in the
South Polar seas, are all that are identical with the species described by Prof. San M
collected by Nansen's Norwegian North Polar Expedition. It would not, however, be
safe to take this list of Prof. Stirs' as the ultimate result of copepod research of the
North Polar seas, and other species may yet be found to be identical

The following table shows the comparative relationship of species of the more
frequently occurring genera :—

N. PoUr. 8. PoUr.

(\ilinut finmarcliieiw replaced by C. propinqwu

„ hyperborevt „ lontut

„ timiUtmtu
„ aeutiu
Mttridia longa replaced by M. gtrlachti

Amallopfiora magna (= Stafki A. magna

calanuM aeroftphalui, San)

froltcMrix brtnevrni* replaced by 8. glanaUt

TarrftfrffffTTfT bortalu replaced by .V. mayntu

„ antarrtirtu

K 2

6 R. NORRIS WOLFENDEN.

N. Polar. S. Polar.

Gaidius tenuispinus') G. tenuispinus

„ Irevispinm] rePIaced b? „ major

„ antarcticus
Microcalamts pusillus (= Pseudo- M.pusillus

/•a In /ii's pygmwus)

Heterorrhaldus norwtgkus replaced by H. austrinus

„ compactus „ longicornis

Emlmta norwegica replaced by E. antarctka

„ glacialis ., austrina

„ larbata „ similis

Haloptilm spinifrms replaced by H. ocellatus

„ spiniceps

S]rinocalanus longicornis replaced by S. antarcticus

Undeuclueta spectabilis replaced by U. major

Oitkona similis replaced by 0. similis

„ helgolandica „ curvata

„ conifera „ conifera

„ notopus „ notopus

„ friffida

In the North Polar Sea, as Prof. Sars remarks, besides the few distinctly Arctic
species are many which extend southwards to the warmer seas, and the North Polar
basin copepod fauna has a pronounced resemblance to that of the North Atlantic
basin, the greater number of species being common to both, and some deep-water forms
of the Norwegian Sea are often surface forms in the North Polar basin. A few forms
regarded as quite southern also occur in the North Polar Sea.

So far as the distribution can be followed from the ' Gauss ' collections, it may be
said that, of the typical Antarctic fauna its representatives diminish gradually to latitude
40° S. (i.e. about the latitude of St. Paul and New Amsterdam) north of which they do
not appear, but extend westwards to those stations situated directly south and westward
to 10° E. as a limit of the Cape of Good Hope, north of which no typically Antarctic
species appears.

North of Kerguelen, i.e. 50° S. lat., no Antarctic species appear to extend, while the
typically subtropical species of the Indian Ocean extend as far south as latitude 30° S.,
where their southern extension appears to be arrested. There is thus a barrier between
lat. 40° and 50° S. and between long. 10° and 80° E. as indicated by the ' Gauss '
collections, at which extension northwards of Antarctic species and southwards of Indian
Ocean subtropical species is prevented, or at any rate, does not occur. While the same
collections indicate that the Antarctic species extend northwards into the Atlantic Ocean
in gradually diminishing numbers, only as far as lat. 40° S., north of which they do not
occur, a few typically Atlantic deep-water species find their way into the Antarctic Sea
(such are Heterorrhdbdus profundus, Labidocera acutifrons, Metridia princeps, Lucicutia
grandis, Gaidius major, Arietellus setosus).

Until the ' Gauss ' collections are fully examined it is of course rash to say that no

OOPKl'ODA. 7

typically Antarctic species ever find their way northwards by way of the deep Atlantic
trough, but there is little evidence of it in the many collections made by the ' Gauss '
throughout its Atlantic traverse. A certain nurntwr of species which arc ubiquitous,
such as Oithona similis, some species of Onoea, Ilnloptilwt hngicornui, Gaidiu*
tenuispinus and major, and Gaetantu (armiger, and possibly caudani), extend from the
Faroe Channel to the southern ocean ; but so far as the evidence at present goes,
the Antarctic Copepod fauna is distinct from that of the Arctic seas, and the species
which are typical of this region, and most numerous, do not extend far into the
Southern Atlantic. As no observations have been made of the Copepod fauna of the
deep water of the Indian Ocean, it is quite possible that Antarctic species may I war a
considerable extension northwards in this direction.

It is curious that no great number of J/arpicticidse appear in the collections of the
'Discovery,' only three examples all told of Harj*icticu# furcifei; which is somewhat
different from any Ifarpaciicu* of the northern hemisphere ; and only five arc descril»ed
from the 'Belgica' collection by Dr. Giesbrecht, two of which (//. brevicornis, II. cMifer),
are identical with northern species. A fair number of species occur in the ' Gauss '
collection, but these have not yet been examined.

The paucity in numbers of the Ifarpactictdse in the ' Discovery ' captures is no
douUt due to the mode of collection.

II.

LlST OF COPEPODS IN THE 'DISCOVERY* COLLECTION.

Kitcfurt't nniartlifa.

„ limiiu.
Mttridui gerlathri.

„ prinetpt.
Culanut of ut us.

„ propinqttvs.

„ tonstu.

„ timilliiiiii.i.
Ctenotalanus ran at.
Oithona similit.

„ friffida.
Oncta etirvata.

Mirrocalanu* piuilliu.

„ aniartlieum.
Xantkacalanut antarftifii*.

Harpof tints furnfer.

Paralabidottra hadgtoni.
Rhintalantu grandix
Clatuotalantu areuiforni*.
Haloptilus ocfllatu*.
FarotUa aniarclifa.
OatUmtu antarrtieut.
Hftrrorrhabdu* longiforni*.

CALANUS (LEACH).

The species first described by Brady as Calantu propinquu* has been subsequently
described by Giesbrecht, who now concludes ('Belgica' report, p. 16) that the copepod
described by himself in 1892 as C. propinquus from the S.W. Atlantic, between 37° and
52° S., is not this species, but one closely resembling it, to which he gives the name

R. NORRIS WOLFENDEN.

C. simillimus. Giesbrecht also suggests that the similarity between the two species
leads to the doubt whether C. propinquus has such a wide distribution as Brady
imagined. This author gave it a distribution throughout all oceans — in the southern
Indian, north in the Pacific (to 35° N.), and in the Atlantic to 30° N.*

I have made a careful examination of dissected specimens from the following

localities :

Lat.' 56° 31' S. Long. 156° 19' 30" E.

Lat. 59° 19' S. Long. 124° 24' 30" W. (28. vi. 04).
Lat. 84° 01' S. Long. 170° 49' E.
Lat. 58° 49' 45" S. Long. 154° 48' W. (24. vi. 04).
Lat. 50° 48' S. Long. 170° 2' E.

and in these hauls occur many examples of a Calanus which answers in all particulars
to Giesbrecht's C. simillimus. As these have well-developed genital segments, in some
cases with spermatophores attached, and differ from C. propinquus not only in size,
being very much smaller than this species, but also in the proportions of the third and
fourth pairs of feet, and the serration and proportions of the fifth pair, there is no doubt
that this entirely agrees with the animal briefly described by Giesbrecht ; and I think
he is correct in regarding it as a species different from C. propinquus, and that the
distribution of the latter is not so extensive as was imagined by Brady. I therefore
describe (7. simillimus (Giesbrecht) as a different species.

A third species of Calanus which does not agree with either of these, being
altogether less robust than C. propinquus, and constantly much smaller, but at the
same time larger than simillimus, but with the basals of the fifth pair without any
marginal teeth, and the distal margins of the basals of the second to fourth pairs
with a row of spines, may probably be C. tonsus (Brady), but the description given
by this author is of the briefest character, and he gives only two figures, which do
not assist in the identification. However, I attach to it Brady's name, as it is probably
the animal meant by Brady.

Four species of Calanus are found in the ' Discovery ' collection : —

1. Cal. propinquus (Brady).

2. Cal. acutus (Giesbrecht).

3. Cal. simillimus (Giesbrecht).

4. Cal. tonsus (Brady).

* Lat. 46° 46' 8. Long. 45° 81' E.
Lat. 64° 87' S. Long. 85° 49' E.
Lat. 47° 25' 8. Long. 180° 12' E.
Lat. 85° 41' N. Long. 157° 42' E.
Lat. 40° 3' 8. Long. 132° 58' W.
Lat. 9° 48' S. Long. 18° 51' W.
Lat. 8° 10' N. Long. 14° 51' W.
Lat. 87° 17' S. Long. 53° 52' W.
Off Kcrguelen Island. Brady, loc. cit.

1. < A I \M S PROPINQUfR.

(Plate I., figs. 1, 2, 8, 3', 4.)

OrffnifTjTij-'-j Hrwlj, Rep. • Challenger ' XIX., Copepoda (1888), p. 31.

Oiwbrccht, Faun* u. Fl. Neap. XIX. (1892), p. 91.

•brecht und Schmeil, Diw Tiem-irh, Copt-poda (189«), p. 1ft.
Oiatbncbt, ' Beljrica ' Report, p. 16.
T. Soott, Trans. Linn. Soc. VI. (1H93), p. 25.

The length of this copepod given by Brady is 5'5 mm.,by fJiesbrecht 4 '9-5 '3 mm.
The majority of the ' Discovery ' and 'Gauss' flpecimcnn measure under 5 ram. The
cephalothorax is rather over three times as long as the nhdomcn, furcal segments twice
as long as broad. Head evenly rounded, without any trace of crest Head separate
from first thoracic segment. Last thoracic segment laterally produced a little, and
ending in short points. Anterior antenna) not reaching beyond the fun -a (thus shorter
than described by Giesbrecht, in these specimens). Relative proportions of 24th to 25th
segments as 18 : 21.

Second feet Re 3 divided into proximal portion 24 : distal 25.
Third „ „ 20; „ 23.

Fourth „ „ ,,36: „ 20.

Hi3 with seven bristles, end saw two-thirds the length of /?«'3. Fifth feet, Ji 1
with inner margin convex below, concave distally, with thirteen or fourteen teeth, and
distally with three teeth, larger than the others. /?/3 with five bristles, two outer,
two apical, one inner. This animal is much more robust than the next special
(simillimwi).

The <J is about the same size as the 9 • mid the chief differences consist in the
structure of the fifth feet and shape of the head and thorax. The latter is very like C.
jinmarchictui, the head oval, rounded, nnd produced, the first thoracic segment with deep
indentation l>etween it and the second, and its posterior margin protruded. The
anterior antennas with the basal joints much coalesced. Posterior footjaw with a long,
stout, densely-feathered dorsal bristle. Fifth left foot much prolonged, /.'• 1 and 2 .
elongated, Re 3 very short and pyriform shape, with short distal bristle. Ri only half
the length of Re, with respectively 1:1:6 short weak bristles. Right foot Ri nearly
as long as Re, with 1:1:6 bristles, lie not much more than half as long as St at
opposite side. First basal with inner margin armed with seventeen to twenty large
teeth ; in its upper part slightly concave, in its lower part slightly convex, the teeth
all of pretty much the same size.

10 R. NORRIS WOLFENDEN.

2. CALANUS ACUTUS.
(Plate I., figs. 9, 10.)
Calanus acutus, Giesbrecht, ' Belgica ' Report,* p. 17.

This copepod was first described by Giesbrecht from the Belgian South Polar
Expedition, and in the ' Discovery ' collection it forms the chief constituent of the South
Polar copepod plankton, along with Euchceta antarctica.

The majority of the animals I have met with both in the ' Gauss ' and ' Discovery '
collections are smaller than the size given by Giesbrecht, viz., 4 '7-5 '3 mm. But few
of our specimens reach 5 mm. length, the majority being from 4 '5-4 '7 mm. The
cephalothorax is not quite four times as long as the abdomen, the head divided from
the first thoracic segment, the last segment of the latter produced laterally, but with
evenly rounded margins and no points. In lateral aspect the head is slightly
produced forwards, and more inclined to be oval than rounded. In the dorsal aspect it
has a distinct triangular appearance, with slight crest in the mid-line. In its broadest
part the thorax is 1 • 15 mm. broad, that is, three times as long as broad.

The furcal segments are a very little longer than the anal, and nearly twice as long
as broad. The anterior antennae vary in length in different animals, in some being only
as long as the furca, in others one or two joints longer, and are distinguished by the
comparative length of the last joint, which is about twice as long as the one before it.

In the second feet the Re is divided into two portions, proximal = 24 ; distal =13.
third „ „ „ „ = 29; „ =15.

fourth „ „ ,, „ = 15 ; ,, = 15.

The Ri 3 has in the second and third feet eight bristles, in the fourth only seven,
and in the fifth only four (with no outer marginal bristle). The end saw of the Re 3 of
the second feet is shorter than the Re 3 ; in the third and fourth pair longer.

The fifth feet are distinguished by the absence of the outer marginal bristle of the
Ri 3, and the total absence of teeth or hairs on the inner margin of the first basal. All
males appeared to be immature.

3. CALANDS SIMILLIMUS.

(Plate I., figs. 5, 6.)
Calanus simillimus, Giesbrecht, ' Belgica' Report, p. 17.

$ 2 '5-2 '9 mm. ; cephalothorax, 1*9 mm. ; abdomen, '6 mm. long.
Head separate from first thoracic segment, evenly rounded, without any trace of
crest. Head not quite as long as the rest of the thorax (as 18 :21). Last thoracic

* " Belgica Report " is throughout this monograph used to indicate " R£sultats du Voyage du S. Y. ' Belgica '
en 1897-1898-1899." Rapports Scientifiques. 1902.

rOPEPODA. 1 1

segment laterally produced somewhat and ending in short points. Anterior antenna)
about as long as the furca, or about one joint longer.

Genital segment as long as the next two. Furca three time* as long as broad, and
longer than the anal segment

>. •••:.•: bet) /•' • '• divided '••. • ':.•• might*] pine Into twoabool • •, -;.ii peri .

Tliinl feet, Jit 3 with eight bristles (four outer, two apical, two inner), /!•• 3
divided into two parts, of which proximal : distal = 17 : 13.

Fourth feet, /fc3 divided by the marginal spine, proximal : distal = 20 : 11.

/?»' 3 with seven Si (two outer, two apical, three inner), terminal saw of /•'• 3 only
three-quarters as long as Re 3.

Fifth feet D 1 toothed, with fourteen teeth on the rather convex margin, and at
the distal end a slight break in the continuity, with three rather larger teeth somewhat
hidden, in front view, by the upper teeth of the marginal surface. Jii 3 with five bristles
(two inner, thin and short, two apical, and three outer). In the second jmir the fti
reaches alxmt the end of the Re 2, in the third pair to the first inner marginal bristle
of the exopodite, and in the fourth pair as far as the second inner marginal bristle, in
the fifth pair beyond the origin of the first inner marginal bristle. The endopodites
are therefore proportionately larger than in C. tonxw*, and the third segment of the
exopodite is not four times as long as broad.

While this species agrees with C. propinquw in many particulars, the proportions
of the third and fourth feet differ, also the toothing and convex margins of the basals
of the fifth feet, and the size. Many of the examples were quite adult females with
spermatophore attached, so there can be no question of their being merely undeveloped
examples of C.propinquu#,&nd,R8 before mentioned, this species has a considerable area
of distribution in the southern oceans.

4. CALAMUS TONSUS.
(Plate L, figs. 7, 8.)

Calanut tonnu, Brady, Rep. XIX., ' Chall.' Report, p. 34.
„ „ Scott, Tr. Linn. Soc. VI. (1898), p. 25.
„ „ Giesbrecht, Fauna u. Flora Neap. XIX., p. 92.

Dahl, Verb. Dcntaches Zool. Ceaelk IV. (1894), p. 77.

Brady's original description of this species is very incomplete, and he merely states
that it is " like C. jinmarchicus and propinguiut, except that the anterior antenna an
almost devoid of setae, except on the three apical joints ; the posterior antenna? are like
those in C. propinquity, the fifth pair without !>asal teeth, and the first alnlomiual
segment large and tumid. The anterior antennae are as long as the body ....
9 Size 3 • 6 mm." He gives only two figures — viz., of the anterior antenna- and the
abdomen.

Giesbrecht includes it under the " Unbestimmbare species," remarking that the

VOL. IV. L

or THE A

NIVERSITY
or /

^•ro**^

12 R. NORRIS WOLFENDEN.

first character probably results from the bad preservation of the animals ; the second is
a character of other members of this genus ; and the third point is characteristic of
gracilis and robustior.

Scott merely mentions that " the large and tumid first abdominal segment seems
to be a fairly good character " (loc. cit. ). However, this is not a characteristic of this
species any more than of C. robustior, and the existence of this species up to the present
time therefore must be regarded as extremely doubtful. However, the ' Discovery '
collections contain several examples of a copepod, which, if it is not Brady's species,
answers fairly well to it so far as his description goes.

? 3 '5-3 '6 mm. long (cephalothorax, 2 '75 ; abdomen, '75. Body broadest at the
end 3f the first thoracic segment (1'lmm. broad). Abdomen short, genital segment
broad, and one-third broader than the following segment. Furcal segments not quite
twice as long as broad, and nearly twice as long as the anal segment. Head evenly
rounded, without trace of crest, separate from the first thoracic segment, last thoracic
segment only slightly produced, and with rounded margins. Anterior antennae only
reaching the end of the third abdominal segment, the only long bristles on the twenty-
third, twenty-fourth, and twenty-fifth segments, the twenty -fourth joint twice as long as
the twenty -fifth. Mouth parts resembling C. jinmarchicus.

Second feet — second basal with four large spines on the distal margin at the inner
side. Re 3 divided by the external outer spine into proximal and distal portions
respectively as 23:16. Ri 3 with eight bristles. The whole endopodite does not
reach beyond the distal margin of Re 2. Re 3 as large as Re 1 + Re 2.

Third feet — four large spines on B 2 distal inner margin, endopodite reaching a
little beyond the distal margin of Re 2. Re 3 divided into proximal part = 32, distal
portion = 16. Ri 3 with eight bristles.

Fourth feet — Re 3 divided into proximal part = 37, distal = 15 ; apical saw only
seven-ninths as long as Re 3. Ri with seven bristles only (three inner, two outer, two
apical). B 2 with one or two spines on distal inner margin.

On the second, third, fourth and fifth feet the outer margin of the second basal is
distally armed with a spine ; in second— fourth, the Re 3 = Re 1 + 2, and is about
three times as long as broad.

Fifth feet— first basal with straight inner margin without teeth or hairs, B 2 with
five spines on the distal inner surface. Ri 3 with six bristles (two inner, two outer,
two apical).

The only Calanus with which this shows agreement is, possibly, Brady's C. tonsus ;
but Brady's description is so fragmentary that it may well be another species. It
occurred in some numbers at Station, 22 -11 -01, Lat. 56° 31' S., Long. 156° 19' 30".
Such males as were observed were all immature.

COI'KPOUA. 13

IMIINCALANUS (DANA).
(Plate II., fig. 6.)

f. yraitdu, Ginbrtcht, • Beltfica ' Kt-p., p. 18.
? Rk. ffiyat, Bradj. ' Clialk-nger ' Rep. XIX., p. 42.

„ Scott. 19th Rep. Scotch Fishery Board (Wl), p. 237.
„ (ii«brecht, Fauna a. Fl. Neap. XIX. (1892), p. 153.

lilt. i/ii/<u was described by Brady as distributed over a very wide area between
long. 53° 32' W.— 130° 52' E. and lat 36° 44' 8 —65° 42' S. Much doubt baa been
expressed by Giesbrecht as to tbc validity of this species, and the figures given by
Brady of abdomen and of the whole animal are those, in Giesbrecht's opinion, of
immature animals, and this author thinks that Brady's figure of the first feet is really
of one of the other pairs of feet

Scott's specimens (Fair Isle and Firth of Forth) are regarded by Giesbrecht as
Jiti. nasutus ( Th. 3 and 4 with dorsal or with a lateral spine, as in nasutus, and a pair of
small dorsal points on the genital segment). Mobius's specimen from the north of
Scotland is also identical with nasutw. Rh. nasutus is very common in the Fane
Channel and seas off the north of Scotland, and occurs abundantly in my collections
made in these regions and along the Atlantic trough, west of Ireland, and also
appears in the ' Gauss ' collections as far south as lat 20° N., while in the eame
collections Rh, grandis (Giesbrecht) appeared. From the remarks of San in
"Crustacea of Norway," VoL IV., p. 15, it might be inferred that Rh. nasutus is of
rare occurrence in the Northern Ocean (" two specimens were taken cast of Iceland,
one specimen by Hjort between Scotland and Norway, and it has not yet been found
in the immediate vicinity of the Norwegian coast") However I have taken it in
abundance on many occasions throughout the Faroe Channel. It is rather important
to establish the identity of Brady's /.'//. gi</<is, and of two preserved specimens at the
British Museum, which I have examined, one measured 5*8 mm. and another 6*0 mm.
Both were immature females with four-jointed abdomen, lateral spines on Th. 3
(small), and on Th. 4 (Urge), with none on the fifth segment, resembling Rh. grandi*,
one dorsal spine on the first atxlominal segment (no dorsal spines on the thoracic
segments), and so far as could be seen without dissection, the first feet had an
exopodite of two segments only, and the fifth pair consisted each of only one ramus
of three segments. These two animals were, of course, very much smaller than
dncribed by Brady (8*5-10 mm.) and were undoubtedly immature, and the species
may well be identical with young Rlt. grandis (Giesbrecht).

L 2

14 R. NORRIS WOLFENDEN.

RHINCALANUS GRANDIS.
R/tincalanus grandis, Giesbrecht, ' Belgica ' Rep., p. 18.

? 7'2-S'O mm. Head produced in front, dorsally roughly triangular in shape,
with large lateral swellings at the base, rostrum not visible from behind. The cephalo-
thorax is over six times as long as the abdomen, which is composed of three segments.
A pair of short spines on the anterior margin of the third thoracic segment, and a pair
of strong and longer spines on the fourth segment, differentiate this species from nasutus,
also the absence of any spines on the abdominal segments. The first pair of feet have
the Ri and Re of only two segments, Re three with two marginal spines ; other feet
(except the fifth) have three-jointed rami.

The fifth pair, of one branch only on each side, with three segments, have on the
second segment a long inner marginal bristle, and on the last segment three bristles of
nearly equal length, two apical, of which the outer is the thickest and the middle one a
little the longest, and one on the inner distal margin. A short spine is present on the
outer margin in its upper third. The anterior antennae are about six joints longer
than the furca. Adult males were absent.

These examples are absolutely identical with Giesbrecht's species.

METRIDIA (BOECK).

One of the most remarkable things about Brady's ' Challenger ' Copepoda is the
omission of mention of any example of this genus from his report. Distributed
throughout the Atlantic from the North to the South Pole, and in the Pacific, and
throughout the track followed in the Atlantic and Southern Ocean by the ' Challenger,'
the absence of mention of any species of this genus is certainly extraordinary. In the
northernmost regions Metridia longa occurs (Sars, Norwegian North Polar Expedition)
throughout the Faroe Channel and the Atlantic trough as far south as Valentia in
Ireland ; and south of the Wyville Thompson ridge, M. lucens, normani and curticauda
(Wolfenden) ; while south of Lat. 40° and throughout the Atlantic occur M. curticauda,
brevicauda, princeps and venusta ; but south of Kerguelen appears a new and charac-
teristic species, M. gerlachei, which replaces all others. This is the representative
species of the South Polar seas, and it appears abundantly in the ' Gauss,' ' Discovery,'
and ' Belgica ' collections, and it is as characteristic of this area as M. longa and lucens
are of the northern cold area. M. princeps occurs seldom, and M. brevicauda as a
straggler, outside its proper area of distribution.

COPEPODA. 15

METRIDIA GKRLACHKJ.

Mttridia ytrlaetti, OhaVMBl, ' Belgica ' Report, p. 27.

9 3* 5-3 '8 mm., very variable in size, occasionally a little larger and often smaller,
but the avenge of size of example* in the ' Discovery ' and ' Gauss ' collections is
rather less than that given by Giesbrecht for the ' Bclgica ' specimens. The ccphalo-
thorax is one and a half times as long as the abdomen, head separate from first thoracic
segment, last segment with rounded margins. The al»domcn has the proportional
length of its three segments as 9 (genital) : fi : 4 (anal), and the fumi is one-fifth longer
than the anal, and three times as long an broad. It is divided into two portions by the
outer marginal bristle, of which the proximal is twice the length of the distal.

The shape of the head and thorax is in this species characteristic, the Iwick la-ing
extremely gibbous, and the head with very bold curve, which makes it easily recognis-
able at sight from any other species of this genus. The anterior antenna- are com-
paratively short and do not reach beyond the genital openings. The first and second
segments are coalesced ; the eighth, ninth, tenth, eleventh, have only faint indications
of separation ; the thirteenth and fourteenth joints are not so clearly divided as the
others. There are strong teeth on one, two, three, five, seven (one each), those of
the third, fifth, seventh joints the strongest, and directed straight forwards. The
Mthetasks are numerous.

The endopodites of the second pair of feet have the usual excavation and hook
process on the first segment, but in this species the inner hook is exceedingly strong.
In the fourth pair the end saw is only two-fifths of the whole length of the lie 3
(shorter than in Giesbrecht's examples).

The fifth pair consists on each side of three segment*, but the distal segment is
more or less completely divided into two, the division however is not complete. The
two basal joints are of about the same length and breadth, each as broad as a little
•over half the length. The third joint, however, is not more than four-fifths as long as
the basals, and only half the breadth. The second joint Ixjars one short distal bristle,
the last joint one outer marginal bristle in the proximal half, and three distal bristles,
of which the innermost is the longest and thickest, the outermost the shortest.

METRIDIA I-RINCKPS.
(Plate III., figs. 3, 4, 5.)

Jfttridia prinrtp*, Gieabrecht, Atti Line. Rend., Ser. 4, v. 5, p. 24.
„ ,, „ Fa 1111:1 a. Fl. Neap. XIX., p. 340.

„ Farran, Ann. Rep. Fish. Ireland 1902-03, PI. II. App. II. (1900).

.' Mitridia macrvra, Hun, Boll, du Mm. Oceanog. Monaco, 1905, no. 40, p. 7.

Though only one example of this species occurred in the ' Discovery ' collection, it
was frequent in the ' Gauss ' collection, not only at several Atlantic stations, but also

16

R. NORRIS WOLFENDEN.

at the southernmost stations. Northwards it ranges to the west coast of Ireland, and,
as has been mentioned, has an extreme southern distribution. Giesbrecht's description
was given from only one specimen, and compared with Sars', very briefly described,
M. macrura.

M. princeps, Giesbreoht.

Cephalothorax one and a half times as long as the

abdomen

Anterior antennae extend beyond the furca
Short teeth on 1, 2, 4, 5, and 6, the 2nd the

longest

Genital segment longer than both the following
Anal segment half as long as the preceding one
Furca twice as long as anal, and 5 times as long as

broad

Size 8 '5 mm.

M. macrura, Sars.
Tail about as long as the anterior division

Longer than the body

Only feeble traces of the strong teeth of princeps

About as long as the two preceding segments

5th feet like princes, but less unequal
Size 10-50 mm.

Of the many examples which have come under my notice I find that the relative
sizes of the abdominal segments and furca are subject to some variation, thus : —

1. G.S. 30. Ab. 2, 16. Anal 6. Furca 28 long, 3 broad. Size 8 '15 mm.

8. 25 3 8 mm.

2.
3.

4.

28.
27.
22.

16.
16.
11.

7.
4.

23

12

3

2

8 mm.
6 mm.

The genital segment is thus always twice as long as the two succeeding segments,
the anal is not more than half the preceding segment, the furca is generally longer
than the two preceding -segments, and usually from seven to nine times longer than
broad. The teeth on the antennae are weak, and entirely resemble the figure given by
Giesbrecht in Plate 33, fig. 3 (op. cit.).

It is difficult to resist the conclusion that the^e are one and the same species, and
not two different species. The male was not described by either Giesbrecht or Sars,'
but I have met with several examples.

? . The largest adults were from 8-9 mm. in length.

The body is very transparent, the head evenly rounded, cephalothorax (4 '15-
4 ' 5 mm. long, abdomen 4 • 0 mm. long), only slightly longer than the whole abdomen,
and a little over one-third as long as broad.

The genital segment is larger than the next two, the anal not more than half
as long as the preceding, often much mofe than the combined length of the two
preceding segments and 6-9 times as long as broad, and divided into two portions
by the marginal bristle, of which the proximal is to the distal as 8:5. The right
furcal segment is sometimes a little longer than the left. Anterior antenna; at least
three joints longer than the furca, the basal joints broad, with short teeth on the
basal seven joints, of which those on the first two joints are the largest. The

17

distal joints ta|»-i ami an- \.-ry slender. Pmportional length of jointa uf anterior
anti-lime: —

l*t|8| 4 | 5|6 | 7 | 8*9 | 10

II

U

i i

15 | 18 | 17 I 1« 19

•.•

n

22 I 28 I 24 I 2fi

4 | ft | 6 | ft | 6

II

.'

I-'

IS | 18 I 14 | 14 | 15

10

I

10 no i 7

The eighth and ninth joints arc quite coalesced, but in some there is a weak line of
division.

The second pair of feet have each the characteristic hooks on /•'/!. and the
outer ono is the longest The surface of the second basal is l>esct with short spines,
hut not the lie 1. The third feet are normal and with shortened end saw. The fifth
pair each consist of four segments, of which the basal is greater than the second,
this longer than the third, and third longer than the fourth and terminal segment
The first joint has on its surface a considerable bunch of long stiff hairs (as in
priwxps), the second joint has a long stiff feathered bristle on the outer distal margin,
and the third joint has a short upright spine on the outer distal margin, in all specimens
(not on the inner side as figured by Giesbrecht), and the end joint has three rather
long fine bristles, of which the innermost is the longest The spine on the third joint
was in one example replaced by two very short spinales on the right foot, while none
were present on the left side.

£ 5 * 8-6 mm. long (cephalothoraz 3 • 25 mm. Abdomen 2 • 3 mm. long).

Relative lengths of the abdominal segments =14, 10, 10, 4, and the furcal segments
13. The left furcal segment is a little the largest and thickest, and six times as long
as broad, and three times as long as the short anal segment!

The anterior antenute extend for alxmt three joints beyond the furca, as in the
female, and the left one is a clasping organ with weak joint between the seventeenth
and eighteenth segments. The segment beyond the ellww is very long and thin, and
as long as the next two distal jointa There are four joints beyond the elbow. The
conjoined first and second (Itasal) joints have two strong teeth, the distal one the
largest and carved slightly forwards. The fourth joint has a smaller tooth. Fifth feet :
The right foot with very long first joint, the second short, but with very strong, br«>ad-
based curved and long hook, the third joint nearly twice as long as the second, and the
fourth and end joint a long simple spoon-process twice as long as the third. In
the left foot the first joint is very small, the second nearly twice as long, the third
a short joint, the fourth a very long simple curved spoon-shaped process. On the
inner margin (proximal) of the fourth, of the third, and the distal foot of the second
arc fine hairs. Both feet arc of nearly similar length.

EUCHAETA (PHILIPPI).

Two representatives of this genus appear in the ' Discovery ' collections —
E. antarctica, and another which appears to have constant differences, and to which I
have attached the name E. sitmli*. I do not in this collection find any example of

18 R. NORRIS WOLPENDEN.

Giesbrecht's species E. austrina, though I have found it in the ' Gauss ' collections.
E. antarctica appears in many stages, extensive captures consisted wholly of
immature specimens, but there are many adult examples. Many males appear amongst
these, and while the females are very distinctively different, I am not able to discrimin-
ate between those males, as to which definitely belong to antarctica, and others
which might belong to similis. Both species are very closely allied, and differ
very considerably from the large species of the northern cold seas, viz., norwegica,
glacialis and barbata.

EUCHAETA ANTARCTICA.

(Plate IV., figs. 5, 6.)
EucJueta antarctica, Giesbrecht, 'Belgica' Report, p. 21.

This is one of the most abundant copepods in the ' Discovery ' collection, appearing
in all stages of growth, and in some samples almost to the exclusion of other species.

Size of adult examples 7 ' 6 mm.-8 mm. Head evenly rounded, without frontal
prominence and with short rostrum directed forwards. Last thoracic segment with
rounded margins, produced forwards and with a bunch of hairs on each side. The
abdominal segments have the postero-distal margins armed with rather strong bluntly
conical and striated teeth, and the two middle segments have on the ventral side
bunches of long hairs. The furca and bristles are the same as in E. similis. The
genital protuberance occupies the lower half of the segment, its upper margin is not
deeply concave like similis, but the whole swelling is directed downwards, and its upper
margin is slightly convex. Above the genital swelling is a secondary prominence,
which in the ventral aspect is seen to consist of two valve-like chitin thickenings. The
lower part of the protuberance has two lateral lobes, the upper are small, and above this
a prominent horn directed straight forwards and never absent in adult specimens of
E. antarctica, making it quite characteristic of the species.

In the ventral aspect the appearance is quite different from that of similis. The
genital opening is oval, almost round, with lateral cushions, and above the upper edge
of the genital opening guarded by a chitin ridge, is the horn.

The whole swelling is quite symmetrical, rather conical, and occupies quite the
lower part of the segment.

The first pair of feet have the outer margin very concave above and very convex
below, with a bunch of hairs on the Re 1 , and a long seta. The seta of Re 2 is also
very long and thin. The Re 3 is only half the length of the coalesced Re 1 and 2.

In the second pair the Re 1 has a very short Se, that of Re 2 is very long and
more than twice as long as the Se 1 of Re 3.

In Re 3, the Se 1 does not reach the origin of Se 2, the Se 2 does not quite reach
the end of the segment, and is three times as long as Se 3 and twice as long as Se 1.
The Sf 1 and 2 are very greatly curved and almost sickle-shaped.

19

III Uie fourth |>air tli. li H not three times a* long as bmad(16:6). Tin-
anterior antrinw arc a little longer than the ccphalothorax.

The & is a littlr -mailer than the 9 , and present* the same sexual difference* M
in other Euduutu-: The bunches of hairs on (he last thoracic segment, HO prominent in
the ? , arc absent in the <J .

The first feet have a three-jointed exopodite, the outer margin of which is not so
concavo-convex as in the ? , and its *SV are short

In the second pair, the Se of the exopodite are also smaller, the Xr of Re'l only
reaching the origin of the Sc 1 of ReS ; the Se2 of AY 3 l>eing little more than half
the length of the distal part of the segment. The fifth feet are characteristic. The
penultimate segment of the left foot is prolonged on the upper margin into a .-nmrjK
toothed process, and has a setose conical unhaired process on the distal margin, tin-
last segment into a long process, narrow and with a strong hunch of hairs at the distal
extremity, and with a large conical and strongly haired process. (This process is
sometimes nearly as long as the principal process of the penultimate joint.)

The first basal is short, the second Ita&nl long, and with very small and rudimentary
endopodite.

The right foot has short first basal, very broad second losal, long first and second
Re (which are coalesced), and with the last segment blunt and rounded.

EUCHAKTA -IMII.lv

(Plate IV., figs. 1, 2, 3. 4.)

This species occurs plentifully in the same samples in which K. antarctica is
present For a long time I regarded them as merely different stages in the history of
the same animal, but the careful examination of a great number of individuals from
different tow-nettings proves the constancy of the points of difference between the two,
and as many of the E. antarcticti and E. *imU'ui have spermatophores or egg sacs
attached, I have come to the conclusion that, though so very similar in most characters,
the two species must be separated on account of the invariably different characters of
the abdomen and genital segment

E. similis is constantly rather larger than E. anlurctiat, 8 • 6 mm.-8 • 8 mm., and
more robust, the head flat and rostrum small, but strong and directed forwards and
rather upwards. The cephalothorax is two and a half times as long as the aMonien.
The last thoracic segment is produced forwards, slightly triangular shaped, with evenly
rounded margin, no spine, but a bunch of long hair- on each side. The abdominal
segments, of which the genital is twice as long as the next, are covered with fine hairs,
nowhere with large bunches, and the posterior distal margins have only very small
teeth, not large, as in antarcticti. Furca with, on each side, a very short dorsal bristle,
the ventral accessory bristle not geniculated at the base, though bent outwards, its
length not more than al» mt half of the two long tail bristles (next to the innermost).

YOU IT. M

20 K. NORRIS WOLFENDEN.

Genital segment with very prominent genital tubercle, occupying half the segment,
(the lower half only), directed slightly upwards in lateral aspect, with apparently three
lobes, an upper ftnd lower, each large with small middle lobe, and without any eminence
on the upper part (of the ventral surface) of the segment as in antarctica, and also
invariably without the ventral horn which is so characteristic of antarctica. In the
ventral aspect the genital tubercle appears to be obliquely placed, directed downwards
towards the left, the vulva guarded by two prominent flaps of which the right one is
below the left one. The whole tubercle occupies more of the left than the right side of
the segment, which in the dorsal -view is only slightly swollen at each side.

The first and second pairs of feet and the number of bristles on the maxilla are the
same as in antarctica.

While the female is so distinctly different from that of E. antarctica I am unable
to find similar distinction in the males. Both kinds appear frequently to occur together
in the same sample, and all the males appear to be alike.

ONCEA (GlESBRECHT).

ONCEA CURVATA.
Oncea curmta, Giesbrecht, ' Belgica ' Report, p. 42.

According to Giesbrecht, this species differs from 0. subtilis in the following points :
Length, 0'6-0'S mm., the body lengthened, the three segments posterior to the genital
are about as long as broad, and comparatively longer than in any other Oncea species
except suljtilis (in which they are longer than broad), all three segments together are
shorter than the genital (in subtilis, the genital segment is only a little longer than the
two succeeding segments), the furca is as long or a little longer than the anal segment
(in subtilis shorter). The posterior antennae in both kinds are similar, the maxillipedes,
however, differ ; the terminal hook which, in subtilis, is thin and unarmed, is strong
and beset with pretty long teeth on the concave side, in curvata. The swimming feet
are similar in both species, except that in curvata the proximal outer marginal bristle
on Ri 3 occurs in all four pairs, while it is absent in subtilis, and the lancet-shaped
apical bristle of the fourth foot is in curvata longer than in subtilis.

The few examples that occur in the ' Discovery ' collections are smaller than
Giesbrecht's examples, none being more than '56 mm. in total length (9s with egg sacs
attached).

The genital segment is a little longer than the next three segments, the furcal and
anal segments of the same length, the second abdominal segment as long as broad, and
rather larger than the third segment, the relative lengths of Ab. 2:3:4 and furca
being as 4 : 3 : 4 : 4, the latter nearly three times as long as broad.

In the posterior antennae, the first basal joint is the largest, and the inner margin
of the second basal has a few fine teeth ; the distal segment has three proximal bristles

21

uf unequal leoglh, tin- middle one very short, and distal to it, a comparatively lung
ami slightly armed bristle, terminally four long and two shorter bristles.

The m.ixillijMjdo is armed with a strong claw bristle, denticulated on the inner
margin, anil tin* -••.•mul basal has two comparatively stout bristles, the proxinuil of
which is armed with wide-apart bristles.

In tin- fourth pair of feet the apical bristle of the exopoditc is longer than the third
exopodite segment by one-third of its length.

The aj-rrrineiit. therefore, between this species and Giesbrecht's examples is very
clan, the only diftnooa being one of size, and there is no doubt that the two species
are identical.

STEPHUS.*

JOKmiu, Givbrecht, Fauna n. Fl. Neap. XIX. (1892), p. 205.

ftp* » ii, Th. Scott, 10th Rep. Scotch Fwhery Board, Vol. X. (1892), p. 245.

Sttpko*, Giesbrecht, ' Belgica ' Rep., p. 20.
Ban, " Crustacea of Norway," Vol. IV. (1903), p. 61.

Since Giesbrecht described the genus ifobianus, which was subsequently identified
with Stfphos (Scott), several other examples of the same genus have been desmU-d.
Scott himself recorded three specimens, viz., S. minor, S. fultoni, and S. yyranx,
supposing the latter to be identical with Giesbrecht's Mobianus yyrans.

Bun has described two new species from Norway, viz., S. lamellatux and
Scotti, which latter is again identical with Stephos gyrans Scott (not Giesbrecht).
Giesbrecht's species (gyransi) is said by Sars to differ in the asymmetrical last thoracic
segment and genital segment, the latter with " a number of irregularly arranged
Hpiniform processes not found in any of the northern species," the last feet of the
male also differing from S. scotti,

The ' Belgica ' report contains yet another species dcscrilicd* by Giesbrecht as
Antarctic, vi:., S. longipes. This species recurs also in both the 'Gauss' and
' Discovery ' collections, and in the latter I find a further and considerably larger
example, to which I have given the name antarcticum.

At these descriptions are scattered over six different volumes, I think it may be
of service to bring them together here.

1. S. GYRANS.

S. ffyraiu, Gieabrecht, Fauna n. Fl. Neap. XIX. (1892), p. 205 ; Giesbrecht, a. Schmeil, DM Tiem-ich.
Copep., p. 29.

Furca longer than broad, anterior antcnute reaching to end of genital segment,
genital segment with a curved hook on ventral side, shorter hook on dorsal, fifth feet

• The author, in agreement with Ban, prefer* the name originally given to the scout by Scott, bat it ii
perhaps better to observe the ordinary rule.— ED.

M 2

R. NORRIS WOLFENDEN.

in female, end segment rather curved and broad basally ; in male, left foot with several
long appendages, right foot, thin appendages at end and rest foliate ; size =0'8-1 mm.
(Naples.)

2. S. SCOTTI.

S. ffymns, Scott, Nineteenth Rep. Scotch Fishery Board (1901), p. 237.
S. scotti, Sars, " Crustacea of Norway," Vol. IV., p. 63.

Slender ; cephalothorax symmetrical ; genital segment without spines, furca longer
than broad, anterior antennae reach end of abdomen 2, Re of posterior, antennae twice
as long as Ri. 9 , fifth, with denticles on last segment, which is elongated and
pointed ; $ , fifth, penultimate joint of left foot tumefied, last segment with about half
a dozen short processes ; last joint of right foot with long, sickle-shaped process.
Size='85--95 mm. (Loch Fyne ; Norway.)

3. S. MINUS.

S. minor, Scott, Tenth Rep. Scotch Fishery Board, 1892, p. 245.

Robust, cephalothorax symmetrical ; genital segment without spines, anterior
antennae about as long as the thorax, furca as long as broad, fifth feet in 9 with
elongated last segment with two little lateral spinules ; in £ , right foot a long foliate
joint at end, left foot with two digitiform processes at end, penultimate joint only
slightly tumefied. Size = 074 mm. (Firth of Forth.)

4. S. LAMELLATUM.
S. lameUatus, Sars, " Crustacea of Norway," Vol. IV., p. 62.

Short and robust, last segment asymmetrical ; right side longest, genital segment
unsymmetrical and rounded prominence on right side, but no spines ; furca about as
long as broad, anterior antennae reach to furca, branches of posterior antennae equal ;
fifth foot in 9 , last joint elongated, with fine spine midway ; in $ , left foot with
much tumefied penultimate segment, with proximally a long spine, and last joint
with a number (about nine) of leaf-like appendages ; right foot not foliate, but last
joint with three or four short, rounded appendages. Size = 1 mm. (Norway.)

5. S. FULTONI.
S. fultoni, Scott, Ann. and Mag. Nat. Hist., 7th series, Vol. I. (1898), p. 185.

Cephalothorax symmetrical ; genital segment with spine and hook ventrally.
Fifth feet in 9 are larger and broader (knife-like) than the other, which is pointed ; in c?
right foot with elongated penultimate segment and short, strong, foliate end segment ;
left foot, penultimate segment tumefied, extremity with five or six leaf appendages,
and bifid claw. Size = 1 mm. (Clyde.)

COPEI'OIIA. 23

6. 8. LOSOIPKS,
H. /../.././^.(Jicrfiredm'Belgiai' Kcp., p. 20.

< Vplialotlmrnx symmetrical ; genital segment swollen laterally and nui^lily
: iMiiuMilur >li;i|>fl. anterior antcnmu not reaching end of thorax, no Hpine.s on genital
segment ; fifth feet in 9 last segment elongated, curved (foliate), with external spine
half as long as in «j ; right foot with two middle segments very elongated, ending
distally in curved hook not articulating ; left foot without tumefied segment, two
middle elongated, last shorter with knob ami spine, but no processes. Size 'S-'O in.
(Antarctic). ('Discovery' and 'Gauss' collections. Wolfcndeu.)

7. 8. ANTARCTICUM.

Robust, cephalothorax a little unsynmirtriml, right side prolonged; genital segment
swollen laterally, with bunch of spine-like bristles each side ; furca as broad as long ;
anterior antenna) reach to Ab. 2 ; Re of posterior antenna) longer than Jit. Fifth feet
in 9 right side longest, each with three end spines, innermost hook-like ; in <f ,
right with third joint elongated and club-shaped distally, with a large, roughly
triangular plate, and last joint a strong, curved hook ; left foot, no tumefied segment,
and last joint with distally a short-stalked haired knob, no appendages. Size = 175-2
mm. ('Discovery' collection.)

STKPHU.S LONGII'KS.

(Plate V., figs. 1,2, 3.)
StrphiM longiptt, Gicabrccht. ' Bolgica ' Rep., p. 20.

9 -75- '80 mm. <J '65-70 mm. Cephalothorax rather more than twice as long as
the abdomen ; head separated from first thoracic segment, but all segments very
difficult to determine, owing to the indistinctness of the lines. Last thoracic segment
with rounded margins and symmetrical. Abdomen in the female four, in the male five
segment*, furcal segments only as long as the anal, as broad as long, and with rounded
margins, each with four tail bristles and a short fifth inner marginal bristle. Genital
segment in the female as long as the next two, laterally with roughly triangular
swellings, and in its greatest breadth, broader than long. Anterior antenna; not as
long as the thorax, and of twenty-three joints, the first, second, eighth and ninth
coalesced, with few bristles, the longest on the seventh and eighteenth joints, but well
supplied with long test he tasks.

Posterior antenna; with exopcdite about one-third longer than the endopodite.
Posterior foot-jaw with rather thick first basal and /?l:Z?2:/ftas8:7:6; mandibles
with broad-ended masticatory plate, one pointed outer tooth, with considerable space
)>etwecn it and the middle stout comparatively broad teeth, and three pointed inner
teeth.

24 R. NOREIS WOLFENDEN.

First feet Ri= 1, Re = 3 ; no Se on Re I, and very short Se on Re 2.

Second feet Ri = 2, Re = 3.

Third and fourth feet Hi and Re — 3.

Fifth, each of three segments, two basal, each short and comparatively thick,
terminal segment longer and thinner, prolonged, with a stout curved hook with short
bristles on the upper margin and an external marginal thin and short bristle.

The male is rather smaller than the female, the abdomen has five segments, the
mouth organs are as in the female, but the fifth feet are transformed into clasping
organs, that of the right side of four segments, the left of five. The second and
third segments of the left are elongated, the distal segment short and broadened
out, the distal extremity ending in a spine, and at the opposite side a short knob
process, apparently without marginal hairs.

The right foot with short broad second basal, and two distal very elongated and
thin segments, with a long thin sickle-shaped process at the end, which appears to be
a continuation only of the joint above it, and though forming an elbow, does not
articulate.

In the general structure this small Copepod bears great resemblance to the
species Stepkus antarcticum, which, however, is twice as large ; the fifth feet in the
female as well as the male are different, and the two species are therefore quite distinct.

STEPHUS ANTARCTICUM.
(Plate V., figs. 4, 5, 6, 7, 8.)

9 1'85-2'0 mm. long, cephalothorax about three times as long as abdomen, and
in its broadest part distal of the middle line one-third as broad as long. Head evenly
rounded, a little produced in front, but without trace of rostrum, a weak line indicating
its division from the first thoracic segment. Last two segments of the thorax im-
perfectly divided, and last segment a little unsymmetrical ; on the right side a little
longer than on the left, produced into a round-ended margin, on the left side more
acutely pointed, which is most marked in lateral view. Abdomen of four segments
respectively proportioned: genital segment, 2:3 and anal as 20:13:8:8; furcal
segments as long as the anal, and as broad as long. Genital segment laterally swollen
in the upper part (genital protuberances) and again slightly swollen laterally in its lower
part, with on each side a bunch of rather long spines, none dorsally. Furca with four
tail setae on each side, outer margins haired and with, on each side, a short lateral
spine instead of the usual bristle, and on the ventral side a very short accessory
bristle. Of the tail bristles, the two middle ones are much the longest and thickest,
and those of the right side more so than those of the left.

Anterior antennae reach in both sexes to about the end of the second abdominal
segment, having twenty-four segments, the eighth and ninth coalesced. In the posterior
antennae the exopodite is longer than the endopodite. The mandibles with branches

25

nearly of same length, masticatory plate with strong teeth.. The posterior foot jaws
with first aii'l -.•••.-nd !,.i-.il and Hi in proportion of 24 : 1 1 : 20 ; nmxilhr. /»'•_' with 5 ;
JRt with 10; /.'< 1 with I, A'' - with 8, Hi 3 with six bristle*, Li 2 and 3 present, and
Lf I with fight l'ii-tlc-.

Tip --t have one jointed Ki and three jointc<l Iff, without Se on Rt 1.

The second feet have two jointed Hi and three jointed /iV.

The third and fourth feet have lioth rami three-jointed.

l-'ifth feet comparatively large, that of the right side a little longer than the left.
Each of three segments, of which the two basals arc equal in size, but the last segment
on the right foot is a little longer than in the opposite foot. Each has terminally
three spiny processes, the innermost comparatively thick, curved, and hook-like, and
with hairs only on the outer margin. The two outer spines are neither much more
than half the length of the inner one, and only half as thick.

The ^ is a little smaller, 1 '75 mm., and of slenderer build ; the asymmetry of the
last thoracic segment is only slight, and on neither side is it so prolonged as in the
female. The abdomen consists of five segments, the first segment in more laterally
swollen than in the female, and is broader than long ; the second and third segments
alMiut equal in length, and much longer than the anal, which is very short. The
antennae, oral organs and feet are the same as in the female, with the exception of the
fifth pair, which are converted into two extraordinary appendages. Arising from a
common basal, the right leg possesses four segments, the left five segments. In the
right leg the first joint is short and rather broad, the second elongated, rather longer
and thicker than the third, which is a long thin joint with club-shaped distal extremity,
and having attached to the joint it makes with the last appendage a broad, nither
triangular plate covered with fine hairs and a few rather strong spines. The last joint
is represented by a comparatively strong and large curved appendage, armed along ita
inner margin with short stiff bristles, these two terminal appendages row-milling an
awkward-looking pair of shears.

The left foot, of five segments, has the first and second comparatively shorter but
broader than the third and fourth. The fifth segment short and broad at the distal
end, has externally a short curved tooth-like ending of the distal margin, and at the
inner end an upright knob-shaped appendage, strongly armed all over, and especially at
the base of the stalk, with short stiff bristles.

The fifth feet of the <J and 9 possess no near resemblance to similar organs in any
other species of Stephus ; the size of the animal, moreover, is comparatively large for
any representative of this genus, and it must therefore be regarded as a new species.
Several examples occurred in two or three of the ' Discovery ' collections.

26 E. NORRIS WOLFENDEN.

PARALABIDOCERA.

There is no mention of any example of the genus Ldndocera in Giesbrecht's
' Belgica ' report. In the ' Discovery ' collection there are a great number of specimens
of an animal superficially resembling Labidocera, but which does not agree with any
known species of that genus, though bearing some relation to L. wollastoni. Between
60° and 70° S. Labidocera acutifrons appears in the ' Gauss ' collection, but is absent
from either the ' Discovery ' or ' Belgica ' collections, and this genus is thus very
sparingly represented in the Antarctic area. The copepod referred to below differs
distinctly from any known Labidocera in the character of the swimming feet of the
female and the five-jointed abdomen of the male, and the fifth pair of feet, and I have
thought it better to create a genus for it.

Characters of the genus. — Very similar in appearance to Labidocera, but a total
absence of " ocelli," and of very unsymmetrical shape, the swimming feet without
spines on the last segment of the exopodite, and the male abdomen of five segments.

PARALABIDOCERA HODGSONI.
(Plate VI., figs. 1-13.)

9 T55-2 mm. ; <J 1*6 mm. long. The head is evenly rounded, produced
forwards a little, and in front are two delicate rostral filaments. There is no trace of
eyes, either dorsal or ventral, but in some males there are two dark spots laterally on
the head, and in a few females a dark pigmented spot on each side, which may possibly
have been ocelli. But considering the mode of preservation, which included freezing
and thawing, and a long sojourn in spirit, these organs may very well have been
present at some time, and subsequently vanished. The head is quite without any
trace of side hooks, and separated from the first thoracic segment ; last two segments
coalesced and produced on each side into lateral expansions, but bluntly ended.
Abdomen of three segments, the genital with large lateral outgrowths, and also
dorsally and ventrally swollen a little ; spines entirely absent ; next segment also
laterally enlarged, and anal segment small ; furcal segments a little unsymmetrical,
the right a little longer and broader than the left ; all tail bristles comparatively
short, consisting of four apical and one lateral marginal (situated a little distal of the
middle), all slightly thickened at the base. There is also a small accessory dorsal
furcal bristle on each side. Anterior antennae shorter than the cephalothorax and with
only twenty-two distinct joints, very densely covered with long bristles.

PROPORTIONAL LENGTH OF ANTENNAL JOINTS.

1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22

10 I 10 | 2 I 4 I 1J I 1 I 2 | 3 | 3 | 8 | 4 j 4 j 4 | 6 | C j 6 | 6 j 7 j 5 j 10 j 6 j 8} '

27

with AV mm-li longer tinin A'., the former with six long bristles
on the first segment, and seven and six bristles on the distal segment AV very
indistinctly segmented, tin- proximal joint very elongated.

Mandibles with /.'/ longer than AV, B2 with one marginal bristle, masticatory
plate broad, with one rather large tooth and a good space between it and the next
five conical short tooth

Maxilla witii /•'•_' and Ri lient outwards; B2, lii, and fie coalesced and almost
indistinguishable ; /.< 1 with seven rather long and stout hooks and two shorter
l.ii-' • / _' :i large loin? with three bristles, Li 2 with one bristle; A'/ indistinctly
segnu nii'<l <>n the outer margin with five apical bristles; /fc scarcely segmented, with
i.iiK two bristles; Lei with seven long and very thick bristles, and three shorter
brisi

Anterior foot jaws with lotas much compressed, and terminal five bristles longer
than the proximal and also much thicker.

Posterior foot jaw very similar to that of Anomalocera. B 2 is short, but rather
thick, and with one short bristle ; Ri short, unsegmented, with only three terminal
and short bristles.

One to four pairs of feet, with Ri of only two, Re of three segments. In the first
pair the external marginal setae of Re are long and thin, in the others the external
spines are short, and in all there is only one marginal spine (apical) on Rf 3.

First pair of feet, // 1 and B 2 with only slightly convex inner margin and no Si.
Rf I longer than AV 2 + AV 3, outside margin haired and long marginal bristle. Re 2
and Re 3 with similarly long marginal bristles. Se = 1:1:2 and Si = 1:1:5. Ri2
nearly twice as long as Ri 1 and with 5 Si.

In the second pair B 2 is broader than long, the inner margin convex and without
hairs or bristles ; D 1 is also without Si. Ri 1 prolonged, Ri 2 shorter (as 7:10). Ri I
with 2 Si, Ri 2 with 7 Si.

Re 1 : 2 : 3 as 1 1 : 6 : 9, with respectively 1:1:1: 6V, that of AV 1 the largest and
of /.'• 3 the smallest ; the end saw about two-thirds as long as the whole AV ; a small
curved spine distal margin of AV 3 just external to the saw. 5 Si on Re 3.

3rd feet B 1 and B 2 without bristles or hairs, Ri as in the preceding pair, Re
as in the preceding pair and with 1 : 1 : 1 Se only.

4th feet B 1 and B 2 as before. Ri 1 with three Si, Ri 2 with only six Si The
three joints of Re rather more equal in length, Sf 1 : 1 : 1 as in the other feet Ri 1 is
more elongated than in the other feet and twice as long as Ri 2.

In the second to fourth pairs the Ri is more than half as long as the Re and
extends beyond the distal margin of Re 2. In all feet the number of external marginal
spines is certainly peculiar, the usual rule being three marginal spines on the Re 3, so
that the animal differs from any true Lnbidocera.

The 5th feet consist of a common Iwisal and one ramns on each side of two
segments B2 and AV. A'i represented only by a spine. The first and second baxnls arc

VOL. IT. N

K. NORRIS WOLFENDEN.

nearly equal in length and each about as broad as long. The second basal segment has
at its distal inner margin a very stout articulating spine, Ri four-fifths as broad as this
joint is long. On its outer surface, near the distal and outer end, is a delicate bristle.
The third segment is nearly twice as long as the second basal, tapers to a point, and just
below the distal end is a delicate little bristle. Near the end of the joint and on the
inner side is a very stout broad-based spine, not articulating, and nearly half as long as
the whole segment. Frequently the foot of one side is a little longer than of the
other.

The whole animal .is very unsymmetrical, especially in the shape of the last
thoracic segment and the genital segment of the abdomen. The characters of the
swimming feet, as to proportions, and especially as to the absence of spines on the last
joint of the exopodite, and the absence of anything like the usual ocelli of Labidocera,
are points which appear to remove it from that genus. The abdomen of most females
is more or less enveloped in a mass of colourless, structureless membrane. The $ is
distinctly five-jointed in the abdomen, whereas in Labidocera this is four-jointed. The
right anterior antenna especially also differs from other species. From the characters
'of the female feet and the abdomen of the males I have thought it justifiable to create
a new genus for this animal. I name the species after Mr. Hodgson, the naturalist of
the ' Discovery ' Expedition.

$ smaller than the 9 (1*6 mm). Cephalothorax with head separate from next
segment, two dark lateral spots, but not ocelli, in front of the head. Abdomen little
more than half as long as the thorax. Head evenly rounded, without side hooks, last
thoracic segment rounded and not produced. Abdomen of undoubtedly five segments,
of which the second is about as long as the third and fourth together, the first is very
short, the fourth is twice as long as the anal, which is a short segment ; the furcal
segments, of which the right is a little larger than the left, are twice as long as broad
and three times the length of the anal segment.

Right anterior antenna a clasping organ, the middle joints swollen, the joint
before the elbow with a marginal row of small teeth and with only two distinct
segments beyond the elbow, of which the- distal is very long and thin (over three times
as long as broad), and in its distal part carrying on the inner margin a very long spine
tapering to a fine whip-like extremity, but broad in its basal portion. This appendage
is half as long again as the whole joint. The joint immediately distal to the elbow has
on its margin proximally a short but thick spine. I find it very difficult in any of the
specimens, of which there arc several, to agree with any degree of accuracy upon the
exact number of segments in this antenna owing to the fact of its being curled up
in every case. First to fourth feet and mouth organs as in the female.

5th feet, powerful clasping organs, the right one of four segments, the left of
three, with a common basal. The first segment of the right foot has on its inner
margin a small knob projection, the second has two short, thick spines, the third a fine
spine, and the last joint is curved into a strong hook, without any appendages.

Tip- in-i juint of tin- l<-tt l->"i li.i- .1 -Ill-it spine on the extcrnnl margin, an<l tin-
last joint is broad proximal ly and folinecous, and on its inner surface is a row of fine
bristles, with three or f«-ur stiff ami lunger than the rest

EUCALANUS (DANA).

That any species of Euatlanua should be found at extreme southern latitudes is
certainly curious. E. elongntux certainly occurs south of lat 40°, and aUmt 40° W.
lung., ' Gauss' collection ; and in the ' Discovery* collection I have found al>out half a
dozen examples of a Knfalitnu* which I regard as a variety of E. xiibttnuiti or mucro-
natti*. This occurred at station marked 21. x. 01, lat 57° 25|' 8., long. 151° }' E.,
and station lat 56° 31' 8., long. 156° 19' 30", 22. xi. 01 ; in both cases a long
• h'tance outside the Antarctic Circle,

The ? (no males were found) is 4 mm. long. The head is very triangular,
elongated, and produced in front into a blunt point slightly bent downwards; there are
lateral swellings as in attenuatus, the part behind is not, however, indented. The la.-t
thoracic segment is rounded. The abdomen has three segments, and one tail bristle on
the left side is a little thicker and longer than the rest The genital segment is
laterally swollen and broader than long. Posterior antenna- with first and second
joints of the exopodite coalesced, the first joint of the endopodite alx>ut three times as
long as broad, and about the same length as Ri 2. The mandibles with very short /.'/.
the proximal part about three times as long as the distal, the whole Hi very much
shorter than the distal part of the basal, and with four bristles and two short marginal
bristles. Maxilla, B 2 with five, Hi 1 with four, Ri 2 with four, Hi 3 with five bristles.
With some resemblance to E. subtenuis, piUattu, and mucronatus, it is larger than any
of them. The five bristles on the B 2 of the maxilla cause it to differ from either
mucnmaltu or *ut»tenui*t and it has considerable differences from piUatun in SIM,
posterior antenna and mandible. The shape of the head is certainly not that of
subtenuis, nor is it so triangular and pointed as mucronatus.

CTENOCALANUS (GIEHBRECHT).
ORNOCALANUS VAXUS.

•

Ctaucaianiu ranu*. Oiabrecht, Atti Ace. Lincei Rend., 8er. 4, 1888, p. 886.
„ „ Fauna a. Fl. Neap. XIX. (1892), p. 194.

This is extremely abundant in the ' Discovery ' collections, but does not differ in
any material particular from the species well known in the Atlantic. Its range of
distribution is very great, extending from the Faroe Channel (Wolfenden) throughout
the Atlantic to the southernmost parts of the Antarctic area, i.f. to the ice region.

N 2

30 R. NORRIS WOLFENDEX.

CLAUSOCALANUS (GIESBRECHT).

CLAUSOCALANUS ARCUICORNIS.

Clausocalanus arcuicornis. Giesbrecht, Atti. Ace. Lincei Rend., Ser. 4., vol. 4, p. 334.

Giesbrecht, Fauna u. Fl. Neap. XIX. (1892), p. 50.
„ „ Giesbrecht u. Schmeil, Das Tierreich, p. 27.

That this species should occur so far south is rather peculiar. It was found in the
collections made at : —

Lat. 49° 40' S. Long. 172° 18' 30" W.

Lat. 55° 44' S. Long. 95° 43' 30" W.

Lat. 56° 12' 45" S. Long. 136° 18' 30" W.

Lat. 57° 25^' S. Long. 151° 43' E.

Lat. 58° 49' 45" S. Long. 154° 48' W.

Lat. 59° 19' S. Long. 120° 24' 30" E.

Lat. 63° 5' S. Long. 175° 43' E.

Lat. 84° 01' S. Long. 170° 49' E.

and does not differ essentially from the species common in the Atlantic. It has a
considerably greater range than was thought, since I can record it from the Irish coast
to nearly the Antarctic Circle.

GAETANUS (GIESBRECHT).

GAETANUS ANTARCTICUS.

(Plate III., fig. 6.)
Qaetanus antarcficus, Wolfenden, Plankton Studies, Part I. (1905), p. 7.

Size 8 mm. The body is very robust and dorsally very gibbous. The head and
first thoracic segment are coalesced, and together much longer than all the rest. The
last thoracic segment carries two short stout curved spines, directed backwards. The
head is in its upper part quite square, and with short stout curved spine, directed a
little forwards. The abdomen is not a quarter the length of the cephalothorax.

Anterior antennae not as long as the body, of twenty-three segments, with the
eighteenth, nineteenth and twenty-first segments longer than the twentieth, and all
joints with very few setae. Hi of the posterior antennae more than half the length of
Re. Posterior foot jaws with lamellar process on the first basal.

Maxillae ; Li 2 and Li 3, each with four bristles ; B 2 with five, Ri small and two-
jointed. Re small, and less than half the length of B 2.

First feet, Re of three segments with three marginal spines, the segmentation being
complete ; Ri of only one segment.

SI

Second feet, Jti .lutinrtly two-joiut«Hl.

Third ami fourth feet, Hi and Re of three joint* each. D 2 of the fourth feet with
tuluil bristles as in Gaidin-.

The extraordinary rise of this animal makes it the largest known specie* of
Gaetanut. It occurred once only in the ' Discovery ' collection, and also appeared in
the ' Gauss ' collection, and is probably Antarctic in its habitat Several new special of
Gaetanus have been described recently, and it may serve some useful purpose to
recapitulate here the different species discovered since Giesbrecht and Schmeil's last
work (Tierreich, 1898).

1. GABTANUS MAJOR.

0. maftr, Wolfenden, Proc. Zool. 8oc., London, Feb. 8, 1903, in Dr. Fowler's paper.
„ Farran, Ann. Rep. Fifth. Ireland, 1902-OS, Part II., App. II., 1905.

Size 5 mm. and over. Anterior antennae larger than the body by al>out one joint ;
lamellar appendage of posterior foot jaws absent ; Re of first feet of three segments,
and with three Se. Cephalic spine short, and as in G. armiger.

2. GABTANUS CAUDANI.

Oattanut catidiwi, Canu, Ann. Univ. Lyon, V. 26, 1896.

„ .. Wolfunden, Jour. .Mar. Biol. Assoc., 19u4, p. 24.

.. (?) pdeatu*, Farran, ibid.

Like G, mils*, but anterior antennae only one-and-half times as long as the body ;
lamella of posterior foot jaw like that of G. miles. Re of first feet, three segmented
basals of fourth feet with tulwl bristles, 5 mm. and over. Cunu's original description
was of one immature <? . Farran's were also immature specimens.

3. GAKTANUS HOLTI.

Oattantu holti, Farran, ibid.

„ lattfrmu, Sara, Boll. Mns. Monaco, No. 26, March, 1905.

„ longispinu*, Wolfenden, Plankton Studies, Part I. (1905), p. 7.

Cephalic spine strong and directed l>ackwards with long interval between the
frontal part and base of the horn. Anterior antennae not as long as the body-spinet of
the last thoracic segment, strong, long, and directed backwards. Small lamella on
posterior foot jaw. First feet with three segments and 'three Se. Fourth feet with
basal tubal bristles. Size 4*74 mm.

4. GAETAMDS ANTARCTICUS.
Gaetanii* antarctieut, Wolfenden, Plankton Studies, Part I. (1905), p. 7.

Thorax gibbous, very stout short curved cephalic spine directed forwards, head
square, not like G. armiger. Abdomen short and thick, not a quarter of whole length.

K. NORRIS WOLFENDEN.

Anterior antennae not as long as the body. Posterior foot jaw with lamella. Re of
first feet with three segments and three Se. Fourth feet with tubal bristles. Size
8 mm.

5. GAETANUS MINOR.
Gaetamis minor, Farran, Ann. Rep. Fish. Ireland, 1902-03, P. II., App. II. (1905), p. 34.

Cephalic spine as in G. armiger, long slender spines of last thoracic segment.
Anterior antennae reaching only to genital segment. First feet with Re of two
segments and only two Se. Second feet with one-jointed Re. Size 2'4 mm.

6. GAETANUS ROBUSTUS.
Gaetamis robustus, Sara, Bull. Mus. Oceanographique Monaco, no. 26 (1905), p. 11.

Robust, cephalothorax a little swollen in middle. Cephalic spine small, curved,
or sometimes absent. Spines of last thoracic segment very strong and divergent.
Abdomen thick and about one-third of whole length. Anterior antennae scarcely as
long as the. body. Size 8 mm. (? possibly the same as G. antarcticus, but the
description of G. robustus is insufficient.)

7. GAETANUS INERMIS.

Gaetanus inermis, Sars, op. cit., p. 12.

Body very thick, anterior division swollen. No trace of cephalic spine, last
segment of thorax rounded and without spines. Abdomen very short, not a quarter of
whole length. Anterior antennas not longer than body. Structure of other parts not
different from other species of this genus (?) Size 6 • 30 mm.

(In the absence of cephalic and thoracic spines, which are constant in this genus,
this is probably not a Gaetanus.)

8. GAETANUS CURVICORNIS.
Gaetamis cvrvicornis, Sars, op. cit., p. 11.

Body like G. miles (Giesbrecht), short curved cephalic spine. Spines of last thoracic
segment moderately large and divergent. Very short abdomen only a quarter the
whole length. Anterior antennae scarcely longer than the body. Size 4 • 35 mm.

9. GAETANUS KRUEPPI.
Gaetamis kruppi, Giesbrecht, Mitt. Zool. St. zu Neapel, XVI. (1903), p. 202.

Like G. armiger, but larger, viz., 3 '6-4 mm. long, thorax shorter. Anterior
antennae reach three joints beyond furca, twenty-second segment longer than in G.
armiger. Feet like G. miles. $ 3 '7 mm. long, thoracic spines shorter than 9 and
antennas shorter than body. Se of Re 2 of first foot rudimentary ; fourth feet without
tubal bristles. Fifth feet Ri of one segment, Re of right foot of two ; of left, of three
segments. Mediterranean.

COI'EI'UDA 33

XANTHOCALANUS (GIKBRBCHT).

! '..• i:.- ..nl\ lu>. .-\:i in ].!••- ••: tii;~ •_• Mill- in !:!•• • I •: . .,\, ; \

the jml-lication of Giesbrecht and Schmeil's " Copepoda " (in " Tierreich ") the genus has

...! ll: ll.V l.MlM'.ll il -]»•, -1. - T.. I'M "I iLMII.lIK dntribed -)"'!.-. B . . A". / , '

and .V. minor (Gittbrecht), are now added .V. Ixwilis (Sara), X. propinquu.* (Saw),
A'. mttticitA (Sara), X crixtatu* (Wolfenden), X. subcrixtatus (Wolfondcn), X. simplex
(Wolfeuden), X. mag mi* (Wolfenden), X. calaminus (Wolfenden), X. atbmtictw
(Wolfenden), X. greeni (Farran), X. pinyui* (Farran), and A'. tJ>tnxtt.<t (Farran) ; and,
as they are described in scattered publications, it may be well to recapitulate the
characters here.

1. X. AOILIS.
A', ayilu, Giesbrecbt, Fauna a. FL Neap. XIX. (1892), p. 286.

Size 2'4 ; furcal segments broader than long, abdominal segments very hirsute ;
anterior antennn reaching end of furca ; fifth feet three segment*, beset with spines
and teeth, and with three apical teeth.

In the <J only one (the left) foot five-jointed. (Mediterranean.)

2. X. SUBAOILIS.

X. tuboffilu, Wolfenden, Jour. Mir. Biol. Awoc., VII. (1904), p. 118.

Size 2'6 mm., resembling A', ayilis, but abdominal segments not hirsute.

Fifth feet with three segments, the basal with strong teeth, the middle with only
hairs, the distal spinulose and with three long apical spines.

4 with a pair of fifth feet nearly equal, the right of four, the left of five
segments. (Mull of Galloway.)

3. X. BUREAUS.
X. bortati*, Sara, Crustacea of Norway, Vol. IV., p. 46.

Size 3 '50 mm. ; furcal segments as broad as long, anterior antenna) reach end of
genital segment ; fifth feet of three segments, proximal two, with teeth on inner
margin ; last with two apical and two lateral spines.

<J with a pair of fifth feet left of five segments, right very short and of only three
segments. (Polar Seas. Norway.)

4. X. PROPINQUUS.
X. propinqutu, San, l«e. at.

Size 1 75 in in. ; furcul rami longer than broad ; anterior antennae slender and
reaching only to second abdominal segment, posterior antennae with lie much longer

34 R. NORRIS WOLFENDEN.

than Hi ; fifth feet of three segments, last much smaller than proximal two, only basal
with marginal teeth, end segment with three short outer and one long inner spine.

<J a pair of fifth feet, the right very rudimentary and short, of three segments.
(Polar Sea. Norway.)

5. X. CRISTATUS.
X. cristatus, Wolfenden, Jour. Mar. Biol. Assoc., 1904, p. 119.

Size 5 'Omm.; head triangular and with prominent crest, anterior antennae
reaching to end of furca ; fifth feet of three segments, all densely spinulose, with two
short apical spines. <J not known. (West of Ireland.)

6. X. SUBCRISTATDS.
X. subcristatus, Wolfenden, Plankton Studies, Part II. (1906), p. 81.

Size 7'0 mm. ; head with crest, abdomen very hirsute, furcal segments very short,
anterior antennae reaching end of genital segment ; fifth feet three segments, the distal
long and tapering, with two short apical spines ; all these segments densely spinulose.
(J not known. (South Polar Sea.)

7. X. MAGNUS.

- *•
X. magnus, Wolfenden, op. cit., p. 32.

Size 8 '8 mm. ; head rounded, not clearly separated from next segment; furcal
segments very short, anterior antennae reaching the genital segment ; abdominal
segments very hirsute ; fifth feet of three segments, very small ; all segments very
spinulose, with two apical and two lateral spines on the last segment. <J not known.
(South Polar Sea.)

8. X. SIMPLEX.
X. simplex, Wolfenden, op. cit., p. 30.

Size 1'45 mm. ; whole surface of thoracic segments covered with fine prickles,
anterior antennae very short, posterior antennae with Re nearly twice as long as Ri ;
anterior foot jaws with only vermiform processes, posterior foot jaws with long, thin
B 2 without bristles ; fifth feet very small, of common basal and two segments, the
distal one very small, with two apical spines on the left and only one on the right foot.
<J unknown. (West of Ireland.)

9. X. CALAMINUS.
X. calaminus, Wolfenden, op. cit., p. 34.

Size 5 '5 mm. ; furcal segments as broad as long, posterior antennae with rarni of
equal length ; anterior foot jaw with powerful toothed hook on fourth lobe, and two
slenderer hooks on fifth lobe ; brush and vermiform processes ; posterior foot jaw short
and stout, the bristles of the endopodite very peculiar and like quills, with broad
chitin expansion with serrated edge ; fifth feet very small, of three segments ; distal
segment with short apical and two short marginal spines. (Bay of Biscay.)

OOPVOD*.

10. X. ATLANTICCa
X. ttlantietu, Wolfenden, Jour. Mar. Biol. Amoc., April, 1904.

nun. ; ant. ii.. r antennas much shorter than thorax and very thick Iwsnlly,
abdominal segments prickly and hirsute, feet very spiuulose ; fifth of three segment*,
distal the largest, with four large articulating spines, two apical, two lateral ; all
segments very spiuuloae ; furcal rami as broad as long, and lie of posterior antennao
much longer than Hi. (West of Ireland.)

11. X. oBTUsrs.

X. obtuna, Farntn, Ann. Rep. Fish., Ireland, I'.uii-oS, pt. ii., App. II. (1005), p. 40.

Size 2*4 nun. Furcal rami little longer than broad, anterior antenna; reach genital
segment, feet very spinulose. Fifth feet, three segmented, spinulose, and with two
terminal and two lateral spines on last segment ; second joint longest, and spiuulose on
both margins. (Atlantic. Ireland.)

12. X. PINOUIS.

JT. jmyvu, Farran, Ann. Rep. Fuh., Ireland, 1902-03, pi. ii., App. II. (190.1), p. 40.

Bin 4*5 mm. Head imperfectly separated from thorax ; lateral processes of
last segment blunt ; furcal segments slightly longer than broad ; anterior antenna)
short, not as long as thorax ; rami of posterior antenna) short and broad ; feet spinuloHe.
Fifth feet of three segments, and two lateral and two apical end spines ; surface of
third and margins of first (inner) and second (outer) spinulose. (Atlantic. Ireland.)

13. X. GREEN I. X. MUTICUS.

A'.ffrtmi, Farran, Ann. Rpp. Fish., Ireland, 1902-03, pt. ii., App. II. (190:.), p. 40. .V. mutietu, Hare,
Bull. Mas. Monaco.

Size 5'30-G mm. ; last two segments of thorax united, last segment with obtusely
triangular margins ; furcal segments short, broader than long ; anterior antenna? little
longer than body. Fifth feet small, two-jointed (Sars), or three-jointed (Farran), without
spines on inner margin ; last joint with three small apical spines. (Atlantic. Ireland.)

14. X. TYPICUS.
Amallophora typita, Scott, Tr. Linn. 8oc. (2), VI. (1894), p. 54.

Only the 4 known ; 27 mm. long ; anterior antenna), twenty segments ; right
fifth foot short and three segments, left like that of .V. agUis.

TOI- IT.

36 R. NORRIS WOLFENDEN.

XANTHOCALANUS ANTARCTICUS.
(Plate VII., figs. 10, 11.)

9 2 '5 mm. long (cephalothorax 2'0 mm., abdomen 0'5 mm. long). Abdomen,
therefore, only one-fourth as long as the thorax. Head rounded and rather oval in
front, with bifid rostrum, composed of two chitinous plates each with a long and rather
thick filament. Head separate from first thoracic segment, last two segments separate,
and distal segment on each side produced into lappets ending in rather pointed but
rounded margins. Furcal segments half as long again as*broad. In the middle, the
thorax is broad, about half as broad as long. Abdominal segments with row of
pectinations laterally, and over whole dorsum at the margin of the very short anal
segment. Anterior antennae twenty-four segments, and short, only reaching to the
end of the cephalothorax.

Posterior antennae with Re a little longer than Ri.

Mandibles, Ri and Re about equal ; B 2 with three marginal bristles ; Re elongated
and narrow, with seven bristles. Maxilla, B 2 and Ri elongated and narrow, imperfectly
segmented; B2 with four, Ril with 1, Ri2 and 3 with six bristles, Li 1 nearly
twice as long as broad, with long, thin hook bristles. Anterior foot jaws with very
convex B 2, the proximal margin of the basals much embayed. Strong toothed and
curved hook on the last lobe ; a number of brush processes, with small heads, and two
vermiform processes" distally.

Posterior foot jaws, having proportionately B 1 : B 2 : Ri = 40 : 30 : 20, the
second basal about three and a half times as long as broad, and a brush process on the
first basal.

1st feet Ri = 1 segment. Re = 3 segments with three long, thin marginal spines.
2nd feet Ri = 2 segments, Re = 3 segments. Ri 2 with a strong corona of
spines.

3rd feet with a few delicate spines on surface of Re 2.

4th feet Ri = 2 segments, Re = 3 segments. Ri 2 with a few spines on the
outer margin. The exopodites of the 3rd and 4th feet not spinulose.

5th feet small. A common basal and each three segments ; the first segment
rather longer and broader than the second, with convex inner margin, and several
(about eight or nine) strong teeth on the margin ; the second segment with two or
three shorter spines on the proximal part of the convex inner margin ; the third
segment with two short apical, and one outer and one inner marginal spines — four in
all — each articulating with the segment, and on the inner margin a bunch of small
teeth. A cluster of five spines on the distal surface of the last segment, and two very
small spines on the outer margin of the middle segment distally. No spines on the
surface of the twp proximal joints.

' "I'KPODA.

XANTHOCALANUS MACHOS.

(Plate VII., figs. 1-9.)

maynut, \V,.lf, u.l. n, Plankton Studiw, Part II. (1906), p. 82.
9 ii 0 nun. long. Head dorsally with line of separation from* the first segment,
n MI n« led aid without any trace of crest ; produced in front into a chitiuoun lamella
with two pointed mini. Last thoracic Hegmeut on each side a little produced. Al>domen
short, die ccphalothorax In-ing three and a half times its length. Genital segment
protulieraut ventrally and longer than the next two, anal segment very small, and
furcal segments much longer than the anal.

Anterior antennae, reaching about the end of the genital segment, of twenty-four
segment*, with thiek Iwsal joints, the eighth and ninth coalesced, the last segment very
small. Posterior antennae with Ri longer than Re, the first joint of the latter with
strong rounded projection of the lower and inner margin. The masticatory plate of
the mandibles with strong teeth, the two outer longer than the inner ones, which are
short and all of the same size. Anterior foot jaws short, but strongly built, the outer
margin very convex, the last lobe bearing a very strong thick basally and curved sickle-
shaped hook, tapering distally ; all the bristles of Ri represented by sensory brush and
vermiform appendages. The posterior foot jaws somewhat extended, the first liasal
comparatively thick and with a brush appendage, the second basal elongated and thin,
with very short marginal bristles ; Hi also elongated, the first and second joint* long,
and its bristles comparatively short Maxillae very like the preceding species, but B 2
with five, Ri with ten bristles, Re elongated and narrow.

1st feet, Re with three distinct segments and three external spines ; Ri only one
segment

2nd feet, Re of three broad segments very spinulose on the surface, and with short
external marginal spines, Ri of two segments with prominent bunches of spines on the
surface of Ri 2.

3rd and 4th feet, each ram us of three segments, the surfaces spinulose.
5th feet small, of three .segments more or less covered, especially the last segment
and margins, with comparatively long, spine-like bristles ; the terminal segment with
two short terminal and two very short marginal spines, not articulating.

This is an adult female, and resembles the animal I have descril>ed as Xaiithocalanu*
magnu* (Plankton Studies, Feb., 1906) so closely that I think they must be regarded
as the same animal. The ' Gauss ' animals are, however, very much larger (up to 8*8
mm), but the only essential differences which I can detect are the much greater size
of the latter, the rather more pointed dorsum of the head, and the more hirsute
abdominal segments. In these collections I have found many examples which appear
to differ only in size, and I am inclined to think that this ' Discovery ' example is
ni.-ifly a smaller one of the same species. The 5th pair of feet are strikingly different
from the northern species.

38 R. NOERIS AVOLFENDEN.

HETERORRHABDUS (GIESBRECHT).

Only one specimen of this genus occurs in the 'Discovery' collection, which is
only what might be expected from the fact that the collection is practically only
epiplanktonic, whereas Heterorrliabdus is without doubt one of the most confirmed deep-
water genera of Copepoda. The species //. auslrina (Giesbrecht), which occurs in the
' Belgica ' and ' Gauss ' collections, is absent from those of the ' Discovery,' and the only
specimen of the genus occurring in the latter is, I think, referable to II. longicornis.

HETERORRHABDUS LONGICORNIS.

Heterochata longicornis, Giesbrecht. Atti. Ace. Line. Rend., Ser. 4, v. (1889) p. 811.

„ „ „ Fauna u. Fl. Neap. XIX. (1892), p. 373.

„ „ Wolfenden. Jour. Mar. Biol. Assoc., Vol. VII. (1904) p. 124.

? Heterocfueta major, Dahl. Verb. d. Zool. (resells., 1894, p. 79.
Heterorhabdus major, Wolfenden. Plankton Studies, Part I. (1905), p. 11.

I first described the male of H. longicornis, hitherto unknown, in 1902. Since
then I have found it frequently throughout the Atlantic, extending to the Antarctic
area. Dahl's description of H. mnjor is very scanty, and the only essential point of
difference between it and longicornis appears to be one of size. The specimen in the
' Discovery ' collection is a $ of 4 • 5 mm. length, but there is no essential difference
between it and smaller males from the Faroe Channel. I suggest therefore that Dahl's
H. major is really H. longicornis, and I now think that the species which I described in
" Plankton Studies " as H. major may best be described as //. longicornis (Giesbrecht).

The diagnostic points of H. major (Dahl) are, according to this author, " anterior
antennae very long, the posterior foot-jaw with only weak median bristles; the
penultimate lobe of the anterior foot-jaw a long ' tap ' lobe, the mandible teeth but
little differing in thickness, nearly the same distance apart ; the exopodite of the third
feet like those of the second and fourth, the size over 5 mm. long."

Except as to size, it will be observed that none of these points differ from those of
II. longicornis, the largest examples of which are not, however, more than 3 • 5 mm. long
in the North Atlantic.

The ' Discovery ' specimen is a $ of 4 • 5 mm. length, the anterior antennae several
joints (about six) longer than the whole body ; the geniculating antennae with six
joints beyond the elbow. The right furcal segment is much longer than the left. The
fifth lobe of the anterior foot-jaw has a very thick-based stout curved hook, without
teeth or bristles except for a few bristles at the proximal end ; the lobe itself is very
large. The two other bristles are long and thin. The sixth lobe has a long and thin
hook, also uncombed. The bristles of Ri are extremely long. The posterior foot-jaw
has a long thin second basal, three times as long as broad, and only two weak bristles
in the middle. The mandible has a large simple conical tooth on the outside, and

these out. i t.-.-tli an- n..t in cither mandible thickened. The thin! fail Nttmble tin-
fourth. The riv'lii fifth foot him a long upright procem on D 2, haired marginally ; the
Ri 2 hronil an. I with a marginal protuberance on which are two or three short teeth
and a -mall liuiu-h of hair-, flattened long spine distal to it Re 3 a long curved spoon -
shaped segm.-nt. with a stout-based apical spine, shorter spine on the inner aspect ;
the right Ri with the second segment elongated and narrow, the third segment
comparatively broad and short, the inner marginal bristle of Ri 2 thickened.

Tin- left foot has a haired marginal projection as Ji 2, Its'.} with a long stout
apical spine, three-quarters as long as the segment, and with a short marginal spine on
the inner side, Hi 2 broad, with thickened bristle. A specimen of //. imijicomi* from
the Faroe Channel measured 3*5 mm. long ; Ksterly records it from Diego, California,
3 mm. long. The Southern Ocean species evidently reaches a much greater length
(4*5 mm.).

FAROELLA (WOLFKNOKN).

In the course of my cruising in the Faroe Channel in 1901 I captured a
copepod which differed from any known species, to which I originally gave the name
Pfieitdaetidiw multiscrrata, in the paper read at the British Association, 1902. In 1903
appeared Bars' supplement, in which he described a new genus, AZtidiojtMi*, which
appeared to be the same animal ; and as I had already recognised that this copepod was
distinctive from others closely allied (PseuJcetitleu*, Chiridiii*, GaiJhi*), I had created
for it a new genus, Fart*-lla. My paper had been in the printers' hands for some time when
Professor SOTS' supplement appeared with the description of sKtiiliofuti*. Consequently
I do not know to which name priority should be given, nor do 1 feel yet certain that
the genus described briefly by Sara is identical with the FnrnfUa ilescriU-d by me in the
J. J/. B. AM. of 1904. Certainly the Farotlla of the Antarctic Sea has some dihVrrinT-,
and I therefore retain the name for the genus which I originally gave, more especially
as Professor Sars, who has examined some Irish specimens, states, as I am informed,
that they are not identical with his.

FAROELLA ANTARCTICA.
(Plate II., figs. 1, 2, 3, 4.)

? size 4*3 mm. (cephalothorax 3*3 mm., abdomen 1*0 mm.). The fore-body is
therefore over three times as long as the abdomen. The head and first thoracic segment
are united, the two last segments of the thorax separate, the anterior segment over twice
as long as the four last segments ; the most posterior of these is well defined from the
one in front, small, and laterally prolonged into stout spines which are about three-
quarters as long as the genital segment In dorsal aspect the head is rather triangular-
shaped, and on each side below the level of the posterior antennas, laterally expanded.
In the lateral aspect the head is evenly rounded, oval, and with stout two-pointed

40 R. NORRIS WOLFENDEN.

rostrum directed forwards, with slight curve downwards, the rostral spines not at all
divergent, as in Sars' picture of ^Etidiopsis. The whole cephalothorax is studded with
fine and closely-set prickles. Abdomen of four segments, slender, the genital only a
little larger than the next, with strong ventral protuberance, > Ab 2 > Ab 3 > Ab 4.
Furcal segments as long as the anal, and twice as long as broad. Tail bristles four ou
each side, with very short and delicate ventrally placed accessory bristles. Anterior
antennae reaching just beyond the end of the genital segment, the first two joints
comparatively large and as long as the next five joints, the combined eighth and ninth
joint as long as the two joints either proximal or distal to it, the eighteenth and nine-
teenth joints longer than those proximal or distal, and the twenty-fourth separate from
the twenty-fifth. All joints only sparingly setiferous.

Posterior antennae with exopodite a little longer only than the endopodite.

Anterior foot-jaws with the outer margin of the basal only a little convex, the
fifth lobe longer than the four proximal and nearly equally-sized lobes, the Ri small
but distinctly segmented. Each lobe with three bristles, two each on the first, second,
third, and one on the fourth, being stout, long, and with wide apart stiff marginal
bristles. Bristles of Ri long, not feathered, but slightly serrated marginally.

Posterior foot jaws having proportionately -B 1 : B2 : Ri = W : 12 : 5. J32
therefore a little longer than JSl, and over twice as long &sRi; B 1 two and a half
times as long as broad, with two small lobes with respectively two and three short
bristles ; B 2 four times as long as broad, its marginal bristles very small and distal of
the middle. Ri short and distinctly five-segmented.

Maxillae, Le 1 with nine bristles and its outer margin nearly straight ; B 2 with
five, and not segmented from Ri with thirteen bristles ; Le 2 a small lobe, but without
bristles ; Re small, longer than broad, and with ten bristles ; Li 1 with nine large hooks
and four bristles ; Li 2 and 3 well-formed lobes.

First feet. Ri one-jointed ; Re three-jointed, with three long thin marginal
spines.

Second feet. Ri two-jointed, Ri 1 short, Ri 2 very elongated, and nearly four
times as long as Ri 1. The whole Ri only extends to the distal end of Re 2. On the
distal part of the surface of Ri 2 is a bunch of fine hairs ; Re 3 is as long as Re 1 + 2,
and its end saw is longer than the Re 3 and beset with a great number (exceeding
fifty) of closely-set teeth, of which those in the middle are the largest.

Fourth feet. Ri and Re of three segments each. Ri proportionately longer than
in the other feet, and the third segment as long as Ri 1 + 2, and a little over three
times as long as broad, with fine hairs on the surface distally. Re 3 much longer than
Re 1 + 2 and over three times as long as broad. Its end saw three-quarters as long
as the Re 3, and with over fifty closely-set teeth. No fifth feet.

The chief points in which this Antarctic species differs from that of the northern
seas are in its greater size, the greater strength of the rostrum, the rather different
body proportions, and the more equal size of the rami of the posterior antennae. The

41

chititi every when- in tlir < cphalothorax is almost covered with prickles, and the
whole animal is more robust

MICBOCALANUS (SAR.S).

MlCROCALANUS PU8ILLU&

(Plate II., fig. 5.)

Micntalmu* jnuXlvt, 0. 0. Sara, CnwUc. of Norway, IV. (1908), p. 15(1.
PHHimlanttM pyymtmt, San, Norwegn. N. Polar Expdn., Vol. V. Crustacea, 1900.
„ „ Gioibnvht, ' Belgica ' K.-|*>rt, p. 20.

Sars originally descritad a small Calanoid, PieiulocalanuA pyynueux, from Nunsen's
Polar Expedition, which he subsequently re-named Mlcrncalainis, and included in the
new genus a second and still smaller form of M. jnisillu*. Giesbrecht described a .small
Calanoid from the Antarctic seas, which agreed generally with Bars'. Polar species, except
for very small differences, e.g. the length of the anterior antennae, and the length of the
outer marginal spines of the exopodites of the feet. The size varied, mostly from
0' 7-0 '75 mm. ; some were even smaller. P. pyymseu*, Sars (= M. pygmaus, San*), i«
a little longer, viz. ' 86 mm. (Sars).

The ' Discovery ' collection contains many examples of a very small Calanoid
which agrees so closely with Sars' M. pusilltut, that I regard them as identical ; and
Giesbrecht's Pseudocalanus pyymsews must, I think, l>e also regarded as identical. This
small Calanoid is one of the few examples of complete agreement in form and structure
between the Polar and Antarctic forms, and on that ground is of interest.

The female is O'GO mm. long. Cephalothorax two and a half times longer than
the abdomen ; the head and first segment united, the former evenly rounded and with
short, delicate rostrum ; the greatest breadth not quite half the length ; the last
thoracic segment with rounded and only slightly produced margins ; abdomen of four
segments. The genital segment is nearly twice as long as the next, which is rather
larger than the distal segment, and the anal as long as the segment preceding it.
Furcal segments as long as the anal and longer than broad, with four short terminal
bristles. The genital segment is very swollen laterally, but symmetrical, and rather
tumid vi-nt rally.

Anterior antennae reaching about the end of the genital segment, and of twenty-
three joints.

Posterior antennae with Re about one-third longer than the /•'/'.

Mandibles with Ri much longer than Re, both rami fully segmented.
Masticatory plate with strong cutting teeth seven or eight in number, and distally
nearly half as broad as long.

Anterior foot jaws with five well-formed lobes, of Calami* type, with well-
segmented Ri.

42 R. NORRIS WOLFENDEN.

Posterior foot jaws with segments proportionately Bl, B'2, Ri = 9, 8, 9. Ri
elongated and thin, with five distinct segments.

First feet, Ri of one, Re of three segments, the first without Se ; Ri with four Si.

Second feet, Ri of two, Re of three segments ; no Si on B 1 or B 2.

Fourth feet, Ri and Re of three segments each. The end saw extremely long, and
longer than the whole Re, broad and coarsely serrated marginally.

In the second to the fourth feet the Re 3 has three outer marginal spines. No
fifth feet.

The few males present were apparently immature.

HALOPTILUS (GIESBRECHT).

One species of this genus appears to be characteristic of the Antarctic fauna, since
it is present in very many of the ' Gauss ' samples, but curiously enough, occurred only
once in the ' Discovery ' collection. It is large, and distinguished at a glance from any
other examples of the genus by the prominent black ocellus dorsally placed, an organ
not possessed by any other species of Haloptilus.

HALOPTILUS OCELLATUS.
(Plate III., figs. 1, 2.)
Haloptilus ocellatus, Wolfenden, Plankton Studies, Part I. (1905), p. 14.

9, of length, from the end of the frontal spine to the end of the furca, 875 mm.,
with cephalothorax over five times as long as the abdomen ; the conjoined head and
first segment much longer than the remaining segments of the anterior body (about
one- third) ; the last two segments of the cephalothorax united, and with rounded
margins. On the second segment, in the centre of the dorsum, or a little to the right
of the centre, is a prominent and very black rounded pigmented ocellus, standing out
in clear contrast to the rest of the very transparent animal. The frontal spine is long,
tapering, and usually a little curved downward, and often laterally, towards the tip.
The distance from the tip of the spine to the base of the anterior antennae is equal to
the distance between the latter and nearly to the distal end of the second cephalic
segment. Abdomen of four segments, with the anal as long as the two preceding, and
the furcal segments over twice as long as broad.

Anterior antennae a little longer than the whole animal, reaching beyond the furca
by about three or four joints, and sparingly setiferous.

Posterior antennae with endopodite very long and exopodite very short, the latter
of six joints, with doubtful division of the last, which would be the seventh joint, the
basal or first segment very long, and nearly as long as the joints distal to it. lie not

COPEPODA. 43

more than one quarter as long as /?; 1. AV 1 elongated and seven or eight times M
long as broad.

Anterior foot jaws with a rather stout, but unarmed hook on the fifth lobe, not
longer, however, than the other bristles.

Posterior foot jaws thick, with Ri of similar thickness to D 2, and of five
segments ; the five stout curved hook bristles of nearly equal length, the two terminal
only a little the longest and thickest

Mandibles with Ri very long and Re only as long as Ri 1 ; masticatory plate with
outer stout, brond-liased, conieal and curved tooth; three pointed abort teeth internal
t«« it, rather like //. mucranatu*.

Maxillae. — First inner lobe with six bristles, of which only two of the distal ones
arc stout hooks ; second inner loin; with one stout long bristle ; third inner lol>c with
one stout elongated and two short thin bristles ; B 2 about as broad as long, with four
elongated and thick bristles and one thin, short proximal bristle, Ifi longer than broad,
and about three-quarters as long as Z?2 and only half its width, and with five bristles ;
Rf very long, twice as long as Ri + Ji2, and nearly twice as long as broad, with eleven
bristles, of which the three innermost arc short and thin.

All feet with three jointed rami, Re of fifth pair only five-sevenths as long as lie
of fourth pair ; AV of fifth pair only as long as Re. 1 + 2 ; Re 3 longer than Re 1+2,
twice as long as broad, with three inner bristles, two outer spines, and end spine nearly
as long as the last segment. I have not yet seen the <J of this species.

OITHONA (BAIRD).

Two species of this genus occur in these collections, one of which, viz., Oithona
fiimilu, is of world-wide distribution, and occurs with great frequency in Antarctic
collections ; the other, to which the name Oitfuma frtgitla has been given by
Qieabneht (and which has been fully described by him in the ' Belgica ' report, v'ule
ante), occurs very sparingly in the ' Discovery ' collection.

HARPACTICU8 (DANA).
HARPACTICUS FURCIFER.

/farjxiftieus furrifer, Oicsbrecht, • Belgica ' Report, p. 37.

The 9 of this species was first described by Giesbrecht in the ' Belgica ' report ;
and in the ' Discovery ' collection, marked 4. i. 02. W.Q., occurred three specimens of
the male, though female examples were conspicuously absent

The 9 , according to Giesbrecht, is 1 '55 mm. long ; the rostrum small, fhe
scries of points on the abdominal segments not numerous, the furca as long as both last
abdominal segments, narrowing distally and alx>ut three times as long as broad ; the

VOL. IV.

,1

44 R. NORRIS WOLFENDEN.

anterior antennae nine-jointed, the exopodite of the posterior antennae like //. chelifer,
but smaller, the second basal of the mandible like H. brevicornis ( = H. fulvu*), and the
exopodite scarcely half as long as the endopodite ; both rami of the maxillae are about
equal ; the first lobe of the anterior foot jaws has three bristles, the fourth lobe is long,
and its hook short, the posterior foot jaw is much thinner and weaker than in //. chelifer
and //. brevicornis and more like H. flexus. The first feet have thin and weak terminal
claws, both rami of only two segments, and the endopodite is short, the joints of both
branches being broader than in flexus ; the endopodites of the second and fourth feet
are larger in proportion to the exopodites than in chelifer and brevicornis, and in the
fourth pair reach to the middle of the last joint of the exopodite ; the bristles on
the second endopodite joint are, however, two, instead of one as in chelifer ; the last
joint of the fifth feet is comparatively small, and is scarcely half so broad as this ;
its last joint has five, the process of the basal joint, four bristles.

The striking feature of H. furcifer is the length of the furcal segments, which are
usually very short in this genus, and though related to //. flexus, it differs in the size,
which, in the latter species, is only • 64 mm. in length, compared with 1 • 5 in
H. furcifer.

45

I XPLANATION OF THE PLATES.

PLATE I.

Calm*u pnpitHfinu, fig. 1. Whole animal, 9 . Oc. 8, obj. 2 in.
fig. 2. Whole animal, $ . Oc. 8, obj. 2 in.

» ., fig. 8 and Sa, $ . 5th pair of feet Oc. 8, obj. } in.

fig. 4. 9 . Ban! joint of 5th foot Oc. 8, obj. \ in.
simillimui, fig. 5. Whole animal, 9 . Oc. 8, obj. 2 in.
» n fig. 6. Ban! joint of 5th foot. Oc. 8, obj. J in.

tonnu, fig. 7. Whole animal. 9 . Oc. 8, obj. 1 in.
„ » fig. 8. Bawl joinU of 5th foot, 9 . Oc. 3, obj. | in.

arutu*, fig. 9. Whole animal, 9 . Oc. 3, obj. 2 in.

,, fig. 10. Batul joints of 5lh foot, 9 . Oc. 3, obj. J in.

PLATE II.

FantUa antarttifa, fig. i.
fig. 2.
fig. *.

n „ |f. 4,

JUirrofolaniui piitil/u*, fig. 5.
lltiinralaniu yrandif, fig. 6.

9 , whole animal. Oc. 3, obj. 2 in.
9 , Posterior foot-jaw. Oc. 3, obj. 1 in.
9 , lit foot. Oc. 3, obj. 1 in.
9 , 2nd foot Oc. 8, obj. 1 in.
9 , wliole animal. Oc. 8, obj. \ in.
9 , whole animal. Oc. 8, obj. 2 in.

PLATE III.

Haloptihu ortllaitu, fig. l.

fig. 2.

Mrtridia print«p», fig. 8.
fig. 4.

» n fig. 5.

Gotten tu antarcfieiu, fig. 6.

9 , whole animal. Oc. 8, obj. 2 in.
5th foot, 9 . Oc. 3, obj. 1 in.
Whole animal, 9 • Oc. 3, obj. 2 in.
5th pair of feet, 9 . Oc. 8, obj. } in.
2nd pair of feet, 9 . Oc. 3, obj. 1 in.
9 , whole animal. Oc. 8, obj. 2 in.

Kiifhmta fimilit, fig. 1.
fig- 2.
fig. 3.

„ anlarrtifa, fig. 5.

n ,, fig. 6.

PLATE IV.

9 , whole animal. Oc. 8, obj. 2 in.

9 , l»t foot Oc. 8, obj. 1 in.

9 , 2nd foot Oc. 8, obj. 1 in.

9 , aUloiwn and genital segment Oc. 8, obj. 2 in.

9 , abdomen and genital Rcgment. Oc. 3, obj. 2 in.

9 , abdomen, lateral view. Oc. 3, obj. 2 in.

P 2

46

R. NORRIS WOLFENDEN.

PLATE V.

Stephus longipes, fig. 1. Whole animal, ,$ . Oc. 3, obj. £ in.

„ „ fig. 2. Whole animal, 9 • Oc. 3, obj. J in.

„ „ fig. 3. Whole animal, 9 • Oc. 3, obj. £ in.

„ antarclicum, fig. 4. Whole animal, 9 • Oc. 3, obj. 1 in.

„ „ fig. 5. Whole animal, 9 > dorsal. Oc. 3, obj. 1 in.

„ „ fig. 6. 5th feet, 9 . Oc. 3, obj. J in.

„ „ figs. 7, 8. 5th feet, j . Oc. 3, obj. i in.

Faralabidocera hodysoni, fig. 1.

fig. 2.

fig. 3.

fig. 4.
.. .» fig- 5.

fig. 6.

fig. 7.
,, ,, fig- 8.

PLATE VI.

9 , last thoracic segment and abdomen. Oc. 3, obj. | in.

Whole animal, 9 • Oc. 3, obj. J in.

5th foot, 9 • Oc. 3, obj. £ in.

4th foot, 9 > exopodite. Oc. 3, obj. J in.

1st foot, 9 • Oc. 3, obj. i in.

2nd foot, 9 . Oc. 3, obj. £ in.

Posterior antennae, 9 - Oc. 3, obj. £ in.

Maxilla, 9 . Oc. 3, obj. £ in.
fig. 9. Mandible, 9 . Oc. 3, obj. £ in.
fig. 10. Anterior foot-jaw, 9 • Oc. 3, obj. £ in.
fig. 11. Terminal lobes of post-footjaw, 9 . Oc. 3, obj. ^ in.
fig. 12. Whole animal, $ . Oc. 3, obj. i in.
fig. 13. 5th feet, 9 . Oc. 3, obj. \ in.

Xanthocalanus magnus,

PLATE VII.

fig. 1. W.bole animal. Oc. 8, obj. 2 in.

fig. 2. Rostrum. Oc. 3, obj. 1 in.

fig. 3. Posterior foot-jaw. Oc. 3, obj. t in.

fig. 4. Anterior foot-jaw. Oc. 3, obj. £ in.

fig. 5. Posterior antenna. Oc. 3, obj. 1 in.

fig. 6. Maxilla. Oc. 3, obj. £ in.

fig. 7. 1st foot. Oc. 3, obj. 1 in.

fig. 8. 2nd foot. Oc. 3, obj. 1 in.

fig. 9. 5th foot. Oc. 8, obj. \ in.
Oc. 3, obj. \ in.

antarcticus, fig. 10. 5th foot.

fig. 11. Posterior foot-jaw. Oc. 3, obj. \ in.

6

Antarctic I Discovery )]K\p.

Copepoda pi. I.

C»l*nin propinquut (1—4).
.. limillimm (5. 6).
„ tonsu> (7, 8).
, acuto* (9. to)

Antarctic (Discovery) Exp.

Copepoda pi. II.

Karoella amarctica (1-4) Mtcmcalanns posillu* (3)
Khioocalanus gnutdlt (6).

OF
IV

OF

Antarctic (Discovery) Exp.

Copepoda pi. III.

Haloptilua ocelUlui (i. a).
Mnrxlia princep* (j — )).
(•actanu« antarrticu* (6)

^i ymA Jfp=

OF THE

UNIVERSITY

9

a

Antarctic (Discover)-) Exp.

Copepoda pi. IV

Kuchacu simile ii

antarctic* (j. 6)

OF THE

I UNIVERSITY

OF

Antarctic (Discovery) Exp.

Copepoda pi. V.

SMphut lonjcipa (i— 3).

aaurcticurn (4 — 8).

OF THE

UNIVERSITY

OF

Antarctic (Discover)-) K\p.

Copcpoda pi. VI.
Pandabidocera.

Antarctic (Discovery) K\p.

Cope|x>di pi. VII

Xanthocalanus magnus (1—9).

antarcticus (10. n)

ECHINODEBMA.
I.

By F. JKKFRKY BKLL, M.A.

Emeritus Professor and Fellow of King's College ( University of Lotulon).

(5 Plates.)

HAD it not been for the earlier return of the 'Gauss,' this report would probably
have hud the distinction of signalising the re-discovery of the interesting Crinoid
genus, Prontachocrinu.*, which, as its name denotes, was one of the prizes of the voyage
of the ' Challenger.'

As in the collection of the ' Southern Cross,' I am again able to will attention to
some remarkable variations within what arc obviously the limits of single species.
The specimens which exhibit these variations could not have been obtained but
by very careful collecting, and in the case of Cycethra it is of importance to note
that they were all taken in a comparatively small area ; the examples of this genus
received by two naturalists, who made a numl>er of species with them, were all taken
from stations comparatively close to one another,* but it is to be hoped that such a
course will never be taken again ; evidence as to the variability of species of
Echinoderms is now beyond question.

A.-ANACTINOGONLDIATA.

I. HOLOTHURIOIDEA.

The collection of Holothurians is small, and the points of greatest interest are
revealed by Prof. MacBride and Mr. Simpson in their valuable report on the larvae,
which follows this memoir.

CHIRIDOTA.

I submitted an example of a form taken at 100 fms., off Coulman Island, to
Prof. Ludwig, who has made Antarctic Synaptids one of his special domains. Owing,
apparently, to the unfortunate use of formol the spicules are so disintegrated that a
definite judgment is impossible, but it seems probable that the species Is C. pisanii or
allied thereto.

• It U true that C. timplejc was found at Trinidad Channel, and the • Challenger ' trpeciniem on the eastern
ride of Patagonia ; bat he who will look at a map of South America will xinile at a " Chorological Hynopnu of
the specie* " which give* three to the Atlantic and one to the Pacific.

F. JEFFREY BELL.

CUCDMAKIA CROCEA.

Holothuriti (Cucumaria) crocea, Lesson, Cent. Zool. (1832), p. 153, pi. lii. fig. 1.

Cucumaria crocea* Ludwig, Hamb. Magalh. Sammelreise, Holothurien (1898), p. 15 ibique citata.

In the succeeding memoir by Prof. MacBride and Mr. Simpson there will be
found an interesting account of the brood-pouches and young of this form. It was
taken at Coulman Island, 100 fms., and on various dates at Winter Quarters down
to 41 fms.

CUCUMARIA LAEVIGATA.

Pentactella laevigata, Verrill, Bull. U.S. Nat. Mus. i. 3 (1876), p. 76.
Cucumaria laevigata, Ludwig, op. cit., p. 32 ifrique citata.

The characters of these two species have been fully discussed by Prof. Ludwig,
and it is quite unnecessary for me to enter again on the subject.
Taken in Winter Quarters to 41 fms. and in McMurdo Bay.

Several distorted Cucumariae and some young were taken at various dates while
at Winter Quarters, and at 78° 25' 40" S., 185° 39' 06" E.

PSEUDOPSOLUS FERRARI.

[ (Plate V., fig. 3.)

It is unfortunate that this new species should be represented by a single specimen
only, but there is no doubt that it is unlike any form that has as yet been described.

Tentacles ten, two much smaller than the rest ; trivial pedicels in three complete
longitudinal series ; dorsal perisome devoid of large scales or grains, flexible and soft
to the touch, and having only a few trellis-like spicules. There are some distinct
pedicels scattered on the dorsal surface. 73 mm. long, 32 '5 mm. greatest width.

The important differences between this species and the only form belonging to the
genus which has yet been described are of varying value ; the distinctness of the dorsal
pedicels is of great significance, for their presence in P. macquariensia is so inconspicuous
that Prof. Dendy denied their existence, on which Prof. Ludwig has remarked
" Zu einer Entschuldiguug will ich aber gleich hinzufiigen, dass die dorsalen Fiisschen
sehr gering an Zahl und sehr retraktil sind, sodass sie sich der Beobachtung leicht
entziehen kb'nnen." f

The spicules are more delicate and trellis-like in the new species, and there is a

* In my ' Southern Cross' Report I cite Wyv. Thomson as the first author of the combination " Cucumaria
crocea " ; he wrote Cladodactyla crocea in 1876 (not 1878, and not on p. 65) ; as all my blunders were made by
Dr. Lampert in his " Seewalzen," I stand convicted of neglecting the legendary advice of the venerable gentleman
who was once President of my College at Oxford.

t Hamburger Magalh. Sammelreise Holothurien (1898), p. 49.

BCHINODERMA. 3

very considerable difference in size. AH there in but a single specimen, I thought
the question whether or no it is hermaphrodite might )>e left over for the present

The only suggestion that I have to make is that Prof. Ludwig's definition of the
genus (<y>. cit., p. 50) should be altered by the omission of the words " Sehr wcnigen,"
us applied to the dorsal " FUsschen."

As the type of the genus came from Macquarie Island, the two are not widely
separated ; no information is given as to the depth of the more northern species ; the
Antarctic form was taken at 20 fms., while at Winter Quarters.

THYONK sp.
A young specimen was taken off Cape Adare.

MRSOTHURIA MAOKU.ANI.

Holathuria magtllani, Ludwitr. Bcr. Oberheas. (ten. xxii. (1883), p. !"'».

Mfsothnria magtllani, Ot«tcrgrcn, Fintskrift for Lilljeborg (1890), p. 850; Ludwig, Ergebn. Hamburg.
Sammelreiae Holothnr. (18U8), p. 8.

Two much-injured specimens are assigned to this species ; I was unable to find
spicules, but I do not always expect to find specimens taken out of ice-cold water and
from 100 fms. equal to the Ixjst products of the lalxiratories of Naples or Plymouth.

Coulman Island, 100 fms.

B.-ACTINOGONIDIATA .

IL CRINOIDEA.
COMATDLIDAK.

PROMACHOCRINUS KERGUELENEXSIS.

(Plate I.)
Promaehocnnus kergutlenxi*, P. H. Carpenter, Proc. Roy. Soc. zzviii. (1879), p. 885.

This is certainly one of the most interesting finds of the ' Discovery,' for the genus
was, till the German South Polar Expedition brought l>ack specimens, only known from
the collections of the ' Challenger,' which found two species in the Southern Seas, and
one a few degrees north of the Equator. The only addition to our knowledge since
Dr. Carpenter's report has been the note by Dr. Minckert of Greifswald,* while Dr.
Bather has suggested that it may be a " permanent meristic variation."! I give a figure

• Zoo]. Anzeig. xxviii. (1906), p. 400.
t TreatUe on Zoology iii. (1900), p. 195.

F. JEFFREY BELL.

of a self-dissected specimen which is, I think, clearer and more illuminating than the
originals of the ' Challenger ' report. It would be of particular interest to discover
fossil examples of this ten-rayed form ; at present, all we can say as to the
" permanency " of the form is that it has endured for thirty years ; as to the
polyphyletic origin of the genus which Dr. Bather suggests,* we can only say that
we have only such data as are given by the distribution of the genus. Though
Dr. Minckert had made two genera, he has not contributed to the elucidation of
Dr. Bather's suggestion.

Taken at East end of Barrier, 100 fms. ; off Coulman Island, 100 fms. ; and at
Winter Quarters.

A?TTEDON ADRIANI.

(Plate II.)

This new species is represented by a number of specimens ; as I have already
named a member of the genus A. wilsoni, I take the specific name from the second
Christian name of Mr. E. A. Wilson, M.B., surgeon and naturalist to the Expedition.

It was taken at various depths, down to 130 fms., at Winter Quarters, and at
500 fms. off Mts. Erebus and Terror.

At first sight this species does not recall the northern A. eschrichti, but it shows,
on analysis, some striking points of resemblance — the shape of the centrodorsal, the
length (but not the number of the cirri), and the shapes of their joints, as well as
the form of the first brachial ; but what distinguishes the creature from all Antedons
known to me is the saw-like appearance of the arms when viewed from the side,
and well shown in the figure ; this is due to the presence of a well-marked square
protuberance on the dorsal face of most of the arm -joints.

Arms ten, centrodorsal semicircular, with three rows of cirri ; these are from
50-60 in number, may be 60 mm. long, and may have as many as 60 joints ; the
axillary has a prominent knob, and the outer edge of the first brachial is twice as
long as the inner ; the arm-joints are stout, laterally compressed, and nearly all have
one or two dorsal spinous processes ; syzygies are extraordinarily rare in the
proximal part of the arm, where, too, the pinnules are rather short ; further out
they are better developed. No specimen is anything like perfect

ANTEDON ANTARCTICA.
Antedon antarctica, P. H. Carpenter, Chall. Rep. Comatulae (1888), p. 144.

I have compared a single specimen taken at Winter Quarters with those
collected by the ' Challenger ' at Heard Island, and have no doubt of their identity.
Some young specimens, which appear to belong to this species, were also taken.

* Treatise on Zoology iii. (1900), p. 195.

l.< IUNODERMA. 5

II L ECHINOIDEA.

ClDARIDAB.

ClDARIS CANALICULATA.

Ttmnofidari* eanalieulata, A. Agaasiz, Bull. M.C.Z. i. (1868), p. 18.

Goniocxhirt* tanatietiiata, itl. Rev. Bchin. (1872), p. 131 ; Wyv. Thomson, Journ. Linn. Roc. xiii. (1876),

p. <'•'• ; LOTCII, Bib. Svnisk. AluuK Udlg. xiii. iv., 1. p. 5 ; Again., Mrin. M.C.Z. xxxi. (1'JOt), p. 4.
OUtrit (Doroeulari*) ranalieulata, I)6derlcin, Jup. Secigel, i. (1887), p. 10.

Cidaru canalirulata, Mi-inner, Ergebn. Hamb. Mugalh. Sainmelreisc, v. (1900), 1., p. :i ibiqvt ritaia.
Sttrtofidarit canaticulaia, Mortciiacn, Ingolf Echinoid. (1903), p. 29.

I have given the name of C. eanalieulata to a number of specimens of an
Echinoid, which were mainly collected at a depth of 100 I'm-., but I must own to
grave doubts as to the correctness of the name. It seems to have escaped notice
that this consensu omnium (with the exception of Dr. Mortensen*) circumpolnr
Antarctic form was first descril>ed from the " Caroline Islands," which Caroline Islands
we surmise to be those- in the Pacific, as in the Revision of the Echini we find
M Caroline and Sandwich Islands " ; elsewhere, Zanzibar and the Navigator Islands arc
given as habitats, "if the localities are to l>e trusted" ; that is to say, the species
was founded on specimens said to IT found within the tropics.

I am not going to join those who claim that forms must differ specifically, either
because they are found at different spoto or different depths ; but, as I showed ninny
years ago, there is a distinct intertropical fauna in the Great Ocean ,f and the fact, if
fact it be, that a member of that fauna is also a circumpolar Antarctic form ought to
rest on the firmest possible basis. Unfortunately, the National Collection has no
examples from any station further north than Tasmania.

Again, the original diagnosis, based on the Caroline specimen or specimens, is too
short for a form which every student who has examined it, except Dr. Mortensen, allows to
be eminently variable ; with the exception of Prof. Doderlein, none of these students has
given a serious diagnosis of the species, and even his is not altogether what one wants.

Yet another difficulty remains to be mentioned ; it is generally agreed that
C. nutrix J and C. vivipara are synonyms of C. eanalieulata ; in other words, the form
has a marsupial habit, but I cannot detect signs of it in the specimens l>eforc me ;
it may be, of course, as it curiously is in the case of Hemiaster caoernosu*, that this
collection consists of males only ; but Mr. Hodgson tells me that he observed no signs
of viviparous habit; our experience, in fact, is the same as that of the late I'Vlix
Bernard § : " Quoique G. eanalieulata soit signal^ comme vivipare, je n'ai pas constate
aucun jeune sur le corps des adultes."

• Op.cit., p. 27.

t For tome modification of this doctrine, we Prof. Koehlcr in " Die Fauna 8udwe»t-Au*tralien» " I. IT.
(1907), p. 242.

J Prof. Lyman Clark doubU this (see f.-n. next page) ; of C. vivipara of Studer he makes no mention.
f Boll. Mas. Parix, i. (1896), p. 272.

VOL. IT. Q

6 F. JEFFREY BELL.

Specimens were taken at Coulman Island, 100 fms. ; Winter Quarters, 100 fms. ;
East end of Barrier, 100 fms.

This species is one that has lately been the subject of some dispute ; Prof. Agassiz,*
who originally described it, suggests that it be placed in a new sub-genus Centrocidaris,
while Dr. Mortensen f finds it to consist of several species, but he does not appear to
be sure of the exact number. J

ECHINIDAE.

ECHINUS MARGARITACEUS.

Echinus margaritaceus, Lamk. An. s. V. iii. (1816), p. 47; Bell, Coll. 'Southern Cross' (1002), p. 219

ibique citato,; Mortensen, Ingolf Echinoid. (1903), p. 101.

Echinus diadema, Studer, MB. Akad. Berlin (1876), p. 456 ; AI. Ag. Chall. Rep. Echin. (1881), p. 117.
Echinus horridus, Al. Ag. op. cit. p. 117 ; Mortcnsen, op. tit. p. 102.
Sterechinus antarcticus, Koehler, Echiu. Voy. 'Belgica' (1901), p. 8.

This appears to be a circumpolar species, as I remarked when I reported on the
collections of the ' Southern Cross.' It was taken by the ' Discovery,' not only at
several dates in Winter Quarters, but at the East end of the Barrier Reef, 100 fms. ;
at Cape Wadsworth, 8-10 fms. ; off Coulman Island, 100 fms. ; and South of
Antarctic Circle, 254 fms.

HRMIASTER CAVERNOSUS.

•

Spatangus (Tripylus) cavernosus, Philippi, Arch. f. Nat. xi. (1845), p. 345.

Hemiaster cavernosus, A. Ag. Rep. Ech. (1872), p. 132 ; Meissner, Ergebn. Hamburg Magalh. Sammelreise,

v. (1900), p. 13 ibique ci/ata; Bell, Coll. 'South. Cross' (1902), p. 219.
Abatus cavernosus, Loven, Bih. Svenska Vet. Akad. Hdlgr. xviii. 4, no. 1, p. 3.

This species is not very well represented in this collection, and, curiously enough,
all the examples are males.

If the late M. Bernard § was right in regarding Tripylus excavatm as a synonym,
the name of the species ought to be excavatus, as that was the first of the three species
described by Philippi ; for myself I am inclined to abide by Prof. Agassiz's view.

Taken at Coulman Island, 100 fms. ; Winter Quarters, 20 fms. ; and East end of*
Barrier Reef, 100 fms.

IV. ASTEROIDEA.

The collection of starfishes was rather large, and contained some very fine
specimens ; but, as will be shown later, there is very great difficulty in coming to
definite conclusions regarding them, and I have preferred to be vague rather than
dogmatic in treating of them.

* Mem. Mus. Comp. Zool., xxxi. (1904), p. 32. t Ingolf Echinoid. (1903), pp. 25-7.

J As the final proof was passing through the press I received Prof. Lyman Clark's latest memoir on the
Cidaridae (Bull. Mus. C. Z. li. (1907), no. 7) ; he has instituted a new genus, Austrocidaris.

§ Bull. Mus. Paris, i. (1895), p. 247.

I.. HINnDKK.M \ 7

AXTERIAS BRAN DTI.

Attrruu braiuiti. Bull, Proc. Zool. Hoc. (1881), p. 91.
Antiritu tuylttta, id., op. cit. pp. 94 und .".of,.

Dr. Mcissm-r has suggested that A. neglecla, which I descrilied at the same time
as A. brandti, is the same species. I think he is right; A. belli of Prof. Studer and
.-1. ffl&merata of Sladen appear to be also synonymous ; with regard to the first three-
Prof. Perrier suggested the same view in 1891.

My species were described on pp. 91 and 94 of the Proc. Zool. Soc. for 1881 ;
Dr. Moaner's synonymy will be found on p. 7 of the Ergeb. Hamburg Magalhan.
Snmmelreise, Lief. vii. (1904). I suggested in the report on the 'Southern Cross'
collections (p. 215) that the species was circumpolar in its distribution, and so it
certainly is.

A large number of specimens were collected; the largest, from 4-10 fms.
McMurdo Bay, was the " mother of a number of young after capture " * ; and some other
large specimens were taken at No. 3 hole, and some smaller at No. 11. Flagon Point
(10-20 fins.), Cape Wadsworth, and Hut Point also yielded examples ; the large;
examples marked 48 and 49 are only said to be from Winter Quarters.

From 100 fms. (East end of Barrier) and 130 fms. (No. 2) specimens were
collected which, but for the formidable synonymy of the species, I should be inclined to
regard as distinct

A.STERIAS LONOSTAFFI.

This fine species is represented by a single specimen, which presents well-marked
features. I name it after the generous friend of science whose contribution to the
funds of the Royal Geographical Society made the voyage of the ' Discovery '
possible.

A large stout form with R about equal to 3 r ; the whole of the dorsal surface is
uniformly and densely covered with paxilliform spines of some size. The madreporitc is
of moderate size ; the podia are enormous, the adambulacral spines are long,
cylindrical, blunt at the tip, so irregularly set that it is difficult to say whether there
are two or three rows ; beyond them there are shorter and more delicate, but otherwise
similar cylindrical spines very closely set ; there are two quite well-marked seta of
marginal plates ; the inferomarginals are very narrow in the angle of the arm, while
the superomarginals in the same region are very short ; both sets are covered with
coarse granules, among which no pedicellariae are to be detected. The podia within
the oral circlet are of enormous size. It = 210 ; r = 70.

W.Q. ; 10-20 fms.

•

As there is only a single specimen, no comparisons arc possible ; but it may be

* The»e have been made the subject of an intere«ting separata report by Prof. MacUridc and Mr. J. < '. Siiiipnon.

Q 2

F. JEFFREY BELL.

pointed out that the closely packed arrangement of the podia, and the distinctness of
the two rows of marginal plates are, on Mr. Sladen's basis of classification of Starfishes.
mutually destructive ; but the well-known observations of Prof. Ludwig on Echinaster
sepositus justify us in supposing that we have here a case of retarded disappearance of
the marginals ; the crowding of the podia appears to be a much more important
morphological character ; but the union of these two strikes, I think, a final blow at
the current classification, against which Prof. Ludwig has already raised his voice, and
the adoption of which by MM. Delage and Herouard in their " Zoologie Concrete "
came as a great surprise to me.

HEURESASTER* HODGSONI.
(Plate III.)

Two specimens of a very fine starfish were taken at about 25 fathoms while in
Winter Quarters ; they appear to me to form the type of a new genus, as to the
general position of which there can be little doubt ; a still larger specimen was taken
from McMurdo Bay at 2 fms.

It has somewhat the appearance of Porania, but has, in the larger examples, spines
in the interambulacral actinal areas. Prof. Perrier defines the Poraniidae as follows :
" Squelette masqu£ par les te'guments ; marginales apparentes, mais formant au corps
un bord tranchant ; squelette ventral forme de series de plaques allant de chaque
adambulacraire a une marginale, squelette dorsal reticule."

I have invented for it a name which will remind the student both of the name of
the ship, and of its indefatigable biologist.

The larger specimens may be thus described : Arms long, tapering to a rather fine
point, R is about = 3. The upper surface is smooth and soft to the touch, and has
papulae, in ill-defined areas, spread over the whole of it ; the edge is quite sharp,
forming almost a ledge, and made up by a large number of small supero- and infero-
marginals. The lower interambulacra covered with some eight rows of small regularly
set plates covered with rather coarse granules, and, in the angle, with short spines
which give a hairy appearance to these areas. The ambulacral groove is bounded by
rows of four or five spines, of which the outermost is small and the innermost spatulate
and fluted at its free end. The podia are stout. At each oral angle there is a huge
spine, the distal third of which is glossy. These specimens are flat, but the smallest
example has the disc arched, and this is probably more natural—

R = 200 r = 70.
R = 130 r = 50.
R = 90 r = 30.

Accepting Prof. Perrier's family diagnosis of the Poraniidae, we may make the
generic diagnosis of Heuresaster : Abactinal surface only invested by integument,

* EiptcK, discovery.

milNOHKUMA. 9

margin very sharp, the inner artiiml ambulacrals with spines, the oral angle provided
with a large spine. While the specific characters would appear to lie that the IOIIL'«T
nulius is nnm- than twice as long as the shorter ratlins, the marginal plates arc very
numerous, and the innermost amlmlacral spine is spatulatc and fluted at ita free end.

The smallest specimen has not quite acquired the generic characters of the larger ;
the abactinal integument is not so thick as to altogether hide the superficial granules,
and the granules on the actiual internmbulacra have not developed into spines, so that
t lii-re is no marked difference between the outer and the inner parts of these areas.

PKXTAOONASTKR INCERTTS.

The single small specimen, is, I think, an ally of the Australian species of Penta-
gonattter (sens, lot.), but the arms are proportionately longer than they generally arc
in this genus ; it is, possibly, an immature specimen in which It would gradually
increase in proportion to r. If it should prove to )>c an adult, its proportions may l»-
compared to those of P. duetx-ni and /'. yunni; it is, however, to IHJ distinguished by
the fact that there are no large plates on the actinal inter-radial areas, the plates being
of the character of, and a little larger than, the small squarish granular plates which
bound the marginals; these last number about 12/14 for the side of each arm, and
are completed by a large terminal ; there are two rows of well-developed spines at the
sides of the ambulacra ; those of the inner row are nearly twice as long and as
numerous as the outer. I propose to call this form Pentagonaster incertus ; it was
taken at 96-120 fins., in MacMurdo Bay.

LEPTOPTYCHASTER KERUUELENENSIS.

Lf/tfoptycfuuter kfrgtulenmsu, E. A. Smith, Phil. Trans. lf.8 (1879), p. i'7s. pi. xvii. 2; Sladen, Chall.

Rep. Ast. (I** li). p. 18t; Bell, Mar. Invert. 8. Africa iii. (l»0.r>) p. 2»L'.
Leptoptyfhaster antarctic UK, Sluden, op. eil. p. 190.

I must own to some temerity in associating a specimen in which R = 212 and
r = 58 with a species whose type had /? = 38 and /• = 12*5, and a representative
of which, hardly much larger, was found to be bearing young ; but even the most
recent writers on Echinoderras have not yet promulgated the doctrine that difference
in size is a specific character, though I am riot quite sure that in practice they do
not sometimes act as though they had. However, one has only to get a clear idea
of the essential characters of this genus to feel sure that one has it here ; as to spo« ifi.-
characters, it is first to l>e said that most of the L. kerguelenensi* material is badly
preserved, while the condition of L. antarcticus is particularly good. Though the
differences between the two species appear, from Mr. Sladen's lengthy description, to
!»• considerable, it will, I think, be found on examination of the specimens preserved
in the Museum, that L. antarcticus is but the expression of some early stages of
L. kergueleiicnsis. It will Ixj rememl>ered that both "species" come from closely
adjacent localities. At any rate, we now know that the specimens of L, antarcticu*

10 F. JEFFREY BELL.

are small, and that the known adults of the genus are large ; it is safe, therefore,
to conclude that this specific name may be made a synonym

Mr. Hodgson collected some quite young examples chiefly in W.Q., and, with
the specimens sent by the Government Biologist at the Cape of Good Hope, we
have now a very satisfactory series ; it was also taken at McMurdo Bay, 96-100 fms.

CYCETHRA VERRUCOSA.
(Plate V., fig. 1.)

Ooniodiscus verrucosus, Philippi, Arch. f. Nat. 1857, p. 132, teste Meissner, Zool. Anz. xxi. (1898), p. 394.
Cycelhra simplex, Bell, Proc. Zool. Soc. 1881, p. 96 ; id. Coll. 'Southern Cross,' 1902, p. 215 ibique dtata.

I was much to blame for neglecting or forgetting, in 1902, Dr. Meissner's very
useful note ; the alteration of the name is not, of course, of the least importance to any
sensible human being, but the knowledge that the species extends nearly as far north
as Valparaiso is of great interest and significance, for it shows us that the distribution
is far wider than we supposed ; had Dr. Leipolt (Zeit. f. wiss. Zool. lix. (1895) p. 602)
known of it, he would have spoken with less point than he did, when he doubted
Prof. Studer's locality of 38° 10' S.

The variations of this species are truly bewildering ; I have had an extreme form
figured, and offer a slight account of it ; it was long before I could convince myself
that it is C. verrucosa, and I am doubtful as to convincing others.

The upper surface is uniformly covered by delicate spines which end in a few, not
more than ten, paxilli ; the space between these spines is soft and membranous ; the
lower surface is densely covered with spines of moderate length and sharp at the tip ;
the spines bordering the ambulacral grooves are somewhat irregularly arranged, but are
longer and blunter than those that crowd the interambulaeral spaces ; there is a
single glossy spine at each angle of the mouth. The two rows of marginal plates are
well defined, and the paxilli are somewhat longer and more numerous than those on the
upper surface of the body. The madreporite is very prominent and is not far from
the centre ; close to this last there is a small tuft of white spines which, I suppose,
guard the anus. Colour, light creamy yellow ; body flat ; M = 55 ; r = 20.

Examples were taken on various dates at Winter Quarters ; off Castle Rock,
14 fms. ; Cape Wadsworth ; Coulman Island ; McMurdo Bay, 20 frns. It does not
appear to inhabit deep water.

HENRICIA ORNATA.

Echinaster (Cribella) ornatm, Perrier, Ann. Sci. Nat. xii. (1869), p. 251.
Henricia ornata, Bell, Mar. Invert. 8. Africa, iii., (1905) p. 250.

H. ornata is recorded by Sladeu from Campbell Island ; his H. simplex, which
is a synonym, was taken off the Crozets, Marion Island, etc.

One specimen was obtained off Coulman Island, 100 fms. ; one from 4-10 fms.
McMurdo Bay ; and three from 96-120 fms., in the same bay.

l.t'lliNODKKMA. 11

SMI \-iii: ... i. .i:\i-i \n v
Sfhuitr octortuliattu, Lodwig, Voy. ' Belgica,' Secrternc (1908), p. 25, pi. iii.

A single specimen of this species was taken at 100 fins., off Coulninn Island. As
the ' Belgica ' took it at four stations between 80° and 88° West, it is, perhaps, a
circumpolar species.

Two specimens of Starfish remain to be noticed, which I shall not uamc, as I
feel confident that they have undergone some change either in formalin or otherwise,
or are abnormal specimens which will not be again found ; one English naturalist has
so burdened the literature of Starfishes with generic names based on immature
specimens that I may be pardoned for not following in his footsteps.

Both of these specimens are very soft to the touch, owing to the numerous large
papulae and the apparent absence of any hard parts on the upper and lower surfaces
of the Starfish, with the exception of a large and prominent madreporite, and of a
few spiculcs round the vent ; the ambulacral grooves are wide, and the pod in of large
size ; there are two well-marked rows of spines, the more proximal of which are
blunter and more flattened than the more distal ; beyond the outer row there is some
slight difference in the two specimens, for in one the integument rapidly becomes
smooth, while in the other regular rows of spines may extend as far as the margin.
1 have not seen any pedicellariae.

The " register numbers " of these two specimens are : 1906. 1. 22 ; 2 & 3.

V. OPHIUROIDEA.

Though the collection of Brittlestars is large, there are not many species ; the
two new genera signalised by me in the report on the ' Southern Cross ' collection
are both well represented, and of Ophiosteira there is so remarkable a variety that I
am constrained to add another figure to those that I have already published. There
are many small immature forms which will be of much use in studying the evolution of
species, but to which it would be most unwise to give definite systematic names ; as it
is, I am perhaps a sinner, though not so great as some, for I should be hard put to it
to defend myself against the suggestion of MM. Delage and Hdrouard,* that Ophiocrene
is a young Aslrophyton.

OPHICRA KOEHLERI.

Like 0. flagellata of Lyman, in having the disc covered with smooth skin, in
which, in neither of the two specimens, are small scales distinctly visible ; the lower
arm spines are deeply imbedded in the skin, and thick skin obscures the forms of the

• Zool. Concrete iii. (1908), p. 159.

12 F. JEFFREY BELL.

oral plates. The diameter of disc is 27 or 20 mm. ; no arm complete ; width of arm at
base about 4 mm., upper arm plates broader than long ; about 7 arm spines, the upper
free and pointed at their free ends, the lower shorter, imbedded in thick skin, and
blunted at the tip ; lower arm plates with narrow distal and proximal edges. Arm
insertion of disc well marked and guarded by about 15 spines, of which the median are
distinctly the larger.

Colour in spirit : disc sickly white, arms more the colour of human flesh.

Taken at 67° 21' 46" S., 155° 21' 10" E. 254 fms.

I am sorry that both the specimens of this interesting species are a good deal
broken, as I should have liked to have had something better to offer to the honour of
the distinguished French naturalist who has done so much for our knowledge of
Ophiuroids.

OPHIOZONA INERMIS.
Ophiozona inermis, Bell, Rep. 'Southern Cross' (1902), p. 217.

This species, which was not well represented in the earlier collections, was found
on numerous dates during the stay of the ' Discovery ' at Winter Quarters.

It exhibits a very wide range of variation, so great indeed as to have been at first
quite bewildering ; and it may be noted that Mr. Hodgson did not recognise any of
them as belonging to an already known Antarctic form.

In the ' Southern Cross ' report I figured some of the remarkable variations in the
arrangement of the plates of the dorsal surface of the disc of Ophiosteira antarctica ;
a similar, but less striking, variation obtains in this Ophiozona. In 0. antarctica I
described the serrated and keeled appearance of the upper surface of the arm ; this is to
be seen also in some examples of 0. inermis. The most striking differences are to be
seen in the height of the disc, and the size of the so-called radial shields ; so that it
is by no means always true that the shields are inconspicuous.

Winter Quarters, 10-178 fms. (various dates) ; MacMurdo Bay, 96-120 fms. ;
East end of Barrier, 100 fms. ; off Barrier, 300 fms. ; Coulman Island, 100 fms.

In some specimens the dorsal plates of the disc are deeply incised ; this appears to
be due to want of sufficient calcareous matter, but as others are infested by a sponge it
is possible that it is the cause.

OPHIOSTEIRA ANTARCTICA.

(Plate V., fig. 2.)
Ophiosteira antarctica, Bell, Rep. ' Southern Cross' (1902), p. 218.

It appears that the ' Southern Cross ' did not exhaust the variations in the disc
plates of this wonderful species ; the upstanding plates shown in Plate V., fig. 2, are
worthy of record, and it may be noted that the single arm left to this disc has the
dorsal serration extremely well marked.

BCHINODERMA. 13

In the original diagnosis of the genus I spoke of "a large keel-like plate on the
disc, which completely overshadows the radial shields," although I WHS aware of and
figured (Plate XXVI., fig. 3) an angle of the disc in which there appeared to l>e two large
radial shields ; in this case, however, three of the angles of the disc conform exactly to
the diagnosis.

I have now before me a specimen in which all five angles have a pair of plates.
Were it not for the ' Southern Cross ' specimens it would not be possible to put this
specimen with the genus Ophiosteira at all.

When we are asked whether " specific characters are useful," we may retort that
generic characters even are not always constant.

Winter Quarters, 13 fms. Coulman Island, 100 fms.

•

OPHIONOTUS VICTORIAE.

Ophionotiit vicloriat, IU-11, Hop. ' Southern Cross ' (1902) p. 219.

This species was not so abundant ; it was taken of}' the Barrier at 300 fins., near

Franklin Island, and from 254 fms. at an unrecorded locality.

t

OPHIACANTHA IMAGO.

Ophiaf an/ha imago, Lyman, ' ('ball.' Rep. Ophiur. (1882), p. 186 ibi'jttf citatum.

I hope I am right in referring to this species specimens from Winter Quarters,
30 fms. ; Hut Point, 77° 12' 12" 8., 167° 27' W., 77° 50' 30" S., 165° 40' R, 100 fms. ;
but, in sooth; some of the ' Challenger ' types of Ophiuroids are hardly mature.

OPHIACANTHA VIVIPARA.

Ophiarantha vivipara, Ljungman, Ofr. Vet. Akad. Forh. 1870, p. 471 ; E. A. Smith, Phil. Trans. 1G8
(187«), p. 278, pi. xvii. fig. 8.

A numl •<•]• of authors have mentioned this species, but, since Dr. Liitken was cited
by Mr. Edgar Smith, I.S.O. (Phil. Trans., vol. 168, p. 278), as doubtful of Ljungman's
original locality (" Altatam urbem mexicanam "), none seems to have remarked on
the apparent peculiarity of the distribution of this species.

Prof. Theel has been so good as to let me see Ljungman's Altaian specimen, and
at the same time to inform me that the types of Ljungman's Ophiacantha vivipara were
brought home from two quite different localities, Altata and Falkland Islands.
Ljungman does not give (op. cit.) the latter locality, but Liitken (Zool. Record 1872,
p. 448), threw grave doubts on the Mexican origin of the specimens, and suggested that
Patagonia was the " true habitat of the typical specimen." What is really more
interesting is the question whether the brood-pouch habit of a given species is found
in the confines of the tropics as well as in the colder waters of the globe.*

• For a list of the viriparon* Echinodermi of wanner waters, and for their preponderance in Arctic and
Antarctic Seaa, <•/. Ludwig, Zool. JB. Snppl. Bd. vii. (1904), p. 684.

TOL. IT. R

14 F. JEFFREY BELL.

The single* six-armed specimen taken by the 'Discovery' at 8-15 fms. off Cape
Wadsworth is a good deal stouter than Ljungman's types, but I can see no reason for
making it a new species ; the arms of the young may be seen projecting from some of
the bursae.

OPHIACANTHA COSMICA.

Ophiacantha cosmica, Lyman, Chall. Rep. Ophiur. (1882), p. 194 ibique citatum.

This widely spread species was taken on various occasions, but many of the
specimens have been marked by me " immature " ; in connection with this I should
like to quote some words of Mr. Lyman : " The specimen just described is unusually
large for this division of the genus, which leads me to think that the bulk of those
now known are immature, and therefore to be treated with all the more caution."

Winter Quarters-. Off Coulman Island, 100 fms. Off Barrier, 300 fms.

OPHIOCONIS ANTARCTICA.
Ophiocoiiis antarctica, Lyman, Chall. Rep. Ophiur. (1882), p. 107 ibique citatum.

If I have correctly determined some specimens from Winter Quarters, the types
of Mr. Lyman 's are very far from reaching the size to which this form attains.

AMPHIURA BELGICAE.

Amphiura beli/icat, Koehler, Res. Voy. 'Belgica' (1901), p. 27.

The Antarctic area appears to be the home of large specimens of Amphiura ; not
only did the ' Discovery ' bring home several examples of what appear to be the
species obtained by the ' Belgica,' but there is a single specimen off the Barrier at
301 fms., which has a disc diameter of 18 mm.,f and which can hardly, I think, be
the fully adult example of the somewhat smaller specimens ; but, as it is solitary, I
hesitate to regard it as an undescribed species.

Several specimens of A. belgicae were taken at Cape Wadsworth, 8 to 15 fms., and
off the Barrier at -300 fms. ; the latter is, as already said, the locality of the single
larger form. Like Solaster octoradiatus, A. belgicae is probably a circumpolar species.

YOUNG OPHIURID.
(Plate IV.)

I give some figures of a remarkable Ophiuroid, to which it is very difficult to
assign a systematic place ; it appears to be a young form in which some of the arm-
plates are still not consolidated. The exceedingly large size of the bursal slits is
perhaps only due to the mode of preservation.

Two specimens, both broken, were taken in Winter Quarters on 2. 1. 04, and two
off Barrier.

* As there is but a single specimen this collection throws no critical light on Prof. Koehler's just-made
suggestion that this species exhibits sexual limorphism ; cf. Bull. Sci. France xli. (1907), p. 322.

t That is exactly the same as the disk diameter of A. magnified, of which Prof. Koehler has lately published
a brief diagnosis. See Zool. Anz. xxxii., p. 146 (17 Sept. 1907), but the species is not the same.

EOHINODERMA. 15

AsTKOTOMA AOASSIZI.
Antrotoma agattuu, Lynwn, III. Ciital. MUB. C. Z. No. viii. II. (1H75), p. 24.

Mr. Lynmn calls attention to the coarse granulation on his single specimen,
which came from the Straite of Magellan ; the series now collected shows that the
granulation may be more or less obscured by thick skin. It is remarkable that the
species was not collected by the ' Gauss.'

McMurdo Bay, 96-120 fins.

Specimens from off C'nulman Island, and from 77° 12' S., 167° 27' E., 2 fins, seem
to In- sutliciently different to IKS regarded as varieties.

R 2

16 P. JEFFKEY BELL.

DESCRIPTION OF PLATES.

ECHINODEKMA.
PLATE I.

Promacochrinus Tcerguelenensis.

FIG. 1. — Disc and base of arms from the side to show the insertion of the cirri, and the form of the

cirrus pit, and the shape of the disc, x 2.
FIG. 2. — Disc and base of arms from below ; the central boss is more distinctly seen than in fig. 1.

X 24.
FIG. 8. — An arm from the side, x 3 ; the roughened edges of the joints are well seen in this figure ; the

proximal pinnules only are completely shown.

PLATE II.

Antedon adriani.
General view of this Crinoid. x 2.

f-

PLATE III.

Heuresaster hodgsoni.

X not quite if, seen from below, to show the breadth of the anibulacral grooves, their size and
arrangement.

PLATE IV.

The young Ophiurid (see p. 14).
FIG. 1. — Aboral view of disc and arm. x f>
Fio. 2. — Oral view of same, x f .

FIG. 3. — View of mouth to show arrangement of the oral and adjacent plates, x 8.
FIG. 4. — 1st, 2nd, and 3rd upper arm-plates, x 10.
FIG. 5. — View of arm from above, at about its middle, x 10.
FIG. 6. — 1st, 2nd, and 3rd lower arm-plates, x 10.
FIG. 7. — View of arm from below at about its middle, x 10.

PLATE V.

FIG. la. — Variety of Cycethra (see p. 10) from above ; fig. Ib from below ; both slightly reduced.

FIG. 2a. — Disc of an Ophiosteira (see p. 12), magnified to show keel-like plates ; 2a, seen from above ;

2b, seen from the side.
FIG. 3. — Pseudopsolus ferrari. X f.

:

Antarctic (Discovery) Exp.

Echinoderma pi. I.

lw>)«>« 4.1 H«u*rwortk. tr

X 2

Antarctic (Discovery) lixp.

Echinoderma pi. II.

!>•')•••

•' OF THE

: UNIVERSITY

OF

Antarctic (Discovery) Exp.

Echinoderma pi. III.

» J B T.rtl a. I Imiur »orlh M.

OF
^41 -'

6

Antarctic (Discovery) K\p.

Kdiinoderma |>l. IV.

1«

AnUrctic (Discover)-) Exp.

Tint. 4*1. H«lh 4.1 fliitorank. M.

Echinoderma pi. V.

£ ECHINODERMA.

n.-ECHINODERM LARV^.

By E. W. MAcBwDE, M.A., D.Sc., F.R.S., Professor of Zoology, McGill University; and
J. C. SIMPSON, B.Sa, Demonstrator of Zoology, AfcGill University.

(1 Plate.)

THE collection of larval Echinoderms brought back by the National Antarctic Expedition,
when received by us from the authorities of the British Museum, was found to consist
of forty tubes of material collected during the years 1902-1904. Of these forty tubes,
however, by far the greater number contain nothing but early segmenting eggs showing
no distinctive features ; the present paper is, therefore, in reality based upon the
contents of eight tubes only.

A preliminary examination showed that, though the collection contained but
four species of Echinoderms, these differ widely from one another, there being one
representative of each of the four great classes of the sub-phylum Eleutherozoa. These
four species are : Asterias brandti Bell, Cucumaria crocea Lesson, a new Echino-
plutew, and an Ophiopluteus which may also be new.

At first sight the Aster ia* seemed to promise much, for there were two tubes
containing a couple of hundred specimens which had been l>orn after the capture of the
mother. More careful examination, however, disclosed the fact that they were all
of approximately the same age, only two closely related developmental stages being
represented. In the case of the Cucumaria, besides a number of new-born young, the
collection contained an adult female which was the mother of some of them. To this
specimen considerable interest attaches, from the fact that in it an unsuspected brood-
pouch containing a great number of embryos was found. Of the two plutei it need only
l>e said in this place that, since they are probably the first free-swimming Echinodenn
larvae found within the Antarctic Circle, they are not without considerable importance.

In the matter of preservation the material leaves much to be desired ; but when
one considers the difficulties of collecting under Antarctic conditions, it could hardly be
expected that this could be otherwise. Of necessity the only preservatives used were
formalin and spirit, and as neither of these is at all suitable to Echinoderm larvae, the
specimens show very little histological detail.

In examining the material it was found that the preservation of the plutei was
such as to make sectioning useless, so they were studied /// toto, first unstained, and
then stained in eosin and mounted in balsam. This method was found to give very
satisfactory results. The Asterias and Cucumaria, however, were treated quite
differently. Selected specimens of these were first carefully drawn at a magnification

2 E. w. MACBRIDE AND j. c. SIMPSON.

of 12£ diameters by the aid of the camera lucida, and then, after double imbedding in
celloidin and paraffin, were cut into serial sections and stained with Delafield's
hsematoxylin.

In the succeeding pages the four species will be taken up and described in detail.

1. ASTERIAS I5RANDTI.
Asterias brandti, Bell, P.Z.S., 1881, pp. 94 and 506.

The young of this species contained in the collection consist in a brood " born
after the capture of the mother," which was taken in McMurdo Bay, February 13th,
1902.

The parent was identified by Professor F. Jeffrey Bell as A. brandti, but as it was
not sent with the embryos, nothing definite can be said as to its " brood-care," except
that this species must now be added to the list of viviparous Asteroidea. It is
probable, however, that according to the analogy of its family, the young are attached
to the lips of the mother during the earlier stages of development.

All the embryos of this brood are of approximately the same stage ; the difference
between the youngest and oldest being simply a matter of size. The youngest is
shown in Fig. 1 , which gives a ventral view. It measures 2 • 4 mm. from the centre of
the disc to the tip of the arms, while the radius of the disc is 1 • 5 mm. (R. = 2 • 4 mm.,
r. = 1*5 mm.). The only external trace of the embryonic condition is in the closed
mouth and the small pre-oral lobe or larval organ. The oldest embryo is much larger
than the preceding (R. = 3 ' 2 mm., r. = 1 • 6 mm.), but apart from the smaller size of the
pre-oral lobe there is little difference. Internally, too, there is little difference in the
development of the organs, so that they may very well be described together.

As may be seen from the figure, metamorphosis is complete in these young starfish,
and an examination of serial sections shows that the internal organs have nearly all
attained the adult relations. As these relations are, moreover, practically the same as
those in a young Asterina gibbosa of the same stage, it will only be necessary to mention
here the few points which show peculiarity.

In the first place it will be noticed that the great size attained before the mouth
and anus are formed must be looked upon as an expression of the favourable conditions
under which development of this species proceeds. Another indication of the same
thing may be seen in the amount of yolk which still remains in the closed gut, and
which may even be seen outside the gut in the space between its dorsal wall and the
inner wall of the coelom.

The pre-oral lobe (larval organ) in the oldest stage is little more .than a vestige,
but in the youngest its connection with the inner perihsemal ring can still be traced.
The water-vascular system is well developed, and one can follow it quite easily in
vertical sections. The madreporic pore, pore-canal, and stone-canal may be said to
have assumed the adult relations. The axial sinus can be seen adjacent to the stone-

i.\i:

canal, and can Ixs traced from its owning into the latter to its junction with the inner
perihtemal rim; In n«.n«> of the sections, however, can one be sure of the al>oral sinus
or of ill-- nm<lrr|M.ric vesicle, though this is doubtless due to the preservation.

Several [Hiinis iii tin- structure of these young A ate rids suggest a relation to an
embryonic starfish studied in this laboratory some time ago by Dr. E. H. Henderson (4).
These were large yolky eggs of an unknown species, belonging, probably, to the genus
AiMstfruu, collected off the Franklin Islands. In the remarkably close resemblance to
the development of Anterina gibboaa of the same age, and in the similarity in the
appearance and distribution of the yolk, we have indications of a near relationship
Aoen this species and Asttrias brandti, but the great difference in age between the
two lots of embryos makes it difficult to institute a closer comparison.

2. OPHIOPLUTKUS sp.n. (?).
(Fig. 2.)

Two specimens of this presumably new Ophwpluteus were taken at the Winter
Quarters of the Expedition ; the first on December 15th, 1903, from No. 12 hole, in
8 fins, of water; the other on January 4th, 1904, from No. 13 hole, in 6 fins. The
first is in a comparatively good state of preservation : the second is Iwdly disintegrated,
and from it alone little could be made out It is, however, of approximately the same
stage of development, though a little 'larger, than the first The following description
is, therefore, based upon the letter preserved specimen.

The body, exclusive of the arms, measures 2*3 mm. in length, and is marked by
a strongly developed conical papilla at its posterior end.

Three pairs of arms are developed ; of these the antero-laterals and post-orals
are only slightly developed, but the postero-lateral pair are already equal to the body
in length and are widely divergent This development of the postero-lateral arms
reminds us of the great size of the corresponding amis in the larva of Ophwthriz
fragilis, which is such an early and marked characteristic of this species. In fact, the
whole appearance of our specimen is strongly reminiscent of the larva- of Ophiothrix.
The ciliated band can be traced quite distinctly bordering all the arms.

In our better specimen the coelom can just be made out. On the left side it has
already divided into anterior and posterior coeloms, but on the right side it is still
undivided.

The gut is well developed and shows the characteristic division into oesophagus,
Urge spherical stomach, and intestine opening on the ventral surface.

No trace of the skeleton is preserved in these specimens, and this, together with
their young stage, makes classification impossible. Judging from their habitat,
however, and from their general appearance, they are quite possibly examples of a new
•paries.

E. W. MAcBRIDB AND J. C. SIMPSON.
3. ECHINOPLUTEDS ANTARCTICUS.

(Fig. 3.)

The material contained only two specimens of this larva : one taken in
lat. 78° 29' 10" S., long. 103° 38' 18" W., on January 25th, 1902, and labelled "30,"
the other marked " 42," taken in the D-net at Hut Point, near the Winter Quarters
of the Expedition, on February 19th of the same year. Both specimens are of
approximately the same age.

In the stained, and mounted condition the specimens measure 072 and 0'84 mm.
in length, including the arms ; but in each case the body, which is sub-spherical in
form, is 0'43 mm. in diameter.

Four pairs of arms can be made out ; of these, the post-oral pair are the longest,
being a little more than two-thirds as long as the body. Of the others, the antero-
laterals are about half and the postero-dorsals about one-quarter the body-length.
The pre-oral arms are very rudimentary, appearing only as small outgrowths at the
bases of the antero-laterals. All the arms are close together and are approximately
parallel.

The ciliated band, which borders all the arms, extends back between the post-oral
and postero-dorsal arms of either side, so as to form lateral bays running about half
way along the body.

Three pairs of epaulettes are present, and are arranged so as to form two
incomplete ciliated bands encircling the body. The anterior circle consists of a
ventral pair situated at the bases of the post-oral arms, and a dorsal pair in a
somewhat corresponding position on the dorsal surface of the body at the bases of the
postero-dorsal arms. The posterior circle is made up of a single pair only, and these
are smaller and much narrower than either of the other pairs, and are laterally placed
near the posterior end of the body.

We find that three Echinoplutei have been described in which the presence of
six ciliated epaulettes is noted. These are : E. coronatus, E. theeli, and the pluteus
of Echinus esculentus. From each of these, however, the present larva differs
considerably. In Echinus esculentus the presence of the posterior pair of epaulettes
was first pointed out by MacBride (8), and a great deal of material of the larvae of this
species has been at our disposal for comparison with our new type. Selecting an
individual which, judging by the development of the so-called " Echinus-rudiment,"
is of the same age as our specimens of E. antarcticus, we find that the size of the
characteristic larval arms is quite different. The post-oral and postero-dorsal pairs
are of about the same length, and are already about twice as long as the body, instead
of two-thirds and one-quarter the body-length, as they are in E. antarcticus. The
antero-laterals are also of great length, being more than twice as long in comparison as
are the corresponding arms in our new species. The most marked difference, however, is

BCHINODERM

iu the pro-oral pair ; for while these m more rudiments in E. antarcticiu, they are
almost equal to the length of the body in K. fsculrntu*.

The other two larva1, E. coronatu* and E. thfeli, were described by Mortcnscn (9).
The first of these is a larva of about the same or a slightly older stage than our
specimens. In it, however, the post-oral and postero-dorsal iinns arc about one and a
half times as long as the body, while the antcro-laterals and prc-orals are equal to it
in length. Besides this great difference in the arms there is also a difference in the
posterior pair of epaulettes ; for Mortensen describes these as " very broad," while in
K. <intn ret it-it* they are much the smallest of the six.

In E. theeli we are dealing with a larva which is much younger than the others
just mentioned, but even here we see certain features in the arms by which to
differentiate the two ; for while three pairs of the arms are of ul>out the same
proportionate length as in our specimens of K. antarcticus, the fourth, the pre-orals,
are long and well-developed, not at all resembling the mere rudiments by which they
are represented in the latter specks.

As in the case of the Ophiopluteus, so here nothing can l>e made out of the
skeleton, the result, probably, of the method of preservation.

On the left side of the larva a well-developed " Echinus-rudiment " can IK? seen,
but of this no details can be made out in either of the mounts.

4. CCCUMARIA CROCKA.

NolotMuria (Cueumaria) erotta, Leason, Cent. Zool. (1832), p. 153, pi. lii., fig. I.
Cttfumaria trofta, Lam pert, Secwalzcn (1885), p. 214.

Our collection contains some two hundred specimens of the young of this
Holothurian, " born after the capture of the mothers," and one adult female, the
parent of some of these young ; all taken at Hut Point, near the Winter Quarters of
the Expedition, on the 13th and 28th of February, 1904.

The young of this species were first dredged by the ' Challenger ' at the Falklands
in January, 1876. Since that time nothing, so far as we can learn, has been added to
our knowledge of their development, and as the specimens then found were of a
considerably older stage than those brought back by the ' Discovery,' it may be well
to recall certain of the observations made by Sir Wyville Thomson at that time, which
have a special bearing upon the developmental history.

In his " Voyage of the ' Challenger,' " Sir Wyville says : "In a very large
proportion of the females which I examined, young were closely packed in two
continuous fringes adhering to the water-vascular feet of the dorsal ambulacrum.
The young were in all the later stages of growth, and of all sizes, from 5 mm. up to
40 mm. in length ; but all the young attached to one female appeared to be of the
same age and size. . . . All the young I examined were miniatures of their parents ; the
only marked difference was that in the young the ambulacra of the bivium were quite

VOL. IT. 8

6 E. w. MACBRIDE AXD ,T. c. SIMPSON.

rudimentary." They " attached themselves by the tentacular feet of the trivial
ambulacra, which are early and fully developed." He says, however, that they " were
too late at the Falklands (January 23rd) to see the process of attachment of the
young in their nursery .... There can be little doubt that, according to the analogy
of the class, the eggs are impregnated either in the ovarial tube or immediately after
their extrusion, that the first developmental stages are run through rapidly, and the
young are passed back from the ovarial opening, which is at the side of the mouth,
along the dorsal ambulacra, and arranged in their places by the automatic action of
the ambulacral tentacles themselves." (Vol. II., p. 215 et seq.)

Now the young brought back by the ' Discovery,' with two exceptions, all
measured from 2^ mm. to 3 mm. in length, and had been born in the interval
between capture and preservation ; it seems, therefore, that we had here a key to the
solution of the manner in which the early developmental stages are passed over, and
that we might thus get a more complete history of the " brood-care " of this species.

A careful examination of the adult female specimen, made with this object in view,
showed two prominent openings, each about 3 mm. in diameter, situated in the dorsal
interambulacrum, a short distance behind the peristome (Fig. 4). Looking into these
openings, one could see a number of embryos of about the same size as the new-born
specimens of the collection. At first sight these openings seemed to lead into the body-
cavity, but a careful dissection revealed the fact that they were in reality the openings
of two large brood-pouches situated in the dorsal body-wall (Fig. 5). Owing to the
extreme contraction of the specimen, the extent of these pouches can be made out only
approximately. The right, which is by far the larger of the two, extends from the
raid-dorsal line downwards to the right ventral interambulacrum, and from a point just
behind the peristome backwards for about one- third of the body length. The left sac,
however, is only about half as large. Both pouches are divided up by a number of
pillars representing the podia, which thus seem to traverse the sacs. In reality,
however, the podia are probably only surrounded by the pouches as they develop.
Some idea of the extent of these spaces may be gained from the figure, which is drawn
of the natural size. No communication was found between the two sacs, nor was any
opening into the body-cavity or genital ducts made out.

In these two pouches 140 embryos were found — 110 in the right and 30 in the
left. These, however, are all of the same size, and of a comparatively late stage of
development, with the exception of two specimens which are much less developed than
the rest.

Through the kindness of Professor F. Jeffrey Bell we have also been able to examine
three other specimens of C. crocea belonging to the British Museum. It was hoped
that they might show some trace of these brood-pouches, but the most careful search
revealed nothing.

If we now try, from the facts before us, to get a connected idea of the " maternal
care " of this species, we shall probably find that its course is somewhat as follows :—

KCHIXonERM LARVA 7

The eggs, either l>eforc or after fertilization, are shed into the water and are attached
to the mother's back, possibly by means of a mucoid secretion. They are soon,
however, surrounded by the brood-pouches, which probably develop by a sinking-in of
the dorsal ectoderm, and here they are retained until they have reached a condition
where the tentacles are able to serve as organs of attachment to the mother. This we
know must be when the little animals are between 3 mm. and 5 mm. in length. They
then pass out of the brood-pouches and are arranged along the dorsal ambulacra of the
mother as Thomson has described.

The presence in the sacs, among a large number of older embryos, of two whi«-h
are comparatively young, naturally brings up the quest ion of the number of broods in
this species. Are all of the young brought forth in a single brood, or is there a
succession of broods such as we find in Synapta vivipara, Clark (3) ? If these two
embryos represented the beginning of a second and younger brood, one would expect
to find the gonad of the mother showing some sign of activity ; but, as this is not the
case, we must conclude that they are members of the same brood as the larger embryos,
and that their development has been, for some reason or other, retarded. What
evidence we have, therefore, seems to be in accord with Thomson's ol>servation that all
the young attached to one mother were of the same age and size.

Coming now to the embryos themselves, it is somewhat disappointing to find that
among the large number brought back only two stages arc represented. Turning our
attention to the younger of these, we see that of the two specimens in the collection
one was somewhat distorted through the pressure of the surrounding embryos in the
brood-pouch ; but this does not at all affect the disposition of the internal organs as
seen in sections. The other specimen, apparently quite normal, is oval in form, and
measured before sectioning 1*8 by 1*2 mm. No evidence of any external opening
could be made out

When sections of this stage are examined it is seen to be a typical pcntactula, and
the only points which need be specially mentioned arc in connection with the water-
vascular system. The ring canal is well developed, and the radial canals coming off
from it can easily be followed as they arch back and run towards the aboral pole. As
yet there is no sign of the development of podia, but the tentacular canals can be seen
to originate one on either side of each radial canal, close to where it comes off from the
ring canal, and to run out into the tentacles which do not reach the surface of the
embryo, but still project into the "atrium." Even at this stage valves may be seen
guarding the entrance to these tentacular canals. The stone-canal is well formed, and
runs out in the dorsal inter-radius to open into a large, clearly-defined axial sinus.
From the latter the pore-canal can I ><• traced outwards for a short distance, but we were
not able to follow it to its external opening in this stage, though in the older embryos
it is very easily followed. No Polian vesicle has yet been developed.

Of the later stage there were a great number of specimens, for it was found that

all the embryos in the brood-pouches (with the exception of the two already mentioned),

- -J

8 E. w. MAC-BRIDE AND j. c. SIMPSON.

and a second brood dredged a fortnight later, are of approximately identical
development. These embryos all measure from 2 • 5 to 3 mm. in length, and have
assumed the typical ten-tentacled form. The tentacles are still simple, unbranched,
peg-like outgrowths (Fig. 6), but there is some slight indication of a differentiation in
size, the two ventral tentacles being just a shade smaller than the rest.

In sections the digestive tract is seen to be well developed, though the ossophagus
is still solid, and the anus has not yet been formed. The stomach walls are thrown
into heavy folds and the intestine shows the characteristic double twist.

The water-vascular system in this stage shows many advances. The pore-canal
can be distinctly seen to open to the outside, and can be traced into the axial sinus,
which runs for a short distance in the body-wall before it communicates with the stone-
canal. The presence in Holothurians of an axial sinus opening, on the one hand, into
the stone-canal, and on the other to the outside by means of the pore-canal, has been
described in two cases only — once by Bury (2) and once by Ludwig (6). As Bury's
observation, however, has been very generally questioned, and as Ludwig, though
noting the presence of these structures, does not attach any very great weight to them,
the present discovery becomes of extreme interest and importance, as a confirmation
which places the matter beyond doubt. As has been noted above, in the younger
embryos the pore-canal cannot be traced throughout its whole length, but this is in all
probability due to bad preservation.

A large Polian vesicle is now seen in the left dorsal interradius, a position which
Ludwig (5) says it occupies in some Cucumarise, though he does not mention C. crocea.
Tube feet have made their appearance, and we now find four arising from each radial
canal, though they have not as yet reached the surface.

The state of the preservation prevents any very minute study of the nervous
system, but one can readily see the well-developed nerve-ring and the five radial
nerves, as well as the branches to the tentacles.

The epi-neural ring and canals are also clearly seen, and we find that the ring also
gives off branches which accompany the tentacular nerves.

Of the condition of the mesenteries little can be made out. A bunch of cells in
the dorsal mesentery close to the stone-canal may very possibly be the rudimentary
gonad, but of these points it is impossible to be sure.

In conclusion, it may be said that the collection, though small, has yielded many
interesting points. Chief among these may be mentioned the discovery of two
pelagic Echinoderm larvae within the Antarctic Circle, and of the brood-pouch in
Cucumaria crocea. The first are important because they disprove the generally
accepted theory that none ot the Antarctic Echinoderms have free-swimming larvae,
the second because it fills a gap in our knowledge of the life-history of a well-known
form. The well-marked axial sinus and pore-canal in the Cucumaria embryo is also

ECHLXODEKM LAKVJS. 9

a point of importance, since it gives positive evidence of the existence of thoao
structures among the Cucumariithe, a fact which hat been asserted, but never
conclusively proven.

Zoological UU.ratory, McGill University, April 1st, 1U07.

LIST OF PAPERS QUOTED.

1. I'.KI.I.. P. J. — " Echinoderms " in Report on the collcctioni of the 'Southern Cross,' Ixjmloii, l'Mr>,

p. 214.

2. BUBY, H.— "Studies in the Embryology of the Echinoderms." Q.J. Micros. Sci., vol. 29 (1H89).

pp. 409-449.

3. CLAKK, H. L.— " SynapU vivipara : a contribution to the morphology of Echinoderniata," Boston,

Memoirs Soc. Nat. Hut., 5 (1898), pp. 5.-J-88, 5 pU.

4. HKXDERSOX, E. H. — •'Some observations on the development of an Ante-rid with large yolky sggi

from the Franklin Islands," Ann. Mag. Nat. Hist., vol. 1C (l'JO'>).

6. LUDWIO, H.— " Die Secwalzen " in Broun's Klasseu u. Orduuiigcn d. Thier-Rcichs, '2 Bd., 3 Abth.,
1 Buch. (1889-92).

6. LUDWIG. — " Zur Entwicklungsgeschichtc dcr Holothnrien." Sitzungsber. I'renss. Akad. d. Wit*.

Berlin, 1891, p. COS.

7. MACBBIDB, E. W.—" Development of Astcrina Kibbosa." Q.J. Micros. Sci., vol. 88 (1896),

pp. 389-411.

8. MACBRIDK. E W.— "The Development of Echinoids," pt. 1. "The lame of Echinus miliaris and

E. esculentns." Q.J. Micros. Sci., vol. 42 (1899), pp. 835-339.

9. MOBTKKSKX, TH.— "Die Echinodermenlarven dcr Plankton-Expedition d. Humboldt-Stiftung."

Bd. II. J., Kiel and Leipzig, 1898.
10. THOMHOX, C. WYVILLB.— " The Voyage of the Challenger— the Atlantic," London, 1«77.

TABLE OF PLATES.

FIG. 1. Atttruu brandtt 'Bell, x 12J. Youngest embryo of brood.

Fio. 2. Ophiopluteus sp. n. ? x Ion.

Flo. 8. te/iinopltilfiix antarcticiu sp. n. x 100.

Fio. 4. Cticumaria trocea Lesson, x 1. Adult female, showing openings of the two brood-|iuuchet.

Fio. &. Cueumaria crocra Lesson, x 1. Another view of the same specimen, with the Iwdy-wall cut

through so as to expose the interior of the brood-pouches.
Fio. 6. Cttcvmaria crocfa Lesson, x 12£. Young embryo, born after capture of the mother, which is

shown in figs. 4 and 5.
Fio. 7. Cueumaria erotta Lesson. Transverse section through the anterior region of embryo shown in

fig. 6.

Tic 2 0»mo*LuTtus in »» *

.

FlG I ASTCMIAS BKANDTI.

Yov.»»<j»»l «mbryo of brood

Bruotl-pvvt,

FlC 3.ECHINOPLUTCUS ANTARCTICUS.

*IOO

Tmtmtltt.

FlC 5. CUCUMAMIA CMOCCA
Adult f«m«l» — to Show
interior of brood-pouch«».

K/

. VrntroJ Tmta

FlC 4 CUCUMARIA CROCCA.

Adult f*m»l» — to «how
openings of brood-pouch**.

P*r* Cm* * I

, Axiul Sinui.

. Sr»n« tanat

•Kptn»u:-ul Catttl.
Kmdim.1 fitrvo.
-Radiml Ca»ml

Polian Vrtitlt

FlC 6. CUCUMANIA CROCCA.
Young tmbryo born tflir
c«ptur* of mother.

Riny Canal

FlC.7. CUCUMARIA CROCCA

Tr»n»v«r§* ttction through
jnt«nor r»g.on of •mbrjro
• hewn in FIIJ 6.

To W b«Md «itk pUir

eu3TujqoiHqO Soil
00\x

z»A:>iio<i- booi ft -

A300SO AIHAMU3UO

wode oJ -8lsm»>
>o
.t x

A330flO AlflAMUOuO .6
woHa 0J -.slsms^ JIubA
asHauoq-booid >o

»<r«»Vi. JnjfcoR -7

A330HO AIHAMUOUO T.OI?
oiJosi saisvensiT
>0 noijjsi 10i isJns
•8 pi 3 ni nvvorlz

q/3

! Itvii..! ,-,,,.|q djiv/ h,,,,, ,

Antarctic (lH*cov«y) Exp.

I. C. «. 4M IHIIIHHlfc. M

Kchinoderms.

MYZOSTOMIDAE.

BY DR. RUDOLF RITTKK v. STTMMKR-TRAUNKKUS (Graz, Austria).

(1 Plate.)

ElNLEITONO.

DAS von der ' Discovery ' heimgebrachte uiul mir von Prof. F. Jeffrey Bell in
liebenswiirdigster Weise zur Bearbeituug iibergebeue J/y.?os/o»uj-Material umfasst
mir fine relativ geringe Anzahl von Exemplarcn, die sich auf zwei Species, eine
freilebende und eine encystirte, verteilcn. Die erstere ist eine neue Art : .'/.
antarcticum, die letztere halte ich fur identisch mit dem schon von Prof. Graff
beschriebenen M. cysti&tlum.

Wenngleich also die Ausbeute nur eine geringfiigige Erweiterung unserer
beziiglichen systematischen KenutnlH.se zu bringen vermochte, so erscheint sic doch
in tiergcographischcr Beziehung recht bemerkenswert : einerseits, weil die bciden
Art en .-in- dem siidlichsten bekannten Gebietc des antarctiscben Uzeans stammen,
aus welch' letzterem bishcr noch nie Myzostomcn heimgebracht worden sind ; *
und andererseite, weil dadurch ein neuer Beweis fiir den Kosmopolitismus
dea M. cyxticolum geliefert wird. Diese Art kennt man nun von drei geographwch
ausserordentlich weit von cinander entfernten Fundorten (tropischer West-Atlantic,
Ostkiiste v. Japan, Ross-Meer in der Antarctis), cin Befund, welcher durch das
hohe geologist-he Alter der Gruppe sowie durch die bekanntlich sehr uniformen
Lebensl>ediugungen des Stillwassers und der Tiefsee seine Erkliirung findet.

Dfc, wie schon gesagt, das mir zur Bearbeitung iil>crgebene Material nur eine
geringe systematische Auslwutung zuliess, BO beachrankt sich die vorliegende Arbeit
haupteachlich auf die anatomist-he Beschreibung der beiden Arten. In dieser Hinsicht
vermochte ich insbesondere das M. cyxticolum genauer zu untersuchen, a Is dies bisher

• DM tudlichste Oebiet, ana welchcm man bis jetzt Uyzoctomen kanntc, nanilich die See urn die Prince
Edward- und die Croiete-Inwln, liegt noch in der SubanUrctu u. zwar nur wcnig lUdlich von der Nordgrence
der treibenden EUberge and der Sadgrenze te Wcltverkchrs. Von hier warden durch die ' Challenger '•
Expedition folgendc Arten fpwammelt :—

it. eomfn-ruum Graff . . .3, pag. 42.
U. coronatum Graff . . 3. ,, 60.

Steltthopu* hyocrini Graff . . 3, „ 70.

(Die fettgednickten Zahlen zu Beginn jedes Literaturnachwei*e< beuehen sich auf die enUprechenden
Xummem dea der vorliegenden Arbeit beigegebenen Literaturverzeichnisse*.

Die Angaben fremder Autoren encheinen in folgendem durchwegs mit : pag., aba., U/., fig. ; die Venretamgon
Ml Bhilai te rorliegenden Abhandlung mit : Pag., Textfig., Fig., citiert, wobei unter der letcteren Bezeichnung
die TO dieacr AiM>pjjMBJ» V^Ujesm m imMkiu rind.)

DR. RUDOLF HITTER v. STUMMER-TRAUNFELS.

geschehen, und es gelang mir dadurch auch emeu, allerdings noch nicht voll-
kommenen Einblick in die Sexualverhaltnisse dieser Species zu gewinnen, der aber
immerhin geeignet scheint, auch die Geschlechtsverhaltnisse der von Graft' fur rein
dioecisch gehaltenen Arten unter den cystenbewohnenden Myzostomen in eine neue
Beleuchtung zu rticken.

GENUS MYZOSTOMA. F. S. LEUCKART.

1. MYZOSTOMA ANTARCTICUM.

(Textfig. ; Plate I., fig. 1.)

" Winter Quarters, 8. iv. 1903, 124 fms."

Von dieser ansehnlichen auf Antedon adriani Bell (diese Reports, vol. iv.,
Echinoderma, p. 4) freilebenden Form lagen mir 18 zumeist verschiedenalterige
Exemplare vor. Die Lange des jiingsten betrug nur 0'7 mm., die des altesten
dagegen 6'0 mm. Zwischen diesen beiden extremen Gliedern der Reihe war noch
eine ganze Anzahl von Zwischengrossen vertreten, von welchen die auf einander
folgenden in Bezug auf ihre Lange zumeist nur durch Bruchteile von mm. differirten.
Die nachstehende Tabelle giebt eine Ubersicht iiber die Dimensionen der einzelnen
Grdssenclassen.

Classe.

Exemplare.

Lange in mm.

Breite in mm.

a

2

0-7

b

1

1-0

Breite gleich der Lange ; Korperumriss

c

3

1-1

fast rein circular.

d

3

1-2

e

1

1-5

Breite nnr wenig grosser wie die Lange.

f

1

1-75

2-0

9

1

2-5

2-75

h

1

2-75

8-0

i

2

3-25

3-75

j

1

4-0

5-0

k

1

4-5

5-5

I

1

6-0

7-5

Die Untersuchung erstreckte sich auf je ein in Xylol aufgehelltes Exemplar der
Classe a, d, e, g, und /, ferner auf das in Langsschnitte zerlegte Exemplar der Classe /.

Wie schon Wheeler (8, pag. 228, abs. 3) hervorgehoben hat, sind die Jugend-
stadien der einzelnen Myzostoma-Arten von einander nur wenig verschieden,
wahrend die artlichen und generischen Merkmale erst bei den erwachsenen Individuen
deutlich zum Ausdrucke kommen. Dieser Erfahrung zufolge stiitzt sich die nach-
stehende Beschreibung nur auf die bei dem grossten Exemplare (Classe I) gemachten
Befunde.

MYZOSTOMI1UB.
\l.U;KMKINK MoRl'HOUMIIK DBS KoRPKRH.

Wie aus don in iler obigen Tatalle angefUhrtcn Dimcnaioncn hervorgcht besitzt
1'ci il-'ii <_'!•• ivM-ren Excmplaren einen qucrovalen Urarisa. An seiner Ventralseito
t'.illi cine von der Leiltesmittc bis zur Reihc der Parapodicn rciehcnde com par te un<l
fast Bcheibenfonuige centralc Partie auf (" Bauchscheibe," Graft', 2, pag. 41, aba. 3),
welche sieh von einer viel zartcren und durchscheinenderen Randzone (Fig. 1, sr)
ziemlich scharf absetzt. Boi einer Lange von 6 mm. und einer Breite von 7 • 5 mm.
btb&gt die I >:•!%•• de.M K"r|"-r- in .!>T Mil'.' . -in i •_' nun. Si.- v.-niiin.|.-rt nA im
Bereiche der Bauchscbcibe bis zur Reihc der Parapodicn hin relativ nur wcnig, wird
jedoch l>eim Ubergange von diescr in die Randzone cine weitaus geringere. Die
letzterc lauft gegen den Korperrand bin in einen allcrdings nur schmalcn jcdoch
ganz deutlioh erkennbarcn, schwacb gewellt«n Randaaum aus, welcher in der Mcdianlinic
sowohl am vorderen als auch am hintcren Korperendc cine schwache Einbuchtung
zeigt Vom Randsaum entspriugen je<lerseits 10 bilateral-symmetrisch angeordnete,
relativ kurze (circa 0'75 mm. lange) Cirren (Textfig., c), welche mit ihrer Wurzel ein
wenig auf die Doraalfliu-he der Randzone hinaufgerilckt sind (Fig. 1 , c).

Die Riickenflftcbe des Korpers ist sanft gewolbt, die Bauc"hseitc dagegen, dadurch
daas sich die l>eiden latemlen Kdrperhiilften ventral wiirts etwas einbiegen, in geringem
Masse ooncav gestaltet. Dorsalseitig zeigt das Integument cine eigentiimliche, aus
zahlreichen kleinen, dicht neben einander gelegenen Wiirzchen bestehende Sculptur,
(Fig. 1). Diese durch rinnenartige Vertiefungen von einander getrcnnten Erhabeuheiten
sind auf der Mitte des Riickens am bedeutendsten entwickelt und nehmen an Grosse
von hier aus gegen die Randzone hin allmiihlig ab. Im Bereiche der letztercn sind
sic nicht mehr warnehmbar. Die Ventralflache des Korpers ist im Allgemeinen glatt,
bis auf die Bauchscheibe, an welcher der Verlauf der sehr stark cntwickeltcn Muaculi
centrales des Hackenapparates durch leistenfbrmigc Vorspriinge dea Integuments
deutlich ausgepragt erschcint.

Die 10 Parapotlien (Textfig., par) sind kriiftig dabei alwr relativ schlank und
wie ihre sehr entwickelte Musculatur vermuten lasst, recht beweglich. Sie konnen
in eine Art von Scheide zuriickgezogen werden, welche von einer grosscn, stark
vorspringenden Ringfalte des ventralen Integuments gebildet wird. An den beiden
mittleren Fussstummeln jeder Seite offnen sich die Ausfiihrungswegc des miinnlichen
Geschlechtsapparates durch eine dem Parapodium von aussen her ansitzendc konische
Genitalpapille (Textfig., p).

Die 8 langlich ovalen, im eingezogenen Zustande nicht iiber die VentralHache
vorspringenden Seitenorgane (" Saugniipfe ") liegen bilateral symmetrisch angeordnet
im Gebiete der Randzone auswarte von der Reihe der Parapodien und in den
Intervallen zwischen je 2 der letztercn (Textfig., *>).

Wie schon friiher erwShnt ist der Vorderrand des Korpers in der Medianlinic
etwas eingezogen. Die^c KinUichtung schliigt als rinncnfonnigc Vertiefung auf die

TOU IT. T

4 DE. RUDOLF BITTER v. STUMMER-TRAUNFELS.

Ventralseitc tiber, um ungefahr in der Mitte der Randzone in die Mundoffnung zu
fiihren (Textfig., in).

Die Kloake offnet sich an der Spitze einer medianen flach-konischen Papille
(Textfig., dp), welclie ebenfalls ventralseitig und in einiger Entfernung vom hinteren
Korperrande gelegen ist. Auf derselben Papille liegt ausserdem unmittelbar dorsal
tiber der Kloakenoffnung die weibliche Geschlechtsoffnung (Uterusausmiindung).

f

Halbschematisehes Ubersiohtsbild der Organisation des M. antarcticum,
nov. sp. Vergr. : 11-fach. Ansicht von der Ventralseite aus ; die
Organe sind in ihren Konturen sitmtlich in die Bildebene projicirt. —
bm, Bulbus musculosus ; c, Cirren ; cl, Kloake ; dp, Kloakalpapille ;
TO, Mundofinung ; mg, Magen ; n, Nephridien ; o, Ovarien ; oe, Oeso-
phagus ; p, Penis ; par, Farapodium ; r, Rectum ; rp, Russelpapillen ;
rl, Hiisscltascho ; so, Seitenorgan ; na, vorderer Uterusabschnitt ;
lib, hinterer Uterusabschuitt ; I, n, in, Hauptstamme der Darm-
verzweigung ; 1, 2, 3, Hauptstamme der Uterus- (Leibeshohlen-) Ver-
zweigung.

ERNAHRUNGSAPPARAT.

Die Mundb'ffnung fiihrt in eine, bei eingezogenem Pharynx vorne rohrenartig
gestaltete Riisseltasche (Textfig., rt), in deren glockenformig erweiterte Caudalhalfte
die Spitze des Pharynx hineinragt. Dieser tragt an seinem Vorderende einen
Doppelkranz von kleinen Tastpapillen (Textfig., rp), und zeigt im iibrigen, mit
seinem relativ schwachen Bulbus musculosus (Textfig., Inn), keine bemerkenswerte
Ausbildung. Das Pharyngealrohr offnet sich am Hinterende des Bulbus musculosus (bei
eingezogenem Riissel !) in einen dorsoventral compressen, beiderseits aber taschenartig
ausgebuchteten Raum (Textfig. oe), der nach hinten durch eine (von Graff, 2, pag. 49
letzter abs., pag. 50 abs. 1, als " Klappe " bezeichnete) Ringfalte vom Magen
abgegrenzt wird. Es ist sehr warscheinlich, dass dieser Raum bei vollig protrahirtem
Riissel sich rdhrenformig streckt, also einen Oesophagus im Sinne Graff's darstellt.

MY7.nsToMII»AK 5

Der grosse und weite Mageu (Textfig., my), ist tounenfurmig gestaltet und vom
kurzen Rectum (Textfig., r) durch einen Spbinkter abgesetzt

Ventralwarts entepringen von ihm beiderseits je drei Hauptetamme der Darmver-
zweigung (Textfig., I, II, ill), von welchen sich ein jeder nach kurzcm Verlaufe in zwei
HauptJi-sto teilt Von diesen sechs Hauptiisten verlaufen in jeder Korperhalfte zwei
nacb vorne, zwei nach der Seite und zwei uach hinten und zwar ungeteilt bis zur
Randzone dea Korpere, wo sie sich dichotomisch in zahlreiche kleinei-e Verzweigungen
auflosen.

DlM sehr kurze und cnge Rectum (Textfig., r) geht alslwld in die relativ
langgestreckte kolbenformige Kloake (Textfig., <•/) iiber, in deren brcitere Vorder-
halfte sich von beiden Seiten her die Nephridieu (Textfig., w) mit liingsspaltigen
Nephroporen iiffnen. Nach hinten verechniiilert sich die Kloake allmahlig und
miindet, wie schon gcsagt, durch einen an der Spitze der Klofikulpnpille gelegenen
Porus nach aussen.

WEIBLICHER GESCHLKCHTSAPPARAT.

Am sogenannten " Uterus " lassen sich zwei, durch ein enges und kurzes medianes
Verbindungsrohr communicierende Abteilungen, cine vordere und cine hintere
unterscheiden. Die eretere (Textfig., ua) ist kurz, dagegen l>edeutcnd in die Quere
entfaltet und liegt ol>erhalb des ersten Magendrittels. Sie teilt sicli lateral jetlemeitig
in drei Canale, welche als Hauptstamme der Leibeshohle (Textfig., ;, s, 3) dorsal von
den drei Hauptetammen der Darmverzweigung verlaufen und nach entsprechender
dichotomischer Teilung auch deren weitere Verastelungen begleiten. Der hintere
Uterusabschnitt (Textfig., tilt) ist gegeniiber dem kurzen vorderen bedeutend in die
Lange gestreckt, rostral zwar stark verbreitert, caudal jedoch rohrenartig verschmalert
Er miindet durch die knapp oberhalb des Kloakalporus noch auf der Kloakalpapille
gelegene weibliche Geschlechteoflfuung aus und stellt das eigentliche Reservoir fiir die
reifen flier dar, welche man im vorderen Uterusabschnitte nur ganz vereinzelt antrifft

Els sind jederseit.s zwei Ovarien (Textfig., ») vorhauden, von welchen je eines an
der Wurzel des zweiten und des dritten Hauptstammes der Leibeshohle gelegen ist.
Diese beiden Ovarien sind einander jedoch sehr genahert, so dass Hire freien in den
vorderen Uterusabschnitt hineinragende Spit/en fast zusammenzufliessen scheinen.

MisNLicHER GESCHLECHTSAPPARAT.

Die Hoden sind nach dem fiir die freilebenden Arteu charakteristischen verzweigten
Typus angeordnet*

• Bel den venehiedenen Myioitoma-\rien kann der m&nnlicbe GeniUlapparat in dreierlei Modificationen
aoftnten, die anliMlich cincr kiinftigen Zcrteilung des Genus als Gattungscharaktere Verwendang findrn
konnteo :—

(a) Vertwtigttr Hoden (" (he typical ramified form " : Graff, 8, pag. 11, ab*. D—

Typen : 11. eirrifentm (Graff, S, pag. 69-61).

11. attrriae (Stummer, 6, pag. 570-683).

l '_•

DR. RUDOLF BITTER v. STUMMER-TRAUNFELS.

EXCRETTONSAPPARAT.

Die beiden Nephridien beginnen mit je einem kleinen und engen Nephrostom,
welches an der caudalen Wandseite des vorderen Uterusabschnittes unweit von dem
die beiden Uterushalften verbindenden Rohre gelegen ist. Von diesen Nephrostomen
aus verlaufen die anfangs sehr engen aber spater weiteren Nephridialcanale (Textfig., n)
zuerst in lateraler Richtung an der Dorsalseite des Magens, um dann an dessen
seitlicher Circumferenz ziemlich vertikal nach abwarts zu ziehen. Nach Umgreifung
des hintersten Hauptstammes der Darmverzweigung schlangeln sie sich dann langs
der Ventralseite des Magens zur Kloake bin, in welche sie schliesslich von beiden
Seiten her mit je einem spaltfb'rmigen Nephroporus ausmiinden.

VERWANDTSCHAFTLICHE BEZIEHUNGEN zu ANDEREN

Die Griinde, welche Graff seinerzeit (3, pag. 22, Abs. 2) veranlasst haben, von
einer Untertheilung des provisorischen Sammel-Genus Myzostoma abzusehen, bestehen
auch noch heute, da fiir die Hauptmasse der von ihm beschriebenen Arten noch
immer eingehende anatomische Untersuchungen fehlen, ohne die eine sy sterna tische

Die einzelnen Hodenfollikel sind von einander relativ weit entfernt und durch mehr oder minder breite
Bindegewebspartien oder durch andere Organteile getrennt. Die Spermiducte sind im Allgemeinen verlangert
und in einzelne Abschnitte (Vasa efferentia und deferentia, Samenblase, Ductus ejaoulatorius), differenziert.
Der mannliche Geschlechtsporus ist ventralseitig, zumeist ziemlich weit nach innen vom Korperrande und fast
regelmiissig an der Spitze einer kegelformigen von der Basis des mittleren Parapodiums nach aussen hin
vorspringenden Genitalpapille (Penis) gelegen.

(6) Compacter Hoden (" compact roundish glands occupying definite areas in the lateral part of the body " :
Graff, 3, pag. 11, abs. 1)—

Typus : M. cysticolum (vorliegende Arbeit, Pag. 13-14).

Die Hodenfollikel sind auf einem deutlich umgrenzten rundlichen Bezirk in den Seitenteilen des Korpers
concentrirt. Sie liegen enge neben einander und werden nur durch dunne Bindegewebsbiilkchen oder -lamellen,
nie durch andere Organteile getrennt. In der Mitte des Hodens bildet sich durch Platzen der dortselbst
gelegenen bindegewebigen Follikelwandungen ein centraler Sammelraum, in welchem sich die reifen Geschlechts-
producte anhaufen. Dieser steht entweder vermittels eines ganz kurzen unbedeutenden Ductus ejaculatorius
oder auch direct mit der mannlichen Geschlechtsofihung in Verbindung. Letztere ist am Seitenrande oder in
unmittelbarer Nachbarschaft desselben gelegen. Eine Genitalpapille scheint nie ausgebildet zu werden.

(c) Den dritten von den zwei vorigen Modificationen principiell verschiedenen Typus der Hodenanordnung
hat Wheeler (8, pag. 247, 248) bei dem M. belli u. bei dem M. cryptopodium beobachtet.

Bei diesen zwei Arten soil die Gesammtmasse der Hodenfollikel dorsal von den Darmiisten in der mittleren
Korperregion gelegen und durch zwei dorsoventrale Septa in drei Lappen (einen medianen u. zwei laterale)
geteilt sein. Die Hodenfollikel springen in den dorsalen Theil der Leibeshcihle (" Uterus ") vor, in welchen
auch die fertiggebildeten Spermien entleert werden, die dann offenbar (Wheeler spricht sich dartiber nicht aus)
durch die weibliche Geschlechtsoffnung (oder Rloakaloflhung ?) ihren Weg nach aussen finden.

Bei der Isoliertheit dieser Beobachtung und bei der Wichtigkeit, welche dieselbe hinsichtlich der Leibes-
hohlenfrage Bowie der Genese der mannlichen Keimzellen bei den Myzostomen besitzt, erscheint eine
Nachuntersuchung dieser beiden von Wheeler nur oberflachlich beschriebenen Arten dringend geboten.

Anhangsweise mochte ich hier noch die sogenannten siibectodermalen Hoden erwiihnen, welche Nansen
(5, pag. 78 u. 79) bei M. gigas, M. giganteum, M. graffi, u. M. carpenteri beschrieben hat. Diese liegen als
vollkommen von dem normalen verzweigten Hoden getrennte Follikel dicht unterhalb des Integumentes. Sie
scheinen — dies lasst auch ihre eigentiimliche Structur vermuten (Nansen, 5, pag. 79, zeile 3-5)— abgetrennte
und in Biickbildung begriffene Terminalfollikel des normalen Hodens zu sein.

MY/.M>T"MII>\i: 7

Bearl>cituDg der Myzoetomiden umlurchfiihrbar iat. Dieser Mangel lasst daher im
Allgemeinen vonlcrhand nur eine additionelle Einreibung einer ucuen Species in

.!..•-•• <;.iU:ni'_' ll»'l>oten <T-, -Iiriih-n. lin S|M-,-i,-Ilrn \\inl in.ni -i. li <lai:iu! '!„•-, in. ii L, n

miissen die neubeschriebene Form init den wenigen Arten zu vergleichen, dercn
Organisation genauer bekannt iat und ihre Beziehungen zu ihncn festzustellen.

Von diesem Standpunkte aua betrachtet stcht das M. antarcticum jener Gruppe
von Arten am nachsten, zu welcher das M. cirriferum (Graff, 2, div. pag., Nansen,
5, div. pag., Wheeler, 8, pag. 229-236 ; pag. 276, abs. 3), ferner das M. giga*, das
\f. giganteum, das M. graffi, und das M. carpenter! gehoren. (Die anatomische
Bcschreibung filr die vier letztgenanntcn Arten bat Nanscu 5, pag. 69-70 u. ff. pag.,
geliefert.)

2. MVZOSTOMA CY8XXOOLUM.

(Plate, fig. 2-10.)
Mytottoma cyttifolum, Graff, Chall. Rep. (1884), p. 66.

" Mte. Erebus u. Terror; 22. i. 1902 ; 500 fins."

Unter dem von der ' Discovery ' heimgebrachten Crinoiden-Matcrial fandcn
sich an den A mien eines Exemplares von A it ted on adriani Bell zwei Cysten vor,
welche eine Atyzostoma-Pal beherbergten, die ich fur identisch mit dem M. cysticolum
Graff (3, pag. 66-68) halte.

HlSTORISCHES UNO KRITI8CHE8.

Diese cystenbewohnende Species wurde von dem genannten Autor nach einer
Serie von Exemplaren aufgestellt, die aus dem westlicben Gebiete des tropischen
Atlantic's (" Cabo Frio," Brasilien ; " Insel Grenada," kleine Antillen) stammten. Als
Wirt dereelben fungierte durcbwegs Actinometra meridional in var. cnrinata P.H.C.
Graff hat aus Griinden, welche er in der Einleitung zu seinem ' Challenger '-
Report (3, pag. 22, abs. 2) auseinandersetzt, bei der Beschreibung der l>etrcffenden
Art fast ausschliesslich nur ihre aiisserlich erkennbaren Mcrkmale Howie ihre
biologischen Eigentiimlichkeiten beriicksichtigt. Seine Angaljen iiber die innere
Organisation dieser Form beschranken sich auf einige die Darmverzweigung sowie den
Genitalapparat betreffende Befunde. Obwohl sich dieselben durchwegs als richtig
beobachtet erwiesen haben, so hat sich inzwischen doch herausgestellt, dass jene
unter ihnen, welche sich auf die Geschlechtsorgane des sogenannten " Weibcbens "
beziehen, von Graff in Ubereinstimmung mit den damals noch herrschenden
Ansichten iiber die Morphologic des weiblichen Sexualapparates missverstandlich
gedeutet worden sind. Dies gilt insbesondere von den angehlichen " Hodenrudimenten,"
welche, wie spater (Pag. 27) dargelegt werden wird, als die functionierenden Ovarien
aufgefasst werden miissen.

Seit dem Erscheinen der ' Challenger '-Publication ist das M. cysticolum, soweit
mir bekannt, nur noch einmal und zwar an der pacifischen Kiiste von Japan in einer

8 DR. RUDOLF RITTER v. STUMMER-TRAUNFELS.

Armcyste von Antedon discoidea P.H.C. aiigetroffen worden. McClendon, welcher
die beiden in der Cyste vorhandenen Individuen untersucht hatte, beschrieb dieselben
als eine neue Varietat : M. cysticolum var. orientate (4, pag. 120-121), und zwar mit
der Begriindung, dass :

(a) die Wirtsformen der Graff'schen und jene der japanischen Exemplare

eine verschiedene systematische Stellung einnahmen ;

(6) die betreffenden Fundorte in tiergeographischer Hinsicht difFerierten, und
(c) die japanischen Exemplare gegeniiber den Graff'schen Typen einige

unwesentliche morphologische Unterschiede zeigten.

Abgesehen davon, dass das McClendon vorgelegene Material (eine einzige
Cyste !) an Zahl weitaus zu gering gewesen ist, um die Aufstellung einer besonderen
Varietat zu rechtfertigen, so halte ich die hierfiir angezogenen Argumente des
genannten Autors nicht fur einwandfrei.

(ad a) Beziiglich des ersten derselben erinnere ich an die durch Graft" (3,
pag. 21) bekannt gewordene Tatsache, dass ein und dieselbe Myzostoma-Art auf
1-4 Crinoidenspecies vorkommen kann (in einigen Fallen sogar auf Vertretern
zweier Genera). Graff, gevviss der erfahrenste Kenner der Myzostoma-Systemntik,
hat sich jedoch nicht veranlasst gefiihlt, auf Grund dieses Umstandes verschiedene
Varietaten der betreffenden Species zu unterscheiden.

(ad b) Was die von McClendon hervorgehobene zoogeographische Differenz der
Fundorte anbelangt, so diirfte dieses Argument gerade bei den Myzostomen nur
mit grosster Vorsicht anzuwenden sein, da das Verbreitungsgebiet dieser Parasiten
noch sehr wenig bekannt ist, aller Voraussicht nach aber mit jenem ihrer Wirte
zusammenfallt, von welchen weitaus die meisten sich als charakteristische Vertreter
der einen cosmopolitischen Charakter besitzenden " Still wasserfauna " erwiesen haben
(Doflein, 1, pag. 251 u. pag. 272-273). Im iibrigen kann man der tiergeographischen
Differenz der Fundorte als solcher allein keine Bedeutung fur die Aufstellung von
Varietaten beimessen, wenn sich nicht die verglichenen Localformen auch in
morphologischer oder in biologischer Hinsicht als solche documentieren. Diese
Bedingung erscheint meines Erachtens nach bei den Exemplaren McClendon's nicht
erfullt,

(ad c) Die Besonderheiten, durch welche sich die letzteren vor den Graff'schen
Typen auszeichnen sollen, beschranken sich auf die bedeutendere Grosse sowie auf
den etwas differenten Bau der Cyste, ferner auf die hervorragenderen Dimensionen der
Parasiten selbst, schliesslich auf eine geringfiigige formale Differenz der sogenannten
" Mannchen."

Die Grosse der durch das M. cysticolum hervorgerufenen Cyste ha'ngt in erster
Linie von der Grosse des ihr Lumen fast vollstandig ausfiillenden " Weibchens " ab.
Da dieses jedenfalls von kleineren Dimensionen zu grb'sseren heranwachst und auch
sonst in dieser Beziehung individuelle Verschiedenheiten zeigen kann, so werden

MY/.OSToMM.AK, 9

demuach aucb versrhieden grosae Cysten vorkommen.* Da wir nun derzeit iil»-r die
umximalc Grosse, welche das M. cyxticolum crrcichon kann, noch vollkommcn im
Unklnron Kind, so erscheint es vorderhand untunlich cine itbcr die Kcknuntcn Mas.se
hinauareichendo Cystengitae al« Kriterium zur Aufstellung cincr neuen Varietal zu
verwenden.

Ebensowenig geeignet sind zu diesem Zwecke auch kleinere, den formalen Typun
der Cyste nii-ht beeintrachtigende Diiferenzen im Bane derselben. Den physiologischen
Aulass zur Bildung einer Cyste gibt allerdings das Myzostoma, da es entweder durch
seine Bisse oder warscheinlicher durch Absonderung toxischer Substanzen, wie es z. B.
die Excretstoffe sein konnen einen localen Reiz auf den WirtekOrper ausiibt, auf
welcben dieser dann durch Wucherung des gcschadigten Gewebes reagirt. Da wir al ><T
den haupteachlichsten, das ist den formativen Anteil an der Cystenbildung dcr
Wirtsform zuschreiben milssen, so werden wir kleine, unwesentliche Verilnderungeu
in der Morphologic der Cyste in erster Linie ala deu Effect einer individuellen
Verschiedenheit des Wirtes zu betrachten baben uud erst in zwciter Linie, wenn sicb
eine solche ausschliessen lasst, an eine VerHnderung des reizenden Substrates denken.
Voraussetzung fiir diese Erwagung ist natilrlich, dass sich die Cysten an verscbiedenen
Individuen ein und derselben Wirtsspecies vorfinden.f In weit hoherem Grade
milssen wir jedoch die Wirtsform fur etwaige Verftnderungen im Bau der Cyste
verantwortlich machen, wenn ein und dieselbe Afyzostotna-Art verschiedene Wirts-
species befallt. Es kommt denn eben die artliche Diiferenz der letztereu aucb im Baue
der auf ihnen entstehenden Cysten zum Ausdrucke. Von diesem Gesichtepunkte aus
muss auch die von McClendon beschriebene Cyste beurteilt werden. Sie fallt
hauptsachlich dadurch auf, dass sie im Gegensatzc zu den mit einem einzigen Porus
versehenen Exemplaren GrafTs, zwei derartige Offnungen und zwar eine an jedem ihrer
Schmalenden besitzt

Als morphologischen Unterechied zwischen den von Graff und deu von McClendon
untersuchten Parasiten erwahnt der letztgeuannte Autor vor allem die ttedeutendcre
Grosse seiner Individuen. Diese kann jedoch ebensogut durch individuelles Wachstum
oder durch besondere Ernahrungsverhaltnisse bedingt sein. Weiters hebt derselbe
Autor die Beobachtung hervor, dass bei dem m&nnlich functionierenden (" small
individual ") seiner Exemplare die zwischen der Korperperipherie und der Reihe der

• Die* ist aach bei den von Graff untenuchten Cysten der Fall geweten. Verglcicht man die auf die
Cyvtonlinge bezttglichen Angaben deg genannten Anton mit dem von McClendon gegebenem Mane, §o ergiebt
«ch, dan die Ordwendiflierenz zwuchen dem kleinsten nnd dem groraten Exemplare Graff's genan dietelbe let,
wie zwuchen dem letztcren und dem Exemplare McClendon's : —

/ kleinste* Exemplar . . 1 mm.
Graff.cherTyp: \ ^^^ J . . , mm < Diflerenz = 2 mm.

McClendon's Exemplar ..... 6 mm. < "

t Die Cysten von M. cyttifolum, welche Graff nntenucht hatte, fanden sich sammtlieh an Actinomrtra
mrridioruMt var. earinata P.H.C. Obgleich einige von ihnen kleine Verschiedenheiten in ihrer Gestalt and
ihrer Lage am Crinoidenarme aufwiesen, so ist es dem genannten Autor doch nicht beigefallen, die Insamen der-
selben als diflfcrente VarieUten der tj-puchcn Form zu beschreiben.

10 DR. RUDOLF RITTER v. STUMMER-TRAUNFELS.

Parapodien gelegene Randpartie (" edge of the disc ") sich weiter gegen das Centrum
der Bauchseite bin ausdehnte, als dies bei der von Graff gelieferten Abbildung (3, taf.
xiil. fig. 5) der Fall ware. Diesem Umstande vermag ich durchaus keine Bedeutung
zuzumessen. Jeder Beobachter, welchem zahlreiche Exemplare einer Myzostoma-Sipecies
zu Gesicht gekommen sind, weiss wie modulationsfahig hinsichtlich seiner centripetalen
Ausdehnung gerade dieser KOrperabschnitt ist, dessen Areale durch starkere Con-
tractionen der " bauchstandigen Muskelmasse " (Graff, 2, pag. 41-42) vergrossert, durch
Entspannung derselben verkleinert werden kann.

Von den eben dargelegten Erwagungen ausgehend glaube ich also dass em
zureichender Grund zur Aufstellung einer neuen Varietat fiir die von McClendon
beschriebenen Exemplare des M. cysticolum nicht vorliegt. Ich hoffe durch die
Beschreibung der von der ' Discovery ' heimgebrachten Individuen, fiir welche alle die
von dem obgenannteu Autor zu Gunsten seiner Auffassung ins Treffen gefuhrten
Argumente in gleicher Weise, ja vielleicht in noch hoherem Ausmasse passen, wie fiir
die von ihm untersuchten Exemplare, einen weiteren Beweis fiir meine Ansicht
erbringen zu konnen.

Von der inneren Anatomic des M. cysticolum berucksichtigt McClendon im
wesentlichen nur den Geschlechtsapparat. Er berichtigt und erganzt die beziiglichen
Angaben Graffs und bestatigt auch fiir die genannte Species das Vorhandensein von
weiblichen Gonaden bei einem jeden der beiden, friiher als " Weibchen " und als
" Zwergmannchen " unterschiedenen Individuen des die Cyste bewohnenden
Parasiten-Parchens.

BESCHREIBUNG DER ' DISCOVERY '-EXEMPLARE.
CYSTEN.

Die beiden mir zur Untersuchung iibergebenen Cysten waren als eiformige
Auftreibungen des Wirtsintegumentes ihrer Lange nach dcm Crinoidenarme etwas
seitlich von dessen Ambulacralrinne angeschlossen (Fig. 3).

Sie massen 6 '2 mm. beziiglich 5 mm. in der Lange und 3'2 mm. beziiglich 2' 5
in der Breite. Die Wandungsdicke der grosseren Cyste betrug durchschuittlich
0'15mm., war also verbal tnismassig viel geringer als bei den Cysten, welche Graff
beschrieben hatte (3, pag. 67, zeile 9-10 ; pag. 68, abs. 2 ; taf. XIIL, fig. 4).*

An jedem Cystenende zeigte sich in gleicher Weise wie bei den Exemplaren
McClendon's eine ins Innere fiihrende Offnung, von welchen die von der Mundscheibe
des Wirtes abgewendete in beiden Fallen viel grosser als die andersseitige war und
sich vor dieser ausserdem noch durch eine schwache Aufwulstung ihres Randes

* Dies hangt mit der ansehnlichen Grosse der antarctischen Myzostoma-IndiviAuen zusammen ; derm eine
Cyste, welche ein grosseres Myzottoma beherbergt, wird auch eine ausgedehntere, daher relativ diinnere Wandung
besitzen als eine solche, welche einen kleineren Parasiten umschliesst. Auch unter den von Graff beschriebenen
Cysten besass die kleinste (3, ' Blake '-Exemplar, pag. 68, abs. 2) eine dcrbere Wandung als die (ibrigen.

MV/OSTOMIDAB. 11

ausxcirhnotc. l>i>' kirinere der Cystenoffnungen wurde erst dann aichtbar, wciui man
die in der Cyste enthnltonen Parasiten entfcrnt hattc, so dans Licht durch den
punktfonuigen Porus fallen konnte.*

AUSSERE MORPHOLOGIE DER PARASITEN.

Jede Cyste umschloss, wie bei den Exemplarcn Graffs u. McClendon'a, je ein
-: — tea uni.ii.il mi. I .-in tidaara nlnnlioh-fbnetioittenndM ln.i:\i.lii:iin In
folgendem will ich uacb dem Beispiele der friihercn Autoren der bequemeren
Ausdrucksweise halber das erste als " Weibchen," das letztere als " Maunchcu "
banicbnen, trotzdem beide, wie spiiter nachgewiesen werden soli, a I- morphulogiscbe
Hermapbroditen organisiert sind.

A. Das Weibchen (" female " : Graff, 3, pag. 67 ; " large individual " : McClendon,
4, pag. 121), (Fig. 2 u. 4), war der Lfinge nach in der Cyste gelagert, deren Innenraum
es dabei so vollstandig ausfilllte, dass sein Hinterende in ihm keincn Platz mehr fund
und aus der groaseren der beiden ('ystenoffnungcn etwas herauazuragen gezwungen war.
Auch der Breite nach entsprach das Cystcnlumcn nicht den Dimensioneri des Tiercs,
weshalb dessen laterale Korperpartieu, wie dies schon die beiden vorgenannten Autoren
gescbildcrt batten, dorsalwarts aufgcrollt waren, so dass die Seitenrandcr in der
Medianebene nicht bloss zur Beriihrung sondern auch stellenwcise (am Hinterende und
in der Mitte des Kdrpers) znr Uberlagerung gelangten (Fig. 7 u. 8).

Wie ein Blick auf das in Fig. 4 von der Ventralseite her abgebildete Weibchen
zeigt, ist der (in der Fig. nach oben gerichtcte) Hinterrand desselben medial
ziemlich tief eingezogen. An dieser Stelle springen die benachbarten hintcrsten
Abschnitte der aufgerollten Seitenrander nach riickwarts vor und legen sich zugleich,
wie dies aus Fig. 6 zu erkennen ist, nach Art von zwei einunder iiberdeckenden
Fliigelklappen von seitwarts und hinten her, Ul>er den trichterfOrmig erweiterten
After (ao), sowie ttber die unmittelbar oberhalb von- dicsem gelcgene weiblichc
GeschlechtaOffnung (t>H). Sic schliesscn dadurch cine Art von Kloakalraum ab, der
seinerseits nur vermittels eines sehr klcinen terminalen Porus nach aussen miindct
Zum Zwecke der Defaecation verm5gen sich diese " Analklappen " zeitweise durch
Contraction der entsprechenden Hautmuskelbezirke aufzuwulsten und so die im
Rectum angesammelten Faeces ins Freie gelangen zu lassen ; (dieses Verhalten ist
aus Fig. 4 ersichtlich, wahrend Fig. 6 eine der geschlossenen Analklappen im
Ijangsschnittc zeigt). Wenngleich nun die Analklappen in geschlossenem Zustande—
und dies scheint ihr normales Verhalten zu sein — die unmittelbare Verbindung
des weiblichen Geschlechtsporus mit der Aussenwelt aufheben, so beeintrMchtigen
sie doch keineswegs die standige Communication desselben mit dem von Graff
als "broodpouch" (3, Erklarung zu taf. MM., fig. 4) bezeichneten Behalter.

• Dieces Verhalten konnte tnoglicherweise die Ang»bc Graft's crklaren, dug die von ihm untenuchtcn
Cytten nor eine einzige Oflhong beMnen bitten (8, peg. 66-67). Bei der Kleinheit des betreffenden Poms Ut
denelbe ••Mnrdentlich leicht in Ubenehen I

VOL. IV.

12 DR. RUDOLF HITTER v. STUMMER-TRAUNFELS.

Dieser wird ja dadurch gebildet, dass einerseits von rechts und links her die
aufgerollten lateralen Kdrperabsclmitte, andererseits von unten her die mediale
Riickenpartie des Myzostoma einen dorsal von der letzteren gelegenen, im Durch-
schnitte dreikantigen Raum umschliessen (Fig. 6, 7, 8, br), in welchen die reifen
Eier abgelegt werden, und in dem die ausschliipfenden Embryonen warscheinlich
einen Teil ihrer Entwickelung durehmachen. Mir scheint nun die Einrichtung der
Analklappen, abgesehen davon, dass sic der Afterb'ffnung einen ausgiebigen Schutz
gewahrt, auch fur die Sicherung des Transportes der reifen Eier in die Bruttasche
eine hervorragende Bedeutung zu besitzen. Denn nur dadurch dass der weibliche
Genitalporus in der oben geschilderten Weise nach aussen hin abgeschlossen wird,
vermb'gen die Eier in den Brutraum iiberzutreten, im gegenteiligen Falle miissten
sie, da das Hinterende des Tieres aus der Cyste herausragt, im Momente des
Austretens aus dem Uterus notwendiger Weise ins Freie gelangen.

Wie der Hinterrand des Weibchens so ist auch der Vorderrand desselben
medial etwas eingebuchtet, jedoch bei weitem nicht so stark wie der erstere. Die
Mundoffnung liegt frei, fast am vordersten Korperende, erscheint jedoch etwas nach
der Ventralseite geriickt (Fig. 6, m).

Auf der Dorsalseite des Tieres, also noch innerhalb des Brutraumes offnet sich
ferner jederseits ein kleiner Porus (Fig. 7, $o), welcher unweit vom Korperrande in
der die mittleren Parapodien verbindenden Transversanebene gelegen ist. Diese
zwei Offnungen sind, wie spiiter nachgewiesen werden soil (Pag. 17 und 18), den
beiden mannlichen Geschlechtsoffnungen des Mannchens (Fig. 5, <Jo) homolog, wenn
sie auch beim Weibchen als solche nicht mehr functionieren.

Schliesslich findet sich auf der Ventralseite des Korpers, direct unterhalb des
Afters jedoch bereits ausser dem Bereiche der Analklappen, noch ein sehr kleiner
medianer Excretionsporus (Nephroporus), (Fig. 6, np), der allerdings nur im Wege der
Schnittmethode sichtbar gemacht werden kann.

Seitenorgane (" Saugnapfe") fehlen vollstandig.

Die Parapodien treten als ausserst kleine rudimentare, mit freiem Auge nicht
erkennbare Warzchen an der Ventralseite des Korpers auf (Fig. 2, 4 ; Fig. 7, h). Es
sind ihrer jederseits fiinf in einer genau zwischen dem aufgerollten Seitenrande und
der ventralen Mittellinie verlaufenden Langsreihe gelegen. Sie bergen einen zwar
deutlichen und vollstandigen, jedoch infplge der Puickbildung der Bewegungs-
musculatur kaum functionsfahigen Hackenapparat.

Die Aussenflache (Ventralseite) des aus der Cyste herausgenommenen Weibchens
erscheint makroscopisch im Allgemeinen glatt. Bei dem einen der zwei mir vorge-
legenen Exemplare (Fig. 4) waren Andeutungen von Querfalten zu bemerken, die
jedoch warscheinlich auf Contractionszustande zuruckzufiihren sind, welche durch die
Conservierung hervorgerufen wurden. An jedem der beiden Exemplare fand sich
weiters aussenseitig dort, wo zwischen ihnen und der Cystenwandung das kleinere
Mannchen gesessen hatte, eine durch den Gegendruck des letzteren hervorgerufene

MTZ08TOMIDAE. 13

Macho Delle (Fig. 2, </; Fig. 4), welche furmlich ein Negativ von dessen Dorsalflache
darstellte. l>i.-«'i Kimlnn-k wir.l -;• ': MB lobioden Itei nil da ...,:.;
stettfindenden temporarcn Platzvcrttnderung de« MftnncheiiH entaprechcnd verschieben.
Am conscrvicrtcii In.lividuum crecheint cr durch die vollzogene Httrtung dcr
{ ,, A. '•., lixicrl

B. Das Mdnnchen (" dwarf-male " : Graff, 3, pag. 67 ; " small individual " :
McCieudon, 4, pag. 121), (Fig. 5, 9, 10), saw in der von den l>eiden Autoren
gnehildertcn Weise zwischen Cystenwand und dem Wcibchcn, wobei os dieaem
ungefiihr in der Mitte von dessen Ventralseite rait der eigenen RUckenHache engc
angcpresst war. Kin entsprcchend durch die nicht entleerte Cyste gefUhrter
Querschnitt lieferte daher ganz ein ahnliches Bild wie die von Graff auf tat. xm.
seines ' Challenger '-Reportes dargestellte fig. 4.

Bezilglich der auxseren Configuration des Manuchens kann ich mich in der
Hauptsache (Gestalt, Umriss, Cmisistvnz, Seitenraud, Cirrcn, Parapodicn, Sciten-
organe) auf die von den beiden ohgenannten Autoren geliefertc und auch fiir die mir

\ . .1 •_•. !• •_•• : • !! ' ''• • ' '• • \ • •!!:-• •! • n.. : ]• .--!-:• I '••• • : i . : i • , ', •_• '' •• . :• ii. n. I '.-i • I ; •• n 'A .11 ,-n

nur noch hinsichtlich der Kiirperoffnungen folgende Ergauzungen Injizufiigen :

Wie schon McClendon lieobachtet hat, erscheint der Vorder- und Hintcrrand de«
Mannchcns in eincm schmalen mcdianeu Bezirke ziemlich weit nnch dcr Kttrpcrmittc
hin cingezogcn. Diese Einbuchtungen sctzen sich in ebenso nchmale rimuMiforinigc
Vertiefungen fort, von welchen die vordere und kiirzere ventralscitig, die hintere
etwas langere dorsalseitig vcrlauft. Am Hinterende der ersteren liegt der Mund
(Fig. 9, TH), am Vorderende der letztercn der After (Fig. 9, ao).

Oberhalb von diesem mtindet der dem sogenannten Uterus dcs Weitx:hen8
enteprechende mediate Dorsalabschnitt des Coeloms durch eine klcine Offnung au«t die
wir demgemass als ein Homologon der weiblichen Geschlechtsiiffnung betnehten
milssen (Fig. 9, 60). Die beiden mannlichen Genitalporen (Fig. 5, 4o), welche in der
die mittleren Parapodien verbindenden Transversanebeue gelegen Hind, offncn sich
nicht, wie Graff und McClendon fiir ihre Exemplare nngeben, marginal, sondern auf
der Doraalseite des Tieres, sind i in less dem KOrperrande sehr nahe gerlickt Von
diesem aus zieht zu einem jeden von ihnen eine kurze rinnenfbrmige Vertiefung des
Integuments, welche moglicherweise von den beiden obgenannten Autoren als Ductus
ejaculatorius aufgefasst worden ist, woraus sich ihre von meinem Befundc diffcrierende
Angaben erklareu liessen.

Ein Excretionsporus ist auch beim Mannchen nachzuweisen. Er liegt wie l>eim
Weibchen median auf der Ventralseitc dicht vor dem After (Fig. 10, np).*

* Wie am dem Vontebenden m entnehmcn ist, finden sich uiutlichc aiu Kr.rpor den Wcibcheni
motretende Offinangen Mich in gleicber Lage beim Minnchen vor, cin Vorhaltcn, welche* niit dem Hinweu auf die
•pater ra Wy»ilii'iii1iin morphologuchen and biologischen Bcziehungen der beiden Ge«chlechUformen ichon

U 2

14 DR. KUDOLF EITTER v. STUMMER-TRAUNFELS.

INNEEE ORGANISATION.

Wie in der Lage der Korperoffnungen so besteht auch hinsichtlich der inneren
Organisation beim Mannchen und beim Weibchen trotz deren aiisserlichen Grb'ssen- und
Formdifferenz eine principielle Ubereinstimmung. Diese aussert sich vor allem am
Sexualsysteme, das bei jeder der zwei Geschlechtsformen durch zwei Apparate, einen
mannlichen und einen weiblichen, vertreten ist, von welchen aber beim Mannchen nur
der erstere, beim Weibchen nur der letztere reife Sexualproducte liefert. Die
Differenzen zwischeu den Geschlechtsorganen des Mannchens und jenen des Weibchens
sind also hauptsachlich functionelle und dies gilt im Grunde genommen auch fur alle
iibrigen Organe. Diese erscheinen bei den zwei Geschlechtsformen nach demselben
Plane angelegt und unterscheiden sich nur durch ihr'en Ausbildungsgrad, der seiner-
seits von ihrer Functionsintensitat abhangt. Diese wird wieder durch die physiologische
Sexualitat des betreffenden Individuums bestimmt.*

Aus diesem Grunde halte ich es fiir zweckmassig in Nachfolgendem die einzelnen
Organe gemeinsam fiir beide Geschlechtsformen zu beschreiben und zwar in folgender
durch obigen Gedankengang begriindeten Reihenfolge :—

(1) Weiblicher Geschlechtsapparat.

(2) Mdnnlicher Geschlechtsapparat.

(3) Ernalirungsapparat.

(4) Excretionsapparat.

Auf eine Schilderung des Nervensystems beabsichtige ich aus dem Grunde nicht
einzugehen, weil sich dasselbe in seiner groberen Morphologic nicht von jenem der
anderen Myzostomen unterscheidet, und weil andererseite eine Untersuchung seines
feineren Aufbaues, der bei dem Mannchen und dem Weibchen moglicherweise
interessante Differenzen vergeben hatte, wegen der quantitativen und qualitativen
Unzulanglichkeit des Materials aussichtslos schien.

WEIBLICHER GESCHLECHTSAPPARAT.

Derselbe wird bei den Myzostomen, wie bekannt, von einem Teile der Leibes-
hohle (dem Uterus mit seinen Divertikeln), und weiters von bestimmt localisirten
Wucherungen seiner peritonealen Auskleidung (den Ovarien) gebildet ; (vergl.
diesbeziiglich : Stummer, 6, pag. 574 u. ft'.).

Mannchen : Bei diesem erscheint der zu den weiblichen Geschlechts wegen
umgebildete Leibeshohlenanteil, — im Gegensatze zu den ausschliesslich mannlich

* Durch die physiologische Unisexualitat werden abgesehen vom entsprechenden Geschlechtsapparate
hauptsachlich die Stoflwechselorgane (Darm- u. Excretionssystem), in ihrem Ausbildungsgrade beeinflusst, und
zwar in der Weise, dass die Functionsuitensitat dieser Organe beim Weibchen gegenuber jener beim Miinnchen
erheblich gesteigert erscheint. Der miltterliche Korper, welchem die Ausbildung der grossen und dotterreichen
Eier obliegt, besitzt eben ein weit grosseres StoffwechselbedtirfniB als der vaterliche, der durch die Production
der winzigen Spermien diesbeziiglich nicht so sehr beansprucht wird (vergl. Pag. 20 und 21).

MY/OSTOMIDAE. 15

functiunieremlcn Iiidividueu andcrer Afy;o*t»ina-Arten, bei welr.hcn sich Usher nur
spurlichc Andcutungen dieser Coelompartie uachwciscn 1 lessen, — wohl entwickclt ; er
ist zwar hiiimlich uioht so ausgedehnt, wie bcim Weibchen, zeigt aber dieselbe Untcr-
glicdcruug wie bei dieseiu. Er setzt sieh aus eiuem dorsalen, zwei twit lichen und

fin. -in ventral- n Abschnitte zusammen.

lK-r Cl-StV rill -|, 11, ill ,|,-r 1,,-i :ili.l,-|vn !MTIIi:i|'ll|-'»llti--il t Ull' I I' -II 1.' I <M I- 1-1 1 M> . ' BM

Arten gemciniglich als Uterus bezeichneten Leibeshtihlenpartie. Er stellt ein mediaues,
dorsoventral abgepluttetes Rohr dar, welches ungefahr obcrhalb der Magenmitte
beginnt und %^>n hier aus unUT allmahliger Verschnialerung nach rilckwiirts bis zur
weiblichen Geschlechtsoffuung verlalift, durch welche es ausmiindct (Fig. 9, ua + ub)
Bald hinter seinem verbreiterten Vorderende erscheint seine Seitenwandung rechte und
links in einen kurzen Zipfel ausgezogen, an dessen Lateralende sich je ein Nephrostom
(Fig. 9, rw) tfffnet

Vorne steht der Uterus jederseite durch einen kurzen und breiten Querast mit den
1'i-i.leii Seitenabschnitten der Leibeshohle in Verbindung. Dicse verlaufen y.uniidi.st in
dorsoventraler Richtung langs der rechten und liriken Seitenwand der vordereu
Magenhalftc nach abwarte, um sodann l>eiderseits nach Umfassung der Wurzeln des II.
und in. Hauptdarmastes in den Ventralabschnitt der Leibeshohle iil>erzugehen.

Letzterer besteht aus zwei symmetrischen, durch cine schmale suboesophagealu
Quer-Commissur verbundene IlSlften (Fig. 10, t7). Eine jede von diesen beginnt als
ein langlicher, unter und seitwarts von den lateralen Mtigenpartien gelegener
Hauptraum, von dem weitx;rhin drei Divertikel gegen die Hauptdarmftste ausgehen,
welche die letzteren untcr entsprechender Teilung lings ihrer ganzen Verzweigung
ventralseitig l>egleiten (Fig. 10, 1,3,3).

An der Wurzel des ersteu und dritten dieser Divertikel liogt nun jc ein aus einer
localeu Wucherung des Peritonealepithels hervorgehender Zellhaufen, welcher von
seitw&rts und von unten her in den Hauptraum einspringt Im Ganzen sind daher
vier derartige Gebilde vorhanden (Fig. 10, 0). Ihrer Lage und ihrem histologischen
Aufbaue nach miissen wir sie in Ubereinstimmung mit McClendon (4, pag. 121, abs. 3)
mit den von Nansen (5, pag. 78, abs. 4) bei anderen Myzostomen entdeckten
" problematischen Orgauen " homologisieren, die spftter von Wheeler (7, pag. 178,
abs. 2) als die echten Ovarien erkannt worden sind. Dass sie Ix-im M&nnchen von
M. cysticolum tatsachlich als aolche frliher oder spilter functionieren lasst sich jedoch
nicht nachweisen. Sie machen im Gegensatze zu den product ivcn Ovarien des
Weibchens einen kiimmerlichen Eindruck, so dass man sie entweder als noch
unentwickelte oder als riickgebildete Organe auffassen kann.

Reife Eier oder auch nur solche in der ersten Entwickelung 1ml »• ich innerhalb der
ganzen am Aufbaue des weiblichen Geschlechtsapparates beteiligten Leibeshohle
nicht beobachten konnen. Dieselbe erscheint, — abgesehen von den die Ovarien
darstellenden Wucherungen, — in alien ihren Abschuitten von eincm glcichmassigen
Peritonealepithel ausgekleidet. .

16 DR. KUDOLF BITTER v. STUMMER-TRAUNFELS.

Beim Weibchen zeigt der weibliche Geschlechtsapparat gegeniiber jenem des
Mannchens eine bedeutende durch sein Functioniereii bedingte Ausgestaltung. Vor
allem erscheint die Leibeshohle viel starker entwickelt. Dies gilt insbesondere von
ihren Seitenabschnitten, welche sich lateralwarts ausserordentlich verbreitern, ventral-
warts mit den Hauptraumen des Ventralabschnittes sowie mit den von diesen
ausgehenden Divertikeln verschmelzen und sich so als machtige, die reichverzweigten
Darmaste umhiillende Sacke in den seitlichen Partien des Korpers fast bis ans
Integument hin erstrecken (Fig. 7, 8).

Infolge dessen bleibt das Korperparenchym, welches beim ManiTchen noch in
relativ grossem Massstabe samtliche Zwischenraume zwisclien den einzelnen Organen
ausfiillt, beim Weibchen hauptsachlich nur im Bereiche eines medial en, etwa ein
Fiiuftel der Korperbreite einnehmenden Leibesabschnittes in Form grb'sserer zu-
sammenhiingender Bindegewebspartien erhalten. Diese umhiillen den Pharynx, den
Magen, den Euddarm, den Uterus, die Nephridien und das Bauchmark (Fig. 8).
In den Seitenabschnitten des Korpers hingegen erscheint das Parenchym auf diinne
die primaren und secundaren Septen begleitende Schichten sowie auf einen relativ
schmalen unterhalb des Integuments verlaufenden Streifen beschrankt der nur dort,
wo die Parapodien zu liegen kommen, zu etwas bedeutenderen Complexen auswachst.

Von jenem beim Mannchen als Ventralabschnitt der Leibeshohle bezeichneten
Raum bleibt beim Weibchen nur die suboesophageale Quercommissur der beiden
Hauptriiume gesondert erhalten. Hinter dieser finden sich jedoch noch einige
breitere, zwischen Magen und Bauchmark verlaufende Querdivertikel der lateralen
Leibeshohlensacke, durch welche die letzteren mit einander in unregelmiissiger Weise
verbunden erscheinen (Fig. G, 7, vl).

Auch der Uterus (Fig. 6, 7, 8 IM, ub) zeigt im Allgemeinen grossere Dimensionen
wie jener des Manncliens. Man kann an ihm einen kiirzeren vorderen und einen
langeren hinteren Abschnitt unterscheiden. Der erste (Fig. 6, 7, ua) ist der breitere
und besitzt im Querschnitte die Form eines gleichschenkeligen Dreieckes, welches
mit seiner Grundlinie der -dorsalen Magenwand fast unmittelbar aufsitzt. Er liegt
ober den Abzweigungsstellen der Hauptdarmiiste und steht beiderseits fast
seiner ganzen Lange nach in continuirlicher Verbindung mit den lateralen
Leibeshohlensacken. An seiuem hintersten Ende ciffnet sich in ihm von jeder
Seite her ein Nephrostom (Fig. 8, ns). An diesen beiden Stellen erscheint
jedoch seine Wandung nicht in derartige trichterfb'rmige Zipfel ausgezogen wie
beim Mannchen. Hinter der Einniiindung der Nephrostome geht er in den
schmaleren hinteren Uterusabschnitt iiber. Dieser (Fig. 6, ub) besitzt im Allgemeinen
einen rundlichen Querschnitt mit radialen gegen. seine Langsaxe einspringenden
Epithellamellen und verlaiift unter allmahliger Verringerung seines Kalibers ziemlich
gestreckt oberhalb der hinteren Magenhalfte und oberhalb des Rectums bis zur
weiblichen Geschlechtsoffnung. Ein die letztere abschliessender besonderer Sphinkter
ist nicht ausgebildet

MYZOSTOMIDAE. 17

l>i' Ovnrien I Kit.'. 7. t li> ••_<• n Mdimill /u iwd --m il.'i \.'!iti.il. n Bdtl •!• i
LeibeshtthlensScke unweit dea Magens und zwar an genau densclben Kflrperstellen,
wie IxMin MMnnchen, n&mlich das vorderc unter und hinter dem i., das ruYkwartige
hiiit'-i dem II. Hauptdarmastc ; (in der Fig. 7 ist das rtickwartige (Jvarium dcr
linken Seite zu erkcnnen). Sic sind viel mftchtiger entwickelt (ungefahr 15-20 mal
so grow) wie jenc des M&nnchens und bestehen auch aus ciuer viel grttesereu
Auzahl von Zellen. Unter diesen kann man tareite sehr deutlich die kilnftigen
Kixi'llen mit ihren groMen blaschenfbrmigen Kcrnen von den kleinc und dunkle
Kerne l>esitzeuden Nfthrzellen ("accessory cells": Wheeler, 7, pag. 178, abs. 2)
unterseheiden.*

Die Innenwandung der Leil>eshtihlensacke und ihrer Divcrtikcl wird teils von
••in. -in Peritoneal-, teils von einem Pseudoepithel (Stummer, 6, pag. 578) gclnldct
Dieae Auskleidung zeigt iil)crall einen reichlichen Belag von sich daseHwt entwickelndcn
Eiern, welche bier in alien Bildungsstadien von der eben vom Ovarium losgelosten
Zelltriade (Wheeler, 8, pag. 233, abs. 2, " triplet-cells ") an bis zum liefruchtungs-
fahigcn la vertretcn sinti

MANNLICHRR API-ARAT.

Beim Miinnchen ist derselbe licreits von Graff und McC'Icndon insoweit gciuiu
beschrieben wortlen, dass ich hierzu nur folgende unwesentliche Erganzungcti zu lieferu
vermag. In jedem der )>eiden der compacten Form (vergl. Pag. 6, Amncrk. 1, li)
angehorenden Hoden liegt ein deutlicher ccntralcr Sammclraum, der direct mit der
GeschlechtHoiTnung communiciert und sowohl Spermien aln auch cine Mcnge von
sich nicht weiter entwickelnden, vielleicht auch degenerierten Spermat<x;yten
(Stummer, 6, pag. 583) enthalt.

Beim \\reibchen fand ich die beiden Hoden genau in derselbcn I^age uud in der
formell gleichen Ausbildung, wie beim MSnnchen. Sie liegen jederseits in der Mittc
des aufgerollten Seitenrandes, dicht neben dem mannlichen Geschlechtsporus, und
stellen eine deutlich vom iibrigen Korperparenchym abgegrenzte Mas.se von reticu-
Ifirem Bindegewebe dar, dessen Intercellularraume zu grosseren Blasen ausgedehnt
sind (Fig. 7, /). Die letzteren fliessen in der Mittc dieses Complexes zu einem dicht an
die mannliche Geschlechtsdffnung anschliessenden centralen Sammelraum zusammen.
Wahrend aber die Intercellularraume beim Mannchen dicht mit Keimzellen bezie-
hungsweise mit deren Teilungsproducten erfiillt sind, enthalten sie l>eim Weibchen
relativ nur wenige Zellen. Von diesen lassen sich zweierlei Formen unterscheiden :
Die einen gleichen auffallend den von Graff (2, pag. 64, Abs. 1) als " Rundzellen "

i«t, dmss schon Graff die Ovarien bci If. ryititolum gecehen hat. Nur warden lie Ton
ihm zufolge der dunali noch hcmchenden An»chaaung, dM« die Ventweigungen dm Uteru« (Leibethohle and ihre
Dirertikel) d»» Oruiam bei den Myzortomen dmteUten, als rudimentare Hoden gedentei (Graff, 8, pa«. 87,
at*. 3). Nach der Abbildnng, welche der genannte Autor von dieien Gebilden geliefert hat (8. taf. sin., fig. 4, /).
•Mfcul"* es jedooh zweifelloi— anch McClon.lon ist dieaer Meinong (4. pag. 131. aba. 8)— dan lie mit den oben
geschiUerten eohten Orarien identificiert werden mttsaen. Die Prioritat, die wahren Mytnttoma-Ovuien roent
goehen xn haben gebuhrt daher Graff, und nicht Kanaen, dem tie biaher mgeKhrieben worden Ut.

18 DE. RUDOLF RITTER v. STUMMER-TRAUNFELS.

beschriebenen freien Bindegewebselementen (Stummer, 6, pag. 589, letzt. Abs.). Die
anderen machen jedoch vollstandig den Eindruck von degenerierten Zellen. Sie
besitzen einen ralativ kleinen unregelmassig gestalteten Kern, der sich mit Haema-
toxylin intensiv farbt und dabei neben einem homogenen Aussehen auch ein
eigentiimliches Lichtbrechungsvermogen entwickelt. Ihr Cytoplasma 1st nur in selir
geringer Menge vorhanden und im giinstigsten Falle als ein kaum tingierbarer,
schmaler, den Kern umhiillender Saum zu erkennen. Von Spermien fand sich
innerhalb des ganzen Bindegewebscomplexes keine Spur. Der centrale Sammelraum
desselben enthielt dagegen bei jedem der mir vorgelegenen Exemplare eine feinkornige
oft fadig angeordnete und auch aus der Geschlechtsoffnung dringende Masse, in
welche zahlreiche der oberwahnten anscheinend degenerirten Zellen eingebettet
waren.

Ich stehe nicht an, die eben geschilderten Organe auf Grund ihrer Lage und ihrer
Ausbildung als functionslose Hoden zu betrachten, die jenen des Mannchens vollstandig
homolog sind. Es ist infolge des ungeniigenden Materials selbstverstandlich sehr
schwer zu entscheiden, ob sie als Riickbildungen von mannlichen Gonaden aufzufassen
sind, die in einem friiheren Lebensalter des Individuums functioniert haben, oder ob
sie auf einem friihen Entwickelungsstadium stehengebliebene, niemals zur sexuellen
Betatigung gelangte Apparate darstellen. Meiner Meinung nach diirfte die erstere
Annahme die zutreffende sein, insbesondere deshalb, weil sich ahnliche Verhaltnisse
(Proterandrie mit spaterer Riickbildung des Hodens) bei zahlreichen frei lebenden
und auch bei einigen entoparasitischen Myzostoma-Aiten, mit Sicherheit haben nach-
weisen lassen (Wheeler, 8, pag. 288-289). Schliesst man sich dieser Ansicht an, so
konnte man von den oberwahnten, in den Intercellularraumen vorhandenen Zellen,
die von mir als anscheinend degenerirt bezeichneten Elemente als riickgebildete
Keimzellen, jene aber, welche den " Rundzellen " Graffs ahneln, als secundar in die
Follikel eingewanderte freie Bindegewebszellen aufFassen. Die feinkornige, im
Sammelraum enthaltene Masse wiirde dann auf Reste von abgestorbenen Spermien
oder von' zugrundegegangenen Sperm atocy ten etc. zuruckzufiihren sein.

Derartige functionslose Hoden sind weder von Graff noch von McClendou bei den
von ihnen untersuchten " Weibchen " beschrieben worden. Ich glaube aber, dass diesbe-
ziiglich ein Ubersehen von Seiten der genannten Autoreiuvorliegt, da diese Gebilde bei
dem vollstandigen Fehlen von tatigen Keimzellen und von Spermien nicht sofort als
Hoden zu erkennen sind. Graff diirfte ihnen schon deswegen keine weitere Beachtung
geschenkt haben, weil er iiberzeugt war, in den von ihm beim Weibchen beschriebenen
"collections of small cells closely resembling the immature testicular follicles . . . ."
(3, pag. 67, abs. 2) Hodenrudimente vor sich zu haben. Dass diese seine Auffassung
eine irrtumliche war, und dass diese Gebilde die Ovarien darstellen, habe ich schou
friiher (Pag. 17) erwahnt.

MYZ08TOM 1 1 • \ I 19

ERNAHRUNUS-APPARAT.

Deraelbe 1st nach <li-ni s.-li.m von dm iihrigen MjSOCtomen her bekannten TypUH
gdbaut

Wegen der vi.-l rinfaoheren Verhaltnisse, welche er IKMHI Mdnnchen aufweist, aei
er ziu-rst an diesem gcschildcrt : Rezuglich von Mund und After verwewe ich auf die
8chon friiher (Pag. 13) gelieferten Angaben. Der lange, cylindrisohe, an seiuem
Vorderende nicht mit Tastpapillcn ausgertistete Pharynx (Fig. 9, Inn), steht
\.-imitt.-l- «-inrs kur/ou "Oesophagus" mit dem Magen in Verbiudung. Dieser
(Fig. 9, mg), zerfttllt durch cine seichte, quere Einschniirung in einen vorderen
weiten und einen hinteren engeren Abschnitt Vom ersteren gehen ventral jederseits
<livi Hunptdarniaste (Fig. 9, I, II, in) ab, welche sich nach kurzem Vcrlaufc teilen
und mit ilircii wenig zahlreichen, stumpf endigenden Divertikeln bis zur Rand/tone des
Ktfrpers reichen. Von dieaen drei HaupUlnrmasten verzweigen sich jedereeite die zwci
ereten in der vor dem Ploden gelegenen, der dritte in der hiuter diesem )>cfindlichen
Kurjxjrpartie. Der hintere Magenal>schnitt ist kleiner als der vordere und geht
vermittels eines nur am contrahirten Tiere (Fig. 5) sichtbareii Absatzes in das
rohrenfbrmige dorsoventral compresse Rectum (Fig. 9, r) tiber. Die AftcrUffnung
kann durch einen sich aus dem Hautmuskclschlauche ditferenzierenden schwai-li<-n
Sphiukter geschlossen werden.

Der gesamtc Venlauungsapparat ist beim Manuchcn cnteprcchend dessen
geringem Ernahrungsltodiirfnissc recht eiufach gebaut Seine einzcluen Abschnitte
setzen sich aus einer weit geringeren Anzahl von Zellelementen zusammen wic jene des
Weibchens. Auch seine Muscularis ist nur schwach entwickelt und an manchen
Stellen Uberhaupt nicht nachzuweisen.

Beim Weibchen hingegen dessen Stoffwechselbediirfnis infolge der reichlichen
Eierproduction ungemein gesteigert ist, zeigt der Erniihrungsapparat einc vi.-l
machtigere Anagestaltung : vor allem ist der Magen (Fig. G, 7, 8, HJ^) relativ
bedeutend voluminoser wie IMMIU Mamie-hen. Er wird seiner ganzcn Ausdehnung nach
sanit dem Oesophagus (Fig. 6, <>>) von einem sehr auflallendcn Wimpercpithel
ausgekleidct, dessen zahlreiche und dicht stehende Cilien wcgen ihrer ungewohnlichcn
Lange bemerkenswert sind (Lange der Epithelzellen : 0'OG8 m., jeue der Cilien:
0*1 m.). Die Hauptdarmantc sind distal wiirte ausserordentlich reich verzweigt
Ihre Divertikcl liegen vollkommen in die lateralen LeilM»bohlens&cke eingcbettet
(Fig. 7 u. 8, da), und verbreiten sich in grosser Zahl insbesondere innerhalb der
aufgerollten Seitenpartien des Korpers, wo sie sich mit ihren Termiualenden bis
knapp unter den HautmuskeLschlauch erstrecken. Das Hinterende des Magens
eracheint seharf von Rectum (Fig. 6, r) abgesetzt Das letztere verlauft in ziemlich
gerader Richtuug bis zum After. Es wird von einem Wimpcrepithel ausgekleidet,
das in radiar in sein Lumen einspringende LMngsfalteu gclegt ist Eine
durch aiissere Ring- und innere Lftngsfasern hergestellte Muscularis findet sich

TOU IT. X

20 DE. RUDOLF EITTER v. STUMMER-TRAUNFELS.

sowohl am Magen als auch am Rectum. Sie erscheint insbesondere am ersteren
kraftig ausgebildet ; an letzterem ist sie jedoch schwach entwickelt, ihre beiden
Schichten werden aber hier noch durch eine periphere dritte Lage vervollstandigt,
die aus locker verfilzten Fasern besteht. Ein besonders difFerenzierter Sphinkter
findet sich weder an der Grenze zwischen Magen und Eectum, noch um die
Afterb'ffnung. Fiir die letztere wird die Function eines solchen wohl von den benach-
barten Partien des Hautmuskelschlauches iibernommen. Ausserdem erscheint aber
der After fur gewohnlich auch noch durch die friiher (Pag. 11) beschriebenen
Analklappen verschlossen.

NEPHRIDIEN.

Dieselben beginnen jederseits mit einem sich in dem vorderen Uterusabschnitt
bffnenden cylindrischen Nephrostom (Wimpertrichter) (Fig. 8, 9, ns). Dieses setzt
sich in einen engen, beim Mannchen kiirzeren, beim Weibchen, wegen dessen
bedeutender Dicke, langeren Nephridialcanal fort, welcher in ziemlich geradem
Verlaufe langs der Seitenwand der hinteren Magenhalfte nach ab warts zieht (Fig. 8),
um sich etwa in halber Magenhohe plotzlich (diese Stelle ist in Fig. 8 und 10 mit *
bezeichnet) zu einer grossen, vielfache Windungen und Einschniirungen zeigenden
Endblase zu erweitern. Die letztere (Fig. 6, 8, 10, rib) besitzt in ihrem proximalen
an den Nephridialcanal anschliessenden Abschnitt eine bedeutende Ausdehnung,
verschmiilert sich aber distalwarts allmahlig zu einem relativ langen, ziemlich gerade
nach hinten verlaufenden Canal, welcher mit jenem der anderen Kb'rperseite gegen die
Medianebene convergiert und sich schliesslich mit ihm zu einem kurzen unpaaren
Endstiick (Fig. 6, 10, nu) vereinigt, das durch den median, knapp unter der
Afterb'ffnung gelegenen Excretionsporus (Fig. 6, 10, np) ausmiindet.

Wahrend der Nephridialcanal von einem verhaltnismassig niedrigen Wimper-
epithel ausgekleidet ist, findet sich in der Endblase bis zum Excretionsporus hin ein
deutliches Driisenepithel, dessen hohe keulenfbrmig gestaltete Elemente (aus Fig. 8 ist
die Lagerung und die relative Hbhe derselben ersichtlich) der Wimpern entbehren.
Eine besondere Muscularis liess sich nur. an der Endblase nachweisen, aber auch
hier ist sie sehr schwach ausgebildet uud besteht lediglich aus einer Ringfaser-
schichte.

So wie der Ernahrungsapparat, zeigen auch die Nephridien des Weibchens
gegeniiber jenen des Mannchens eine viel machtigere Entfaltung, die sich insbesondere
in der relativen Lange des Nephridialcanals sowie in der Grbsse der Endblase aussert.

SEXUALVERHALTNISSE UND DIMOBPHISMUS DER GESCHLECHTSFORMEN.

Graff (3, pag. 11-12) unterscheidet unter den von ihm beschriebenen cysticolen
Myzostomen zweierlei in Bezug auf ihre Sexualverhaltnisse divergente Formen und
zwar rein dioecische und rein hermaphroditische.

MYZOSTOMIIUE. 21

Zur enteren Gruppe gehtfren : M. tenuispinum, M. willemoeai, M. inflator und
M. murrayi. Bei diesen Species findcu sich ID je eioer Cyst* zwci auffallcnd dimorphc
Imlividuen vereinigt, cin klcines, nur mHnnlich organisirtes " Mannchen " und ein
grosses, nur weihliche Charaktere besitzendes " Weibchen." *

Die zweite Gruppe wird durch M. pentacrini und M. deformator rcpraesentirt,
Die Individuen dieser beiden Arten leben entweder einzeln in je einer Cyste, oder
aber sic wenlen, falls sic zu mehreren (zwei bis drei) eine solche bewohnen, in dieser
durch cntaprcchende Scheidewande von einander getrennt. Zur innigen Beriihrung
zweier Exemplare, wie dies bei den rein dioecischcu Species die Rcgel i«t, kommt es
daher l«-i ibnen nicht. Ebensowenig bestebt bier ein DimurphiHmus unter den InsoHscn
einer und derselben Cyste. Jedes Individuum zeigt sowohl mannliche \\\< aurh
weibliche Charaktere und functioniert auch dementsprecheud als Hcrmaphrodit.f

Das M. cysticolum betrachtete Graff als eine in sexueller Beziehung intermediate
Form zwischen den rein dioecischen und den rein hermaphroditischen Species (3,
pig. 67, Abs. 2), da er vermutete, dass bei ihr das " Weibchen " Hodenrudimente
betJBHc.

Auf Grand unserer heutigen Kenntnisse iiber die Anatomic dieser Art diirfen wir
jedoch behaupten, dass ihre beiden dimorphen Geschlechtsindividuen morphologisch
als Hermaphroditen organisirt sind, in physiologischer Beziehung jedoch nur
eingeschlechtlich und zwar entweder nur als Mannchen oder nur als Weibchen
fniii-ti'niieren.J Von einem Sexualdimorphismus in der gebrauchlichen Bedeutung
kann daher bei M. cysticolum nicht die Rede sein. Man wird vielmehr die bei
dieter Species herrschenden Geschlechtsverhaltnisse als das Endergebnis einer
physiologischen Sexualdifferenzierung betrachten miissen, die sich im Laufe der
Ontogenese auf (morphologisch-) hermaphroditischer Grundlage vollzieht und dabci niit
einer Dimorph'osierung der sich entwickelnden Geschlechteindividuen verbunden i-t.

Der typische Afyzostoma-H&lritus und die relativ geringe Grosse des Mannchens,
ferner die verhaltnismMssig sehr einfache Ausbildung seiner inneren Organe sowie
der primitive Aufbau seiner Korpcrgewebe lassen dasselbe viel urspriiuglicher und
unveranderter erscheinen als das Weibchen, weshalb wir annehmen miissen, dass die
Dimorphosierung hauptsachlich an dem letzteren wahrend seiner Entwickelung voll-
zieht Es wird hicr die allmahlig sich eiustellende physiologische Praeponderanz
der weiblichen Gonaden nicht nur den hoheren Ausbildungsgrad des weiblichen
Sexualapparates (LeibeshOhlc), sondern auch die functionelle Ausgestaltung jener

• Mar einnuU wurde von Onff (8*, pag. 8) in einer durch M. urillemoeri herrorgerufcncn Cyste blow ein
einzige* und swu minnlich functionierendcs Individuum rorgefunden.

t In dieM Oruppe gehoren noch : il. beardi and If. platypu* (Oraff 8ft, p«g. 18 a. 14 ; far M. plat y put niche
•uch : Wheeler 8, p»g. 314), ferner U. belli, M. cryplopodium and M. errmita (Wheeler 8, p»g. 246 261) lowie
wancbeinlich »uch du If. clarki (McClendon 4. pag. 121-122), obwohl e» dem letztgenannbm A a tor nicht
geglackt ist, M in Ton ihtn antenachten Individuen Hoden oder Be«te von nolchen •nfniflnflan

t Ich vennate, dmM ihnlichc Yerhaltniise auch bei den Ubrigcn von Oraff all " dioccuch " beztichneten
Arten hcmchcn. Eine diesbesttgliche Entocheidung ist allerding* nor init Hilfe der Schnittmethode en
erbringcn, die der genannte Autor, wie au» dem betreflienden Terte and aat den von ihm gelieferten Abbildungen
, M OMB Species nich angewandt hat.

x 2

22 DR. RUDOLF RITTER v. STUMMER-TRAUNFBLS.

Organe bedingen, welche fur die Eierproduction eine nur indirecte Bedeutung haben :
Darm und Excretionsapparat (Stoffwechselorgane). In zweiter Lime wird der
correlative Einfluss der auftretenden physiologischen Unisexualitat bei dem sich zum
Weibchen entwickelnden Individuum auch aiisserlich und zwar an der relativen
Dimensionierung, der Consistenz und der Form des Kbrpers, ferner am Bewegungs-
apparate zum Ausdrucke kommen : Die reiehliche Anzahl der producierten Eier,
welche innerhalb der Leibeshbhle infolge der Aufnahme von Dotter zu einem
Vielfachen ihrer urspriinglichen Grb'sse heranwachsen, bedingt einen verhaltnismassig
umfangreichen miitterlichen Kb'rper mit derber dem bedeutenden Innendrucke
gewachsener Wandung. So kommt es denn schliesslich dazu, dass die Dirrrensionen
des Weibchens jene des Mannchens weit iibertreffen. Die Breitenausdehnung des
ersteren wird sogar eine so grosse, dass ihr das Cystenwachstum nicht zu folgen
vermag, weswegen die Seitenrander des Tieres gezwungen sind, sich median warts
in der friiher beschriebenen Weise einzurollen (Graff, 3, pag. 12, abs. l). Da ferner fiir
das Weibchen die Bewegungsmoglichkeit innwhalb der dem Kb'rper enge anliegenden
Cyste verloren gegangen ist, so werden die Parapodien als fiir die Eiproduction
iiberfliissig gewordene Korperteile eingeschmolzen. Es findet sich "von ihnen nur
mehr der Hackenapparat, wahrend die hautigen Teile des Fussstummels fast
vollstandig in die Leibeswand aufgenommen erscheinen. Dementsprechend ist auch
die Stummel- und Hackenmuskulatur rudimentiert.

Bei dem sich zum Mannchen entwickelnden Individuum braucht sich die
Sexualdifferenzierung fast nur auf die functionelle Ausgestaltung des mannlichen
Apparates zu beschranken ; denn der Darm und der Excretionsapparat werden
hier jenen Ausbildungsgrad kaum iiberschreiten, welchen sie bereits bei dem der
Geschlechtsform unmittelbar vorangehenden Jugendstadium besassen. Beim
Mannchen erfahrt eben das Stoffwechselbediirfnis durch die Erzeugung von Sperma
keineswegs eine so wesentliche Steigerung, wie beim Weibchen durch die
Versorgung der Eier mit Dotter. Die Production der kleinen, wegen ihrer Form in
compendibsester Weise zusammenpackbaren Spermien benbtigt nur relativ geringe
Korperraume, wogegen fiir die Absatzmbglichkeit des Samens eine unbehinderte
Bewegungsfahigkeit des Tieres vorteilhaft erscheint. So besitzt denn das
Mannchen einen kleinen, zarten und abgeflachten Korper mit wohlentwickeltem
Parapodialapparate. Es spricht vieles dafiir, dass sich dasselbe auch im Habitus
nicht wesentlich von jener Jugendform unterscheidet, die ihm in der Ontogenese
uumittelbar vorausgegangen ist.

Die Entwickelungsstadien, auf welchen sich die Sexualdifferenzieruug bei den zwei
Geschlechtsformen herausbildet, sind uns leider noch vbllig unbekannt und daher bleibt
die Frage noch offen : Gilt auch fur das M". cysticdum die Teorie Wheeler's von den
successioen Sexualphasen (8, pag. 288-289), oder herrschen diesbeziiglich bei dieser Art
andere Verhaltnisse ?

Nach den Beobachtungen des ebengenanntcn Autors ist die Sexualdiffereuzierung

MY20STOMIDAE. 23

dcr Myzoatontcn cine IDfleeeoivo, das heisst, CH tritt ini I^cbcn cinoH jedcn
nach Alilauf einea geschleditli. h inditlerrntrn .lu^tMidstadiunis zucrst cine niMnnlirlie,
spa tor hingegen cine wcibliclic Geschlechtsperiodc auf. Bei den mciston Artcn fand
Wheeler, daas diese bciden " Sc.xtialpluiseii " diirdi i-in.-n intermediaren functioncll
hermaphroditischcn Zustand verbunden sind ; cr beolwchtete jedoch auch Specie* (z. B.
^f. pulvinar, 8, pag. 289), bei welcheu .die«e Zwischenperiode ausfallt, so da»H die
uiannliche und die weiblicbe Geschlechtsphase direct an cinander nnschliessen.

Ich halte es nun durchaus nicht fiir unwarscheinlich, dass dienc Teorie Wheelers
auch fiir das M. cysticolum Geltung besitzt Das Vorhandeusein von Ovarien Iteini
Mannchen und von Hodenresten l>eiiu Weibchen lassen diesc Annahnie zu, in»d>eHondere,
wenn man gcneigt i-i . die Ovarien den Mannchens als sich erst entwickelnde, aber noch
nicht functionsfahige Organe aufzufassen. Wenn wir demgcniass das .Mannchen als
ein frlihes Entwickelungsstadium des Weibchens lx;trachten, so ware aber auch der
auffallende Umstand zu erklaren, dass in alien bisher beobachteten Fallen (10)* in
jeder Cyste nur die beiden extremdimorphen Individuen, aber keine ontogenetiHchen
Zwischen- oder Ubergangsstadien, deren Vorkommeu man ja in Consequenz der obigen
Teorie voraussetzen muss, angetroflfen wurden. Wheeler, welchcr da* Bediirfnis
diesbeziiglich zu einer plausiblen Vorstellung zu gelangen wohl gefUhlt hat tr. gpricht
sich darilber folgendermassen aus : " v. Graffs supposition that the young Myzostomes
associate in pairs and together take part in forming a gall seems to me hardly
plausible. Judging from my observations on M. glabrum and pulvinar, both of which
show a distinct tendency to occur in pairs, each consisting of a senior and junior
individual, I believe that in the case of the cysticolous species the gall must be formed
by a single individual, and that later a young Myzostome, when it abandons its
pelagic trochophore stage, must enter through the aperture of the gall and settle down
to a quiet life with the senior individual. The latter prolmhly dies at the end of it«
female stage, and, undergoing decomposition, may perhaps serve as food for its still
vigorous junior partner. This one in turn may thereupon become the senior partner
of another young Myzostome, and so on. According to this view, all the cysticotoiut
Myzostomes of a given species would not be cyst-producing, but only those which,
instead of entering the orifice of a pre-formcd gall in their juvenile stage, happen to
settle between the arm-joints or contrive to work their way into the calcareous skeleton
of the Crinoid " (8, pag. 259 und 260, anmerkg. 1 ). Ich glaube dass es zur Erklarung
des beredeten Umstandcs einer derartigen etwas phantastisch anmutenden Annahme
nicht bedarf. Ohne mich diesbeziiglich auf Beobachtungen stiitzen zu ktfnnen, halte
ich es doch in h5herem Grade fiir warscheinlich, dass das mannliche Individuum cines
encystirten Part-hens nach dem Absterben des durch die reichliche Eierproduction
srhliesslich erschOpften Weibchens die Cyste durchbricht, urn dann einige Zeit,
wah rend welcher es eine functionell hermaphroditische oder vielleicht cine geschlechtlich

• 18 Cyrten (19 TOO der « Huiler '-Expedition, 1 ron der • BUke '•Expedition). beoUehtete Onff
(3, p«g. 00 and 68) ; 1 Cyrte kg McClendon vor ; 3 Cy*ten warden von der ' DUcovery ' heimgebrmcht.

24 DR. KUDOLF R1TTER v. STUMMER-TRAUNFELS.

unproductive Ubergangsperiode vom mannlichen zum weibliehen Zustand durchlauft,
frei auf den Armen seines Wirtes weiter zu leben. Man kann sich nun leicht
vorstellen, dass bei dieser Gelegenheit eine von den herumschwarmenden Larven
derselben Art in ahnlicher Weise, wie es bei andereu Myzostomen der Fall ist, sich nach
Abstreifung ihres Wimperkleides als ktinftiges Mannchen auf den Riicken des
betreffenden jetzt zum functionierenden Weibchen werdenden Individuums anzusiedeln
vermag. Das letztere bohrt sich dann nach vollstandiger Riickbildung seiner mann-
lichen Keimzellen an einer passenden Stelle in das Integument des Wirtsarmes ein,
von dem aus eine neue Cyste gebildet wird, welche das junge nun aus einem
functionierenden Mannchen und Weibchen bestehende Parchen allmahlig umgiebt.

Wenn wir uns also vorderhand beziiglich der Sexualverhaltnisse des M. cysticolum
der eben besprochenen Teorie Wheelers als der bis heute noch die meiste
Warscheinlichkeit besitzenden Erklarung anschliessen, so ist immerhin auch die
Moglichkeit nicht von der Hand zu weisen, dass gerade bei der vorliegenden Species
mit ihrem hochgradig ausgebildeten Sexualdimorphismus sich der ontogenetische
Entwickelungsgang der GeschlechtsdifFerenzierung anders abspielt, als bei den von
dem genannten Autor untersuchten Formen. Es ware ganz gut denkbar, dass von zwei
friihzeitig associirten und urspriinglich als Hermaphroditen angelegten Individuen sich
das eine unter Sistierung der weibliehen Fortbildung * zum Mannchen, das andere
unter Riickbildung der mannlichen Keimzellen zum Weibchen entwickelt, so dass sich
die Sexualphasen nicht successive an einem einzigen Individuum abspielen, sondern
sich ziemlich gleichzeitig auf zwei zu einem Parchen vereinigte und gleichalterige
Exemplare aufteilen wiirden. Der Gegensatz, in welchem das M. cysticolum unter
dieser seinerzeit schon von Graff (3, pag. 11-12) ausgesprochenen Annahme zu den von
Wheeler beobachteten Myzostomen zu stehen kame, hatte nichts Auffallendes an sich,
da wir ja bei vielen anderen Parasiten die maunigfachsten Anpassungen hinsichtlich
ihrer Sexualverhaltnisse vorfinden, so zwar, dass die betretfenden Einrichtungen auch
bei verschiedenen Arten einer und derselben Gattung differieren konnen.

Die Geschlechtsverhaltnisse der Myzostomen scheinen uberhaupt verwickelter zu
sein als man bisher angenommen hat, und die Teorie Wheelers von den successiven sich
an einem und demselben Individuum abspielenden Sexualphasen wird moglicherweise
nicht fur alle Arten aufrecht erhalten werden konnen. Leider ist bisher fast jeder
diesbeziiglichen Eutscheidung der Mangel an ausreichendem insbesondere verschieden-
alterigem Vergleichsmaterial hindernd in dem Wege gestanden. Die histologische
Untersuchung der Gonaden bloss eines einzigen oder nur weniger Individuen vermag
eben nur in den seltensten Fallen iiber diese Frage Aufschluss zu geben vor allem aber
dann nicht, wenn — wie dies zumeist zutrifft — der Erhaltungszustand des Untersuchungs-
Materials kein ganz einwandfreier ist.

* Man wird dann die Ovarien des Miinnchens als rudimentare Organe aufzufassen haben.

25

LITBRATURVERZEICHNIS.

1. F. DOKI.KIX. Ortaiienfahrt. Ixif&ig und Berlin, 1906.

2. L. v. GRAFT. Du Oeniu Myxortoma (F. 8. Leackart). Leipzig, 1877.

8. I., v. GRAFT. Report on the Myzoatouiida collected during the Voyage of 11. M.S. ' Challenger ' during
the years 1873-70. Rep. 'Challenger' Exped. vol. x. (1««4), pog. 1-H2, tab. i.-xvi.

3a. I., v. GRAFF. Supplement ta dem vontchcndcu Report. Rep. 'Challenger' Exped. vol. xx. (1887),

pag. 1-16, tab. i.-iv.
4. I. F. MoQumnr. The Myiostomea of the ' Albatross * Expedition to Japan. Hull. Ainer. Museum

of Natural History, vol. XXIH. (1906), pag. 119-180, tali, xv.-xvu.
6. F. NAJTWDC. Bidrag til Myzostomernes Anatomi og Histologi. Bergen, 1X85.

6. R. RITTKR v. SruxxKR-TRArxFKLS. Beitragc zur Anatomic und Hiotologie dcr Mjzostomeu :

I. Mrzoctotua Mteriae Marenz. Zeitachr. f. wiraenach. Zoolog. i.xxv. (19U8), pag. 4U&-59.1), taf.

XXJCIV.-XXXVIII.

7. W. H. WHKKL.KR. Protandric Hermaphroditiam in Mrzostoma. Zoolog. Anzeigcr, xvu. (1894),

pag. 177-182.

8. W. M. WUKEI.KR. The Sexual PhaMfl of Mvawionw. Mitth. Zoolog. St. Neapel. XH. (1890),

pag. 227-802, tab. x.-xn.

ERKLARUNG DER ABBILDL'NGEN.

FCR ALLE FIGUREN GILTIGE BUCHSTABENERKLiBUHO.

aS, After. ' o,

Aw. Bulbus mnsculosns. or,

br, Bruttasche. & 6,

f. Cirrus. 5 ...

m, Banchinark. r,

i/, dellenformiger, durch das Mtinnchen veran- rt,

lasster Eindruck. w,

dfl, Darmast. '.

A, Hackenappurut. HO,

m, Hand. ub,

my, Magen. r/,

m«, Mnkelseptnni. v,

n, Nepbridium. I, II, III,

nb, Endblaw d« Nephridiums. 1, S, 3,

np, Ncphridialporus.

tu, Nephrostom. *,

mi, nnpmares Endsttick der Nepbridien.

In

Ovarium.

Oawphagos.

in.iiiiilichc (k«chlecbtaofTnung.

weibliche Gcuchlechtaoffnnng.

Rectum.

KunelUache.

lUndzone.

Hoden.

vorderer Uteru8abw;hnitt.

binterer Uteriuabacbuitt.

Ventralabechnitt der Leibothohle.

aufgerollte Seitenpartie de> Korpen.

HaupUUmme der Dannverzweignng.

Hauptattkmme der ventralen Leibeahdhlen-

vertweignng.
Ubergangwtelle zwisclien Nephridialcanal

und EndbUae.

Fig. 5-10 ist mil rotrr Farbt angelegt : der ErnithriingBapparmt ; mit gtlbtr t'arbe angelegt : der
Excretiouappmlk

!• 1%. 6-10 iat durch KWMM Farbt gekennzeichnet : der weibliche Geachlechtaapparat.
In Fig. 6 encheint aoHerdem anch der Brutraam (br) in iceuter /Vw6# gehalten.

26 DR. KUDOLF RITTER v. STUMMER-TRAUNFELS.

FIGURENERKLARUNG (PLATE).

FIG. 1. — M. antarcticum, n.sp. Von der Dorsalseite her abgebildeter Korpersector zur Demonstration seiner
Oberfliichensculptur und der Randzone. Vergr. 20-fach.

FIG. 2-10. — M. cysticolum Graff.

FIG. 2. — Weibchen (aus Cyste a) ; von der linken Seite gesehen. Das in der Abbildung obere Ende ist
das Caudalende des Tieres. Vergr. 6-fach.

FIG. 3. — Cyste (a) an einem Armstiick von Antedon adriani ; von der linken Seite gesehen. Vergr. G-fach.

FIG. 4. — Weibchen (aus Cyste l>) ; von der Ventralseite gesehen. Man erkennt die dellenformig einge-
buchtete Stelle, an welcher das Miinnchen gesessen hatte. Vergr. 10-fach.

FIG. 5. — Miinnchen (ans Cyste a) ; von der Ventralseite her abgebildet. Das Exemplar ist rostro-caudal
etwas contrahiert. Vergr. 27j-fach.

FIG. 6. — Weibchen (aus Cyste a). Medianer Langsschnitt. Vergr. 11-fach.

FIG. 7. — Weibchen (aus Cyste V). Etwas schief zur Liingsaxe ausgefallener Querschnitt in der die mittleren
(dritten) Parapodien verbindenden Transversanebene. Von den beiden Ovarien der linken
Korperseite ist nur das riickwartige, von den beiden Hoden nur der linke getroffen. Vergr. 32-fach.

FIG. 8. — Weibchen (aus Cyste V). Hinter dem vorigen, in der Gegend der Nephridialcaniile gefiihrter
Querschuitt. Nur der rechte Nephridialcanal ist in demselben getroffen. Vergr. 32-fach.

FIG. 9 und 10. — Miinnchen (aus Cyste V). Aus Fliichenschnitten combinirte und etwas schematisirte
Teilbilder. In Fig. 9 sind die mehr dorsal gelegenen Organe, wie der Erniihrungsapparat, der
miinnlicbe Geschlechtsapparat und der Dorsalabschnitt des weiblichen Geschlechtsapparates, ferner
die Nephrostomata dargestellt. Fig. 10 zeigt den Ventralabschnitt des weiblichen Geschlechts-
apparates mit den Ovarien, ferner die Endblasen mit dem unpaaren Endabschnitte der Kephridien.
Zwischen den Nephrostomatis (ns in Fig. 9), und den in Fig. 10 mit * bezeichneten Durch-
schnittsstellen durch die Nephridien verlaufen in dorsoventraler Richtung die kurzen Nephridial-
caniile. Vergr. fiir beide Fig. 79-fach.

ub

oo'

Aa'. del

MYZOSTOMA.

SIPUNCULOIDEA.

By W. F. LANI-HKSTKR, M.A..
King's College, Cambridge.

I'll AM oU)SOMA SOCIUM.

THIS collection consist^ of some thirty specimens, all of small size, \\lu.-h \\,-\-,-
mostly taken )>y means of a net from holes in the ice, while the ' Discovery ' was
in Winter Quarters. One specimen, however, was capture*! with the dredge, the
depth Iteing given as 100 fathoms. That the hulk of them l>eloug to a single
specie* of Phascoloaoina is quite certain, lx>th from their general apiH>arancc and
from a study of the relations of the various organs, though it is to lie noted that
these relations have Wen found to lie far from invariable. Some half-a-dozen,
however, have l>een the cause of considerable difficulty in the exact determination of
their specific identity ; yet it has seemed best to include them under the aliovc
species. In the first place, they agree in the main details of their structure with
the typical forms ; and I have not IKHMI able to place them without hesitation
under any of the other closely allied species (sub-species or varieties) which centre
round the Arctic PA. margaritaceum, namely, PA. antarcticum, ctipsiforme, ftiHCiim,
yeoraianutn, or lay fuse. In the second place, I find in some of them no structural
differences from the typical form ; while in others, such differences are hardly greater
than those found within the limits of the typical form ; while in regard to the most
obvious distinction that separates them, namely, difference of general appearance,
I have considered that, in so far as colour is concerned, the difference may l»c due to
methods of preservation (in one case at least corrosive sublimate is given as the
preservative), and that, in so far as the greater thickness and opacity of the Ixxly-
wall is concerned, this is quite conceivably due to the greater contraction of the
muscles that limit it interiorly. One has, of course, been influenced in forming an
opinion on this matter by a variety of other considerations. A certain small
acquaintance with the various types of Sipunculids enables one to estimate, by a
method which is licyond description, except in so fur as it may be called the method
of past experience or the results of practice, how far one may venture to overlook
those differences of detail which occur as l>etween individual specimens. Again, one
is able to recall that, while in many Sipunculids certain outstanding features appear,
which enable one with little difficulty to place the individual considered, yet in others
these same features are too indefinite or variable to be of much use. And the actual
truth is that, speaking generally, the main features which we rely on as affording
VOL, nr. V

W. F. LANCHESTER.

differentiating specific characters in this group are ultimately of a very vague and
unsatisfactory kind ; while they appear clear enough in many cases, yet in
others we at once realise how indefinite they are. We are dealing with very contractile
animals, and yet use such points as the relative lengths of the body and the
introvert ; nay, the very limits between these two are generally indefinable, and even
were they not, still, one portion may be relatively more contracted than the other.
The length of the segmental organs relative to other structures has its value, yet
I cannot help feeling that a certain amount of extensibility must be conceded to them ;
and the question of their colour surely affords an absolutely valueless criterion.
So with the number of coils of the gut ; extremes of number indicate a
difference undoubtedly, but one is naturally suspicious of such distinctions as
lie between 16, 18, 20, and so on. Thus the present specimens (out of those
that are undoubtedly similar) give us a length of segmental organs varying from
l'5-4 mm., and number of gut-coils varying from 12-25. So that when Dr.
Michaelsen tells us that two of the differences between Ph. antarcticum and Ph.
fuscum are that the latter has " etwa 18" and the former " ungefahr 20 Doppd-
windungen," and that in the latter the segmental organs are " weit langer als die von
Ph. antarcticum," we feel that the first distinction is valueless, and that the value of
the second must depend on the amount of the other distinctions, which are as a
matter of fact admitted to be slight as between the two species. I trust I may not
be taken to mean that the characters usually quoted as of specific worth have no value.
My object is, in the main, to indicate the considerations that have induced me to join
up in a particular instance what others might have been inclined to separate ; though
it is naturally obvious that I tend to consider that certain details that are universally
quoted in specific descriptions of Sipunculids should be subjected to very careful
consideration before they are accepted as having any absolute value. The absence of
longitudinal muscles, and, in most species, of hooks within the genus Pkascolosoma is
doubtless one of the reasons why it is more unsatisfactory to deal with than the other
genera of the family.

The present species then presents in the main the chief features of the Ph.
margaritaceum group. From Ph. margaritaceum and Ph. capsiforme it seems to be
distinguished by having the introvert not much shorter than, instead of being only
half, or a little more than half, the length of the body ; yet here I would like to point
out that I have examined three specimens of Ph. capsiforme which were kindly sent
to me, through Prof. F. J. Bell, by Dr. Michaelsen. Now in one of these, in which the
introvert is fully extended, it appears from careful measurements that the introvert is
" approximately " equal in length to the body. I say " approximately " only, because
of two difficulties : (a) that the introvert is twisted round on itself ; and (b) that,
simple as it seems to distinguish the confines of introvert and body when regarding
the object with the ordinary eye, yet the application of a measuring instrument at once
reveals the difficulty of deciding on a real line of demarcation between the two.

-II'ONCTLOIDEA. 3

However, it in clear that iu this case, at any rate, the introvert in relatively much
longer than in the cases originally described. I turn, however, to a second of the
specimens lent me. which I have cut open, and find that the uneztendcd introvert
which lit - in-iilr the Uxly is now markedly less than half the Ixxly length. Now it
aeemH not at all unlikely, on ii priori grounds (I have no actual observations to offer in
this case), that tin- introvert, when pushed in amongst the organs of the body, should
contract to a considerable extent, so as to occupy as little room as possible; and that,
therefore, the normal length (if one can speak of any such thing) in this case may be
considered as being about half the body length. But over and al>ove this, we find
that the numl>cr of gut-coils in this specimen is something between 30-40, while
capsiforme is laid to have only about 20. Since, however, I have been kindly
permitted to examine then specimens pretty closely (I have not been able to open the
other two), I am able to. state that I see no reason to doubt the accuracy of their
identification with capsiforme, despite these differences. That is to say that I,
assuming a much more careful examination on the part of the original identifier, am
not willing to controvert his decision, and that my own observations tend to support
'it And so I provisionally assume that considerable differences of this nature may be
expected to occur, at any rate within the margaritaceum group.

Ph. papillosum, Ph. capense, and Ph. hanseni appear to be sufficiently distinct
from the margaritaceum group to render any detailed comparison with this species
unnecessary. On the whole then, it seems reasonable enough to consider this form as
a type differing from Ph. margaritaceum and Ph. capsiforme ; and for much the same
reasons that we can so separate Ph. antarcticum and Ph. fuscum. But the comparison
with the two latter forms is a much more difficult matter. In the first place, much the
most striking feature in the majority of these specimens, as viewed with the naked eye
or the lens, is (a) the thinness and semi-transparency of the skin, and (b) its extreme
smoothness, the papillae only being barely visible with the aid of the lens (they are all
small specimens) on the hind end, or sometimes also at the base of the introvert.

In these features of their general appearance, in their light, somewhat straw-like
colouration, coupled with the abruptly pointed termination of the Ixxly, and the
localisation of the papillae, they recall the general facies of Ph. vulgare more than that
of any other species with which I am familiar. And it is in just these features that
they seem to differ so markedly from Ph. antarcticum and Ph. fuscum, in which the
body-wall is relatively thick, the colour is dark, and a system of cross-striation is
present at the hind and front ends of the body, which appears to l>e at best but weakly
represented in this form. We may then proceed on the conceivable presumption that
thmr are a single (i.e. from the specific point of view monovalent) group of points ; in
other words, that we are dealing with one specific point only, and that that point may
be either variable according to size and age, or variable within the species. Passing
on then to other points, we find a clear resemblance to Ph. fuxcum in the close
approximation of the openings of the segmental organs to the line of the anus.

Y 2

4 W. F. LANCHESTER.

Further, on opening up the body it is found that the ventral retractors arise
from one-third to nearly one-half of the distance between the segmental openings
and the end of the body, which again suggests Ph. fuscum ; but that the segmental
organs are very short and do not reach as far as the origin of the ventral
retractors, which accords with the arrangement in Ph. antarcticum. The latter
point is complicated by the fact that in three of the specimens the segmental
organs do over-reach the ventral retractors by as far again, yet are not so much
as half the length of the body in Ph. fuscum. And the number of coils of
the gut points to neither the one nor the other, as they vary from about fifteen to
about twenty-five (about eighteen in Ph. fuscum, about twenty in Ph. antarcticum}.
Turning lastly to the papillae, we find that the description of them in the other two
species accords fully with their appearance in this, except that here they are variable in
one point. Dr. Michaelsen tells us that in Ph. antarcticum they are • 027 mm. wide
and • 08 mm. high (i.e. three times as high as wide) ; in Ph. fuscum, he only says that
they are "bis '07 mm. lang, also nicht ganz so lang wie die entsprechenden von
Ph. antarcticum" but the comparison of absolute lengths where the difference is so
small is valueless, assuming the proportions to be the same; so that we can only'
conclude that they are practically identical in the two species of Michaelsen in regard
to this point. Now in our species the papillae in some individuals show this ratio of
3 : 1, but in others the ratio of the height is less (2'5 : 1, 2*25 : 1, 2 : l).

In Ph. geoi-gianum the introvert is only half the length of the body, and the
papillae are much longer than in the other species quoted ; and these points, taken
along with other smaller differences, seem to clearly separate our form from it.
From Ph. lagense Fischer, these specimens are very little distinct ; but this form is
itself hardly to be distinguished from Ph. antarcticum. The thinness and clear
colouration seem to be the main distinction in most of the specimens ; and in all
of them the relative unimportance of the system of striation on the hind end. In
Ph. lagense, moreover, the segmental organs are as long as half the body.

I subjoin a list of measurements taken from a few of the specimens ; the
sign " S.O." is meant to indicate the "opening of the segmental organ," and the
measurements are in millimetres and taken as accurately as the conditions would
permit.

The letters A, B, etc., correspond to different localities, while the figures
1, 2, 3, etc., merely indicate different specimens. At the same time it may be
pointed out that the A and B specimens are all obviously of one type, viz.,
thin-skinned and straw-coloured, while those marked E and F are those that I
have already mentioned as differing from the rest in general facies, viz., dark in
colour and apparently thicker-skinned.

The most remarkable difference is to be found in the specimen marked E 2.
Here the extended introvert is much longer than the body, 26 '3 mm. and 8 '5 mm.
On the other hand, in the one marked E 1, which presents all the general facies

HPUVOULOIDEA.

—

From and to 8.O.

Lamihol-
: Iplrowrt.

I >i«Unco of rttractor origin*
Numterol puu,,,. „,
ury 1 ' i *••• tnn B>o<
Doiwl. Ventral

A 1 1'J

22

8-R 8 19 1

A2 1«'5

17'ft

2 7 21

A

:-

II

2-25 7-75 .20 1

A4

11*1

18-S i' 47.'. 17 -25

Al

l.v.'.

17-6 2-ft 4-76 1M l

A

10

18 2 5 25 -26

A i lo*.»

8 1-25 4-25 17 -2&

Bl 21

22 8 9'5 Ifi -5

B 12

10 2 R 20 -5

E 1 0

10 2 fi'2."> ?

IS

- .

•j •

1-75 8-5 21

Fl

14

11-5

1-25 4 11-12

and the same internal structure as in E2, together with closely similar papillae,
the retracted introvert is shorter than the body, 14 mm.: 11*5 mm. In El,
moreover, the dorsal retractors originate from a point half-way Itetween the opening
of the segmcntal organ and the ventral retractor, while in E2 they originate at a
position only one-third approximately (as in the other specimens) of the dixtance
Itetween these two points. It is a little difficult to see how to deal with those
anomalies satisfactorily ; and I have preferred to consider these two provisionally
as within the limits of the present species. In two of the specimens I find the
segmental organs reaching aUmt as far again beyond the insertion of the retractors.
\ -[liiiilli- muscle, not attached to the hind end, is generally present, though in one
»r two cases I have not been quite able to decide as to its presence or absence.
In two cases I have found two distinct muscles attaching the front portion of the
gut to the body-wall ; in the other cases I have seen no muscles at all, though it
seems quite possible that they may easily lie torn, and so escape notice, in these small
specimens.

The following list of localities and notes may be added :—

A. 10 tpedmoM .... Winter Quarters. 4. ix. 08. No. 12 Hole. D net.

8. of Hot Point. 2&-30 fas. 100 yd*.

B. 2 «pecin*eM . . „ „ 28. ii. 02. Up to 10 fnu.

C. 8 tpedmens . 11. ii. 02. Duet. Hat Point.

6 W. F. LANCHESTER.

D. 1 specimen .... Winter Quarters. 17. i. 03. Flagon Point. 102

fms. Dredge.

E. 4 specimens .... „ „ 29. viii. 03. Same as A.

F. 4 specimens .... „ „ 3. ii. 04. D net. Hut Point.

G. 1 specimen .... ,. „ 19. iii. 02. D net. Hut Point.
H. 3 specimens .... „ „ 19. xi. 02. I) net. Hut Point.
K. 1 specimen .... „ „ 19. iii. 02. Up to 10 fms.

Finally, reference should be made to a very brief preliminary description of
Phascolosoma charcyti, obtained by the Charcot Expedition, and described by
M. Marcel Herubel in the "Bulletin du Mus. d'Hist. Nat.," 1906, p. 127, together
with Ph. fuscum and Ph. antarcticum. Since Ph. charcoti is there said to resemble
Ph. georgianum closely, I do not think our species can be synonymous with it. On
the other hand, we are told, in regard to the skin of Ph. charcoti, that it is
" a peu pres lisse," with " quelques papilles espacees et petites " ; in which it both
resembles our species and differs from Ph. georgianum, of which Michaelseu says,
" Der ganze Korper ist mit grossen dunklen Papillen besetzt." This discrepancy
in one point may not invalidate its general resemblance to Ph. georgianum, but,
combined with the brevity of the diagnosis, it makes it impossible to assert
similarity or the reverse as between our form and Ph. C?uzrcoti.

COELENTERA.

IV.-ACTINM;.

I

Bv JOSEPH A. CLUBB, M.Sc.
(3 Plates.)

THE Actinian collection of the ' Discovery ' Expedition consisted of upward* of fifty
specimens, which have l>een allocated to eight species, representing six genera and four
families. With two exceptions, all were taken at or near to the Winter Quartern,
McMurdo Bay. The two exceptions* are Port Harris, Falkland Islands, containing
two species, and Enderby Island, Auckland Islands, containing one species, and from
both places a considerable number of specimens, more than half the total, was obtained.
The specimens were for the most part in a fair state of preservation, but I must express
regret that no memoranda were made of the colours of the living nuimals. In the
preserved condition alwolutrly no trace of colour remains, and while admitting that
anatomical characters are the only sure guide in Actiuological studies, the colours of
living Actinians, when known, serve at least as a valuable clue to identification with the
species of the older authorities, who give very little but external characters as their
diagnoses.

The work has been carried on in the lalioratories of the Liverpool Museums,
and I have again to express my thanks to the Museums Committee for permission to
use apparatus, etc., and to Dr. Forbes, the Director of Museums, for other facilities.

FAMILY ACTINIID^E, Gosse, 1858.

ACTIXIARIA with an adherent base ; column wall smooth or provided with verrucaj,
but never with hollow vesicular outgrowths ; sphincter muscle endodermal, generally
diffuse and weak ; tentacles simple ; margin smooth or provided with simple acroragi ;
mesenteries in several cycles, of which usually more than one is perfect ; longitudinal
muscles of tentacles usually diffuse ; no acoutia.

Genus PARANTHEOPSIS, McMurrich.

Actiniidae without acroragi, but with conspicuous verrucas on the distal portion of
the column ; no collar, but the margin a distinct parapet, within which is a well-marked
fosse ; tentacles of moderate length and rather slender, capable gf being concealed in
contraction ; sphincter diffuse.

• ThcM (onus are *o interacting that, after MM con»i<leration, I have allowed their entrance into thii
report. — ED.

JOSEPH A. CLUBB.
PARANTHEOPSIS CRUENTATA.

Actinia crumtata Couthouy in Dana, U.S. Exploring Expedition, Zoophytes* (1*16), p. 188, PI. 3,

Figs. 23 and 23A.

Cereus crmntatm Milne Edwards, Hist. Nat. Corall. (1857), 1, p. 2C8.
Bunocks critentata Gosse, Actinologia Britanuica (1860), p. 194. Verrill, Trans. Connect. Acad. (186!)),

1, p. 467. Andres, Fauna und Flora des Golfes von Neapel (1884), p. 215.
CondylMtis crumtata McMurrich, Scientific Results of Explorations U.S. Fish Commission Steamer

'Albatross' (1893) p. 150. Carlgren, Zoantharien, Hamb. Maghal. Sammel. (1898), p. 10, Figs. 13

and 14.
Parantheqjsis crumtata McMurrich, Zoolog. Jahrb. Suppl. VI. (1905), p. 233.

I have adopted the course taken by McMurrich (10, p. 233), and place this species,
previously known as Condylactis cruentata, under this genus.

Six specimens were taken by the ' Discovery ' at Port Harris, Falkland Islands,
along with a number of specimens of Bunodes octoradiatus. In my description of the
latter species (p. 7) I refer to the extreme similarity in external appearance of these
two species.

The largest specimen measures 2 • 5 cm. in height of column and 1 • 5 cm. in
diameter. Proximally the body-wall is longitudinally grooved, corresponding to the
insertion of the mesenteries ; distally the corrugations are replaced by distinct rows of
verrucse. The papillae or pseudoacroragi terminating the rows have no trace of
nematocysts, and are simply outgrowths of the body-wall, being essentially identical in
structure (PI. 1, Fig. 3) with it. They vary considerably in size in different individuals,
evidently depending on their state of contraction. In the specimen from which the section
figured (PI. 1, Fig. 1) is taken the pseudoacroragi are much inflated. Considerable
variation exists in the number of tentacles. The largest specimen has 48 tentacles
(6 + 6 + 12 + 24), the smallest has only 26, and an intermediate size 32. This
certainly suggests that the adult condition is hexamerous, and that any octamerous
condition is simply a stage in development and transitory, as Carlgren supposes. No
sphincter muscle is developed, there being no indication in transverse sections through
the parapet and adjoining body-wall of any modifications more than the ordinary
muscle bands found generally (PI 1, Fig. 1). The circular muscles of the body- wall
are moderately developed, and the mesogloeal layer is relatively thick (PI. 1, Fig. 1).
The longitudinal muscles of the tentacles and oral disc are ectodermal, and the
longitudinal and parieto-basilar muscles of the mesenteries are well marked.

FAMILY PARACTIDJ5, R. Hertwig, 1882.

ACTINIARIA with a mesodermal sphincter muscle ; no acontia ; mesenteries arranged
in several cycles, of which usually more than one is perfect ; longitudinal muscles of
the mesenteries usually diffuse ; column wall usually smooth, sometimes with sub-
marginal plication and solid ridges ; no acroragi.

* Dana (Amer. J. Sci. (2) i. (1846), p. 178) cites his Zoophyte work as of this year ; the copy in the Zool. Dept.
of the Museum bears date 1848. — ED.

\. ll\l I. 3

MIS PARACTW.

l'u nii.hr with thin .MI.) -month ••••Immi wall : tentacles of moderate length and
nf uniform thickness throughout; margin not lobcd ; individual mesenteries uf ea<-li
jwir c«{ually developed.

PARACTIS POLARIS.

One specimen lalx-lK-.l • Winter Quarters, 24. 8. 03," anil taken in 25-30 fathoms
(PI. I. Kg. 2).

I Light of column •_'•:; cm.; diameter of foot-dine 3- 5 cm., diameter of oral disc
2'5 cm. Tentacles of uniform size, about 0*3 cm. long, in four cycles
12 + 12 + 24 + 48 = 9G. The Itase is adhesive and thin, the insertions of the
mesenteries l»eing visible through it. It in produced all round wider than the column,
which narrows considerably. The Imdy-wall in firm, and near the parai>ct somewhat
thick, where it is slightly puckered by contraction. The mouth opening is visible in
the centre of the tentacles, crowded together by contraction. The a'sophagus is
plicated and the two siphonoglyphs arc well marked.

The sphincter is fairly strong, mcsogloeal, and produced to a fine termination.
A transverse section (PI. 1, Fig. 3) shows that it lies nearer to the endoderm, and is
reticulate in appearance, giving indistinct traces of layering. (Ireater details of
structure are given in Fig. 4.

The mesenteries are membranous, and the longitudinal muscles are somewhat
diffuse, but well marked (PI. 1, Fig. 5). The mesoglooal layer is thin, except close to
the body-wall, where the pnrieto-lwsilar muscle arises, but the lamclhu of the muscles
are conspicuous and branched. Two cycles of the mesenteries arc complete, and these
are fertile, including the directives. The specimen is female, ova in various stages of
development lx?ing seen in sections.

The radial muscles of the oral disc and the longitudinal muscles of the tentacles
are mesoglocal (PI. 1, Fig. 6). In regard to this character McMurridi (8, p. 161),
in his comments on the genus I\inicti*, states that of two forms of the ' AHutross,'
collections which, according to the generally accepted definition of the genus, must be
assigned to l'<tmcti«, one had these muscles ectodermal and the other mesogloeal,
and while raising the question as to whether this character is worthy of generic
di-tinction, he places l»oth provisionally under Paractis. Of the ' Discovery ' specimens,
in both the one under consideration and the succeeding one the said muscles arc
mesogloeal.

PARAITIS PAPAVKR.

Aflttiia papaetr Dnytoo in Dana, p. 148, PI. I. KIL-. 2'.», op. cii.
Pmrmctit JMJMMT Milne Edwmnto, p. »49, torn.

A single specimen labelled "Mi.Mmdo Bay, Winter Quarters, 20 fathoms," only
slightly contracted, with tentacles, oral disc and oesophageal opening plainly visible,
the last-named being 1 *5 ciu. in diameter.

VOL IV.

JOSEPH A. CLUBB.

The height of body-wall is 4'5 cm., diameter of column 5 '5 cm., and of oral
disc very slightly less. The body-wall is quite smooth and extremely membranous,
giving the animal quite an inflated appearance. The upper margin is slightly
crenulated by contraction of the sphincter ; the pedal disc is adherent and a little
firmer and thicker than the body-wall ; the oral disc has slight radiating folds running
outwards, up to and between the tentacles ; the oesophagus is strongly corrugated,
and the two siphonoglyphs are visible, but not markedly distinct. The tentacles are
stout, in four cycles (12 + 12 + 24 + 48 = 96); in the innermost cycle about 1 cm.
long in the contracted condition, and diminishing slightly in length in the outer cycles.

The sphincter is mesogloeal, occupying almost the whole of the mesogloeal layer,
and projecting as a distinct collar, with the ectodermal epithelium, and forming a deep
fosse between it and the tentacles. Fig. 7 on PI. 2 shows a radial section through
the sphincter. The muscles of the oral disc (PI. 2, Fig. 7) and tentacles (PI. 2, Fig. 8)
are mesogloeal. The longitudinal muscles of the mesenteries are in the form of slender
strands, which may be seen running more or less parallel to each other over the extremely
delicate and membranous mesenteries. In transverse section (PI. 2, Fig. 11) it is seen
that the mesogloea is extremely thin throughout, but is slightly thickened and bears
short slightly branched lamellae for the muscle strands. The parieto-basilar muscle
is small and insignificant. Two cycles of mesenteries are complete, and the pairs of
directive mesenteries bear the usual relation to the siphonoglyphs. All the complete
mesenteries are fertile, including the directives, and ova in various stages are seen
(PI. 2, Fig. 9), and a figure is given of a section of the somewhat characteristic
mesenterial filaments (PI. 2, Fig. 10).

I have identified this specimen with Paractis papaver, the Actinia papaver of
Drayton, briefly described and figured by Dana (4, p. 143, PI. 4).

Genus ACTINOSTOLA, Verrill.

Paractidae usually of large size, with a firm leathery wall, which may be
corrugated or folded, but with no verrucas. Margin tentaculate. Tentacles relatively
thick, and not enlarged at the base. The pairs of mesenteries in certain cycles are
unequally developed, having one of each pair smaller and narrower than the other.

ACTINOSTOLA CHILENSIS.

Artinostola chilensis McMurrich, Zoolog. Jahrb., Supp. VI. (Fauna Chilensis iii.), (1905), p. 247, Pis. 15
and 16, Figs. 80-83.

Five specimens of large size contained in one bottle labelled " 4.3.04, 254 fathoms,
mud and stones." The locality is not stated on the label, but I am informed that the
position was 67° 21' 46" S. and 155° 21' 10" E. The specimens vary greatly in form
and size. One specimen measures 5 ' 5 cm. in height and 4 cm. in diameter of
column ; another is 4 cm. only in height, but 6 cm. in diameter.

The wall is firm and leathery, smooth to the touch, and no verrucas or tubercles

\in\i.i: 5

•

arc present ; but irregular -liall«>\\ furrows are formetl by eoutractiuti in these preserved
specimens. In all cases the tentacle* «• exposed, aud they agree generally with the
description given by MrMurrirh, in hia account of the species (10, p. 247) from a
-|M-,'!ni,-n iV-. in i 'album, both as regards number and arrangement and relative size.

Tin- irsophagus is generally widely expanded, and in the majority of the
specimens is alm<>-t indistinguishable from the portion of the oral disc inside the
tentacles. But in one of the more contracted specimens the lip of the oesophagus is
easilv distinguished, but the corrugations are continued radially over the oral disc,
running between aud a short distance up the bases of the tentacles. The siphonoglyphs
are in no case well marked. The foot-disc is well-marked and strongly adherent.

The more minute anatomy agrees generally with McMurrich's description. The
mesogloea of the body-wall is variable in thickness, and in one specimen, just
l>elow the sphincter, is nearly 2 mm. thick, from which in an upward direction it thins
rapidly to less than 1 mm. The mesogloeal sphincter exhibits no trace of layering,
having a simple reticulate structure in transverse section. Transverse sections of the
tentacles exhibit the same appearance of portions of the ectodermnl epithelium licing
cut off and apparently enclosed in the mesogloea, as descril>ed by McMurrich, due to
contraction and the unusual thickness of the mesogloeal layer.

The mesenteries in manlier and arrangement agree with McMurrich's descrip-
tion, but I cannot confirm the hermaphroditic condition described for the species.
In the specimen selected for histological work no spermatozoa were found, and the
ova were large and well-developed. Otherwise the ' Discovery ' specimens agree
well with McMurrich's description, and I have no hesitation in placing them under
this species.

FAMILY ALICIID.& DUKRDKN (1895).

A' ri N IAKIA with a large, flat, contractile base. Tentacles simple, cylindrical and
entacmteous. Column wall with simple or compound hollow tultercles or vesicles,
covering the greater part of the column, arranged mostly in vertical rows. Sphincter
muscle endodermal and diffuse, variable in amount of development. No cinclides or
acontia, margin with or without acroragL Mesenteries arranged in several cycles of
which usually more than one is perfect

Genus CYSTIACTIS, M. Edw.

Aliriidiu having the column covered with simple vesicles. Tentacles of variable
length, in one, two, or three cycles. Numerous perfect mesenteries ; sphincter muscle
moderately well developed.

( YVHACTIS ANTARCTICA.

A single specimen, bearing the label " McMurdo Bay, Winter Quarters, 28.2.02."
Depth, 20 fathoms.

z 2

6 JOSEPH A. CLUBB.

|

External characters — Height of body-wall 5 '75 cm., diameter of column 4 cm.,
diameter of oral disc 3 cm., diameter of pedal disc (contracted condition) 2 '75 cm.

There is no trace of distinctive coloration in the preserved specimen, and no
record on the collector's label. The single specimen is poorly preserved, and owing to
the thin and delicate body-wall is quite collapsed. The column is studded with
thin-walled simple vesicles (PI. 2, Fig. 12) communicating with the coelenterou and
arranged in twenty-four vertical and parallel rows, corresponding to the mesenteric
spaces. The attachment of the mesenteries may be seen in places, between the rows
of vesicles. The vesicles of adjacent rows are arranged alternately, and they increase
in size from the proximal to the distal portion of the column (PI. 2, Fig. 12), where
the largest are not much inferior in size to and do not differ much in appearance from
the tentacles in the preserved and contracted specimen.

The tentacles are twenty-four in number, arranged in two cycles of equal size,
short and somewhat club-shaped. The largest measures 0' 5 cm. in length. The
oral disc is slightly furrowed radially and the mouth is large and bears a well-marked
hypostome. The oesophagus is short and strongly plicated, and two deep and
well-marked siphonoglyphs are present, and are continued some distance below the
rest of the ossophagus as distinct lappets.

The coelenteron is large, and twelve pairs of mesenteries, all complete, are present.
Two pairs of directive mesenteries bear -the usual relation with the siphonoglyphs.
The single specimen taken is poorly preserved and much of the histological detail is
lost. The sphincter is diffuse and endodermal in character, with slightly projecting
mesogloeal lamellae shown in transverse section (PI. 2, Fig. 13). The muscles of the
tentacles are ectodermal. The muscles of the body-wall (PI. 2, Fig. 14) are feeble,
and sections through the vesicles show a similar structure. The ectoderm and
endoderm layers consist of relatively high columnar cells (PI. 2, Figs. 13 and 14);
but the mesogloea is thin. The mesenteries are very thin and delicate, but transverse
sections show a weak but well-marked longitudinal muscle (PI. 2, Fig. 14), the
mesogloea in relation with it being thickened and possessing short branching lamellae.
The parieto-basilar muscle is small (PI. 2, Fig. 14).

The specimen is female, and ovaries are found on all mesenteries, including
directives, forming broad bands occupying in their widest portions about a third of
the width of the mesenteries. Ova in alb stages of development (PI. 2, Figs. 14
and 15) are present.

FAMILY CRIBRINHXE, McMurrich, (1901.)

ACTINIARIA with a well-developed circumscribed endodermal sphincter ; simple
tentacles ; without cinclides and acontia ; verrucac usually present ; adherent base ; no
true acroragi, but frequently pseudoaeroragi present.

\< TIM.fi. 7

HUH CltlBRINA.

Cribrinidir with sin mi: '•n.l.xli'rnial sphincter; frequently with pseudoacroragi;
ecUxlermal musculature . .t discs and tentacles not iml>C(l(le<l in the mesogloea ; column
wall provided with verruca} arranged more or leas distinctly in vertical rows ;

•• :,' • '• .:..;.;.

CKIRRINA OCTORADIATA.

Bwttdtt trtoradiaiu* (Vtrlgren, Huntraiy. Maplhaeoi Sammolrciw, K<>anlh. (!«'.»«), p. 20, Fipi. 11 mid 12.
1898.

This special was formed by Carlgrcn (1, p. 20) for a numlx-r of specimens taken
from the Straits of Magellan and other localities in the South Sous. In one of the
' Discovery* bottles, labelled " Falkland Islands, Port Harris," were fourteen specimens
of very similar-looking Actinians. Upon dissection and microscopical examination,
however, I find two distinct species — Bunodr* octoradititux, Carlgren, and Pttrmtktoptu
crutntata, Couthouy. Both Carlgren (1, pp. 12 and 21) and McMurrich (10, p. 2:13)
c-omment on the similarity of appearance of this Crihrinid and this Actiniid, and it is
interesting to note in this instance that they were apparently found in association in
the same habitat. If this is commonly the case it is not surprising that confusion has
occurred in previous descriptions of the Actiuiid. Unfortunately I have no record as
to the colours of the ' Discovery ' specimens, but McMurrich (10, p. 234) descriltes tho
colour of the Actiniid as being very variable ; it is possible that when in association
the two species may also approximate in appearance when living, and if this be so it
at once suggests mimetic resemblance.

In size the ' Discovery ' specimens of Criltritui ttctoradiata agree with I'arlgren's
specimens (1, p. 20). The sixteen rows of verruca; on the distal portion of the column
and the sixteen pseudoacroragi are well seen in all the specimens. Usually gravel
and broken shell are attached to the verruca?. The number of the tentacles
(8 + 8 4- 16 = 32) appears quite constant. The two siphonoglyphs are well marked
and quite regular. The sphincter muscle is strongly circumscribed, and may l>e
described in McMurrich's terms (9, p. 20) as of the pedunculate palmate variety
(PI. 3, Fig. 16). The section figured is through the space between the pseudoacroragi,
I nit sections passing through the latter show the sphincter projecting from the inner
wall, near its base, so that the cavities of the pseudoacroragi project beyond and
the sphincter. The mesenteries number sixteen pairs (8 + 8 = 16), all
to ami fertile, including the directives (PL 3, fig. 18). The longitudinal
"I" the tentacles arc ectodermal and transverse sections (PI. 3, Fig. 17) show
strong mesogloeal lamella}. The longitudinal muscles of the mesenteries and the
parieto-ba&Uar muscles are exceedingly large and strong, and in the contracted
specimens are closely packed into a solid mass, so that the various organs are quite
dovetailed together (PI. 3, Fig. 18). The mesogloeal lamellae of the muscles arc long
and branched. The circular muscles of the body-wall too are very strong. . The
strength of the musculature generally is quite a feature of the sjiecies.

JOSEPH A. CLUBB.

But the most interesting character of these specimens from the Falkland Islands
is the presence of " brood pouches." Immediately below the pseudoacroragi is a well-
marked constriction of the body-wall, also present, it is interesting to note, in Paran-
theopsis cruentata. In this constriction, in the species under consideration, specimens
were found possessing sixteen pores, one pore to each line of verrucse, leading into
distinct invaginations of the body-wall, forming characteristic " brood pouches " in the
coelenteron. These " brood pouches " agree in all essential points with the arrangement
shown to be present in the 'Southern Cross' specimens (3, p. 299). The drawing
(PI. 3, Fig. 16) shows a section of the body-wall, passing through one of these pores,
and the " brood chamber " into which it opens. The three layers of the body-wall may
be distinctly traced as shown, each of them much thinner — the mesogloea markedly so.
Usually each " brood chamber" contains two embryos only, and in the section figured
they are well advanced, showing several of the mesenteries complete and others well
developed. Dissections were made so that, looked at from below, the sixteen " brood
pouches " were seen like so many miniature grapes, lying quite regularly in the body
cavity, between the mesenteries. In diameter they average about 3 mm.

In the short description of the species by Carlgren (1, p. 20) no mention of
them is made. It is somewhat remarkable if none of Carlgren's specimens possess
these " brood pouches," and possibly re-examination may show their presence in some.

CRIBRINA HERMAPHRODITICA.

Runodes hermaphroditicus Carlgren, Hamburg. Magalhacns. Sammelreise, Zoanth. (1898), p. 23, Fig. 18,
1898.

About forty specimens from "Enderby Island, Auckland Islands, 19. 3. 04" of
sizes ranging from 0 • 3 cm. to 2 cm. in height of column. In some cases the oral disc
and tentacles are fully exposed, in others the animals are contracted so that the
tentacles are entirely hidden, and there are many intermediate stages of contraction.

The foot-disc is well marked and strongly adherent. The rows of principal
verrucae, usually twenty-four in number, are well seen, especially on the distal portion
of the column, where also secondary alternating series are visible. The pseudo-
acroragi, terminating each principal row of verrucse, are distinct, emphasizing greatly
the parapet. The tentacles are in four (6 + 6 + 12 + 24 = 48), or five (6 + 6 + 12
+ 24 + 48 = 96) cycles.

The sphincter is endodermal and circumscribed. The oesophagus is longitudinally
folded, and two well-marked siphonoglyphs are present. The mesenteries in the
specimens examined are all hexamerously arranged, and in the larger specimens in four
cycles, of which three are complete. The longitudinal and parieto-basilar muscles
(PI. 3, Fig. 19) are well developed. Two pairs of directive mesenteries bear the usual
relation with the siphonoglyphs. The longitudinal muscles of the oral disc and
tentacles (PI. 3, Fig. 21) are ectodermal. Zoanthellae occur throughout the endoderm,
being especially abundant in the tentacles (PL 3, Fig. 21). Reproductive elements

ACTINM?.

are present in rel;itu'ii with the mesenteries of the first cycle in small specimens, including
directives, and in large specimens, with the mesenteries of two cycles ; hut I was not
al. If to .If!. •num.- to my satisfaction if present in relation with the third complete cycle.
Both ova and spcrruatozm are present, often associated together in relation with the
same mesentery (PI. 3, Fig. 19). Occasionally fertile mesenteries arc found showing
ova or spermatozoa alone, but in all the specimens examined microscopically the
hermaphroditic condition is constant Embryos in various stages of development are
present in the body-cavities of many.

Genus RHODACTINIA, Agassi*.

Cribrinidie (Tealiidae, Bunodida?, Bunodactidse), with well-developed foot disc ;
distinct verruca* present on body-wall ; radial muscles of oral disc and longi-
tudinal muscles of the tentacles neso-ectodermal to mesogloeal ; tentacles simple ;
strongly circumscribed endodermal sphincter.

RHODACTINIA rRA.s«cx>KNis.

Actinia crat*ieornu Muller, Prod. Zool. Dan. (177C), p. 281.

Artinia rl»y<tnti»»ima Brandt, Prod, duacriptionis animaliam ah II. Mertensio ul«crvatomm ( 1885), p. 18.

Jlrif- -f- -'- Hantii Agiasix, L., Coinjm* Rend as, XXV. (1847), p. 677.

Mktfrtinia davuni Verrill, A. K., Mem. Soc. Nat. Hint. Boston (1HG4), p. 18.

I'rlirina rrauieorntt Vi-rrill, A. E., Proc. E«ez Instil. VI. (1%9), p. 4G9.

Leiotralut >i>it;brrgrn*i* Kwietnicwuki, Zool. Jahrb. (Syrt.) (1898), p. 1, p. 121.

Carlgren in 1902, in his report on "Die Actiniarien der Olga Expedition"
(2, p. 39), thought it necessary to revive the old genus of Agassiz, Jt/uxlnctini<i,
and has taken for its type the well-known and widely distributed Actinia Mtusuormr,
Miiller. The principal distinguishing character given for the genus is that the radial
muscles of the disc and the longitudinal muscles of the tentacles may be meao-
ectodermal to mesogloeal. and under /.'. crn**i<#rnij» Carlgren places a number of
synonyms, including Actinia eltgantissimn, Brandt, and Leiott-al i/i *pitibergen&i#,
Kwietniewski. I have previously been struck with the variation in the degree in
whifh the muscles of the tentacles are imbedded in the mesogloca, in specimens of
cnu*i&rrnix examined by me, from a well-known local habitat, Hilbre Island, at the
mouth of the Dee. The text- figures (2, p. 41) given by Carlgren are most
instructive, and from Hilbre Island specimens I can produce examples in which the
longitudinal muscles are quite as much meso-ectodermal as C'arlgren's figure 4, if not
as much as his Fig. 3. In my examination of the ' Southern Cross ' Antarctic
Actinians (3, p. 294), I was convinced of the many close resemblances to the type
species crassieomis, seen in the specimens then under observation ; and notwithstanding
that the longitudinal muscles of the tentacles were not imbedded in the mesogloea,
I decided to include these species under the genus I'rtirina. I feel therefore some-

10 JOSEPH A. CLUBB.

what justified in this course, although I agree that it will be better to adopt
Carlgren's proposal, and include such forms under the genus Rhodactinia, and reserving
the genus Urticina or Tealia for species where the radial muscles of the oral disc
and the longitudinal muscles of the tentacles are always completely imbedded in the
mesogloea.

In the ' Discovery ' collection are six specimens, all taken at the Winter
Quarters, McMurdo Bay, in from 10 to 20 fathoms, and bearing the following dates :
20. 2. 02 ; 17. 1. 03 and 17. 2. 04. They are all of large size, the largest being
6 • 5 cm. in height of column and 6 cm. in diameter of oral disc. The external
appearance of these specimens varies considerably, according to the degree of con-
traction of the oral disc or body-wall. I have had under observation for some time a
number of specimens of crassicornis from Hilbre Island, living in one- of the tanks of
the aquarium of the Liverpool Museum ; they are all in a flourishing condition and
feed voraciously. I have been surprised to note their extreme mobility. Tentacles,
oral disc and body-wall exhibit remarkable degrees of inflation, sometimes all together,
sometimes singly, and sometimes only sections of each of them. In extreme dilation
of the body-wall all trace of verrucas, as such, is lost, and as Teale (11) observes, it
" renders the corium perfectly smooth, so that the small opaque spot alone indicates
their former situation." Correspondingly, in extreme dilation of the oral disc, the
tentacles may be so reduced as to be little more than papillae. On PI. 3, Fig. 22 is
a reproduction from a photograph of the oral disc of a living animal, which well
illustrates this extreme contraction of the tentacles. On the other hand, I have
recorded specimens where the tentacles have measured 4 • 5 cm. in length and 0 • 6 cm.
in diameter near the base. There is always general expansion, especially of the
tentacles after feeding. McMurrich (9, p. 28) discusses this great variation in external
appearance brought about by this great mobility, and the internal anatomy, sphincter,
muscles of the tentacles, mesenteries, and body-wall and other internal structures
must correspondingly vary, according to the degree of contraction or expansion.

In all the ' Discovery ' specimens under consideration the radial muscles of the
oral disc and the longitudinal muscles of the tentacles are meso-ectodermal, and in
contraction are identical in appearance with the figure given by Carlgren for his
variety spitzbergensis (2, Fig. 3, p. 41). The sphincter is endodermal, and strongly
circumscribed, and may be seen in all cases by the naked eye when the parapet is cut.
Transverse sections show a little difference in detail in the arrangement of the
mesogloeal processes, whereas in some the more typical bipinnate arrangement is seen,
in others the plan is similar to the figure given by Kwietniewski (7, p. 14, Fig. 8),
for his Leiotealia spitzbergensis, which Carlgren (2, p. 40) now regards as a synonym
of R. crassicornis. I give a figure of this form of sphincter drawn from a section of a
' Discovery ' specimen (PI. III., Fig. 23).

As regards the verrucse, tentacles, oesophagus, siphonoglyphs, directives and
other mesenteries, and in other anatomical features, the specimens under consideration

ACTINIA 11

agree generally with the typical cra*»icornist to I have concluded to place them all
under Rkodactima crauicorni*, probably more nearly approximating to the variety

MEMOIRS REFERRED TO IN TEXT.

1. CJJUXIREX, 0.— Zoautliarien ; Hamburg. Majralliaens. Sammelreise, Hamburg (1899).

2. CAEtOtaur, 0. — Die Actiniarien der Olga- Expedition, Wisst/nsch. Mecresnnter. Neue Folge, r. Bawl.

Abt. HelgoUad (1902), Heft 1.

3. < I.UBB, J. A.— Actinue. Report on the ' Southern Cross ' Antarctic Expedition. London (1901).

4. UAHA, J. D. Zoophyte. UJ3. Exploring Expedition (1838-1842) ; Philadelphia (1M46).

5. ITKRDIW, J. E.— On the genus Alicia (Cladactis), with an anatomical description of A. rottir. Pane.

Ann. Mag. Nat Hut. (6), Vol. XT. (1895).
0. Goao, P. H.— On the British Actinue. Ann. Mag. Nat. Hist. (3), Vol. i. (1858).

7. KWIKTXIKWSKI, CABIIUR B.— ' Actiniaria von Oat-Spitsbergen,' Zoologischc Jahrbucher (Sjrrt.) xi.

(1898).

8. McMcRRicH, J. PI^TFAIB.— Actiniaj of 'Albatrow' Explorations. Proc. United States National

Museum, Vol. xvi. (1898).

9. McMuRBicu, J. PLAYFAIB.— Report on Uie Hi-xactini» of the Columbia University Expedition to the

Pnget Sound daring the Summer of 189C ; Annals New York Acad. Sci., Vol. xiv., No. 1 (1901.)

10. McMuuRlCH, J. PLATFAIB.— Zoolog. Jahrb. Supp., vi. (Fuuna Chiliensis iii.) ; The Actinia; of the

Plate Collection (1905).

1 1. TBALK, T. P.— On the Anatomy of Actinia eoriaeta ; Trans. Phil. Soc. Leeds, L (1837).

YOU nr.

EXPLANATION OF PLATES.

b.c. brood chamber*.

c. coelentcron.

d. directive mesenteries. ^
6c. ectoderm.

em. em OTTOS.
en. endoderm.

in. ec. inTaginated ectoderm,
in. en. inTaginated endoderm.
in. mg. inTaginated mesogloea.
in. o. opening to brood chamber.
1. m. longitudinal muscle.
1. m. m. longitudinal muscle of mesentery.
1. m. t. longitudinal muscles of tentacle,
mesenteries.

2 A

12 JOSEPH A. CLUBB.

m. f . mesenterial filaments,
nig. mesogloea.
m. 1. muscle lamellfe.

oe. oesophagus.

oe. in. cesophageal invagination of embryo,
ov. ova.

p. pseudoacroragus.
p. b. m. parieto-basilar muscle.

s. siphonoglyphe.
sph. sphincter.

t. tentacle,
tes. testis.
v. verrucse.
x. zooanthellae.

PLATE I.

FIG. 1. Parantheopsis cruentata (Couthouy). — Radial section of the upper part of the body-wall, passing

through a pseudoacroragus.

FIG. 2. Paractis polaris, sp. n. — Drawing of the specimen (natural size).
FIG. 3. P. polaris. — Radial section of upper part of body-wall and sphincter muscle.
FIG. 4. P. polaris. — Portion of sphincter muscle more highly magnified.
FIG. 5. P. polaris. — Horizontal section through the body-wall and mesenteries in connection with

oesophagus.
FIG. 6. P. polaris. — Transverse section of the wall of a tentacle.

PLATE II.

FIG. 7. Paractis papaver (Drayton). — Eadial section through the upper part of the body-wall and sphincter

muscle.

FIG. 8. P. papaver. — Transverse section of wall of a tentacle.
FIG. 9. P. papaver. — Section through ovary.
FIG. 10. P. papaver. — Section of mesenterial filament.
FIG. 11. P. papaver. — Section through portion of a mesentery.
FIG. 12. Cystiactis anlarctica, sp. n. — Drawing of specimen (natural size).
FIG. 13. C. anlarctica. — Section of upper part of body-wall, passing through sphincter muscle.
FIG. 14. C. antarctica. — Horizontal section through body-wall and mesentery, with ovary.
.FiG. 15. C. antarctica. — Section of ovary more highly magnified.

PLATE HI.

FIG. 16. Crilrina octoradiata (Carlgren).— Radial section through upper part of body-wall, sphincter and

opening to ' brood ' chamber.

FIG. 17. C. octoradiata. — Section through wall of a tentacle.
FIG. 18. C. octoradiata. — Horizontal section through body-wall and mesenteries, including the directive

mesenteries.

FIG. 19. C. hermaphroditica (Carlgren). — Horizontal section through body-wall and fertile mesenteries.
FIG. 20. C. hermaphroditica.— Section, more highly magnified, of genital gland.
FIG. 21. C. hermaphroditica. — Section through wall of tentacle.
FIG. 22. Rhodactinia crassitornis (0. F. Miiller). — Reproduction from a photograph of a living specimen,

illustrating an extreme reduction of the size of the tentacles.
FIG. 23. R. crassicornis. — Radial section through the sphincter of a ' Discovery ' specimen.

4.

a.

Antarctic (Discover-.

ERSITV

12

ll

13.

11

9

•

i

ml

15.

Ant,,-i .scovor;.

14.
.iae PI U

DIVERSITY

•

irctic(Discovery) Exp

Actiniae PI. in.

Aut.del Huth so

POEIFERA (SPONGES).

H. TETEAXONIDA, BENDY (8. P. 63).

BY R. KlRKPATRICK.

(Plates VIII.-XXVI.)
I. GRADE TETRACTINELLIDA, MARSHALL.

IN the collection made by the ' Discovery ' there are fifty-nine specimens of
Tetractinellida, belonging to four species of TetiUidae, all of which have been
described before ; and, further, there are new varieties of two of the species.

The bulk of the material consists of numerous and large specimens of Cinachyra
barbata Sol las, a species which was originally obtained from Kerguelen by the
' Challenger.'

The following is a list of species :—

Craniella sagitto (Lendcnfeld) var. mitrosigma var. n.
Craniflla lagitta var. pofhyrrhabdtu var. n.
CnmMa leptodtrma (Sollat).
Cinachyra barbata Bellas.
Cinachyra vtrtex Lendcnfeld.
Cinachyra vtrtex var. montiruliri* var. n.

SUB-ORDER SIOMATOPHORA, Sollas.

FAMILY TETILLIDAK, Sollas.

CRANIELLA, Sollas.

TetiUidae without porocalyces, and with a cortical skeleton of oxeas arranged more
or less radially and vertically to the surface.

CRANIELLA SAOITTA rar. microsigma.
(Plate VIII., figs. 1-13, and Plate IX., figs. 15, 16.)
1907. Ttlhya tagitta Lendenfeld (11 a. p. 306).

Sponge large, pyriform or cocoanut-shaped, narrowed and truncate at lower end,
with finely pilose surface and with root-tuft With a few small circular oscules about
0 • 75 mm. in diameter in the contracted state. The body divided into three zones, an
inferior root-tuft zone, a broad median or equatorial poral zone, and a superior oscular
zone. Surface, where the pile has been denuded, showing obscurely-marked longitu-
dinal anastomosing ridges, much more evident in the poral region. Pores in sieve-like

2 A 2

K. KIRKPATRICK.

groups in grooves between the ridges. Flagellated chambers eurypylous. Cortical
skeleton formed of palisade-like rows of oxeas and fan-like bundles of trichodal
protriaenes, occurring only in the root-tuft zone and poral zone.

Spicules. Megascleres. 1. Somal Oxeas (VIII. 10)* from 5 to 13 mm. in
length and 0 • 06 to 0 ' 08 mm. in 'thickness, straight, fusiform, tapering gradually to
very sharp ends.

2. Choanosomal oxeas lying scattered between the radial fibres, 1-3 mm. long and
0 ' 04 to 0 ' 06 mm. thick.

3. Cortical oxeas (VIII. 13) almost straight or slightly curved, 1 • 12 x 0'04 mm.

4. Anatriaenes of two kinds, each varying in length and tapering to a filiform
extremity. 4a. (VIII. 5, 5a, 6) with pointed conical cladome, with slender sharp-
pointed cladi each 225 p. long, chorda 187 p. long; rhabdome about 12 mm. long, and
averaging about 20 p, broad, thick below cladal origin, then with a slender neck followed
by long thicker portion tapering finally to filiform extremity.

Anatriaenes, 4b (VIII. 7, 7a) common in root-tuft, with thick rounded conical
cladome, with short thick cladi 118 p. long, chorda 118 p, long; rhabdome of nearly
uniform diameter till it tapers off to filiform extremity, length varying from 10 to
40 mm.

5. Protriaenes. Cladome commonly with one cladus 0*135 mm., longer than the
other two, though there may be two equal long ones, or they may all be equal ;
rhabdome fusiform, tapering to a filiform extremity, on an average about 9 x 0'054
mm. in length and thickness.

6. Trichodal protriaenes 218 p. long, with one cladus 28 p. in length, longer than
the other two ; cladal end of rhabdome slightly swollen.

Microscleres. Sigmata 12 '3 p.-l3 p, long, 7 '04 /x broad (when seen in C-like
aspect), and 1 • 5 p thick, surface micro-punctate.

There are five specimens of this sponge, four large and one very small. The
largest (VIII. 1) is a fine example with a root-tuft; this appendage not being present
in the other large specimens owing to having been torn away when the sponges were
uprooted from the sea-bottom.

The body of the largest specimen is 10'5 cm. long and 11 '5 cm. in broadest
diameter, and of the smallest 10x8 mm. in length and breadth. The mass of root-
tuft in the type specimen is about 4 cm. in thickness.

The surface pile of spicules varies considerably in its degree of development ; in
two examples the surface is almost bare ; in two others the pile is soft, and about 3 to
4 mm. in height, with oscular fringes about 4 • 5 mm. in height. The pile is formed
mainly of the projecting triaenes of the radiating fibres of the skeleton, each fibre
spreading out fan- like in an oblique or vertical plane. In the specimens bare of the
pile, the boundary between the poral and oscular zones (VIII. 2) is a fairly well marked
circular line of demarcation situated at the junction of the upper fourth and lower

* Roman numerals followed by Arabic refer respectively to the Plate and figures ; thus (VIII. 10) means
Plate VIII., fig. 10.

T1.TUA.\<>NII>.\ 3

thrre-fiuirths of the body. The surface of the oscular zone is smooth and glistening
ami with faintly marked longitudinal ridges ; whereas that of the poral zone has more
prominent ridges, and a punctate appearance due to the pores in the grooves between the
ridges ; this appearance is brought into relief by drying the sponge, for then the poral
sieves contract and the points of the palisades of cortical oxcas prick up the dermal
membrane ; these palisades are not present in the oscular area, which, in fact, is devoid
of a cortical skeleton of spicules, though it possesses a fibrous cortex.

The OSCUles are small circular orifices about 0 • 75 mm. in diameter, five or nix in
numl>er, arranged in an irregular circle on the broad upper end of the sponge. In the
complete state they each have an outer fringe of protriaenes nlxmt 4*5 mm. high, and
mi inner fringe of trichodal protriaenes, a little )>elow which is a sphincter.

The dermal pores occur as sieve-like groups in the more or less longitudinal
grooves (VIII. 2, 3) lietween the palisade* of cortical oxeas, the meshes of the sieves
l>«-ing supported by fan-like tufts of trichodal protriaenes.

The EctOSOITie. The sponge cuts easily, and a vertical section shows the skeletal
fibres radiating from a central nucleus. No well-defined boundary is visible to the
naked eye between ectosome and choanosome. In the poral zone (VIII. 3 and IX. 15)
the pores lead into sub-dermal spaces bounded by vertical rows of cortical oxeas ; the
rows of cortical oxeas are likewise present at the base of the sponge in the root-tuft
zone. In the oscular zone (IX. 16) there is an ill-defined layer, about 1*5 mm. thick,
composed of loose bundles of fibrous tissue crossing in various directions with rounded
"•llencytes scattered about; also there may IN- a few scattered oxeas.

The Canal System. The sulxiermal spaces beneath the sieve-groups of pores
open into canals descending into the choanosome. The pyriform flagellated chaml>er8,
47 x 42 /x (VIII. 4), are eurypylous ; the figure shows several chambers opening
direct into a terminal exhalant canal.

The chamber system of Cramella simillima Bk. and of other species of Craniellsi,
described by Sollas in his 'Challenger' Report, is aphodal (21. p. 30, PI. II., fig. 19).

The Skeleton. A tangential surface section in the poral region shows l>est the
cortical oxeas, which are mostly arranged in double rows, each row bending outwards
slightly towards a row of another double series, so as to form an incomplete arch over
a sub-dermal space or channel.

A small " silica pearl " was found in one section.

Numerous large ova occur, with oval body and thick pseudopodia-like extensions,
the body being 130 x '65/n, and the total diameter over 300 p. In some instances,
the ova are more rounded and compact, with the cytoplasm more drawn together.

I had described the specimens as representatives of a new species, but I now find
that they belong to CranifUa sagitta (Lendenfeld).

Dr. v. Lendenfeld very kindly sent me the revised proofs of his Report
"Tetraxonia der deutschen Siidpolar-Expedition, 1901-1903," thus enabling me to
make the necessary alterations before it was too late.

4 R. KIEKPATRICK.

The 'Gauss' Expedition obtained six specimens, which are all young and small,
being from 2-10 mm. in diameter.

The difference in the size of the megascleres of the ' Discovery ' and ' Gauss '
specimens may be attributed to differences in the age and size of the specimens ;
but the difference in the size of the sigmata of the two sets of specimens is more
notable — -those of the ' Gauss ' being 14-20/u, and of the ' Discovery ' being only 12 • 3/u
to 13/i in length as viewed in the C aspect. In my opinion difference in size of
sigmata has not the importance that is sometimes attached to it.

All the specimens were dredged in the neighbourhood of Winter Quarters in depths
of 10-130 fins. Also ' Gauss' Expedition, near its Winter Quarters, 350-385 m.

CRANIELLA SAGITTA var. pachyrrhabdus.
(PL VIII., figs. 14, 15 ; PL IX., figs. 17, 18, 19.)
1907. Tethya sagitta Lendenfeld (Ha. p. 306).

This variety is represented by several oblong pieces about 6 cm. in length, deeply
blackened by osmic acid, and forming part of a large sponge, which had been cut up
for preservation in that reagent.

The fragments are sufficient to show that the surface of the sponge was covered
with a spicular pile about 4 mm. in height.

Mr. Hodgson informs me that the bulk of the specimen was lost. He was under
the impression that this sponge, which was obtained from the same locality as the
var. microsigma, and which resembled the latter in general appearance, was identical
with it in all its characters. An examination of a section, however, shows that the
radiating fibres of the skeleton contain numerous large thick strongyles and styles,
about 7 x 0'116 mm. in dimensions (VIII. figs. 14, 15); the sigmata are 18-20 p.
long, and slightly thicker and rougher than in the var. microsigma.

Plate IX. figs. 18, 19 show collar cells, stained in gentian violet. The cells are
here seen to be separate, and the collars apparently not coalescent ; but the tissues are
not very well preserved ; and, though much care was taken in the drawing, a renewed
inspection of the preparation from which the illustrations were drawn has shown that
the figures of the collars are not satisfactory. The flagellated chambers usually contain
only a few collar cells in the normal position (IX. 17), the rest forming detached
masses in the centre of the chambers. The flagellum originates from the end of the
large oval nucleus, which is situated a little below the surface of the body of the cell.

Locality. Winter Quarters, Hut Point, 25 fms.

CRANIELLA LEPTODEKMA.

(Plate XL, figs. 4-14.)
1888. Tetilla leptoderma Sollas (21. p. 3).

Sponge, club-shaped or pyriform ; with verruculate surface ; with one large oscule
near the summit ; with root-tuft. Surface pile of spicules absent in the adult state,
and only slightly developed in early stages, large protriaenes being very rare or absent.

TKTKAXONIDA. 5

Spicules. .Mt^asderee. 1. Large oxeas, 9 • 7 x 0*081 mm.

2. Smaller oxeas, scattered in the choanosome, 540 to 1350/i in length.

3. Cortical oxeas (XI. 8), 770 x 30 ft, straight, fusiform.

4. Anatriaenes of three kinds: a. (XI. 9) Cladi 170 p long x 20ft thick at
!,;.,.. -I, !,'l'-i. sharp-pointed, ohotda I10/i; rhaMonu', nlH)ul 12 mm. long thick ftl
junction with cladome, then tapering, again thickening, and finally terminating in a
filiform extremity.

b. (XL 11) Cladome with thick conical cladi, 140 x 35 /x, chorda 60/*; rhabdome
with a thick neck, and then nearly uniform till it tapers to a filiform extremity.

c. (XI. 10) Cladome with very long, straight, slender cladi, 150 x 10 p., nearly
parallel with, or making an angle of, about 10° with the rhabdome, chorda only 31 ft ;
rhabdome slender (of uncertain length).

5. Protriaenea Rare, usually with three slender equal cladi, 90 /A long ; rhabdome
of nearly uniform diameter ; but slightly diminishing near the cladome.

6. Triehodal protriaeues, usually (especially in the case of those forming the inner
oscular fringe) with one ray greatly prolonged to 140 /A or more, the other two having
almost disappeared ; rhalxlome slightly thickened at cladal origin.

7. Sigmata, 14*25 to 15 ft long, 0*75 ft thick, and 12'3/n broad in the c aspect,
with slightly and finely granulated surface.

There are three specimens of this sponge, the largest (XI. 4) being 12 cm.
in total length and 5 '2 cm. in breadth, with a root-tuft 3' 5 cm. in length;
the second specimen (XI. 5) is oval, 2'9 cm. long and 1*7 cm. broad, and the smallest
is 13 x 12 mm.

The colour of the first two is dirty gray, and that of the la.st pale buff.

The scale-like, occasionally over-lapping, verrucae, attain a height of 2 mm., and
a breadth at the base of 4 mm.

The oscule is oval-shaped, edged with a l»rely visible fringe of trichodal
protriaenes ; it measures, in the large specimen, 7'5 x 5 mm.

The skeleton. The radiating fibres only very rarely project beyond the surface ;
near the oscule are a few tufts of oxeas. Over the rest of the surface of the sponge,
the dermal membrane covers over the summits of the verrucae, the distal points of
oxeas and the cladomes of anatriaenes showing through ; medium-sized in large
protriaenes, so abundant in Cranielln xagllta, are almost absent from the two large
specimens of the present species, though in the smallest specimen there is one to each
coiiule. Tufts of trichodal protriaenes are present in abundance.

The ectosome. On section, the ectosome is scarcely differentiated from the
choanosome, excepting that sub-dermal spaces are visible just beneath the dermal
membrane.

The cortical oxeas are more densely packed than in the preceding species.

A tangential section shows the pores in longitudinal areas between the densely
serried rows of cortical oxeas, and fan-like tufts of trichodal protriaenes supporting the

6 K. KIRKPATR1CK.

sieve-like pore-areas ; the pore-areas are distributed over the surface up to the
neighbourhood of the oscule.

Canal System. The pores lead into sub-dermal channels whence canals descend
into the choanosome. The flagellated chambers are eurypylous.

Small ova, 60 x 30 /z, with pseudopodic-like processes, occur in the largest
specimen.

A young specimen is pear-shaped with a granular surface, and without a
root-tuft. Each " granule " is a sharp-pointed conule, with a protriaene projecting
from it.

The young specimen of Craniella sagitta has a very different appearance, being
quite fluffy on the surface owing to the numerous tufts of protriaenes.

Specimens were obtained from : (1) West of Balleney Island in 254 fms., Type
specimen (XI. 4) ; (2) W.Q., Flagon Point, 10-20 fms. ; (3) W.Q., Hut Point, 25 fms.

The ' Challenger ' dredged specimens off Rio de la Plata, 600 fms.

CINACHYRA BARBATA.

(Plate IX., figs. 1-14.)

1888. Cinachyra larl/ata Sollas (21. p. 23. Pis. III., XXXIX.).
1906. Cinachyra barbata Kirkpatrick (10. p. 6C2. PI. XIV.).

Twenty-seven specimens of this sponge, varying in diameter from a few
millimetres up to 14 centimetres, were obtained. In addition to these, the collection
includes a massive root-tuft no less than 32 cm. across and 12 cm. in thickness,
belonging to what must have been a very large specimen, for the concavity into which
the latter fitted occupies nearly the whole area of the upper surface of the root-tuft.
The largest specimen obtained by the 'Challenger' — from Kerguelen — was 10 cm. in
its longest diameter.

As in the case of the ' Challenger ' specimens there is a considerable variation in
the shape, which is usually spheroidal ; some specimens are spherical, the nucleus of
the radiating bundles of the skeleton being exactly central, others again resemble an
inverted cone, and very young specimens are oval.

An interesting feature shown in adult specimens is the division of the sponge into
three zones, viz., (1) a basal root-tuft zone ; (2) an equatorial zone of porocalyces ;
and (3) a polar zone of oscules.

These zones are clearly perceptible in the fine series of well-preserved Antarctic
specimens, but are not so well marked in those obtained by the ' Challenger.' The
surface pile of spicules in the porocalycal zone is longer and looser than in the oscular
zone, and directed obliquely downwards ; but the pile in the oscular area is shorter,
and more vertical and stubble-like.

The root-tuft. A remarkable feature about the root-tuft is its great size in
some specimens, in which it may be much more bulky than the sponge body ;

I I I KAXONIDA. 7

generally, however, it is of smaller size than the body, especially in large and
old specimens.

The porocalyces. In adult specimens the porocalyces form a broad belt round
the sides of the body. A few, in large specimens, were nearly 2*5 cm. (one inch) in
depth. As I have already stated (10. p. 662) the porocalyces are, in all species of
Cinachyra, inhalant areas perforated by true dermal pores, and are never exhalant or
OMular.

The oscules. In Cinachyra barbnta these structures are arranged in an obscurely
spiral manner on the upper or " polar " surface of the sponge in adult specimens. In
almost every instance, the oscules are tightly contracted, tho spicular fringe l>eing
closed over them in form of a conical stack. It is this strong contract ihility that has
led to their being overlooked in most of the species of this genus, the porocalyces
having been regarded as partly inhalant and partly exhalant.

The oscules are very rarely found open. Where this happens, the oscular orifice
is at the summit of a slight crater-like elevation. Within the circle of long protriacnes
forming the oscular fringe is a circle of fine protriaenes on the edge itself and also
on the gentle plope leading inwards. At a short distance from the edge, the slope
leads to a vertical barrel-shaped tube ending in a well-marked sphinctrate ring, which
is about on a level with the innermost plane of the cortex ; the oscular tube slightly
expands below the sphincter into a sub-cortical cavity in the lateral walls of which are
the orifices of exhalant canals proceeding from just Iwneath the cortex. At its lower
end the sub-cortical space is continued on into a large main vertical canal, which passes
up radially from the central region, receiving in its turn many affluents. The wall of
the oscular tul* is mainly composed of a thick sheath of concentrically arranged
myocytM.

The vertical section of the oscular tube, when contracted, shows the almost closed
lumen of the tube surrounded by a dark zone of sigmata, and with spiral bands of
myocytes passing upwards around the lumen.

Two of the eight known Cinachyra species, viz., C. voeltzkowi Lendenfeld (11. p.
101) and C. malaccensis I. J. B. Sollas (19. p. 219), are described as having scattered
pane in addition to the pores of the porocalyces. Some of these " pores " will certainly
be found to be oscules situated, in young specimens, more or less antipodally to the
porocalyces, others possibly belonging to developing porocalyces ; for as W. J. Sollas
observes (21. p. 29), " in their inception, the porocalyces are simply poriferous areas of
the cortex."

Young specimens. Very young specimens (IX. 3) are oval, broader at the
superior end, and without a root-tuft, though the surface spicules are longer at the
narrow end of the sponge. At this stage the sponge lies with its long axis oblique or
horizontal. The smallest specimen in this collection is 3 x 4 mm. in diameter. It
has one porocalyx situated in the centre of one side, and one osculc a little to one side
of the centre of the summit of the broad end. This nearly bilateral symmetry calls to
VOL. IT. 2 B

8 R. KIRKPATRICK.

mind the species of Fangophilina Schmidt, viz. F. submersa 0. Sch. (17. p. 73), and
F. gilchristi Kirkp. (10. p. 667), in which the ovoidal body of the sponge has one
large porocalyx and one large oscule situated on the upper aspect. As the sponge
grows, the porocalyces increase in number and extend on each side of the original one
till they form a complete equatorial belt ; in the meantime, the main axis of the sponge
gradually rotates from horizontal to vertical, so that the crown of oscules comes in
many specimens to lie in a horizontal plane at right angles to the vertical axis.

The cortex. In young specimens the cortical oxeas are arranged tangentially
in a single layer. In the largest examples, the cortex reaches a thickness of 3 '25 mm.,
the densely packed oxeas being arranged vertically and obliquely to the surface.

The skeleton. The only additional observations to be made here are on certain
bodies which Sollas (21. p. 24) refers to as " Globules ; accessory or accidental forms,
0'0535 mm. in diameter." These bodies, which Schulze terms "silica pearls," are now
known to be not uncommon in sponges (Schulze, 18. p. 6, and Weltner, 33. p. 190).
Spheres may be normal spicules of the sponge, as in Caminus sphaeroconia Sollas. In
many instances, however, spheres or globules result from malformation or incomplete
development, as in cases where a tylote spicule is reduced to a knob, or a sphere may
result from the reduction of an oxea, as in Epallax callocyathus Sollas. In some of
the specimens of C. barbata there are a considerable number of pearls and some of them
are double (IX. 6, 7). One example 114 x 94 p is oval, with two nuclei and with
concentric layers of deposition round each, up to the point where the spheres come in
contact ; later there is a single oval layer common to the two centres, but an annular
depression or kink is always visible in the plane midway between the nuclei of the two
original spheres. Sollas (21. p. 214, PI. XXVII., figs. 8-9) figures composite spheres
occurring in Caminus sphaeroconia, but here the composite sphere does not possess
separate central points round which are deposited concentric laminae, but (apparently)
one centre and an axial line or rod round the end of which the layers are deposited.

Whether these Spheres or silica pearls are always spicules or the result of incom-
plete development of spicules, or whether they are sometimes due to deposition of
layers of silex round some foreign organic or inorganic body, has not been determined.
The nucleus is generally a refringent point, but is sometimes irregularly shaped and
of a faint yellow colour ; attempts to investigate it under high powers generally result
in the pearl being crushed.

Several of the smallest pearls, only 6 or 7 ju in diameter, are associated nvith fan-
shaped crystalline bodies (IX. 8-12) ; these latter proved to be simply crystals
deposited from the sea-water. Sometimes the pearl is in the centre of a spheroidal
mass of crystals. Under polarised light, the crystals are doubly refracting and are
brilliantly defined on the dark field, while the pearls are isotropic, though usually very
faintly visible ; accordingly a.spherical mass of crystals surrounding a pearl has a dark
central space.

In a vertical section of the sponge stained in borax-carmine the spheroidal

TETRAXONIDA. 9

conglomerations of crystals, 30-45 /i in diameter, arc clearly defined owing to their being
much more deeply stained than the surrounding tissues. The appearance of capsules
in PI. IX. figs. 13, 14, is due to an optical effect. The deep staining and the well-
defined spheroidal shape might cause these bodies to be mistaken for organic structures ;
they are barely visible when unstained, and it is not easy to separate them from the
dehydrated tissues in which they he, in order to test the action of reagents ; when
separated they are found to be soluble in water.

Localities. The specimens come from six different localities in the neighbour-
hood of Winter Quarters, from depths ranging from 10 to 25 I'm-. Also Kcrguelcn
Island, 25-60 fathoms, ' Voy. Challenger.'

ClNACHYRA VERTEX.

(Plate X., figs. 1-14.)
1907. Cinafhyra rtrttx Lendenfeld (lla. p. 310).

Sponge, generally oval, ovoidal, or conical, with long diameter horizontal. Surface
with conules, usually oblique and more or less appressed to the surface, arranged in
spiral rows, and terminating in tufts of spicules, the conules and tuft* wrapping round
the body. With well-developed root-tuft.

Porocalyces, varying greatly in size and appearance, being hemispherical pits with
circular orifice, or obliquely directed pockets with slit-like opening ; surface hispidated
with tricbodal protriaenes.

With several oscules situated in a more or less circumscribed area on the side
opposite the area of greatest development of porocalyces, or at or near the summit of
conico-ovate specimens. The oscules were open only in one instance ; they were
3*25 x 2 mm. in diameter, with a marginal fringe, and a platform-like spicule below
the rim.

Ectosome, an outer layer of collenclymae, and an inner of fibrous tissue, with
cortical spicules.

Flagellated chambers eurypylous.

Spicules. Megascleres. (1) Somal ozeas, 5*5 x -057 mm., fusiform, tapering
gradually to sharp points.

(2) Choanosomal oxeas 520 x 18/*, slightly curved.

(3) Anatriaenes ; cladome, more or less hemispherical, often with apical umbo, cladi
at first at right angles, then making a sharp bend ; length 210 n, thickness at base 70 /*,
chorda 300 ft, sage tta 200 /i: rhabdome 6'5/j broad at junction with cladome, then
narrowing slightly, again thickening slightly, and terminating gradually in a fine
sharp point.

(4) Protriaenes, cladome usually with one long cladus 170 ft twice the length of
the other two; rhabdome 7*5 x '019mm. nearly uniform in thickness in distal
half, tapering at proximal end to fine sharp termination.

2 B 2

10 R. KIRKPATRICK.

(5) Trichodal protriaenes with rhabdome 550 p in length, and with longest
cladus 50/i.

Microscleres. (6) Sigmata, 18'36/u in length, 1'8/i in thickness, and 10'5/iin
breadth in the C-aspect ; surface finely granular.

There are twenty -five specimens of this sponge, the largest being 10 -5 x 8 cm.
and the smallest 8x6 mm. The salient external features of the larger specimens
are the spiral series of conules and tufts wrapping round the sponge, and the porocalyces
obliquely displaced by the spiral growth of the sponge. The root-tuft is often much
more bulky than the sponge body. The conules which form a fleshy basal sheath to
the spicular tufts attain a height of 5 mm., and the tufts of an additional 10 mm.
The porocalyces have an average diameter of about 4 mm., the oscules being about
• 5 mm. in the contracted state ; the latter have a fine marginal fringe of small
protriaenes.

On section, the flesh is reddish in colour ; the cortex indiscernible and can be
peeled off in the form of a thin skin, its thickness being about 30 mm. ; the outer
third is formed of large rounded collencytes, 18 ^ in diameter, full of granules,
which stain deeply. Covering the outer surface of the sponge is a well-defined
layer of sigmaspires : a layer of the same spicules also lines the surface of
the canals.

Canal System. Currents pass through the pores in the porocalyces to large
sub-dermal spaces leading to one large inhalant canal.

The flagellated chambers are eurypylous. PI. X. fig. 5 shows numerous apopyles
opening into a commencing exhalant canal.

The collar-cells (X. 6) here have their collars concrescent.

Numerous ova are present in some of the specimens, and in various stages— but
not in the same specimen — some having pseudopodia-like processes, others being
spheroidal and devoid of processes. In orfe unstained section in balsam the large
spherical ova are 330 p in diameter, with nucleus 50 /*, and nucleolus 9 ' 75 p in diameter ;
here the ova can be seen with the naked eye as deep yellow spots, and with a hand
lens the clear nucleus can be distinctly made out.

Young specimens. In early stages there is no root-tuft, but in a series of small
specimens, the conule tufts can be seen to lengthen at the narrow end of the specimen
till they form a root-tuft. At first there is only one porocalyx situated on one side,
and a single oscule, surrounded by a fringe of protriaenes, a little below and behind
the broad end of the egg-shaped specimen. One small specimen (No. 264, 13, ii. 04)
differs from the others in having long tufts each with a long fleshy base, the tufts
all pointing downwards from the flat upper end ; a few small contracted porocalyces
are concealed in the axils of the fleshy conules.

Localities. Winter Quarters. Twenty -five specimens were obtained from ten
localities, and from depths varying from 10-30 fathoms.

TKTRAXONIDA. 11

('is \. MVI:\ vi n \ M

(Plate X., figs. 15, 16. Plate XI.-, figs. 1-3.)
1907. Cinafhyra rtrtn Leodenfcld (lla. p. 310).

Two small specimens of this species, obtained from deep water (130 fins.).
show a remarkable variation in the porocalyces and oscules. The porocalyces instead
of being situated below the general surface are elevated altove it in the form of
rounded warts or monticules alxmt 1 mm. in height ; the oscules likewise form
small cylindrical chimneys about 2 mm. in height.

The larger of the two specimens is 3*5 x 2*5 cm. in its diameters, not including
the root-tuft, which is about 2 cm. in thickness. A large part of the surface is covered
with the little porocalycal monticules, and at one end are several oscular chimneys.
Situated at a varying height on each monticule is a semiluuar slit which opens into
a caeca 1 fold or pocket that has been formed by the upward growtli and protrusion
of the monticule ; when the lip is near the summit of the monticule, it surrounds the
latter like a kind of prapuce. The pore-perforated floor of the porocalyx itself is,
in every instance, closely folded with longitudinal pleats which, however, can l>e unfolded.
The summit of the monticule is formed by the tops of these folds, and bundles of the
hlspidating trichodal protriaenes can occasionally be seen projecting from the summit.
The long slender bundles of protriaenes divide into two layers at the inferior proximal
edge of the semilunar fold, one set being in the fold iteelf and extending to the edge
of the same and sometimes a little beyond, the other set passing up in to the floor and
walls of the porocalyx iteelf. The edge of the semilunar lip is provided with a Iwnd
of myocytes. This curious variation appears to have taken place as an adaptation
necessitated by the spiral mode of growth of the sponge ; if such growth l>ecame
excessive the porocalyces would be in danger of becoming closed altogether from one
lip of the orifice overlaying the other.

Locality. Winter Quarters. No. 10 Hole, 130 fms.

II.— GRADE MONAXONELLIDA, DENDY.

There are, in all, forty-three species belonging to this group. Of these, eight
belong to the Astromonaxinellida, and thirty-five to the Sigmatomonaxiuellida.
Twenty-two species are new, and there are, in addition, seven new varieties of species
already described. There are four new genera, of which one l>elongs to the Axinellidae,
two to the Myctilinae, and a fourth, Pylodenna (Renierinae), has l>een established to
receive Halichondria latrunculioidex Ridley and Dendy.

New and interesting forms of xpicule-t. — The new Mycaline genus Cercidochela is
characterised by the possession of remarkable modified isochelae, which I have termed
canonchelae (I.e., shuttle-shaped chelae). In them the single central teeth from each
end have fused, so that a complete shuttle-shaped spicule results, recalling the melon-

12 R. KIRKPATRICK.

like spicules of Melonanchora, in which, however, three opposite pairs of teeth have-
become united.

Hoplakithara dendyi is remarkable for its exotyles with very large spined heads ;
this species also has fimbriated placocheles.

The spathulate bipocilla occurring in two new species of lophon are worthy of
mention. A bipocillum of this kind occurs, also, outside the genus lophon, viz., in a
new species of Myxilla.

Geographical distribution. — As might have been anticipated, a large proportion of
the known forms have already been recorded from the Southern Hemisphere.

Of the twenty known species, fourteen have been recorded from the South Atlantic
or Indian Ocean, or from the southern portion of the east and west coasts of South
America. Only two species have been previously recorded from the Antarctic region,
viz., lophon radiatus Topsent and Gellius rudis Topsent, both obtained by the
' Belgica ' expedition from the opposite quadrant of the Antarctic circle.

Of four species which have been recorded from the Arctic region, one, viz.,
Sphaerotylus capitatus (Vosmaer), has never been obtained from any intermediate station.
Artemisina apollinis (R. & D.), which was obtained by the ' Challenger ' from Kerguelen
Island, has been recorded by Lundbeck from off Greenland. Stylocordyla borealis
(Loven), recorded from the North Sea, and occurring also in the Antarctic, has been
obtained from several intermediate localities. Of considerable interest is the occurrence
of Esperiopsis villosa Carter, a form frequently recorded from high northern latitudes,
but only from one intermediate station, viz., in deep water off the Azores.

Classification. — The classification of the main groups adopted here is that of Dendy
(6. p. 60, 134). Dendy divides the Order Tetraxonida into two grades, Tetractinellida
and Monaxonellida, and the latter grade into two sub-orders, Astromonaxonellida and
Sigmatomonaxonellida. The use of the last two names implies the theory of the
relationship of these sub-orders with the astrophorus and sigmatophorous Tetragti-
nellida respectively.

The divisions Clavulida and Aciculida (Topsent) of the Astromonaxonellida
(Hadromerina pars Topsent) are adopted ; so likewise is Lundbeck's division of the
Desmacidouidae (Poeciloscleridae Topsent) into two sub-families, Mycalinae and
Ectyoninae, and of the Mycalinae into two groups, Mycaleae and Myxilleae.

A preliminary report giving descriptions of new genera and species has been
published in the Ann. Mag. Nat. Hist. (7), vol. xx., pp. 271 et seq., Sept. 1907.

Below is a list of species of Monaxonellida :—

GRADE MONAXONELLIDA DENDY.
I. Sub-order ASTROMONAXONELLIUA Dendy.
I. Tribe CLAVULIDA Vosmaer.

i. Family Spirastrellidae Ridley a'nd Dendy.

Latruncnlia apicalis Ridley and Dendy var. biformis var. n.
Latrunculia apicalis Ridley and Dendy var. basalts var. n.

I Kl KAXOXIDA. 13

ii. Family P*lymatK,ia* TOON*.

Polymattia MfOffinata Kirkp.

Spkofrotylu* aiitarrlifus Kirkp.

Spluurotylii* capitntus (Yosmner).
iii. Family Subtritidat Schmidt.

*nl>rntft mterottomut Ridluj and Dendy var. tttllatut var. n.

Snbrritft taminal ttx Ridley and Dendy var. fni/iilliitim var. n.

1'ffmintubfrilfs hyaliniu (Ridley and Dendy).

II. Tribe ACICCLIDA.

i. Family Stylufonlylidae.

StyUforilyln boreali* (Ix>vcn) var. attuita var. n.

II. Hub-order SIOMATOMONAXONKI.LIDA Dendy.
ii. Family Atiitellitla* Ridluy and Dendy.
AxineUa gttpraiumesuni Topaent.
SigmaxinyuM phafalliouks Kirkp.

iii. Family Dttmacidonidae Ridley and Dendy.

I. Sub-family Ectyoninat Ridley and Dendy.

Hymftittmia areolata Thielc. *

HyiHftiesmia txigua Kirkp.
Hymtrrhaphia rufa Kirkp.
Ophlitagpongia nultfitata Kirkp.
Litsomyxilla hanitsthi Kirkp.

II. Sub-family Myralinat Thiele.

I. Oroup Myj-illrar Lnndbeck.
lophon radial u» Topeent.
lophnn spatulatus Kirkp.
lophon JiabtUo-digUaius Kirkp.
Littodendoryi tpongioM (Ridley and Dcudy).
Myrilla drtrpta Kirkp.
Ttdania variolota Kirkp. *

Ttdania toulmani Kirkp.

II. Group Myealfoe Lundbeck.

Artfiiiiximi apollinu (Ridley and Dendy).

Ktprrwpsis villosa (Carter).

J/ijfai« magellaniea (Ridley).

Mycait arf rata Kirkp.

Mycalttf.

Dttmacidon krrgtuknensu Ridley and Dendy var. nnturrtifa var. n.

Deimaeidon ktrguelfntnsi* var. eactoidts var. n.

Dttmaeitlon spinigtra Kirkp.

Detmafidon maeandrina Kirkp.

Joyeuxta belli Kirkp.

Crrculochfla lankfxtiri Kirkp.

Hnplakithara dendyi Kirkp.
III. Family Haplottltndtu Topaent.

I. Sub-family GeUiituu Ridley and Dendy.
(if Iliu f rudis Topsent.
Qdliiu fimbriahu Kirkp.

14 B, KIRKPATRICK.

Gellius pilosus Kirkp.

Oellius cuctirbitiformis Kirkp.

Gellius glacialis var. nivea Ridley and Dendy.

Oceanapia tantula Kirkp.

II. Sub-family Eenierinae Ridley and Dendy.

Pyloderma latrunculioides (Ridley and Dendy).
Petrosia fistulata Kirkp.
Reniera scotti Kirkp.
Reniera dancoi Topsent.

SUB-ORDER ASTROMONAXONELLIDA, Dendy.

TRIBE CLAVULIDA, Topsent.

FAMILY SPIRASTRELLIDAE, Ridley and Dendy.

LATRUNCULIA APICALIS var. biformis.

(Plate XV., figs. 1-7.)
Latrunculia apicalis Ridley and Dendy (15. p. 234).

There are three specimens of the new variety. The one selected as the type is'
massive and conical, 9 cm. high and 7 '5 cm. broad at the base. The surface is covered
with the little disk-like poral papillae, but in place of many conical oscules, such as are
found in the typical form described by Ridley and Dendy, there is one large oscule
(16 mm. in diameter and much contracted) at the summit of the specimen. The chief
variation from the. type is in the character of the disc-asters, of which there are two
kinds, one with an apical spike resembling that found in the typical form, but stouter
and shorter, and the other kind without the apical spike. Mr. Highley's figures show
the characters of these two forms ; the first kind is 362 \i long and 200 fj. in greatest
breadth ; the second kind is 1$0 x 150 /i. All the specimens have what appear to be
reduced discasters, i.e., spined styles in which the toothed disks of the discaster have
become reduced to spines. The smooth styles are nearly straight, 325 x 12'5 /u,, and
with tornote pointed ends.

Two of the specimens are chocolate brown in colour, but a third is paler and has
more of the reduced discasters. One specimen is labelled " The green Sponge." The
alcohol in which they have been preserved is dichroic, being amber-coloured by
transmitted, and olive green by reflected, light.

The specimens were all dredged near Winter Quarters in 10-15 fms.

LATRUNCULIA APICALIS, var. basalis.

Lalrunmlia apicalis Ridley and Dendy (15. p. 234.).

There is one small, thin cake-shaped specimen 30 x 20 mm. in area and 6 mm.
thick, of a light-brown mud colour.

There are several of the discoidal raised pore-areas, and one conical oscular papilla.
The flagellated chambers are aphodal, 49 x 19 /*, the aphodus being 32'5 p. long. The

TETRAXONIDA. 15

diacasters vary somewhat, the apparently typical form being 64 p. long and 39 p.
broad, with a small spike at the base, and a still smaller one at the apex ; sometimes
neither the basal nor apical spines are present, and the spicule resembles more nearly
the discaster of Latrunctdia Itrevis Ridley and Dendy. There are two small Im-yil
whorls of spines ; al>ove them is a plano-convex whorl with denticulate edges and
deeply cleft along 3-5 radii ; then follow two smaller whorls pointing upwards.

The smooth styles are fusiform, with tornote point, 557 p. long, 6'5 p. in diameter
at the head, and 10 '5 in diameter at the centre.

The specimen was dredged West of Balleney Island in 254 fms.

FAMILY POLYMASTIDAK, Vosmaer.

POLYMASTIA INVAGINATA.

(Plate XII., fig. IB, and Plate XIV., figs. 5-1 5a.)
Pflymaitia Mvayinaia Kirkpatriek (lOa. p. 271).

Sponge hemispherical, free or attached, covered with a thick pile of pointed
spicules. With one large oscular papilla usually completely invaginated, so that the
summit of the oscule is on a level with, or below the general surface. Under surface
with a fleshy l>asal pad.

Pores in longitudinal meridional groups in the cortex.

Colour, in spirit pale yellow above, and often gray and semi-transparent on the
under surface in free specimens.

Consistence dense and firm.

Flagellated chambers spherical, 30'5 p in diameter, diplodal (XIV. 7).

Skeleton. Choanosomal, formed of fibres curving upwards from the base to the
periphery, penetrating the cortex and forming the thick surface pile ; with stellate
clusters of small tyles between the fibres.

Cortical skeleton formed of a dense layer of vertical tyles of various lengths
embedded in a tough, fibrous layer from *5 to 1*25 mm. thick.

Basal skeleton consisting of spicules transversely arranged, and crossing each other
in an irregular manner.

Spicules. Large, smooth, slightly curved styles, or occasionally strongyles,
2240 x 40/i.

Cortical tyles, with small spheroidal head, short neck, fusiform straight shaft,
varying in length from 140 to 350 p, and in thickness from 12 to 19 /< A few very
slender styles scattered in the choanosome, 70 x 6 p, with head and neck making an
angle with the shaft Some medium-sized cortical tyles in the oscular papilla with long,
oval heads. Tyles of the stellate clusters, slender, with the head making an angle
with the shaft, 200 x 15 p,

There are fourteen large specimens and five very small ones of this species. The
largest is 8 cm. in diameter and 4 • 5 cm. high ; and the smallest, which is triangular,

TOT. IT. 2 C

16 R. KIRKPATRICK.

is 8 mm. high, 8 mm. in length at the base, and 4 mm. in thickness. In some the
pile has been rubbed off, exposing the smooth, fleshy cortex. Some of the specimens
are growing on large stones ; others are free, but with numerous embedded pebbles in
the fleshy basal pad.

In many specimens the oscule is invisible, being entirely retracted within a cavity
below the summit, and with the opening of that cavity, contracted to a point and
concealed by the surface pile of spicules ; a vertical section reveals the oscular papilla
in its cavity. One specimen has two oscular papillae. The general structure of the
skeleton recalls Trichostemma, which likewise has a basal pad, with fibres radiating from
base to periphery, and with a cortex of tyles. T. sarsii also has the stellate groups of
tyles between the fibres of the choanosome. The basal pad is composed of stellate-
celled colleuchyma, the collencytes with their branched anastomosing processes being
embedded in a clear, gelatinous matrix.

Most of the specimens were dredged in the neighbourhood of Winter Quarters in
depths of 10-30 fms. ; one came from near Mounts Erebus and Terror, 500 fms.

SPHAEROTYLUS, Topsent (26. p. 244).

Polymastidae, massive ; provided with tylostyles, and with exotyles in the form
of spherotyles or spherostyles.

This genus was established by Topsent to contain Polymastia capitata, Vosmaer
(32. p. 16). I have slightly extended the original definition by the addition of the
word " spherostyles," because one of the two species of Spltaerotylus in the present
collection has exotyles in the form of spherostyles, i.e., exotyles with the proximal
inner end simply rounded and not enlarged into a knob, a spherostyle being a style
with the distal or outer end enlarged. The difference between a spherostyle and a
spherotyle is hardly of generic importance, consequently Sphaerotylus must include
forms with spherostyles.

Hitherto only three Astromonaxonellid sponges with exotyles have been described,
viz., Tylexocladus joubinii Topsent, Proteleia sollasii Ridley and Dendy, and
Sphaerotylus capitatm (Vosmaer). The 'Discovery' collection contains two species
of Sphaerotylus, a new one S. antarcticus, and one which appears to me to differ in no
important respect from the Arctic species S. capitatus (Vosmaer).

SPHAEROTYLUS ANTARCTICUS.

(Plate XII., figs. 1A-16 and Plate XIII., figs. 1-7.)
Sphaerotylus antarcticus Kirkpatrick (10a. p. 272).

Sponge, dome-shaped or spheroidal, attached or free. Surface beset with a
covering of long spherostyles, and with a dense short pile of cortical microtyles.
With several usually elongated papillae with or without a large terminal orifice.
Dermal pores distributed over the cortex, each pore opening into a single tubular

TETRAXONIDA. 17

canal in the cortex ; the mouth or pore of the pore canal guarded by a ring of
radiating rortinil i : Mil. :{). Flagellated chamlwrs diplodal (XIII. 7).

Skeleton formed mainly of radiating fibres composed of styles, with diverging
brushes of spherostyles near the surface. Cortex with a surface layer of densely
picked tufts of small vertical tyles, and a sub-cortical layer of tangential styles
and tyles.

Spicules. Spherostyles (XII. 8-12) 8 mm. in length by 30 » in diameter in
the middle, and 1 4 p. in the region below the distal knob ; distal knob 28 « in diameter,
hemispherical, with granular surface and with a few square teeth or serrations on
the edge.

Styles straight, fusiform, blunt-pointed, 2*8 mm. in length, 41/i in diameter in
the middle, 23 /i in diameter at the rounded end.

Cortical tyles (XII. 15, 15a), curved, 146/i long, head, 3 '25 n in diameter ; slender
neck, 2'75/* thick with broad oar-blade-like shaft, circular in section, 7/i thick.

Styles of lower cortical tangential layer (XII. 7), also in choanosome, 900 x 20 /*.
Tyles of the same layer, nearly straight, 270 n long ; with head 7 n in diameter,
and relatively thick neck 6'Sn in diameter (XII. 16-16A).

Slender curved tyles, 460 x 10 /*, scattered in choanosome (XII. 13).

Young specimens are oval with one long closed papilla ; the bundles of divergent
exotyles are more or less separate and distinct, and the distal knobs retained and not
broken off ; the radial bundles of fibres radiate out from a point below the centre of
the specimen.

There are five large specimens, of which three encrust rocks and two are free, and
six very small ones, all of which are free. One of the large free ones (XII. 5) is a
globular fluffy ball 6 cm. in diameter, with a few stumpy, much contracted papilla; ;
another free specimen is dome-shaped, with a flat, fleshy base like a thin pad. The
typical shape is that of a dome, with the skeletal fibres radiating upwards and
outwards from the base to the surface. The spheroidal form results from the growing
round of the edges till they meet ; accordingly a section shows a central core with
cortical microtyles still present, but displaced, and with main fibres radiating out all
round the core.

The largest encrusting specimen (XII. 1), which is growing on a boulder of black
volcanic rock, is 10 cm. in diameter and about 3 cm. in height, the papilla;, about
twenty in number, rising to an additional height of 3-4 cm. The papilla: are coated
with the thick pile of microtyles, but there are no exotyles. A few of the papillae
have large terminal orifices with apparently cicatrised edges, as if the breach had arisen
from a bursting away of the summit from pressure from within the papilla ; these
papillae, at any rate, are oscular ; but so also, probably, are some of the closed papillae,
which have irregular cribriform patches around the summit or along one side of the
papilla. The same difficulty of deciding on the function of these organs occurs also in
the case of species of Polymastia,

2 c 2

18 R. KIRKPATRICK.

The radial bundles of styles proceed from base to periphery, and usually do not
penetrate the surface ; the exotyles form pyramidal tufts which originate just below
the cortex, this arrangement being well seen in young specimens.

Specimens were obtained from Winter Quarters, Flagon Point, 10-20 fms. ; from
No. 12 hole, 25-30 fms. ; and from McMurdo Bay up to 20 fms.

SPHAEROTYLUS CAPITATUS.
(Plate XII., fig. Ic, Plate XIII., figs. 8-13, and Plate XIV., figs. 1-4.)

1882. Radiella schoenus Sollas (20. p. 163), Nomen nudum.
1885. Polymaslia capitata Vosmaer (32. p. 16).
1898. Sphaerotyhis capitatus Topsent (26. p. 244).

The single specimen (XII. Ic) is attached to a block of volcanic rock on which
specimens of S. antarcticus and Polymastia invaginata are growing. The sponge is in
the form of a flattened dome 1'5 cm. in diameter and 5 mm. in height, with a small
papilla 5 mm. in length, rising from near the centre of the upper surface. From one
side of the specimen slender bud-bearing stalks arise to a height of 2-4 mm. ;
sometimes the buds are in linear series (XIV. 2), sometimes zigzag (XIV. 4).

The axes of the stalks are the prolongations of radial fibres, and consist of tyles.
The buds, which vary from • 5 mm. to 1 • 25 mm. in diameter, bristle with the sharp
points of small and medium-sized tyles.

The largest bud (XIV. 3) has exotyles, and show* all the stages of their develop-
ment. Firstly, there is a thickening of the distal end of an ordinary sharp-pointed
tylostyle, which becomes cylindrical ; then the end becomes roughened, and finally
clavate.

Merejkowsky (14. p. 4, PI. L, figs. 8, 12, and PI. III., figs. 1-3) figures buds
forming on stalks at the summits of the papillae of Polymastia mammillaris (Rinalda
arctica) from the White Sea. The stalks and buds strikingly resemble those of
Sphaerotylus capitatus, excepting that the latter grow direct from the general surface
of the body and not from the summits of the papillae.

Merejkowsky gives an interesting account of the development of these buds, which
drop off from their stalks, and become flattened and disk-like as they rest on the
bottom. The surface becomes covered with debris which the spines help to retain. He
supposed that the decaying organic debris served as nutriment, which was directly
absorbed by the surface of the young sponge.

The flagellated chambers are diplodal. The colour of the specimen in
spirit is pale yellow. The body presents a surface uniform to the naked eye, but
under a strong lens the summits of the club-shaped exotyles are visible. These
spicules arise in bundles of the shape of inverted cones, the apices of which are a short
distance below the cortex.

The cortical skeleton is formed of slender tyles. There is no sub-cortical layer of
tangential spicules.

TKTRAXONIDA. 19

The rhoanosomal skeleton is formed of fibres radiating upwards from the base -to
the periphery, and spreading out in the upper third of their course ; they do not pene-
trate the cortex.

Spicules. Choanosomal tyles of the radiating fibres, 1120 x 20 /t, straight,
fusiform, attenuating gradually to a sharp point; head long, oval. 20 /t long, 17 fi
broad.

Tyles scattered between the radial fibre*, varying in length from 370 to GOO/x,
slightly curved.

Rgotyli'.H in form ol ^plicrotyfos, 760 ji l"H-j -ni'l i" /< m diameter al th< broad
distal end, which is clavate, ami with a granulated surface ; the heads at the proximal
end (in the interior of the sponge) are 15 /i in diameter.

The cortical tyles (XIII. 11), 218 x 12'5/*, are slender, tapering, curved, and
with small oval heads and a thin neck ; there is a shorter, straight* r variety of cortical
tyle (XIII. 12, 12a) with a relatively thicker neck, and abruptly pointed.

Although there is some uncertainty as to the actual size of the spicules in the
specimen described by Vosmaer, yet the relative proportions and characters of these
spicules are very similar in the Arctic and Antarctic specimens. The Arctic specimen
was globular and 2 cm. in diameter.

Topsent mentions ftadiellti (tchcenus Sollas, as a synonym of Vosmaer's species.
Sollas gives the only names without generic or specific definition.

The specimen was obtained from Winter Quarters, off Flagon Point, 10-20 fms.
It has also been found at Lat 72° 14' 8" N., Long. 22° 30' 9" E., 165 fms., Vosmaer,
Willem Barents Expedition.

FAMILY SCBERITIDAE, Schmidt

SUBERITES MICROSTOMU8 Var. stellatUS.

(Plate XV., figs. 8-13.)

Subtrittt mifrotloiiiujs Ridley and Dcndy (15. p. 199).

There are four small specimens in the form of smooth, pale-coloured, whitish
rounded masses, smooth to the eye but slightly rough to the touch, encrusting the
stems of branching Polyzoa ; the largest is about 4*5 cm. long and 2*5 cm. broad, and
forms a nodulated mass on the branch of the Polyzoon.

There are two or three minute conical pointed oscular papillae alxnit 1 mm. high.
The pores occur in stellate areas among the cortical tyles which occupy the rest of
the surface. The skeleton is composed of radiating fibres and of a dense cortical layer
of vertical or oblique tyles.

Spicules. Megascleres. Large straight fusiform tyles with an oval head,
gradually attenuating to a sharp point, 1130 x 24 /x. The head (XV. 9a) is usually
sub-tylote or even stylote, and has a sub-terminal oval swelling 13/t in diameter. The
cortical tyles vary considerably, the figures showing three kinds. The largest 500 x 30 ft

20 R. KIRKPATRICK.

(XV. 10) with a small spherical head 20 /* in diameter, with a slender neck, and a thick
fusiform shaft, blade-like in optical section, and with tornote end ; a few of these
spicules are scattered in the choanosome between the radiating fibres. A second kind
(XV. 12) 220 x 20 /j. is straight, with a short, thick neck and cylindrical shaft with blunt
point. A third kind (XV. 11) 220 X 7 /u, is slender, slightly curved, and sharp-pointed.

The new variety differs from the typical form in having stouter cortical and
choanosomal spicules. This is at once obvious on looking at the vertical sections of the
two forms, the cortical spicules especially being much thicker and denser in the new
variety. Again, the poral areas on the surface of the type are small and circular
(XV. 14), whereas those of the new variety are stellate (XV. 13).

The specimens were dredged near Winter Quarters, at No. 10 hole, 130 fms.

The typical form was obtained from between Kerguelen and Heard Island from
150 fms. (Voy. 'Challenger').

SUBERITES CAMINATUS var. papillatus.
(Plate XV., fig. 16, and Plate XVI., figs. 11-14.)
Suberites caminatus Ridley and Dendy (15. p. 198).

There are two specimens, each firmly attached to a piece of rock ; one has two
oscular chimneys, the other only one. The body of the larger specimen is 2 • 5 cm. in
diameter and 1 • 5 cm. in height, the oscules being 7 mm. in height.

The upper half or more of the surface is covered with small flattened papillae, each
about 1 to 1 • 25 mm. in diameter, and about '7 mm. high. These structures carry the
pores in stellate grooves (XV. 16). At the base of the sponge is a chitinous lamella
with tangential tyles scattered irregularly and intercrossing. At the base of the larger
specimen the basal lamella is produced into a fringe of little processes formed of
flattened bundles of tyles, the distal ends of which have become rounded. The skeleton
is formed of radiating choanosomal fibres and of a dense cortical layer of dressed
tyles.

Spicules. Straight fusiform tyles (XVI. 14), 1530 x 20 /u, with small oval head
15 x 10 fji. Cortical tyles mostly of two kinds ; one (XVI. 12) with small
spherical head, with relatively slender neck and thick cylindrical shaft, blade-like
in optical section, with tornote pointed end ; the other (XVI. 13) shorter and
stouter and with thicker neck. A fourth kind of spicule, rather rare (not figured), is a
sub-tylote, 1,530 x 20 /j. long, with a very slender curved distal end projecting beyond
the surface, resembling the distal end of the heteroxea of Stylocordyla.

The chief variation from the type lies in the poral papillae. In the typical form
the pores are in stellate grooves. The papillae of the new variety called to mind the
same structures in Trichostemma irregularis R. & D. On examination, the papillae in
this species also proved to be poral, with the pores in radiating grooves.

The two specimens were dredged from a depth of 254 fms., due west of Balleney

TKTRAXONIDA. 21

Island. The typiral form came from off the Rio de la Plata, 600 fins. ( Voy. ' Challenger ').
Topaent records it from near Fayal, 130 metres, and 55 miles N.N.W. of Fayal, 1900
metres. %

l'>KriMISI I'.KI'.I'I K> HYAIIS'I v

(Plate XXVI., figs. 7a-b.)

IK*?. Hynuniafitian ? hyalina Ridley and Dendy (15. p. 168).
1898. Pttitdotubrrite* hyalinut TopKDt (27. p. 108).
1900. PtaubtukrUtt hyaliniu Topsent (28. p. 170).

The single specimen is of a compressed cylindrical shape, 6 cm. long and 8 mm. in
diameter. The colour is white. There does not appear to Iw any area of attachment,
both ends being complete and rounded. Two thin walled oscules, each about 2 mm. in
diameter occur along one edge of the specimen, the exhalant canals proceeding towards
them being clearly visible. On one side of one of the oscules is a wall of spicules
formed by a continuation and flattening out of some of the lateral or secondary skeletal
bundles.

A longitudinal section shows clearly a central loose core of longitudinal fibres each
about 100 ft thick, anastomosing slightly and giving off at right angles bundles of
spicules which support the dermal membrane ; in the thinner parts of the specimen
these bundles are one spicule in length, but at the thickest part they may be two or
three spicules in length and traversed by longitudinal bundle*. The original
specimens obtained from Kerguelen by the ' Challenger ' are " massive and amorphous,"
and the skeleton is more of the " halichondrioid type," but here also the longitudinal
bundles and those supporting the dermal membrane can be distinctly made out In
the Antarctic specimen the cylindrical shape has brought about the more regular
arrangement The Mediterranean sponge, identified by Topsent (27. p. 103) as
P. hyalinu.*, is massive, amorphous, and encloses foreign bodies. The surface is
mammillated and a little hispid, this condition resulting from the prolongation of the
lateral bundles or fibres of the skeleton. The tyles of the Antarctic form are about
1000 x 19 ft, those of the Kerguelen form 1100 x 25 n, and those of the Mediterranean
form from 300 x 10 ft to 1200 x 26 ft.

An interesting feature is that all the spicules point in a direction upwards or
upwards and outwards.

Two other species of Pseudotuberites are P. *vdphureu* (Bowerbank), from the
seas of N.W. Europe, and P. andreicsi, Rirkp. (9. p. 135), from Christmas Island ; this
last species forms a cake-like crust, and its spicules are much smaller than those of
P. hyalinus, being only 350 x 6 ft.

The specimen was dredged from W.Q. No. 10 hole, 130 fms.

The species has been found off 8.W. Patagonia, Voy. 'Challenger,' St 31 1, 175 fms.,
and off the Mediterranean coast of France in 500 to 600 metres, Topaent

22 R. KIRKPATRICK.

TRIBE ACICULIDA, Topsent.

FAMILY STYLOCORDYLID^E, Topsent.

STYLOCORDYLA BOREALIS var. acuata.

(Plate XVI., figs. 6-10.)

18G8. Hyalontma boreale Loven (Ofvers. Vetensk Akad. Forhand., Arg. xxv, No. 2, p. 105, pi. ii.).
1873. Stylocordyla borealis Wyville Thomson (' The Depths of the Sea,' p. 113, fig. 13).

The small specimen representing this variety has a slender stalk 6 • 2 cm. long with
an oval head 7x5 mm. A second broken stem arises from the common base which
has a few grains of gravel and a fragment of the horny tube of a Hydroid attached
to it.

The following spicular elements occur :—

1. Large oxeas with a central swelling (XVI. 10), in the axis of the stem,
1450 x 60/x.

2. Smaller oxeas without central swelling (XVI. 8), in the radiating fibres of the
head, 900 x 20 p.

3. Heteroxeas (XVI. 9) 1000 X 10 p. in the radiating fibres of the head.

4. Microstyles (XVI. 7, 7a) 110 x 4 /*. vertically dressed at the surface of the
stem, and not found in the head. In the typical form there are microxeas, which
may be centrotylote. The variety is named " acuata " from the presence of the
microstyles.

Topsent (25. p. 286) considered that Stylocordyla stipitata Carter, differs from
S. borealis in not having microxeas, which the latter possesses ; and he based his
opinion on the absence of these spicules from certain preparations of S. stipitata
lent to him by Canon Norman ; but I have found the microxeas in abundance in
Carter's type specimen. Accordingly there is probably only one known species of
Stylocordyla, viz., the typical S. borealis, and two varieties, var. globosa R. & D., and
var. acuata nov.

The figures of Loven and Sars do not show any indication of the microxeas being
centro-tylote, but this condition is found in specimens from the Bay of Biscay.

The Antarctic variety has no dermal tangential microscleres in the head, but
microxeas occur in this position in the typical forms from the North Atlantic.

The ' Discovery ' dredged the specimen from off Erebus and Terror in 500 fms.
The species has been obtained from the following localities :—

North Sea (Loven) ; between Scotland and Faroe (Carter, Schulze) ; Grenada
(Schmidt); South of Nova Scotia, 85 fms. (Voy. 'Challenger'); Bay of Biscay, 1710
metres (Topsent) ; Bahia 7-20 fms. (Voy. ' Challenger ') ; var. globusa — Southern Ocean
145 fms. (Voy. ' Challenger') ; off Kerguelen 10-100 fms., Voy. ' Challenger.'

TETRAXONIDA. ;

SII;(IKI>KK HAI.H •HONKKIXA, Vosmacr.
FAMILY AXINKLUD.*, Ridley and Dcndy.

AXINRLLA SfPRATUMKSCKNS.

(Plate XXII., fig. 8, and Plate XXVI., figs. 6a-b.)
1907. Aritulla »upr<ttumt»ttn» Topecnt (31. p. 6).

The specimen, which is a small conico-cylindrical fragment 3'5 cm. long and 9 mm.
in diameter, is the broken-off end of a branch. The surface is finely hispid, and the
broken end shows the dense but flexible central axis.

The straight or slightly curved styles are mainly of two sizes, the larger (XXVI.
Ga) being 594 x 12 '5 /a, and the smaller (XXVI. Gb), which forms the surface tuft*,
306 x G-25/i.

The specimen was dredged near Winter Quarters in 10 fms.

The ' Francais ' Antarctic Expedition obtained numerous specimens from Booth-
Wandel Island and Wiencke Island, from low water up to 30 metres (Topscnt, 31. p. 74).

SIOMAXINYSSA Kirkpatrick.
1907. Sigmafiny»M Kirkpatrick (lOa. p. 272).

Cup-shaped Axinellidse with longitudinal skeletal fibres joined by transverse ones
on the inner aspect, and with tufts given off at right angles to these on the outer
aspect Megasclores, oxeas ; microscleres, sigmata and toxa.

SlGMAXINYSSA PHAKKLLIOIDES.

(Plate XVII., fig. 6, and Plate XXIV., figs. 4a-c.)
1907. Sgmajunysta phaluUioidet Kirkpatrick (lOa. p. 272).

Sponge sessile, cornucopia- or cup-shaped. Inner surface smooth, outer surface
coarsely pilose. Consistence rather hard, but flexible. Colour in spirit, grayish drab.
Inner surface with numerous small oscules, each about 1 mm. in diameter ; outer
surface pilose, with dermal membrane perforated by round pores 95 /* in diameter.

Skeleton. On inner surface formed of close set longitudinal lines joined by
cross bars, and giving off tufts of fibres, which proceed outwards at right angles to the
outer surface, pushing up the dermal membrane, but barely projecting beyond it

Spicules. Oxeas, 835 -5 x 42 • 25 /*, curved at centre, sharp-pointed. Sigmata,
81'25/i long, 35 '75 p broad, and 3'25/i thick, often with an angular bend at centre
of shaft Toxa, 130 /* long, and 3 • 25 /u thick at centre, with smooth surface.

This species bears in its outward aspect a very close resemblance to cup-shaped
species oiPhakellin ; also the skeletal arrangement is like that of Phakellia ; the oxeas,
sigmata and toxa are those of a typical Gellhis. The Axinellid genus Sigmaxintlla
Dendy, which has microsclcres in the form of sigmata, has styles for mcgascleres.

The only specimen was dredged off Coulmau Island in 100 fma

VOU IT. 2D

24 R. KIRKPATRICK.

FAMILY DESMACIDONID^E, Ridley and Dendy.

I. SUB-FAMILY ECTYONIN^;, Kidley and Dcndy.

HYMEDESMIA AREOLATA.

(Plate XXII. , fig. 3-3c.)

1905. Hymedesmia areolata Thiele (23. p. 452).

There are two specimens, one a fine large one, the other a small nodule ; both have
a broken surface, and on the label is the legend " Broken off a stone."

The larger specimen is an oval mass 12 '5 cm. long, 6' 5 cm. high, and 5 '5 cm.
thick.

The species is easily recognised from the areolated appearance of the surface.
There are numerous oval or circular poral pits, 1 • 5 mm. across, with over-hanging
edges ; the concave floor is perforated with pores.

The densely packed ectosomal oxeas are arranged radially round the pore pits.

The oscules are small inconspicuous cones contracted to a point.

The flagellated chambers are aphodal, 36 X 23 /x in diameter.

Thiele's specimens were in the form of incrustations on chitinous tubes.

The ' Discovery ' specimens were dredged near Winter Quarters in 12f fms.

The species has also been recorded from Calbuco, Chile, 40 metres (Thiele).

HYMEDESMIA EXIGUA.

(Plate XXII., fig. 4, and Plate XXVL, figs. 2a-f.)
1907. Hymedesmia exigua Kirkpatrick (10a. p. 273).

Description. The sponge forms a thin translucent grayish-white crust, about
5 mm. in diameter, on a stone. The surface is smooth, and the substance of a fleshy
consistence.

The skeleton. The choanosome contains scattered short acanthostyles, and the
dermal membrane tangential tylotes isolated or in bundles of a few.

Spicules. Megascleres. Acanthostyles, 94 x 19 /i, short, thick, with spines
pointing backwards slightly. Ectosomal tyles, 157 X 3'5 //,, straight, smooth, with
oval heads, 5 /u, long, and 4 ' 5 fi broad.

Microscleres. Pluridentate isancorae spatuliferae : with five foliate teeth, 5 /t in
length, at each end, sometimes with three or four ; shaft deeply curved, 2 • 5 fi thick,
sometimes with central alate expansions.

Sigmata 9 ' 6 /t long, 5 • 6 broad, • 5 p thick, scattered separately in the choanosome.

The new species resembles in several respects H. zetlandica Bowerbank, but the
ancorae of the latter have only three teeth, the sigmata are much longer (51 yu) and
in sheaves; also the ectosomal tyles are much larger, viz., 328 x 3 '25 /x, and
the spines of the acanthostyles more verticillate. Hymedesmia irritans Thiele, from

i i:i i:\.\nMi.\ 25

%
Juan Fcrtiandcz, has nearly the same spicular elements, but of different dimensions, and

has the labis among its microscleres.

The specimen was dredged from 254 fins, due west of Balleney Island.

HYMERRHAPHIA RDFA.*

(Plate XXII., fig. 5, and Plate XXVI., fig. 3a-e.)
JjpMTffcpMingb Kirkpalrick (lOa. p. 274).

Description. The sponge forms a thin mud-coloured brown crust on a branched
Polyzoon. The surface if smooth, and no pores or oscules are visible. The consistence
is rather tough. The skeleton of the choanosome is formed of longer and shorter
acanthostyles dressed vertically, that of the ectosome being formed of tangentially
•Banged auisotoruotes, either isolated or in bundles.

Spicules. Megascleres. Larger acanthostyles, 312 x 25 M, swollen at the head,
spiued all over, with larger curved spines at the head. Smaller acanthostyles,
131 x 18 '75 M. Anisotornotes of ectosome, 344 x 12 M, straight, fusiform, attenuat-
ing gradually at one end but abruptly at the other. Microscleres. — Isancorae, 28 '5 M
long, with three or four teeth at each end ; rarely, the teeth are not developed, the ends
being in the form of hemispherical cups (XXVI. 3e).

Dredged near Winter Quarters, No. 10 hole, 130 fms.

OPHLJTASPONGIA NIDIFICATA.

(Plate XXII., figs. 6, 6a, and Plate XXVI., fig. 5a-d.)
OjMUtupongia nidtficata Kirkpatrick (lOa. p. 274).

Sponge massive, of an inverted pyramidal shape, sessile, narrow and contracted at
its point of attachment Surface uniformly coarsely spinous. Circular oscules ( • 7 mm.
in diameter) on the upper surface, at the bases of the spines. Sub-dermal cavities flat
and shallow. Flagellated chambers diplodal. Colour in spirit, dirty gray, the tips of
the spines being yellowish. Consistence hard and tough.

Skeleton consisting of dense branching axes of styles cemented with spongin and
echinated by smooth styles passing at right angles from the axis to the surface.
Ectosomal spicules in form of slender straight styles. (Considerable tracts of dermal
membrane were devoid of these spicules.)

Spicules. Megascleres, large, straight, smooth styles, on an average about
1000 x 50 M. Also smooth curved kind 625 /LI long. Ectosomal styles, straight,
smooth, tapering gradually to a point, 406 x 9 M- Microscleres, toxa smooth, 638 M
long, 6 '25 A* thick at the centre. These spicules occur in nests or groups of 5-10.

The specimen is attached to a stone by a narrow base about 2 cm. in diameter.
The height is 6*5 cm., and the greatest breadth on the upper surface 7*5 cm. The
upper surface is triangular, each angle being slightly produced and provided with an
oscule, there being also an oscule in the centre.

The spines, which are simply the ends of skeletal fibres, are about 1 cm. in height.

• Inadvertently the specific name ha* been badly chown. the colour— in alcohol— being that of pale brown mud.

2 D 2

26 R. KIRKPATRICK.

At the base of the spines is the thin dermal membrane, generally torn away, thus
exposing the circular openings of inhalant channels.

The new species comes well within the genus Ophlitaspongia Bk., as amended by
Dendy (5. p. 36). 0. seriata Bk., 0. subhispida Carter, and 0. membranacea
Thiele, all have toxa, but the first and third species are encrusting, and the second has
long slender branches. The chief characters of the new species are the strongly spined
surface, and the great development of the spicular core, the spongin not being so
abundant as in other species, and not forming distinct fibres.

The single specimen was dredged off Coulman Island in 100 fms.

LISSOMYXILLA Hanitsch.

This genus was established by Hanitsch (8. p. 194), to include Ectyonine Sponges
with fibres having a core of smooth styles echinated by acanthostyles and with special
ectosomal megascleres ; with or without microscleres. Unfortunately the species he
selected as type of the genus ( Tethea spinosa, Bowerbank) in no way fell in with the
definition, since, as Topsent points out (28. p. 265), this species has neither echinating
spicules nor special ectosomal spicules, and Topsent refers Lissomyxilla to the limbo of
useless names. Among the ' Discovery ' sponges, however, is a specimen which fits in
with Hanitsch's definition of Lissomyxilla, which runs, " Skeleton fibres of the
choanosome, formed of smooth monactinals echinated by spined styles. Megascleres of
the ectosome smooth diactinals or monactinals. Microscleres (isochelae, &c.) may be
present." Accordingly, I propose to revive the name.

LISSOMYXILLA HANITSCHI.

(Plate XXII., fig. 7, and Plate XXVI., fig. 4a-c'.)
1907. Lissomyxilla hanitschi Kirkpatrick (10a. p. 275).

Description. There are two much-worn pieces of this sponge of a dark, dirty-gray
colour, and a third young small specimen, whitish in colour, growing on a specimen of
Hornera. The largest specimen is 4 cm. high and 5 cm. in diameter at the base ; the
dermal membrane is worn away, exposing several openings of exhalant canals, 4 mm. in
diameter. The dermal membrane of the smallest specimen is transparent, smooth,
and raised up at one place into a small conical oscule, with radial tangential spicules in
its walls.

The skeleton of the choanosome is formed of branching fibres on an average
about 150/a thick, echinated by spined styles in an obscurely verticillate manner, the
whole skeleton, as seen in sections, having a somewhat confused appearance. The
ectosomal spicules partly proceed obliquely from the main fibres to the dermal
membrane, and partly lie tangentially in that membrane.

Spicules. Megascleres. Smooth styles, 500 X 19/i, smooth, curved near the
head, sharp-pointed.

TKTKA.\t»XII».\ 27

Echinating acanthostyles, 219 x 18'75/j (without including spines), larger spines
situated on the head 9 /< long.

\.- ' — nial ainpli it ylrs, :!.">(; • I I .. -: MI- :i raj dightly fusiforn ibtylot*- ind
with a small uiucru at each end.

Mi. ros. Im-s, none

Myjcillti victorutwt Dcndy (Ifalicltondria pushtlosa Carter), (5. p. 30), would come
under this genus, although, at the same time, it is in possession of isochelac, and the
heads of the styli of the main fibres occasionally have a slight indication of spination

Two fragments (No. 123) were dredged off Coulman Island in 100 fins. A third
specimen (No. 38) on Horn/era was obtained from east end of Barrier, 100 fins., l>ottorn
mud, stones, and rocks.

II. SUB-FAMILY MYCALINAE, THIKLK.
I. GROUP MYXILLKAK, LUNDBKCK.

MYXILLA DKCKPTA.

(Plate XXII., figs. l-2a, and Plate XXV., fig. 3a-£)
1907. Mynlla dettpta Kirkpatrick (lOa. p. 278.)

Description. There are two very small specimens of this species. One is in the
form of an extremely thin incrustation on a piece of rock ; the surface is pilose owing
to the projection of vertically dressed acanthostyles, each surrounded by tufts of
ectosomal spicules. The other, which incrusts the branch of a Polyzoon, is thicker, and
the surface here is partly smooth, partly provided with minute sharp-pointed couulcs
supported by acanthostyles. The colour of lx>th specimens is reddish brown.

The skeleton in the very thin incrustation, at first sight, resembles that of a
Hymerrhaphia ; each vertical acanthostyle is isolated, and with its head on the base
and its pointed end projecting.* In the thicker specimen it is possible to make out
primary and secondary lines of skeletal fibres.

The cctosomal spicules are partly arranged in paniculate tufts, partly lying
tangeutially in the dermal membrane.

Spicules. MflflMileim Choanosomal acanthostyles, 4G8 x 23*5/i, curved,
spined at the head only, with sul>-tornote points. Ectosomal strongyles 238 x 4 • 6 /*,
straight, smooth, cylindrical, usually with a pointed mucro at one end.

Microscleres. Arcuate isochelse, 19'5/i long, 5'G/i broad, palmate tooth 8 ft long
ovoid, with rounded distal edge ; with thick curved shaft ; with tongue-shaped ale
about 8 fi long.

Isancorse uiiguiferae, 15'3/z long, with slender, curved, sometimes wavy shaft,
with three sharp claw-like teeth at each end.

Chelate bipocilla 8 /* long, with deeply curved shaft with spathulate ends each
with three triangular denticles ; these spicules are fairly common and not accidental.

K. KIRKPATRICK.

There are also several isocheliB arcuatae in which the alse and denticle are replaced
at one end by a spoon-like lamella.

The presence in Myxilla of chelate bipocilla similar in many respects to those
found in the new species lophon spatulatus and /. flabello-digitatus is exceptional ; some-
what similar structures occur, however, in Myxilla iophonoides Svartzevsky (Mem. Soc.
Nat. Kieff, Tome XX., p. 340, PL XV., fig. 27g-l) from the White Sea, but there are no
isochelse arcuatse. In other respects the spiculation of the new species is that of
a typical Myxilla. The isochelae arcuatse, though only half the length, resemble in
shape those of Myxilla nobilis R. and D., from off the Rio de la Plata, and M. digitata
R. and D., from the Cape of Good Hope.

The specimen incrusting the rock was dredged from 254 fathoms, due west of
Balleney Island ; and the specimen incrusting the Polyzoon from Winter Quarters,
No. 6 hole, 125 fathoms.

LlSSODENDORYX SPONGIOSA.

1887. Myxilla spongiosa R. & D. (15. p. 134, PI. XXVII., figs. 3-3f).

1901. Lissodendoryx spongiosa, Topsent (29. p. 18).

One small specimen and three minute fragments were obtained from three localities
in the neighbourhood of Winter Quarters. The complete specimen is cylindrical, 2 cm.
long, and • 5 cm. in diameter, and is attached by part of its length to the branch of a
Polyzoon. The specimen is well preserved. The colour is white, an opaque white axis
showing through the transparent dermal layer.

The spicular elements are similar to those of the Challenger type, and sigmata
occur in all, though in varying proportion. Topsent describes a variety of this
species from the ' Belgica ' collection, viz., L. spongiosa, var. asigmata, which is
wholly without sigmata.

The specimens were dredged near Winter Quarters, (l) from near the ship, 20 fms. ;
(2) No. 12 hole, 25-30 fms. ; (3) McMurdo Bay, 4-10 fms. The species has been
found also off Rio de la Plata, 600 fms., Voy. ' Challenger.' The ' Belgica ' collected
the var. asigmata in Lat. 71° 14' S., Long. 89° 14' W. in 450 metres.

lOPHON RADIATUS.

(Plate XXL, figs. 3, 4, 4a-c, and Plate XXV., fig. 4 a-c1.)

1902. I&phon radiatus, Topeent (29. p. 21, PI. III., fig. 13).

Specimens and fragments come from five different localities. They are mostly in
the form of small cylinders about 4 mm. in diameter and varying in length up to 5 cm.
The longest (XXI. 4) is 5 cm. long and 4 mm. broad, tapering to a point at one end.
The colour is dark brown, a dark axial core of choanosome showing through the semi-
transparent dermal membrane. One specimen, the largest (XXL, fig. 3), has broadened
out into an oblong loose rather flabby mass, torn at each end, 4 cm. in length and
2' 5 cm. in breadth, and with a branch given off on each side near one of the ends.

TKTUAXONIDA. 29

Several <»smle8 about 9 nun. iu diameter occur along the margin ; they an slightly
raised atul with a thin plain rim with sjHvial spirular armature. The pores owupy
irregular oval aren-s about 2 mm. in diameter. The spirules resemble theme of
specimen No. 421 described by Topsent, the dimensions Injing as follows : — The amphi-
t\l. •_•<'. i ; . 1 1. lt : iicanthustyli's -|.iii.-.i ,-ii the !I<M.J tad ai u tlu point, i ir < L6ji
larger anisoeheles 36 p long, a smaller kind 20 M loug ; bipocilla 9*75 M long, usually
\\iili one end pocillate and the other flattened and with three to five minute denticles.

An interesting feature is the presence of embryos '25 mm. in diameter with a
solid cylindrical coreof peculiar spiued amphityles 105 x 5/u reaching from surface to
' nitre, and with anisocheles with a spine at the small end, 15 '9 M long and arranged
in rosettes, the latter being distributed around the summit of the spicular plug or
cylinder.

Specimens were obtained from (l) Winter Quarters No. 12 hole, 25-30 fins. ;
(2) East end of Barrier, 100* fms. ; (3) No. 5, Seal Hole, 178 fins. ; (4) From near the
ship, February 28, 1902 ; (5) Off Couhnan Island, 100 fms.

The 'Belgica' collected specimens from Lat 71° 14' S., Long. 89° 14' \V.,
450 metres, and from Lat 71° 19' S., Long. 87° 37' S., 450 metres.

•

lOPHON 8PATULATD8.

(Plate XXL, figs. 5, 5a-c, and Plate XXV., fig. 5a-d.)
1907. lophon spaJttlatiis Kirkpatrick (lOa. p. 276).

Sponge slender, cylindrical, branched. Colour, pale brown in spirit. Oscules
oval, about 1 x • 6 mm. in diameter, slightly raised ; surrounded by a radiating zone of
tyles. Pores in sieve-like areas 2-3 mm. in length and about 2 mm. in breadth, on
a level with the general surface.

Flagellated chambers 26 x 23 ft. '

Skeleton. Dermal ; a closely packed layer of dermal amphityles.

Choanosomal, a loose network of 3-6 spicules thick, forming the core of the
cylinder ; multispicular fibres, giving off strands at right angles ; the longitudinal
fibres are joined obliquely or transversely by fibres one or two spicules thick, which
support the dermal membrane.

Spicules. Megascleres, smooth styles, 462 x 25 ft, with a mucro at the I M-M I end.
Ectosomal sub-amphityles 225 x 12 '5 ft, fusiform with the swelled ends smooth
laterally, and with the extremities only slightly convex, in fact almost truncate, and
covered with spines.

Microscleres. Palmate anisochelee 18 • 7 ft long, 6 • 2 p. broad (front view).

Bipocilla, rare, IS '6 ft long, I '15 ft thick, shaft deeply curved, slightly twisted,
and with ends almost similar, spatulate, with crenulated edges ; occasionally with five
relatively large denticles in place of the finer crenulation. A second kind of
bipocilla have a longer, less curved shaft with scoop-like rather than spathulatc

30 R. KIEKPATEICK.

ends, each scoop having four or five denticles. One example has three sharp prongs
at one end and the spathulate prolongation at the other.

The species is represented by numerous small, for the most part fragmentary,
slender cylindrical pieces. Many of the fragments came from the same locality as the
specimens of /. radiatus and were mixed up with them. Generally it was easy to
separate the two by the colour, that of /. radiatus usually being of a much darker
brown ; but in one or two instances the specimens of the new species were also dark
coloured. Several of the fragments are branched. The specimen figured (XXI. 5) is
4 cm. long and 5 mm. in diameter. The oscules are provided with a ring of radiating
amphityles, whereas /. radiatus is devoid of such structure. A glance at the spicules
at once enables one to distinguish the two species. The ectosomal amphityles of
/. radiatus have well-marked knobs spined all over, and not merely at the ends. The
choanosomal spicules of /. spatulatus are apparently modified acanthostyles, smooth
along the shaft, but with the spines at the rounded end gathered into one large
spine situated on the summit, thus giving it a tornote aspect. In one instance, the
single spine is represented by a little tuft of the spines. The axial canal terminates
some distance below the base of the spine.

Several spherical embryos 268 n in diameter were present in one fragment, but
there were no special spicules in them.

Since writing the above description, accounts of two new species of lophon from
the Antarctic have been published by Topsent (31. pp. 4-6), viz. /. unicornis and
/. pluricornis. The two new species described in the present Eeport and Topsent's
two species all possess the curious modified acanthostyles. /. unicornis Topsent, has no
bipocilla, and /. pluricornis has bipocilla apparently with pocillate ends ; the ' Dis-
covery ' species have spathulate bipocilla with crenulated ends ; and further, there are
differences, viz., in the shape of the body, the arrangement of the skeleton, and the
character of the ectosomal spicules, which lead" me to regard the 'Discovery' specimens
as belonging to distinct species. A differential diagnosis of the four Antarctic
species with modified acanthostyles is given at the end of the description of the
next species.

Fragments were dredged at No. 11 hole, 28 fms.; at No. 12 hole, 25-30 fms., and
off Coulman Island, 100 fms.

lOPHON FLABELLO-DIGITATUS.

(Plate XXL, figs. 6, GA-C, Plate XXV., fig. 6a-f, and Plate XXVL, fig. la-c1.)
1907. lophon flabello-digitatus Kirkpatrick (lOa. p. 277).

Sponge forming a large palmato-digitate or digitate growth in one plane ; branches
compressed, usually with oscules along one edge. Surface finely verruculate ; with
elongated pore areas supported by fan-like wisps of ectosomal spicules.

Colour dark brown ; consistence soft, the sponge being readily broken.

TBTRAXONIDA. 31

Skeleton typi.alK I'.-imc.l of a network of spiculo-fibre, the primary HUGH of which

M i upwards :m«l outwudi ti»m :.-.•• mic r -mtarc . .t" • • - ular tabM tO '••• MfeM

surfm •«• i>t tin ~|M.ir_'c. where their ends forni the vcrrucae ; the secondary fibre** join
the primary at right angles forming rectangular meshes a)>out 1*25 mm. square. The
thickness of the primary liltivs is al>out '7 mm., that uf the secondary about *5 mm.

The ectosomal skeleton consists of fan-like bundles and wisps of spiculcs, isolated
en priMTi-iling upwards and spreading out from the terminal main fibres.

Spicules. Megascleres. Smooth styles, 590 x 25 M, curved, with a mucro at
the head cud.

•somal spiculcs, 344 x 12'5 M, sub-amphitylote, fusiform, with a marginal
ring of vertical spines at both ends and a terminal central spike at one extremity.

Microscleres. Palmate anisochelae of two sizes, a large kind 35 ^ long and 107 M
broad, with thick shaft; with triangular palmate tooth, 17*5 .« long, at the large end,
not quite as long nor as broad as the alae ; lower margin of alae convex. Lower
central tooth with a curved upper edge produced into a spine.

A small kind 17 '5 ft long, 6*2 M broad ; upper palmate tooth triangular, rounded
above, as broad and as long as the alae ; lower border of alae concave ; lower tooth
with simple rounded upper edge.

Bipocilla varying in length from 5 '5 to 11 M, according to the convexity of the
shaft which is usually deeply curved ; lioth ends spathulate, nearly similar, and with
crenulate edges, or with 5-7 teeth.

The tine specimen (No. 184) which constitutes the type of the new species, is
24 cm. wide and 25 '6 cm. high. Unfortunately, owing to the very soft nature of the
tissues, the sponge has IKJCII broken into many fragments. A short stem expands
into a palmate portion, from the edge of which arise several compressed digitate
branches, usually with round sphinctrate oscules along one edge ; the oscules lead into
a pseudogastral cavity extending nearly to the opposite edge of the branch. The
bipocilla of the new species somewhat resemble those of /. sjmtulntus, but the difference
in the body form, skeletal arrangement and spimlation arc too many to necessitate
detailed enumeration.

A second specimen (No. 202) has two compressed branches rising from a flattened
contracted base, the under surface of which is coated with pebbles. The oscules face
each other along the edges of the entering angle. The branches are each about 10 cm.
in length, the greatest breadth being 4*5 cm.

A third specimen (No. 287) of this species, which might, indeed, be regarded as
a variety, forms a discoid crust on a small Ophiurid. There are embryos, 340 M in
diameter, present, and some, but not all, of these have a plug of peculiar tylote
spicules in them, with a ring of anisocheles round the outer end of the bundle.
The tylote spicules of the embryo, 161 x 7 '8 M, have a smooth rounded end
surrounded by spines, the other end being cylindrical with marginal spines, and a
single central spine.

TOU IT. ml

32

R. KIRKPATRICK.

The anisocheles of the embryos nearly resemble the smaller kind of these spicules
found in the choanosome of the type specimen.

The type specimen (No. 184) was dredged near Winter Quarters, No. 11 hole,
28 fms. ; specimens 202, 287, and 289 were dredged from No. 10 hole, 130 fms.

DIFFERENTIAL DIAGNOSIS OF THE FOUR SPECIES OF IOPHON wrry: MODIFIED ACANTHOSTYLES.

Shape.

Ectosomal Spicules.

Bipocilla.

lophon unicornis .

Massive, of irregular con-
figuration ; listuliir os-
cules rising from upper
surface.

240 X 10 /x.
Heads with a bouquet of
spines at extremities.

Absent.

lophon pluricornis

Plate-like ; oscules appa-
rently level with surface.

280 X 10 /x.
Heads elliptical, with
spines at extremities, or
over . the whole surface
of head.

Apparently with po-
cillate end or ends.

lophon spatulatus .

Slender cylindrical ; with
oscules slightly raised.

225 X 12-5 /^
Ends truncate, with spines
at extremities only.

Having spathulate ends
with crenulated
edges.

lophon flalello-digitatus

Palmato - digitate, with
oscules along the edges
of branches.

344 x 12-5 /*.
Ends truncate, with ter-
minal marginal rim of
spines at both ends and
terminal central spike
at one end.

Having spathulate ends
with crenulated
edges.

TEDANIA VAEIOLOSA.

(Plate XXL, figs. 1, la, and Plate XXV., fig. la-b.)
1907. Tedania varwlosa Kirkpatrick (10a. p. 279).

Sponge in form of a mass of thick flabellate or digitate fronds arising from a
common base ; with circular sphinctrate oscules, each about 1 cm. in diameter, situated
at the summits, or along the upper edges of the branches, the canals into which they
lead extending nearly to the base of the branches. General surface of the sponge
covered with circular pore areas each about 4 mm. in diameter, the oval or circular pores
being about 90 f- in diameter, and the strands of the poral reticulum about 30 jj. in
breadth. Colour in spirit, pale brown. Consistence, soft and fleshy, being easily torn.

Flagellated chambers, 42 x 35 /u, oval, aphodal, with aphodus (in a measured
example), 13/* long.

Skeleton. Choanosomal skeleton formed of loosely agglomerated compound,
longitudinal or main bundles about 1 mm. in diameter, curving out to the surface as

TI:TI: \\<>\li>\ 33

they pass upwards; tin- s(-|>arntc fibres of the main bundles about 80 n thick. The
main bundles joined at right-angles by secondary fibres, 1-3 spicules thick. Bpongin
not percept il>lr. Ectosomal skeleton formed of circles of strongyles, the spicules isolated
or in fan-like wisps, arranged {wrtly vertically, partly tangentially round the pore area* ;
the vertical spiculcs usually isolated, and the tangential ones in wisps. On drying the
spon L.". tin- edges of the pore areas stand up sharply, the areas themselves sinking in,
giving a jHH-k-mnrked aspect to the surface.

Spicules. Megascleres. Choanosomal styles, 402 X 1 3 M, curved at aln>ut one-
fourth of the length from the round end, smooth, but occasionally with a few spines
about the head.

Botosmniil stNBgjla^ Ml >- 8'Sji, «nooih, oeoMMoally aKghtfj .-\\..ll.-n :,t Mch
end.

. Mieroscleres. None.

The single specimen is in the form of a squarish mass of thick fleshy flabcllo-
palmate or digitate lobes ; the height is 18 cm., and the breadth 13 cm. The flabellate
fronds are obviously formed of fused tubular digitations, as can be seen from the
oscules along the upper edge and from the faintly indicated longitudinal grooves down
the sides. The walls of the oscular canals are smooth and lined with numerous
orifices of exhalant canals, about 3 mm. in diameter. The arrangement of the pores
in circular areas each surrounded by a zone of ectosomal spicules is not common in
Tedania ; it occurs in the second new species descried below, and something of the
kind is found in Tedania tenuicapitata Ridley (15a. p. 124), from the Straits of
Magellan. In the present species this feature is so well marked as to give the surface
a pock-marked appearance.

The rhaphides, usually so characteristic of Tedania, have entirely disappeared ;
but the loss of microscleres is of such frequent occurrence that it has not seemed
necessary to create a new genus or subgenus to include such forms ; though, perhaps,
the more or less definite arrangement of the ectosomal spicules might, in the present
instance, necessitate such a course.

The specimen was dredged near Winter Quarters, 10 fms.

TEDANIA COULMANI.
(Plate XXL, fig. 2. Plate XXV., fig. 2a-b!.)

1907. Ttdania toulmani Kirkpatrick (lOa. p. 280).

Description. The single specimen is in the form of a finger-like fragment
5*5 cm. long and 17 cm. in its greatest thickness. The colour is dirty gray, and the
consistence soft. The surface shows the same circular pore-sieve areas as in T. variolota.
Along one 'side of the sponge the surface has been torn away, exposing an exhalant
canal running along the length of the specimen, but the terminal oscule has apparently
been torn away.

2 K 2

34 R. KIRKPATRICK.

Skeleton. Rings of spicules partly vertical, partly tangential, isolated or in
tufts, surround the pore areas.

The choanosomal skeleton is formed of primary longitudinal fibres about 120 /j.
thick, joined by secondary fibres one spicule in length and 2-3 in thickness, joining the
former at right angles.

Spicules. Megascleres. The choanosomal acanthostyles, 475 X 18 /*, curved,
smooth, or with sparse spines, usually on the upper and lower thirds of the
length.

Dermal ectosomal tornotes, 319 x 12 '5 /*, smooth, straight, fusiform, larger at
one end than the other. Under a high power each end shows a rounded shoulder
prolonged into a mucronate spine. Microscleres absent.

The present species resembles T. variolosa in having the circular pore areas,
and in the absence of rhaphides, but differs widely in the character of the dermal
tornotes. Both species differ from all other species of Tedania in having no
microscleres. The nearest species to the present one are Tedania tenuicapitata Ridley,
from the Straits of Magellan, and Trachytedania spinata Ridley, from the same locality ;
both of these have rhaphides, and neither have the circular pore areas, though in
T. tenuicapitata there is a tendency to a radial arrangement of bundles of dermal
spicules. The spination of the acanthostyles recalls a similar character in Trachyte-
dania spinata.

The specimen was dredged off Coulman Island in 100 fms.

II. GROUP MYCALEAE, LUNDBECK.

ARTEMISINA APOLLINIS.

(Plate XX., figs. 4, 4a-c.)

1887. Amphilectus apollinis Ridley and Dendy (15. p. 124).
1894. Artemisina apollinis Topsent (24. p. 13).
1905. Artemisina apollinis Lundbeck (13. p. 114).
1907. Artemisina apollinis Topsent (81. p. 70).

The single example is a massive cake-shaped specimen 8x8 cm. in area, and
4 • 5 cm. in its greatest thickness. The colour is dirty white in alcohol. One of the
surfaces is smooth and opaque, and probably the sponge lay free on the bottom on this
side. There is a large oval oscule (l'5x '5 cm.) on one of the edges, and several
much smaller oscules on the upper surface. The sizes, in fi, of the spicules are :—
curved styles 600 x 16; straight styles 400 X 6; chelae 13; toxa 300 (or less) x 3.
The flagellated chambers, 29 X 23 M in diameter, are eurypylous. An interesting
feature, not occurring in the type specimen from Kerguelen Island, is the presence of
small fusiform villous processes (XX. 4a) on the surface, especially round the margins
of the oscules ; similar villi, but larger, are found in the nearly related sponge
named Esperiopsis edicardii, var. americana R. and D. The finding of this sponge
in the Antarctic adds another to the list of bipolar species.

I I TKAXONIDA. 35

To-wont (31. p. 70) records a new apedei, Artrinisinti (liaiuu; from Booth-Wandel
Island, Autan-tic, apjwrcntly very closely related to A. apollini*. The ' Discovery '
specimen resembles A. diamw in having curved styles of the same length, but they
arc narrower ; and the spincd toxa arc much smaller in the first •peciefc

The specimen was dredgetl near Winter Quarters, off Hut Point, '25-30 fms.

The ' Challenger ' obtained specimens from Kergueleu Island, 20-60 fins. ; and
the Danish Ingolf Exp. from East Greenland, depth ? , Lundl)cck.

ESPKRIOPSIS VILLOSA.

(Plate XX., figs. 3, 3a-c, and Plate XXIV., figs. 9a-b«.)

1874. Kffiriaprit viHota Carter (1. p. 2 IS).
1H87. Btprrwprit rittota Priatvdt (7. p. 451).
1904. Etpmopns viUota Topscnt (80. p. 211).
IMC. J^Nrfpib *ObM Lundbt*k (13. p. 9).

The single specimen, which has been broken off from its attachment, is massive
below, but divides above into two cylindrical lol>cs. The colour is pale brown in spirit.

A few small oscules, evidently much contracted, occur at the summits of the lotas.
The surface is finely villous, from the presence of the projecting point* of fan-like
lines of oxeas.

Here and fhere the flesh has become macerated away, leaving fluffy lines of fine
skeleton fibres.

The only difference worthy of mention that I can find between the Northern and
Antarctic specimens is the absence of the placocheles or isochclac palmatac with broad
shafts. I can only discover two kinds of isochclae palmatae. The sigmata occur
in many sizes. Some have bifurcated terminations, as figured by Fristedt, and
Lundl>eck.

Spicules. Megasclerea Styles 671 x 18 M, fusiform, straight, occasionally with
head slightly bent

The styles in Carter's type from Faroe are almost identical in character.

Microscleres. Larger palmate isochelae, 43 M long, 9*75 M broad, length of
teeth, 16 M.

Smaller palmate isochelae, 1 8 M long, 4 M broad. Sigmata, numerous, varying from
small up to very large sizes.

The distribution of this species is very interesting, occurring as it docs in Arctic
and sub- Arctic waters ; then, as an intermediate link between Arctic and Antarctic, in
deep water off the Azores.

The specimen was dredged near Winter Quarters, in McMurdo Bay, 20 fms.

This species has been found (l) Between Scotland and Faroe, 440 fms.,
Carter; E. Coast of Greenland, 254 m. (140 fms.), Fristedt; off Iceland, Denmark
Strait and Davis Strajt, 44-553 fms., Lundbcck ; Azores 2,252 m. (1,196 fms.),
Topsent.

36 R. KIRKPATRICK.

MYCALE MAGELLANICA.
(Plate XX., fig. 2.)

1881. Esperia magettanica Ridley (15a. p. 117).

1887. Esperella magellanica Ridley and Dendy (15. p. 67).

1905. Mycale magellanica Thiele (23. p. 442).

The ' Discovery ' collection contains one large piece, much damaged. The figure
(XX. 2) shows the smooth surface in contrast to the hispid surface of the new species
M. acerata described below.

The specimen was dredged near Winter Quarters, from No. 10 hole, 130 fms.

The species is also recorded from Sandy Point, 7-10 fms. ; from Otter Island,
Patagonia ; and from east of Cape Virgins.

MYCALE ACERATA.

(Plate XX., figs. 1, la-b, and Plate XXIV., figs. lOa-e1.)
1907. Myxilla acerata Kirkpatrick (10a. p. 280).

Sponge large, massive, with numerous small rounded mammillae ; surface finely
reticulate and finely hispid. Colour, creamy-white in spirit. Consistence, soft, the
tissues being easily torn. The flesh reddish (but soon decolorised), and showing the
glistening white strands of the skeleton.

Oscules in form of wide thin-walled cylindrical chimneys with rather jagged upper
edges, about 1 cm. in height, and 1-2 cm. in diameter.

Skeleton. Ectosomal : — A network of triangular meshes formed by bundles of
oxeas, the strands being from '35 mm. thick and the meshes about '5 mm. across.
Main skeleton formed of long thick anastomosing fibres which attenuate gradually
from 1 • 5 mm. in thickness and break up a little below the surface into .panicles of
much finer fibres which support the dermal membrane and penetrate the strands and
nodes of the dermal reticulum, giving rise to a finely hispid condition of the surface.
Parallel groups of oxeas scattered in the choanosome.

Spicules. Megascleres. Oxeas 850 X 16' 25 M, slightly curved, rather abruptly
pointed at one end, and more tapering at the other. These oxeas form the fibres, and
also are gathered into bundles, one spicule in length, of parallel oxeas, scattered in
the choanosome.

Microscleres. Large anisochelae palmatae, 105 x 50 /u, separate or in rosettes,
usually with an angular bend in the shaft ; with a triangular upper tooth 60 M long,
about the same length as the upper alae, which latter are very wide. With the
lower tooth oblong, 12'8/u high, with a slightly convex edge; in one of the
specimens this edge is produced into a long denticle (XXIV. 10 c).

A smaller kind of anisochelae palmatae (fig. lOd, d1), 47 /* long and 17 M broad, at
the upper end, with a long oval tooth 20 M long, extending below the alae.

Trichodragrnata, 62 X 12 n, the trichites being very fine, sharply-pointed oxeas.

There are three fine specimens of this species, the largest forming a thick

TKTRAXOXIDA. 37

:v. tlal>< ll.it< l....|v 17 ••in. high, 11 cm. broad, and 7 cm. thick; a second
-|H. iinon is massive and spheroidal, being alwut 15 cm. in diameter. The white
inaiuinillated surface coven a reddish flesh in which the glistening fibres of the
skeleton are visiMr. The mammillae are on an average about '75 cm. in height,
and 1 cm. in diameter at the base. The new species bears a very close resemblance
to ifycalc magellanica Ridley, which likewise has a mammillated, finely reticulate
surface and glistening skeletal fibres, but in the latter the surface is smooth and not
lii-ipiil (XX. 2), and the megascleres are styles, or sub-tyles, such as are normally found
in the genus Myccde, The microscleres also are different in the two species.

\ '••••>iul specie! of -'/'/<•.//« uiih ".vat'- iiii"j.i--I'T.'~ i.- .I//,.-.//. i;j/«TMrt/»Vi ((). Srh.),
from East Greenland (16. p. 433) and Thiele (22. p. 381, fig. 12). The specimen
consisted only of a fragment ; the spicules, which are all considerably smaller
than in the Antarctic species, have the following dimensions: — Oxeas, 450/x long,
10-1 2 M thick; large anisocheles, 50-60 M long; small anisocheles, 18 M long.

Specimens were dredged from three separate localities near Winter Quarters,
viz. : (1) No. 6 hole, 125 fms. ; (2) No. 12 hole, 25-30 fms. ; No. 5 hole, seal
hole, 178 fms.

MTCALK, sp.

Several pieces of glistening white skeletal framework were dredged from No. 12
hole. The largest is cylindrical, 7 cm. long and about 3 cm. in diameter. A very
small quantity of the body substance still remains in some of the angles of the
mcshwork.

The spicules are: — Styles, fusiform, straight, abruptly pointed, 562 X 17 '5 M ;
Urge palmate anisochelae, separate or in rosettes, 100 M long and 35 M broad, with
the larger tooth 47 A* long, nearly the same length as the alae, triangular, with
straight lower border ; with the central tooth at the smaller end squarish,
with straight upper edge; small anisochelae, 14x48/u, with oval tooth 28 n long,
extending further than the alae; sigmata 224 M long, 104/x broad, and G'5/x thick.

This species resembles in many respects a form described by Thiele (23. p. 443,
fig. 61, a-d) as Mycale sp., from Punta Arenas, Straits of Magellan ; but the sigmata
are much smaller in the South American form, being only 1 .'»,«. and the small
anisochelae are smaller and with a narrower and shorter upper central tooth. No
specific name is given, owing to the incompleteness of the specimens.

Dredged near Winter Quarters, No. 12 hole, 25-30 fms.

*

DESMACIDON KERGUELENENSIS var. antarctica.
(Plate XIX., figs. 1, IA, and Plate XXIII., figs. la-dl.)
Dttmofidon kurgveUnentit Ridley and Dendy (15. p. 110).

Sponge ovoid or cylindrical, with hispid or finely conulated surface, the conules
surmounted by fine tufts of spicules, conules and tufts rising to a height of 1 to 2 mm.
Consistence, soft ; colour in spirit, dirty white.

R. KIRKPATRICK.

A single round oseule at the summit. Pores round or oval, scattered, 20 M
in diameter. Dermal layer not detachable; sub-dermal spaces 190 M in depth.
Flagellated chambers about 30 M in diameter.

Skeleton composed of rather thick main fibres about 10 spicules broad viewed
en face, about 190Mor more in thickness, with a few loose spicules between the
main fibres, scattered, but with a tendency to be parallel to the surface, and
forming distinct horizontal bundles just below the surface. Spongin is present, but
not easy to detect.

Spicules. Megascleres. Oxeas 676 X 26 M, slightly curved at centre, terminating
gradually in fine points.

Microscleres. Isocheles 19'5/u long, 6'5/x broad in lateral view, with slender
convex shaft without perceptible alae ; with broad saddle-shaped teeth with their distal
edges not far from each other, but not touching or overlapping.

Four small specimens of this sponge were obtained. The largest is 4 cm. long
and 1 • 4 cm. in diameter. The surface has a finely hispid aspect in two specimens, but
is finely conulated in -the third, and smooth and worn down in the fourth. The variety
differs from the type (15. p. 110) in having much larger megascleres, those of the type
being only 350 x 18 M- The isocheles of the type, again, are longer, viz., 28 M- After
a careful examination of the isocheles of the Kerguelen specimen in a good light and
with the spicules slowly rotating in balsam, I found that the median sub-dental
continuation of the shaft did not bend backwards from the tooth as in Desmacidon
(Homoeodictya) palmata ; the appearance of bending back (see 15. PI. xxiii., fig. 3b) being
due to the lateral bend of the palmate tooth ; accordingly, even if Homoeodictya be a
true genus or sub-genus, 1). kerguelenensis would not belong to it.

Three specimens were dredged near Winter Quarters, from No. 12 Hole, 25-30
fms. ; off Hut Point, 100 fms. ; and offCoulrnan Island, 100 fms.

•

DESMACIDON KERGUELENENSIS var. cactoides.
(Plate XIX., fig. 2, and Plate XXIIL, fig. 2a-b1.)

Sponge pyriform, laterally compressed ; surface prickly, with an Acanthella-like
aspect. Consistence rather firm. Colour in spirit, pale yellow.

Skeleton formed of rather stout main fibres proceeding to the surface, where they
conspicuously push up the dermal membrane, with scattered spicules in between the
main fibres.

Spicules. Megascleres. Slightly curved strongyles, but sometimes with rather
blunt pointed ends (Le. oxeas), 768 x 31 '25 p..

Microscleres. Isochelae palmatae 17 M long and 6'25/x broad with the wide
saddle-shaped palmate teeth almost in contact or even overlapping.

The single specimen is attached to a worm tube by a contracted stalk-like base ;
the length is 5' 5 cm., and the greatest breadth 3 '5 cm. It was a matter of doubt
whether to regard this form as a distinct species ; but it will be obvious from the

- J

TETRAXONIDA. 39

•

figures of the spicules that the differences between the var. antarctica and var. cactoides
are mnvlv those of degree; the teeth of the isochelae, for instance, arc more
approximate I in tin- latter than in the funncr; further, the rounding of the ends of
. the oxeas is n<>t an important distinction ; and lastly, the cactus-like surface wuuld
result from a branching of the. ends of the skeletal fibres just below the surface, so
that the dermal layer would stretch over the points of bifurcation.

If the iicu variety were devoid of cheles it would be placed under Bat:ellat thus
jii-tifving Thielc's observation (23. p. 438) that the latter genus might l»e regarded as.
a Desmaddon which had lost irs chelae.

One specimen was dredged from Winter Quarters, No. 12 hole, 25-30 fins.

I)KSMArIIH)N SI'IMCKKA.

(Plato XIX., figs. 3, 3A, and Plate XXIII., figs. 3a-c'.)
1907. Dttmaridon tpinigera Kirkj«lriok (10a. p. 283).

Sponge digitiform, or knob-like ; surface coarsely spinous. ' Consistence rather
hard. Colour pale red. Several small oscules about 2 mm. in diameter.

Dermal membrane spread like a delicate net between the spines, and at »<mie
distance from the floors of the sub-dermal spaces. Flagellated chambers oval,
46 x 32 ti.

Skeleton formed of coarse, longitudinal, main strands, about 1 80 /t thick, radiating
out to the surface, with loose scattered spicules between, united in horizontal bundles
only Iteneath the surface.

Spicules. MyaiJMM*. Oxeas 731 x 26 /x, curved at centre, mostly sub-tornote,
though some attenuate gradually, with sharp pointed ends.

Microscleres, of one kind, viz., isochelae palmatae, 24'64/x long, and 5'28ti broad
on side view ; pointed at each end ; with straight axis ; with palmate teeth 8 * 8 /* long,
and with narrow alae 8 • 8 /x long.

Four specimens were obtained. The type specimen (XIX. 3) from 10 hole,
130 fins., is digitate, 7*5 cm. in length and 2 cm. in diameter. The spines are
•2- '5 cm. long, those at the lower end pointing obliquely upwards, but al>ove becoming _
vertical to the long axis. Another specimen forms a spheroidal knob encrusting the
stem of a zoophyte.

A third specimen is in the form of cylindrical fragments.

A fourth specimen, much macerated, is massive and bifurcated, and about
12*5 cm. long ; further there is a considerable amount of spongin present, sometimes
appearing as distinct fibres, where the spicules have become lost or dissolved away.
The spicules are in all respects identical with those of the foregoing forms.

This species bean much resemblance to Desmacidon setifer, Topsent, obtained
by the ' Belgica ' from the Antarctic. The isocheles are of much the same character,
but those of D. setifer are very much larger, viz., 75-100 M by 18-20M. Further,

TOU IT. 2 F

40 R. KIRKPATRICK.

the consistence of D. setifer is soft, the colour yellowish in spirit, and the surface
hispidation much finer.

The four specimens were dredged : (l) No. 48, from 10 hole, 130 fms. ; (2) No. 124,
near Winter Quarters, 20 fms. ; (3) No. 124b, off Coulman Island, 100 fms. ; (4) No. 126a,
from 12 hole, 25-30 fms.

DESMACIDON MAEANDRINA.

(Plate XIX., figs. 4, 4a, and Plate XXIIL, figs. 4a-b3.)
1907. Desmacidon maeandrina Kirkpatrick (lOa. p. 282).

Description. — The material consists of three sub-cylindrical fragments tapering at
the distal end.

The consistence is hard and dense. The colour in spirit is dirty brownish-gray.

The surface is fairly uniformly level, and presents flattened papillae or meandrine
ridges, slightly roughened at the top by projecting oxeas (best seen on side view with
a lens).

The dermal membrane roofs over the grooves and spaces between the papillae
and ridges. The pores are mostly circular and about 95 /u in diameter.

The small circular oscules, numerous and scattered, are about 1 mm. in diameter.
The skeleton is formed of a thick, main axis, consisting of rather loose, longitudinal
strands ; from this are given off at right angles cylindrical or lamellar bundles of loose
strands, which proceed to the surface and form the papillae and ridges.

SpiculeS. Megascleres. Oxeas, 579 x 39 M, curved (usually) or bent at the
centre, with sharp, pointed ends. Microscleres. Isancorae unguiferae, 26 M long, and
15 ' 8 n broad, shaft strongly curved and 3 • 52 M thick. With usually five teeth or claws,
about 5*28 n long, at each end, viz., a central, single, and two lateral bifurcated teeth.

The largest of the three pieces of this sponge is 6*2 cm. long and 16 mm. in
diameter. The fragments appear to be broken off from some branched specimen, and
I shall refer to them as branches. They are sub-cylindrical, being slightly compressed
in one plane. The chamber system is aphodal, the flagellated chambers (44 x 29 /*)
being pyriform. There is a considerable amount of variation in the teeth of the
isancorae, the number varying from three to six, the most usual number being five.

In some respects the new species resembles Desmacidon (?) ramosa (R. and D.)
(15. p. 107), obtained by the ' Challenger ' from the Cape of Good Hope and Marion
Island. In both species there is a central axis of longitudinal fibres, whence fibres
proceed to the surface at right angles, but here the resemblance ends. In the
' Challenger ' species the radiate bundles branch in a fan-like manner, finally forming
an almost uniform surface layer of vertical oxeas. The microscleres in D. (?) ramosa
are isochelae arcuatae. As in the case of D. (?) ramosa, it is doubtful whether the new
species, with its fiaspailia-like skeleton of axial and radial fibres, should be placed
under Desmacidon or under a new genus.

The specimens were dredged off Coulman Island, 100 fms.

TKTKAXONIDA. 41

JOYRUXIA BRLLL*

(Plate XVI., Figs. l-5a.)
l'.»07. Joynuia Ml, Kirkpatrick (lOa. p. 283).

>|,..nj.- .iit.-i. h. .1. ovoid, \\ii!i a thi'k tiiin iiii'l MttkMUIg a --It pulp. \Vn!i ihorl
conical oscular, and long trumpet-shaped poral papillae. Surface finely pilose. Colour
of surface yellow, of the rind whitish, and of the pith deep yellow.

Flagellated chamlwn 23 X 20 /i ; diplodal.

Skeleton. Cortical skeleton formed of layers of strongyles crossing each other
at right angle*. The walls of the oscular and pornl papillae supported liy n layer
of longitudinal strongyles. The surface of the sponge hirsute with a fine pile of
strongyles standing out at right angles or obliquely. Choanosomc without spiculcs.

SpiculeS. Slightly flex in ins smooth strongyles 850 ft long, lO/i in diameter at
the ends, and 13/i in diameter at the centre.

There is one adult specimen 5 cm. long, 3*5 cm. broad and 3 cm. thick, with
a deep groove on the under aspect, by which it was probably attached to a worm tuln;
or stem of a Hydroid. There is also a small conical specimen 6 mm. high, attached to
a piece of rock.

I was at first disposed to regard this remarkable species as a member of a new genus,
partly on account of its very thick rind, which is in places over a millimetre in thickness,
and partly because of the highly specialised poral papillae ; but apart from these characters,
the new form evidently shows the closest affinities to Joyeuxia. The three hitherto,
described species all have a rind enclosing a soft pulp, the latter being without or almost
without a skeleton ; then too the pulp is highly coloured. Joyeuxia tubulosa Topsent
and J. ascidioides (Fristedt) have fistulae, which, however, appear to be oscular. Two
of the species, J. viridis and J. tubtilosa have strongyles ; J. aacidwide* has tyles and
also cheles. Accordingly Topsent places the genus near Desmacidon.

The poral papillae attain a height of 1 to 1 '2 cm. ; they are expanded at the end,
the margin being sharp, usually a little jagged, and showing the ends of strongyles.
The mouth is closed by a sieve-like funnel-shaped membranous pore-area, which is
supported on its under surface by strands of tissue passing from the wall of the tube to
the poral membrane.

The tube passes through the thick cortex into the choanosome, where it expands
Krt'i.n- branching into four or five inhalant canals.

Between certain parts of the inner surface of the cortex and the choanosome is
what appears to be a space (see XVI. 2) ; but in other parts the choanosome abuts on
to the cortex ; probably these peripheral spaces do not result from contraction of the
tissues, but form part of the exhalant canal system.

• Named in honour of Emeritui ProfcMor F. J. Bell, of the Zoological Department of the Natural History

Hwm and editor of the " Report* on the Natural History Collection* " brought home by the • Dtaoovery '

from the Antarctic.

2 F 2

42 R. KIRKPATRICK.

The inconspicuous oscular papillae are only about 4 mm. high, and are tightly
contracted.

The tissue of the cortex is crowded with branched collencytes immersed in a
gelatinous matrix.

I am indebted to Professor Dendy for pointing out to me certain resemblances
between the poral and oscular papillae of Latrunculia and the present species ; in the
case of both genera, too, the tissues are rich in pigment. Possibly we may have here
a clue to the true position of Latrunculia, the discasters of which appear to be spined
oxeas, or styles, the spines of which have become verticillate.

The specimens were dredged near Winter Quarters, Flagon Point, 10-20 fms.

Joyeuxia viridis Topsent was obtained from the Azores, 454-845 metres ;
J. tubulosa Topsent, also from the Azores, from 200 metres ; and J. ascidioides
(Fristedt) from Baffin Bay, 169 fms. (7. p. 445).

CERCIDOCHELA Kirkpa trick.*

1907. Cercidochela Kirkpatrick (10a. p. 284).

Mycalinae with peculiar shuttle-shaped chelae or canonochelae,f with the single
central teeth from each end of the shaft joined together, and with a semi-circular
vertical lamella extending inwards from the shaft and from the dental bridge, ^so as
nearly to meet.

CERCIDOCHELA LANKESTE^II.J

(Plate XIX., figs. 5, 5 A, and Plate XXIII. , fig. 5a-L)
1907. Cercidochela lankesteri Kirkpatrick (10a. p. 284).

Sponge elongated, slender, fusiform. Colour white ; consistence soft. Surface
smooth to the naked eye, but finely hispid under a lens. With several small scattered
oscules about 1 mm. in diameter, level with the surface. Flagellated chambers
aphodal, oval, 31 x 21 M-

Skeleton formed of long longitudinal lines of spicule fibres about 100/u thick, not
forming a definite central axis, radiating out in plumose manner to the surface ; with
a few isolated spicules arranged in a scalariform manner at right angles to main fibres.
Spongin not perceptible.

Spicules. Megascleres. Oxeas 452 X 19 '5 /u> curved at centre, attenuating
gradually to sharp points slightly planed away on inner aspect. Microscleres.
Canonochelae, somewhat shuttle-shaped,§ 45'5/u long by 22'75/u broad, with the two
teeth fused to form a bridge, and with a semi- circular lamella passing upwards from

* ntpKit, ifiof, a shuttle,
f Kavuvt ovos. & shuttle.

t Named in honour of Professor Sir E. Ray Lankester, K.C.B., F.R.S.

§ The canonochelae are, in this species, shaped like an oval basin with a truncated bottom, but they look
shuttle-like as commonly seen in balsam slides, viz., lying on one side with the lamellae uppermost (see XIX. Of).

•n.ii:\\"\ii>\ 43

the shaft and downwards from the dental bridge, both lamellae being nearly on the
same plane and nearly meeting, the lamellae sometimes with basal tubercles.
Devel"|'iii--ui.il form- in shape of thin oval linear bodies, the oval at first not being
complete.

Tin miii|in- >|M-< inion representing the new genus and species is 12*5 cm. long and
1 cm. in lirr.t.lth at the centre. The body attenuates to fine ends, and apparently
has not been attached to anything.

The remarkable canonochelae recall to mind the spherancorae of Melominchora,
but the latter spicules have three pairs of fused teeth.

The canonochelae are scattered about in the choanosome in considerable nninlxTs.
The shape may IMS compared with an oval basin with the bottom cut out, and with two
semi-drcular lids or lamellae passing horizontally from the upper edge of the basin,
so as to nearly meet ; further it is necessary to imagine such a basin turned up on
its side.

The earliest developmental forms have an elongated Q shape ; then the open Q
becomes a closed oval ; by this time the falx at each end is perceptible, and the
beginnings of the lamellae appear. A further change leads to a marked asymmetry,
the thin oval ring Incoming a broad band by widening in a direction away from the
edges whence the lamellae arise. If the widening of the band were to continue the
growing edges would meet and a sort of lateral dome would result, but growth does
not go so far as this.

It is often difficult to make out the real form of a transparent body, with its lines,
curves, lamellae, Ac., crossing each other in various directions and altering their
appearance with the slightest shifting of position. In the present instance the figures
will show the remarkable variations in form corresponding to change in position,
and it was only after a very prolonged examination that the real shape l»ecamc
apparent, a result finally due to Mr. Highley's ingenuity in devising a medium of
suitable consistency in which the spicules could ta made to slowly rotate on their
long and short axes. Fig. 5 f is the position in which, in spiculc preparations, the
spicules commonly lie, viz., on the detruncated bottom of the t>asin and with the
lamellar edges uppermost. Fig. 5 h, which resembles the Diatom Amphora, shows
the aspect when the shaft portion or the dental portion of the I mini is uppermost and
vi. wed in ite breadth ; the axial <»nal which traverses the shaft and part of the dental
lir'nlge is best seen in this aspect When a spicule in this position continues to rotate
a little more we see either the lamellae along one side or the free growing edges of the
on the other. Fig. 5g shows the growing edge of the dental bridge nearly
the edge of the shaft. Fig. 5k shows half of a spicule broken across, with
the axial canal (which forms nearly a complete circuit) exposed at each lower corner of
the figure. Lastly an end view is seen in fig. 5 L

The specimen was dredged near Winter Quarters at No. 10 hole, 130 fms.

44 R. KIRKPATRICK.

HOPLAKITHARA Kirkpatrick.*
1907. HoplaWham Kirkpatrick (lOa, p. 285).

Mycalinae possessing exotyles with large spherical spined heads, and with fim-
briated placocheles.

HOPLAKITHARA DENDYI.

(Plate XIX., figs. 6, Ga, Gb, and Plate XXIII., fig. 6a-c2.)
1907. Hoplakithara dendyi Kirkpatrick (10JU p. 286).

Sponge in form of a small cushion, attached by a narrow base. Surface smooth
to the naked eye. Colour, pale brown in spirit. Consistence, hard externally, soft
within. Flagellated chambers 32 '5ju in diameter, spheroidal, eurypylous.

Skeleton with protective armour formed by gigantic spheroidal heads of exotyles,
the exotyles being arranged as radiating bundles in form of inverted cones, with the
apices a little below the cortex ; with scattered strongyles.

Spicules. Megascleres. Exotyles f with the heads a little inclined to the
long axis of the spicule, the proximal end (in the interior of the sponge) rounded, the
distal end swollen into large spherical heads, with short cylindrical spines covering the
distal three-fourths of the head. Total length 358 M, the shaft enlarging in diameter
from 6'5M at the proximal end to 16 M just below the head. Head, 55 M in diameter;
cylindrical denticles, 1 '76 to 3 '52 /t* in height, with finely denticulate edge, and with
cup-like depression at the summit.

Strongyles straight, fusiform, smooth, 467 ' 5 M long, 9 • 75 M in diameter at centre,
6 • 5 n in diameter at ends.

Microscleres. Placocheles, fimbriated, 84' 5 M long, 29'25/w broad; length of
tooth, 37-75^.

Sigmata very small, slender, (J-shaped, 8 • 8 /u long, 5 ' 28 M broad, • 9 M thick.

The minute spheroidal or cushion-shaped specimen was 2 ' 2 mm. in height and
3 mm. in horizontal diameter ; it was growing on the side of an Alcyonarian, creeping
over a branched Cellepora. No pores or oscules were discernible. The under surface,
which was narrowed to the point of attachment, was paler in colour than the upper.

The new genus is closely related to the Mycaline genera Rhaphidotheca and
Guitarra, to the former by its exotyles, and to the latter by its fimbriated placocheles.

The distal knobs of the exotyles of R. marshall-hallii Kent, 49 M in diameter, are
smooth and spherical, and those of R. rhopalophora Schmidt (R. affinis Carter, see Thiele
22. p. 383) are 104 M long and 30 M broad and club-shaped. Lundbeck (13. p. 32)
regards these two species as probably identical, and certainly the differences are slight.

In the centre of the heads of the exotyles of //. dendyi is an oval granular zone,
which at first suggests a cavity filled with protoplasm, but that there is no cavity is

* <nrX<i (pi. of oTrXoi/), annour ; uMpu, a. guitar.

f Professor Dcndy, to whom I showed these remarkable exotyles with their large spined heads, regarded them
as possible examples of spicules which might form a surface layer of pseudastcrs by reduction of the shafts.

TETRAXONIDA. 45

evident from the fact that iln- axial i-anal can l>o seen traversing the granular zone.
Some <>f the exotyles have one or more swellings in the length of the slmt't .

The single example was dredged near Winter Quarters, at No. 6 hole, in 130 fins.

FAMILY HAPLOSCLERIDAE TOPSKNT.
SUB-FAMILY GELLIINAE RIDLEY AND DBNDY.

GELLIUS RODIS.

(Plate XVII., figs. 1, IA, and Plate XXIV., fig. 1 a-b.)
1902. 0W/IM rufu, Topscnt (29. p. 14, PI. I., flg. 9, and PI. III., fig. 4).

The type of this species, dcscril>ed by Topsent, was n small globular sj>e<-imen
27 mm. in diameter. The present collection contains six pieces of this sponge, some of
them of large size ; accordingly 1 am enabled to slightly supplement the original
description. The largest fragment, 18 x 5 x 2*5 cm. in dimensions, is masnively
lobate and closely resembles Petrosia similis var. massa R. and D. (15. plate III., fig. <>).*
The other specimens are in the form of thick sub-cylindrical unbranched or branched
fragments, the largest (PI. XVII., fig. 1) being 14 cm. in length and 2*5 cm. in
diameter. Several of the pieces have oncules about 7 mm. in diameter with circular
raised rims.

The surface, where the dermal membrane is intact, is to the naked eye quite
smooth ; though, where the dermis is rubbed off, the surface appears very finely hispid
as Topsent describes, and feels rough to the touch.

A vertical section of a branch in the plane of the axis shows main longitudinal
lines of fibres curving outwards from the axis to the surface as they pass upwards, and
dividing in a more or less paniculate fashion near the surface ; the secondary fibres, one
or two spicules thick in the centre of the specimen, become much thicker a little below
the dermal membrane.

The oxeas (XXIV. la), 403 x 20 /«, are sub-tornote. The C-shaped sigmata have
a uniform curve almost in one plane : they are 37 M long, 19*4 M broad, and *8 M thick,

• In this variety the skeletal arrangement it very similar to that found in Grlliui rutlii ; the terminal
branchlets of the main fibre* are spread out beneath the dermal layer, and anastomose with branchleU from
neighbouring main fibres ; this sub-dermal reticuluro, then, is due solely to the spreading out of the periphery of
the paniculate ends of the main fibre* (as in O. rtniit), and is not as one might at first suppose a special dermal
skeleton ; the central tuft* of the main fibre* proceed vertically up to the surface ; also, not a few scattered oxea*
pas* obliquely or vertically upwards from the subdermal network of spicule fibre* ; accordingly the dermal
membrane is to a great extent supported on the tips of vertical oxeas. Though Ridley and Dendy clearly saw that
the dermal reticulation was only the uppermost layer of the main skeleton (16. p. 11), it was not quite correct to
state that the dermal membrane was not supported on tuft* of spicules ; a vertical section shows that it is sup-
ported in this manner, viz., by the central tuft* of all the main fibres, as well a* by oxea* from the subdermal
reticulnm. The gpicule* of P. timilU var. manu resemble those of O. rudit in shape, but are smaller (vis. 289 x
9'76 p). Accordingly P. rimilit var. matta is closely similar to O. rudit in form, in skeletal arrangement, and
in spknlation ; in fact it very probably represents an example of a Qrlliiu which ha* lost it* sigmata. I have
gone into this matter in some detail because it is always interesting to trace the probable line of descent of •
Benierine sponge from some form with microsclere*. Further, an instance of this kind furnishes MM support to
part of Dendy'* recent scheme of classification, vie., that of making the Gelliinae the starting-point of the

46 R. KIRKPATRICK.

and frequently have a small central knob. (Figures are given for convenient comparison
with the nearly related new species, G. fimbriatus, described below.)

The pores, abundant over the areas roofing over the spaces between the main
fibres, are circular or oval and vary from 30-70 n in diameter ; the subdermal spaces
are about a millimetre in depth ; there are also smaller and shallower subdermal spaces
between the more central terminal branches of the main fibres. The flagellated
chambers, which are unusually small, viz., about 23 /u in diameter, are spheroidal and
eurypylous ; as seen in section they line labyrinthine folds and spaces.

The cellules spheral euses attain the large size of 35 n in diameter, the spherules
also being of considerable size, viz., 8 /*.

Sections of the branches show nests containing 30-50 embryos, forming conspicuous
red patches in the sections in balsam ; each embryo is about 235 M in diameter, and of
yellowish colour, though red in masses. Some contain rhaphide-like oxeas. On the
surface of the figured specimen are several little hemispherical pits from 2-5 mm. in
diameter ; these are probably the exposed surfaces of the embryo-containing cavities,
and not due to parasites.

The specimens were dredged in McMurdo Bay, in depths up to 20 fms. The
'Belgica' obtained a specimen from Lat. 70° 23', Long. 82° 47' W., from 500 metres.

GELLIUS FIMBRIATUS.

(Plate XVII., figs. 2, 2a, and Plate XXIV., fig. 2a-b.)
1907. Gelliusfmibriatus Kirkpatrick (10a. p. 286).

Sponge in form of a thick triangular cake, or conico-cylindrical. Texture soft,
easily broken. Colour in spirit, pale buff.

Surface level, or almost imperceptibly hispid ; showing through the dermal
membrane a somewhat areolated pattern, each areola being formed by the end of a
main fibre giving off fimbriated twigs which roof over the subdermal spaces between
the main fibres. A few circular oscules about 5 mm. in diameter and with slightly
raised rims occur.

Skeleton formed of longitudinal lines of flat, loose, band-like main fibres, with
an irregular and obscure reticulation of single spicules between. The main fibres
spread out in a paniculate manner a little below the dermal membrane.

Spicules. Megascleres. Oxeas, 537 x 16'25ju., slightly bent or curved at
centre, attenuating gradually to sharp points. *

Sigmata varying in size, the largest being 40 p. long, 1 7 • G p, broad, and 1 '76 p.
thick, with one or more angular bends in the curve, usually one end of the shaft with
an angular bend, the other end curved.

The specimen selected as the type of this species has a flattened angular pad-like
body, 10 cm. long. 4'5 cm. broad, and 2' 5 cm. thick; it apparently lay free on the
bottom.

n. n: \\oxiDA. 47

Tin- u|>|HT surface alone shows the arculatcd appearance beluw the dermal mem-
brane, tin- un<ler surface being nearly opaque.

A second specimen is conico-cylindrical in shape, 11 cm. long and 3 cm. in
diainet* T. the lower end being broken off. This specimen also apparently lay on the
IM.MOMI :il..ni: |'.>rt of its length, as the pore surface extends all round only at the
upprr en.l of the specimen.

\ thinl >]MM iiin-n from Coulman Island is fan-shaped, 7 cm. high, 3'5 cm. thick,
7 cm. l'r-M-1 ai tln> iip|>cr edge, and 3 cm. broad at the base, where it is attached to two
-in.ill -tours. Tli. -in laee is quite worn away, and the body of the sponge full of debris.

An "an-ola" with it* central node (the end of the main fibre) and lateral
branchlet* occupies on an average an area of 6 x 4 mm.

Tlti- ends of the main fibres are arranged in linear scries.

Tin- dermal membrane is separated al>out 3 mm. from the floors of the large
sub-dermal spaces, and the band-like supporting pillars are al>out 2 mm. broad. A
vertical section gives the appearance of a miniature "hall of a thousand columns."
The dermal membrane on the under surface of the sponge contains scattered ozcas
arranged tangentially.

The pores vary a good deal in size and shape, 1>eing oval or round, and 45-120 p
in diameter. The flagellated chambers are spheroidal, eurypylous, and about 25 p in
diameter. The triangular cushion-like shape of the type recalls to mind Gelliu*
flagellifer (R. and P.), but there are no flagellate sigmata in the new species.
G. rudis (Topsent) has a much firmer and denser structure ; the oxeas are shorter,
thicker, and with tornote ends, and the sigmata are more slender and with uniform curve.

Dredged near Winter Quarters, off Hut Point in 12-20 fms. ; also off Coulman
Island in 100 fms.

GBLLIUS PILOSUS.

(Plate XVII., figs. 3, 3a, and Plate XXIV., fig. 3a-c.)
1907. QMutpilonu Kirkpatrick (10m. p. 287).

Sponge in form of an erect flattened triangular or elongate lamella divided or
digitate at the upper edge.

Consistence soft, fleshy, compressible.

Colour in spirit, dirty white or very pale yellow.

Surface finely conulose and pilose owing to the projection of the ends of the main
skeleton fibres about 1 mm., the conules being about '6 mm. apart from each other.

A few oecules about 1 mm. in diameter, on a level with the surface.

Skeleton formed of slender main axial fibres on an average about 2-5 spiculea
thick, curving out to the surface where they form the pile, and of secondary fibres,
usually one, but sometiui.-- tw.. or three spicules thick, at right angles to the main
ones, with which they form oblong scalariforni nv -In •-. Spongin well developed at the
nodes of the network.

rou rr. 2 O

48 R. KIRKPATRICK.

Spicules. Megascleres. Oxeas, 537 X 22 '75 /A, sharp pointed, sub-tornote, a
few being distinctly tornote.

Microscleres. Sigmata, very abundant both, in choanosome and ectosome, C-shaped,
39 x 16 ' 25 /A in length and breadth, and 1 • 5 p thick.

There are two specimens. One of them is in the form of a triangular lamella
dividing into two sub-terete branches, the total length being 10 '5 cm., the breadth
3 ' 5 cm., and the thickness 1 cm. The second specimen is 1 1 cm. long and 3 cm. broad,
with little more than a notch at the upper edge, indicating a division into branches.
The- fleshy matter is nearly all macerated out, leaving a flexible skeletal framework.
In the first specimen the surface shows the little conules penetrated by spikes 2-3
spicules thick, with squarish concave depressions about '5 mm. between any four
conules. The dermal pores are imperceptible. There is a thin collenchymatous
ectosome about • 1 5 mm. thick, excavated by shallow sub-dermal cavities ; the
eurypylous flagellated chambers are ovoid, 26 x 23 p.. The new species bears some
resemblance to Gellius flagellifer (E. and D.), but differs from it in the absence of
the peculiar flagellate sigmata. Further, G. flagellifer has an even surface, and
a dermal skeleton network of spiculo-fibre ; but in the new species the secondary
fibres, usually not more than one spicule thick, are often not present at or just
below the surface.

Dredged near Winter Quarters, off Hut Point, 25-30 fms. ; and at No. 12 hole,
25-30 fms.

GELLIUS CUCURBITIFORMIS.

(Plate XVII., figs. 5, 5a-c, Plate XXIV., figs. 5a-b.)
1907. Gellim cucurbitiformis Kirkpatrick (lOa. p. 288).

Sponge small, free, bulbous with fistular prolongations. Surface smooth, showing
under a lens a fine white reticulum. Colour in spirit, pale brown. Consistence,
rather soft.

Skeleton. With a distinct dermal layer of irregularly arranged tangential oxeas.
Choanosomal skeleton a reticulum (with square or triangular meshes) of spiculo-fibre,
the strands 2-3 spicules thick, with a faint indication of main fibres radiating to
the surface.

Spicules. Megascleres. Oxeas, 342 x 9'75/u., slightly curved, sub-tornote.

Microscleres. Sigmata varying in size, the smallest being about 20 p long,
C-shaped and with uniform curve, and the largest 39 p, long, 19'8/x. broad and
1 • 2 p. thick.

There are two small specimens, both of which were found in a tangled mass of
debris surrounding a worm-tube. The larger, the type specimen, consists of a basal
bulbous portion, 13 mm. long, 7 mm. broad, and 8 mm. high, from one side of which
arises a rather thick-walled fistula 13 mm. high and 5*5 mm. in diameter; at the
opposite side is a broken circular area, from which, in all probability, a second fistula

TKTRAXONIDA. I ''

arose ; lastly, between these two, in a small raised knob with a rounded orifice on one
side of it. The narrow oscular canal is not central, but along one side of the thick
walled eomplet-- ti-tula.

The second specimen is tubular, with a slightly enlarged solid base, when. •• arises
a ii-tiilu ; the total length is 2*2 cm., and diameter *6 cm. No pores are discernible ;
the subdermal spaces are about • 2 mm. in depth.

The eurypylous flagellated chambers are 23 /i in diameter. Cellules sphtSruleuses,
8-9 fi in diameter, are common.

Tin-re is no bast-like suMermal layer as in Oceanapia inollis Deiidy (4. p. 248),
a n< I the spicules of the latter are smaller, the oxeas being 200 x 8 /i, and the sigmata
only 16 p. Lundbeck (12. pp. 64, 66) describes two species of Gfllim with h'stulae,
and with a well -developed dermal l«rk, viz. : G. lurulus and G. microtoxa, but l>oth
these species have toxa in addition to sigmata.

The two specimens were dredged off Hut Point (Winter Quarters) in 25-30 fms.

GELLICS GLACIALIS, var. nivea.

(Plate XVII., fig. 4.)
1887. Oflliiu ffladalu vw. nivea Ridley and Dendy (15. p. 42, PI. VIII., fig. 8, etc.).

The one example is in the form of a small spherical knob attached to a branched
Polyzoon, resembling in these respects the ' Challenger ' specimen from Prince Edward
Island. The sponge, which is 17 mm. in diameter, is brittle, and has a well-marked
ectosomal skeleton formed of tangential oxeas. The megascleres and sigmata are
smaller than those of the ' Challenger ' specimen, but are in other respects of
the same character ; accordingly I have not regarded the Antarctic specimen as a
new variety.

The larger sigmata are often united in bundles of sigmadragmata.

The following table gives the dimensions in M of the spicules of G. glacialis, and
of the ' Challenger ' and Antarctic specimens of var. nivea.

G. gUcialU.

var. nivea
•Challenger.'

rar. nivea

• DUcovery.1

Ox«»

670 X 40

704 X 29-25

&22 X 19

SigmaU .

75 X 4ft X 8-5

209 X 74-5 X 5'7

119 X «7 X 6-6

Dredged off Coulman Island, 100 fms.

The 'Challenger* obtained G. glacialis from Agulhas Bank, 150 fms., and var.
niwa from Prince Edward Island, 140 fms.

2 o 2

50 R. KIRKPATRICK.

OCEANAPIA TANTULA.*

(Plate XVIII., figs. 5, 5a, 5b, and Plate XXIV., fig. Sa-e1.)
1907.- Ocmnapia tantula Kirkpatrick (10a. p. 289).

Description. — The sponge consists of five small fragments of tubes, the longest of
which is 8 mm. in length, by 4 mm. in diameter ; three of the pieces are hollow, thin-
walled and tubular ; the other two are solid. One of the solid pieces seems to belong
to the top of a fistula.

The colour is transparent white.

Skeleton. The dermal layer is composed of a chitinous-looking membrane with
strongyles lying tangentially, usually in one layer and densely packed, but sometimes
more or less scattered.

The white strands of the loose subdermal reticulum are visible through the surface.
They are longitudinal, and only anastomose occasionally. The strands are less than
• 1 mm. in diameter. They vary in composition ; in some parts being composed of
strongyles smaller than those of the dermal layer, in other parts of smooth tricho-
dragmata, or again of strongyles, amphityles and trichodragmata. The pale transparent
choanosomal tissues are crowded with small spined rhaphides.

SpiCUles. Megascleres. Strongyles, 437 x 19ft, slightly fusiform, curved once
or sometimes twice. Occasionally one end is pointed, the spicule becoming
a style.

Amphityles, 395 x 7 '25 /A, slightly fusiform, heads 13//, long, 9 '75 mm.
broad.

Microscleres. Long, smooth raphides, separate or in bundles forming part of the
subdermal reticulum, 650 x 2 • 5 /A.

Short scattered spined raphides, usually stylote, 162 /u, long, and about 2 • 5 /x broad.

H. V. Wilson describes (34. p. 128) a species of Oceanapia, viz. 0. bacilli/era
with strongylea, but it has the usual sigmata.

Oceanapia (Phloeodictyori) singaporensis (Carter) has strongyles in the dermal
layer, but oxeas as well as strongyles in the skeleton fibres, and there are no micro-
scleres.

The species of the Gelliine genus Ehaphisia have oxeas, trichodragmata, and, in
one species, toxa ; but there are no fistulae, and there is no subdermal reticulum of
spicular fibres.

It is regrettable that there is so small an amount of material on which to base a
new species, but the marked characters of that which is available seem to render
such a proceeding justifiable.

Dredged near Winter Quarters, from No. 10 hole, 130 fms.

* Tantului, ever so little.

TBTRAXONIDA. 51

SUB-FAMILY RENIERINAE, RIDLEY AND DKNDY.

PYLODERUA.
(paw) Ridley and Dendy (15. p. 6).

Renierinae with a paivhment-likc, easily-separated, dermal mcinltniiic in which
are situate*! closely-packed tangential oxeas, and with distinct round or oval pore areas.

I'M "I-KRMA LATRUNCULJOIDBB.
1887. HMrkontlrM hitrunruliouit*, Ridley and Dendy (15. p. 6, PI. I., li-. :.. A.M.

There in one Haltcllate specimen attached to a small stone. The height is (> cm.,
the width at the upper rim 5*5 cm., and the thickness 1 '8 cm. The osculet*, generally
contracted into little white conules, are scattered about among the pore areas, and not
•!.•'!•• i "ii t :.<• •!•_'• .1 ,n tli-- • i h.ili.'irj- 1- exunplec, Each oKoIar opening lead
into a smooth funnel-like cavity with a sharp-edged circular sphincter-like opening at
the boa.

The strongly marked characters of this peculiar species seem to me to necessitate
its removal from I/alichotulria and its inclusion under a new genus. In the ' Challenger '
Report Ridley and Deudy state that they were at first doulitful whether the arrange-
ment of the pores in definite areas would prove to be a character of generic importance,
and finally decided that this feature was only one of adaptation. Prof. Dendy has
since told me that he now thinks that this species should be placed in a distinct genus,
and in this opinion I concur. There are no microscleres to help in tracing its affinities
and the body skeleton is Reuierinc in character ; accordingly it is for the present placed
among the Renierinae.

Dredged off Coulman Island in 100 fms. The ' Challenger ' obtained specimens
from a depth of GOO fms. off the mouth of the Rio dc la Plata.

PETROSIA FISTULATA.

(Plate XVIII., figs. 4, 4a-b,and Plate XXIV., fig. 7.)
1907. Petrotia fitlulaia Kirkpatrick (lOa. p. 290).

Sponge tubular. Surface smooth, showing the round openings of the inhalant
canals about • 4 mm. in diameter and close together.

Inner surface of the tube of the sponge finely or rarely coarsely pilose, and
showing the round openings of the exhalant canals about 1 mm. in diameter. Colour
in spirit, pale yellow. Texture firm, but slightly compressible. Eurypylous
flagellated chaml>er8 spheroidal, 24 • 5 n in diameter.

Skeleton formed of main fibres proceeding from the inner to the outer surface,
joined by secondary fibres one spicule thick, so as to form obscurely quadrangular or
hexagonal tubes about • 5 mm. in diameter ; ends of spicules cemented with spongin.

Spicules. Oxeas, 492 x 24*4 /*, bent usually, or curved at centre, sub-tornote.

52 R. KIRKPATRICK.

There are four specimens, the two larger being uniformly cylindrical and the
smaller ventricose. The largest is 6 cm. long, the diameter being 2 • 1 cm. and the
thickness of the wall 5 mm.

The dermal membrane roofing over the inhalant orifices is usually supported there
by two or three single spicules radiating to the centre. The pores are '95 M in
diameter.

Small embryos about '76 mm. in diameter occur. The new species comes nearest
to the species from Kerguelen, which Carter (3. p. 287) identified as Thalysias
subtriangularis Duch. and Mich., but which Ridley and Dendy (15. p. 9) regarded
as synonymous with Petrosia similis * (Ridley and Dendy).

The spicules of the Antarctic species are very much larger than those of Carter's,
and partly in consequence of this the skeletal network of the latter is much denser
from a closer approximation of the fibres.

The dimensions of the oxeas of four nearly related species of Petrosia are
as follows : —

Petrosia similis (R. and D.), 225 X 16 /A.
Thalysias subtriangularis (D. and M.), Carter, 190 X 12/x.
Petrosia (Schmidtia) aulopora, 0. Sch., 175 x 7/i.
Petrosia Jistulata, 493 X 24'4/i.

Dredged near Winter Quarters, No. 12 hole, 25-30 fms. ; McMurdo Bay,
96-120 fms.

RENIERA scoTTi.f

(Plate XVIII., figs. 1-2, and Plate XXIV., fig. G.)
1907. Reniera scotti Kirkpatrick (lOa. p. 291).

Sponge consisting of one or more fistulae. Texture very soft and easily lacerated.
Colour in spirit, varying from yellow to pale reddish. Outer surface varying from
being finely hispid to having large conules and meandrine ridges. Inner surface of
fistulae very finely hispid in the spaces between the numerous orifices of exhalant
canals. Flagellated chambers large hemispherical, 60 X 40 p..

Skeleton formed of parallel longitudinal lines of main fibres, about 2-6 spicules
thick, curving outwards from the inner to the outer surface, where they pass into the
conules and ridges ; secondary fibres at right angles to the main ones, one or two
spicules thick. The spicules are not closely united, and spongin is only present in very
small amounts.

Spicules. Oxeas 343 x 14 '6 p, curved or bent at centre, sub-tornote.

* The specimen from Kerguelen Island which Carter identifies as Thalysias subtriangularis Duch. & Mich,
is, I believe, specifically distinct from Petrosia similis R. & D. The fine spicular network of the former, with its
slender main fibres, and still more slender unispicular secondary fibres, contrasts strongly with the thick cable-
like longitudinal and transverse strands of the latter. Carter's specimen appears to me to belong to a new
species.

t This fine species is named in honour of Captain R. F. Scott, R.N., C.V.O., the leader of the Expedition.

TKTKAXONIDA. 53

There are six specimens and fragment*. Tlie outward appearance varies greatly
according to age and size. In one small specimen the surface is finely hispid, in larger

eonuloee, an«l in very large ones conulated and with high meaudrine ridges. The
jKM-imen N". I I - i I'l. \\lll . ii-j. 1 ) ll in iv BoRH "I i \\ i.L- thick \\.-ill-.i till*',
1-J cm. hiizli and G cm. in diameter, and with walls 1 '5 cm. thick, but attenuating towards
the rim <>f tin- tuln?. This specimen is incomplete IH?!OW. The orifice is circular, and
\\iihin the rim is a (liaphntgni contracted to a white line.

The sui: overed with large conulcs and meandrinc ridges rising to a height

of nearlv I cm.

Tin* dermal membrane, in the spaces between the conules and ridges, shows as a
line lace-like reticulum, with circular pores 133 M in diameter, and Itencath it the
i>i Slices (1-1 '5 mm. in diameter) of the inhalant canals arc visible. The cxhalant
orifices on the inner wall of the tube are much larger than the inhalant ; they vary
from 1 to 6 or 7 mm., their edges are smooth and rounded. In the second largest
specimen, these orifices are arranged in longitudinal rows, and are oval with the long
diameter vertical. This arrangement results from the way in which the contraction
of the tubular sponge body is restricted by the main longitudinal skeletal fibres
running beneath the inner surface.

In a third specimen (No. 132), the outer surface of the wall is almost smooth to
the naked eye, though under a lens it is seen to be finely hispid, with the top of oxeas
projecting from minute conules.

In another specimen the Ixasal part of the sponge divides into two tubes, one
of which is nearly smooth, and the other with ridged surface.

The species closely resembles R. spinosella, Thiele (23. p. 459), from Punta
Arenas. In Thiele's species the body is tubular, with conulated surface, and the texture
is rery soft ; but the skeletal framework is irregular, and the oxeas, though similar in
form, are much shorter, smaller and more slender, l>eing only 150-170 /* long and
7-8 /i thick.

Another species showing certain resemblances to R. scot 1 1, viz. R. aquaeductu*
Schmidt, var. infundibular if (R. and D.), has a unispicular skeletal network.

Specimens were dredged near Winter Quarters, in No. 12 hole, 25-30 fms. ; S.E.
of Cape Armitage, 100 fms. ; off Flagon Point, 5-25 fms. ; off E. end of Ice Barrier,
100 fms., mud and stones.

REXIERA DANCOI.
(Plate XVIIL, fig. 3.)
1901. Rmitra dancoi, Topwnt (29. p. 12, PI. II., fig. 1, and PI. III., fig. 8.)

There are two specimens. One is of a compressed digitate form, 10 cm. long,
2 ' 5 cm. wide, and 1 cm. thick. About the middle of one side is a projecting shoulder
with an oscule. Near, but not quite at the summit, on the opposite edge, is
a second oscule. The specimen is contracted at the centre; the alternate oscules

54 R. KIRKPATRICK.

and contracted waist give the impression of a budding of the upper half of the specimen
from the lower.

The colour is pale brown in spirit, and is due apparently to the abundance of
masses of cellules spheruleuses ; the specimens described by Topsent are grayish or
whitish.

The ' Discovery ' specimens have, in parts, a hispid rather than a conulose surface,
the spicules standing above the level dermal membrane like sticks out of water ; but
in other places the surface is finely conulose. The oxeas are 642 x 18 n, those of the
'Belgica' being 630 x 18-20/ot. By deep staining of a portion of macerated skeletal
network a small amount of spongin becomes discernible at the nodes. The flagellated
chambers, 55 x 40ft, though rather large, are smaller than those of Topseut's specimens,
in which they are of unusual size, viz. 70 x 50 p. The second specimen is merely a
shapeless fragment.

Dredged near Winter Quarters, off Hut Point, 25 fms. ; No. 12 hole, 25-30 fms.
The ' Belgica ' Expedition obtained several small specimens frortl 450 metres, in
Lat. 71° 19' S., Long. 87° 37' W.

TKTKAMiNlUA 55

MEMOIRS REFERRED TO.

1. CARTER, H. J. — Descrijitions and tigun« of Deep-sea Sponges . . . from the Atlantic Ocean, dredged

. . . H.M.S. Porcupine. Ann. Mag. Nat. Hist. (4) xiv., 1874.

2. CARTER, H. J.— Some Sponges from the West Indict and Acapulco in the Liverpool Free Mnxeum.

Ann. Hag. Nat. Hist. (.'•) ix., 18M2.

3. CARTER, H. J.— Spongiidae from Kerguelen Inland. Transit of Venus Exp" Phil. Trail*, vol. 138

(extra). London, 1879.

4. DEKDT, A.— Catalogue of the Non-Calcareous Sponges . . . Port Phillip Heads. Pan i. Proc.

Roj. Soc. Victoria, vol. vii., n j., 1895.

5. DSHDT, A.— Catalogue of Non-Calcareous Sponges . . . Port Phillip Head*. Part ii. Proc. Roy.

Soc. Victoria, vol. viii., DJU, 18%.

6. DEXDY, A. — In HERDXAH, Ceylon Pearl Fisheries. Supplementary Report, xviii. London, 1905.

7. FRISTEDT, K. — Sponges from the Atlantic and Arctic Ocean* and IMiring Sea. Vega-F'xpeditioneus

retcnskapliga jakttagelser Arbeten. Bd. iv., 1887.

8. H AX ITHCH, R.— Revision of the Generic Nomenclature and Classification in Bowerbank'g "British

Spongiadae." Trans. Liverpool Biol. Soc., vol. viii., 1894.

9. KIRKPATRICK, R. On the Sponges of Christmas Island. Proc. Zool. Soc. London, 19"0.

10. KIKKPATRICK, R.— On the Oscules of Cinachyra. Annals and Mag. Nat. Hist., 1905 (7). vol. xvi.
10a. KIRKPATRICK, R. — Preliminary Report on the Monaxonellida of the National Antarctic Exhibition.

Ann. and Mag. Nat. Hist. (7), vol. xx., September 1907.

11. I.KM'KXFK.i.i'. R. vow.— Spongien von Sansihar. Abhand. Scnckcnbcrg. Gescllsch., 1899, Bd. xxi.
lla. LESDEKFEIJ), R. VON. Deutsche Sudpolar-Expedition, 1901-1908. Bd. ix. Zoologie I. Tutraxonia.

Berlin, 1907.

12. Lt'KDRECK, W. — Porifera (Part i.). Homorrhaphidac and Heterorrhaphidac. The Danish Ingolf-

Expedition, vol. vi., Copenhagen, 1902.

13. LtrxDBECK, W.— Porifera (Part ii.). Desniacidonidac. The Danish I ngolf- Expedition, vol. vi.,

Copenhagen, 1902.

14. MKRKJHKOWHKY, C. — Ktudes sur Ics KpongcH de la mer Blanche. Mom. Acad. Imp. Sci. (7), xxvi.,

N* 7, St. Petersbourg, 1878.

15. Run.KV. S. O., and DKNDT, A. -Report on the Monaxonida collected by H.M.S. 'Challenger.'

Ix>ndon, 1887.

15a. RIDLEY, 8. 0.- Account of the Zoological Collections made during the survey of H.M.S. 'Alert'
in the Straits of Magellan. Proc. Zool. Soc. London. 1881.

16. SCHMIDT, O. — Die zweite dentache Nordpolarfahrt, 1869-70. Zweiter Band, Zoologic, KieseUpongien.

I^eipzig, 1874.

17. SCHMIDT, 0. — Die Spongien dcs Meerbnsen von Mexico. Jena, 1879.

18. N nn./K. F. E.— Cber die Ableitung der Hexactinelliden-Nadeln vom regnlaren Hexactinc. Sitc-

nngb. Akad. Witts., Berlin, xlvi., 1893.

19. Sou. AS. I. H. On the Sponges collected during the Skcat Expedition to the Malay Penituuln,

1897-1900. Proc. Zool. Soc. London, vol. ii., 1902.

20. SOLUAS, W. J.— The Sponge-fauna of Norway. Annals and Mag. Nat. Hist., 1882, (5), vol. ix.

21. SOLLAS, V.'.J.— Report on the Tetractinellida collected by H.M.S. 'Challenger.' Zoology, vol. 2.r>.

London, 1888.

22. THIELR, J. — Beschreibnng einigi-r nnzurcichcnd bvkunnten Monaxoncn Spongicn. Archiv. Natnrg-,

1903. Band i. Berlin.

23. THIELE, J. — Die Kieeel- nnd Hornschwamnic der Sammlnng Plate-. Zool. Jalirb. Suppl. vi., Dr. L.

Plate, Fauna Chilensis : Bd. ill., Heft 3. Jena, 1905.
VOL. rr. 2 H

56 E. KIRKPATRICK.

24. TOPSENT, E. — Une reforme dans la Classification des Halichondrina. Mem. Soc. Zool. France,

tome vii. 1894.

25. TOPSENT, E. — Resultats scientifiques de la Campagne du ' Caudan ' dans le Golfe de Gascogne, 1895.

Eponges. Annales de 1'Universite de Lyon, xxvi., (4). Paris, 1896.

26. TOPSENT, E. — Eponges nouvelles des A£ores (premiere serie). Mem. Soc. Zool. France, vol. xi.,

p. 225-255. Paris, 1898.

27. TOPSENT, E. — Introduction a 1'etude monographique des Monaxonides de France. Archiv. Zool.

Exp. et Gen. (3) vi. 1898.

28. TOPSENT, E. — Etude Monographique des Monaxonides de France. Archiv. Zool. Exp. et Gen. (3).

tome viii. 1900.

29. TOPSENT, E.— Expedition Antarctique Beige. Resultats du Voyage du S.Y. 'Belgica'en 1897-99.

Zoologie. Spongiaires. Anvers, 1902.

30. TOPSENT, E. — Resultats des Campagnes Scientifiques du Prince de Monaco. Spongiaires des Afores.

Monaco, 1904.

31. TOPSENT, E. — Poecilosclerides nouvelles recueillies par le ' Francais ' dans 1' Antarctique. Bulletin

du Museum d'histoire naturelle. Paris, 1907. No. 1.

32. YOSMAEE, G. C. J. — The Sponges of the Willern Barents Expedition, 1880-81. Bijdragen tot de

Dierk. An. 12. Amsterdam, 1885.

33. WELTNER, "VV.— Susswasserspongien von Celebes. Archiv. fur Naturgeschichte, 1901. Beiheft.

34. WILSON, H. V. — Reports on an Exploration off the West Coasts of Mexico, Central and South

America, and off the Galapagos Islands, by the ' Albatross ' during 1891. The Sponges. Mem.
Mus. Comp. Zool. Harvard, vol. xxx. No. 1. Cambridge, U.S.A., 1904.

OP -HE

Y \

10

Sponges PI

Sponges R 9

OF THE

UNIVERSITY

OF

\

\

•f

, jJ»A«VtS

.

tuf.

.

UNIVERSITY.

CAL FGH£

Antarctic (Discovery) Ejq»

fi.'i PI !1.

Hi>

1

j

^

•.'"-•5

1

'

Antarctic (Discovei-y/ P^

SpoiKjes PI 12.

Be

ro

/2

^covery; Ex p.

s PI 13

P HigKley.

.-

\

10"

,

9 i

/J

/•#

H "A

Antdn.:tic(Discovery) b

S

p.

.

Pl.l-f

PHiqhleyl.onHon

I

->\

*

Sponges PI. 15

10

Antarcfcc( Discovery)Exp

Sponges PI 17

P Highly. London

X*V>I«MA*«

/ /^r , . X

f \ V

I

4 b

Antarctic ( Discovery )Exp

Spongea PI 18

P Highly. London

OF THE

( UNIVERSITY }

OF

Antarctic ( Discovery ) Exp

Sponges PI

r Mighty. London

Antarctic ( Discovery )Exp

Sponges PI 20

P Mighty. London

/^

. ^'VERS;

1 a

Antarctic ( Discovery ) £xp

Sponges PI 21

Lcmdoo

*

Antarctic^ Discovery )Exp

Sponges PI 22

P Highlfy. London

6c*

Antarctic^ Discovery )Exp

Sponges PI 23

P Highly, '..on dor.

Antarctic* Discovery )Exp

Sponges PI 24 .

P Highly London

ft

Antarctic* Discovery ) Exp

Sponges PI 25

'Mi

.CRSITY J

Antarctic ( Disccveiy ) Exp

Sponges PI 26

P Highly. London

POKIFEEA.

III.-CALCAEEA.

By C. F. JKNKIN, B.A.
(12 Plates.)

PART I.

INTRODUCTION AND CLASSIFICATION.
INTRODUCTION.

THE preparation of the Report on the 'Discovery' collection of calcareous sponges
was, in the first instance, entrusted to Professor E. A. Minchin. He had partially
examined the Homocoela and made some drawings of them, when he was unfortunately
obliged to abandon the undertaking, owing to the pressure of other duties. The
collection was then entrusted to the author, who has had the advantage of using
Professor Minchin 's notes on the Homocoela and his lifelike drawings (Figs. 12 and 14),
showing the habit of growth of the two new species of Leucosolenui.

The author desires to record his indebtedness to Professor Minchin for much
valuable advice, to Professor Dendy for valuable suggestions as to the classification of
the new genera, and for specimens of Grantiopsis cylindrica, to Mr. R. Kirkpatrick
for constant help and particularly for permission to examine the British Museum
collection of calcareous sponges, to Professor R. von Lendenfcld for specimens of
Sycon teneUum, and to Professor Jeffrey Bell, to whose kindness he owes the privilege
of undertaking this most interesting investigation.

The collection consists of 109 specimens. Of these, 39 belong to the grade
Homocoela, and are divided among five species, of which two are new to science. The
remaining 70 belong to the grade Heterocoela, and are divided among 18 species
ami one variety of an existing species, all of which are new to science.*

Of the five species l>elonging to the grade Homocoela, two belong to the genus
Clathrina and three to the genus Leucosoltnia.

Of the 18 new species of Heterocoela, five belong to the genus Leucandra and the
remaining 13 are distributed among six new genera.

CLASSIFICATION.

The classification here used, which is shown in the following table, is that
proposed by Polejaeff (4) for the Homocoela and by Dendy (2) for the Heterocoela

• With the pocnible exception of three of the specie*, which on only to cooadered M proruiotnlly MtUed.

2 H 2

C. F. JENKIN.

and adopted with slight modifications by Minchin (1) ; it has been considerably
extended to include the new species. The last column in the table gives the
number of species in each genus found in the ' Discovery ' collection.

Families.

1. Clathriiiidae

2. Leucosoleniida;

1. Sycettidse

2. Grantiidas

3. Heteropidae

4. Amphoriscidae

5. Chiphorida

C Staurorrhaphida.

Genera.

Number of Species

in ' Discovery '

Collection.

GRADE A. -HOMOCOELA.

(Clathrina
. < Ascandra.
iDendya.

(Ascyssa.
'(Leucosolenia

GRADE B.— HETEROCOELA.

;|Sycetta.
. <Sycon.»
I TentJtrenodes

Grantia.f
Ute.

Sub-genus Synute.
Utella.
Anamaxilla.
Sycissa.
Leucandra .
Lelapia.
Leucyssa.
^Dermatreton

fGrantessa.
< Heteropia.
( Vosmaeropsis.

Heteropegma.
Amphomcus.
Sycnlmis.
Lencilla.

(Streptoconus
\Hypodictyon

{Achramorpha
Sub-genus Grantiopsis.
Megapogon .

5 species.

5 and 1 variety.
2

1
1

8

4

18 species and
1 variety.

In the al>ove table the names of the new families and genera are printed in italics.

* The genus Sycantha is omitted for reasons stated on p. 4.
f Grantioptit is transferred to the family Staurorrhaphidce.

CAIS'AUA.

baaed primarily on the

the families are
the spimlation. The familial

New FAMILIES — The two new families, Ckipkondae * and Sumrorrttapkidrj are
to contain the 9 new specie* and 2 old onea which have
* M introduced to denote • •neriil type of qi
the ordinary quadriradiate* both in ahape and in position in the
The <±i~*im+ IB » quadriradiate spieale lying with it* baaal ray directed
outward* (eentrifugaDy) and ha apical ray, which i* bent at ita base so aa
to Ee almost in line with the baaal ray, directed radially inwards (eentripetally)
and projecting into the gartral canty. § This type of spicale has hitherto been
in only 2 species of sponge, each represented by a single apuimfn, tit.,

Polejaeff (4) and Crmmfitftit cyUmdriea, Dendy (7).
indnded in the new gens Mtyepoy**. The latter is transferred
to the new family StamrorriufJud*, the generic name being retained. In order to
make the present report a complete record of aO species in the two new families.
lumt deseriptaone of Meytpoym crmctfera and Gramtwpiu «slnaV«es, with ilii>mgj

In some of the new species thai type of epicnle occurs in eonjuKtion with the
typea, bat in several it forms the whole gnstral and body-wall skeleton, to
i of th? ordinarr amadriradiates and linxmitfs ; ito importanee, therefore,
can hardly be over-rated, and folly justifies the formation of the new famines. AD

way that the Syeettid* do from the Grmmtndm, it seemed better to drride
two

I For.

4 C. F. JENKIN.

NEW GENERA. — The new genera, Hypodictyon* Dermatreton,\ and Tenthrenodes,^
are introduced to contain the five new species which have " linked " flagellated
chambers. The term " linked " is here used to describe the peculiar arrangement
of the flagellated chambers in an open network or honeycomb pattern, so that a
large number surround each of the very large incurrent canals (intercanals). In
Sycon and Grantia the incurrent canals are usually smaller than the flagellated
chambers, and are each surrounded by only three or four flagellated chambers.

Three of the new species have freely projecting distal cones ; two of these are
included in the new genus Tenthrenodes, in the family Sycettidse, and the third,
which contains chiactines, is in the corresponding genus Hypodictyon, in the family
Chiphoridse. The two remaining species have distinct dermal cortices, and are,
therefore, included in the new genus Dermatreton, in the family Grantiidse. The
dermal cortex is not continuous, but takes the form of a network covering the top
of the flagellated chambers, and is pierced by large holes corresponding to the spaces
(incurrent canals) between those chambers.

The " linked " arrangement of the flagellated chambers appeared to resemble
closely the " grouped " arrangement described by von Lendenfeld (6) as occurring in
Sycantha temlla. As doubts existed concerning the accuracy of some of the details
of this description, it seemed advisable to re-examine the specimens. The author was
enabled to make this examination by the great courtesy of Professor von Lendenfeld,
who sent him all the remaining material he possessed. This material is
considerably macerated (as is stated by von Lendenfeld in his original description),
but is in quite good enough condition to allow the general structure to be ascertained
with certainty. The results of the author's examination show that all von Lendenfeld's
figures represent the structure correctly except Figs. 53, 54 and 56, which, though
no doubt accurately drawn, are quite misleading. The interpretation placed on the
figures by him appears, however, to be erroneous, and would seem to be due to the
unfortunate cross-section shown in Fig. 56. The true structure is best shown in
Fig. 57, representing a tangential section near the gastral cortex. This figure shows the
regular rectangular network formed by the chamber walls. The rectangles are alternately
flagellated chambers and canals, arranged like a chess board, exactly in the manner
shown in Haeckel's " Kalkschwamme," Fig. 13, Plate 60, as typical of his sub-genus
Sycocubus (sub-genus 3 of Sycandra). Haeckel's figure shows (correctly) the arrange-
ment of the flagellated chambers and inter-canals of Sycon schmidti ; this arrangement
only differs from that of Sycantha tenella in the shape of the flagellated chambers,
which in Sycon schmidti are approximately square (in tangential section), whereas
in Sycantha tenella, though still rectangular", they are much longer in one direction
(parallel to the axis of the sponge) than in the other (circumferentially). Sycantha
tenella is, therefore, a typical Sycon, remarkable for the size and regular arrangement

* imo, under ; Starvon, a net. t bfppa, skin ; Tptjrus, holed.

t,s, honeycombed.

CALCARBA. 5

tin- tlngellated chamben. The flagellated chambers touch each other only at the
ainl have no intercommunication canal*. Serial tangential sections show that
each flagellated chamber opens independently into the gastral cavity. (See
Plate XXXVIII, Fig. 137.) This may also be inferred by comparing the distances
between the openings into the gastral cavity, shown in von Leudenfeld's Figs. 56 and
58, and the distanres between the flagellated chambers shown in Fig. 57 (rcmeml>ering
that the rectangles are alternately flagellated chambers and canals). The crow-sections
made by the author correspond exactly with the structure above described. It
is, therefore, evident that there is no similarity between Sycon tenellum, as it
should now be called, and the new genera with " linked " chambers.

The new genus Streptoconus* in the family Staurorrhaphidse, corresponds with
Sycon in the Sycettid«,

The new genera Achramorpha f and Megapogon^ in the family Staurorrfuiphidee
correspond with Grantia and Leucandra in the Grantiidse.

Dendy's sub-genus Grantiopns is transferred from the Grantiidse to the Staurorrha-
j'hiilx, because the only species (Grantiopsin cylindrical) contains chiactines.

TERMINOLOGY. — The following terms are used to describe the different pans
and passages by which the water passes through the sponge :—

Ostia. — The pores in the dermal membrane through which the water first enter* the sponge.

Prosopylts. — The openings in the flagellated chamben bj which the water enters them.

Apopyle*. — The openings in the flagellated chambers by which the water leaves them.

Ports. — This is a new term used to describe the openings in the gastral membrane by which the
water passes from the excnrrent chambers into the giwtral cavity.

Port is only used in a general sense for any of the above.

Inairrent rhambrrs.—Thc large spaces sometimes found under the dermal cortex. The wmter
enters them by the ostia and leaves them by the incnrrent canals.

Infvrrrnt canals. — The passages by which the water reaches the ontaidc of the flagellated
chambers. They communicate with the outside water either through the ostia or directly.

Excurrml canals.— The passages by which the water, leaving the flagellated chambers through
the apopyles, reaches the gastral cavity or oscule.

Hffurrent chambers— The spaces sometimes found outside the gastral membrane into which
•eveml excnrrent canals or several apopyles open. The water flows out of them through port* into
the gastral cavity.

0»cwfo.— The opening by which the water finally leaves the sponge. It may be at the end of one
or more excnrrent canals, or at the top of the pwtral cavity.

Oscular collar.— A. thin tubular extension of the body wall, without flagellated chambers, leading
to the oscule.

The most interesting features of the collection are :—

(a) The large number of species (9) containing chiactinc spicules.
(6) The five new species with " linked " flagellated chambers.

(c) A remarkable sponge, Mtgapogon rillosiu, with larger spicules than any hitherto recorded for
a calcareous sponge. The oxea are upwards of 15 mm. long.

, twisted ; ««MW , a cone. t «»>•*, » P*** : »«*** '<*».

gw»t ; »«y»r, • beard.

6 C. F. JENKIN.

(d) The development of the gelatinous mesoderm in Leucimdra gelatinosa.

(e) The duplicate ovum, apparently a new type of egg cell, in Hegapogon raripilvs.

Locality. — All the specimens were taken, at various dates, in Winter Quarters
and most of them from shallow waters.

PART II.

DESCRIPTION OF THE SPECIMENS WITH DEFINITIONS OF THE NEW GENERA AND

SPECIES.

GRADE HOMOCOELA POL.
J?AMILY CLATHKINIDAE MINCHIN.

CLATHRINA PRIMORDIALIS.
Ascetta primordMis Haeckcl (3), Vol. II., p. 1G.

There is one small broken specimen of this species in the collection. In colour
it is yellow as preserved in spirits. The spicules agree exactly with Haeckel's
description.

CLATHRINA CORIACEA.
Ascel/a coriacea Haeckel (3), Vol. II., p. 24.

There are two small broken specimens of this species in the collection. In colour
they are yellow, as preserved in spirits. The spicules agree exactly with Haeckel's
description.

FAMILY LEUCOSOLENIID^ MINCHIN.
LEUCOSOLENIA COMPLICATA.

Leucosolenia complicata Minchin (8), p. 860.

There are five pieces of sponges of this species in the collection. The spiculation
agrees closely with the revised description given by Minchin.

LEUCOSOLENIA DISCOVERY!.
(Plate XXVIII., Figs. 12-13.)

There are about twenty-five specimens of this new species in the collection.
Its appearance and habit of growth are admirably shown in Fig. 12, drawn by Professor
Minchin. Its colour is white as preserved in spirits. The oscular tubes in some
specimens attain a length of 9 mm. and a diameter (flattened) of 1 mm.

CALCAREA 7

Spicules (Fig. 13).
The following description is due to Professor Mim hin :—

ft.™/.— The monaxon spicules of this species are very constant and characteristic.
They are divisible at the outset into (</) ordinary and (b) refringent nionaxons : —

(«») The ordinary monaxons are curved and vary greatly in dimensions, from
about 100 p. to 450 M or more in length, and from 3 ft to 16 M in thick ness,
Every possible gradation is to be found l>etwecn the greatest extremes in
size, so that it is not possible to divide these spicules into different classes
The -haft is thickest near the proximal end, which is also nearly straight
for rather more than half the length; from this point the shaft curvet
evenly, and tapers slightly, towards the distinct lance-head, which usually
shows very plainly the form of a double l>cnd, especially in the smaller
examples of this type of spicule. Among the ordinary nionaxons there are
always a very few P type monaxons : these are very rare, and only one or
two are usually found in a whole spicule slide.

(b) The refringent monaxons are much scarcer than the ordinary type, but they

are easily found, as their optical peculiarity makes it easy to distinguish
them from the others, especially under a low power of the microw-ope.
Like the ordinary monaxous, they show a wide range of variation in size,
from small to very large. Their form in similar to that of the ordinary
monaxons in a general way, but shows certain constant peculiarities ; the
shaft is very slightly curved, sometimes nearly perfectly straight; the
proximal swelling of the shaft contrasts more sharply with the distal
extremity, which is often very slender and tapers down rapidly about half-
way from the proximal end ; and the lance-head tends to lie rudimentary
or even quite absent By these special features the spicules can easily be
recognised as a distinct class, apart from their characteristic appearance
under the microscope.

Triratliata.

(c) The ordinary triradiates are of fairly large size, in form very similar to

those of Laicosolenia crnnplicata, with the unpaired rays markedly longer
than the paired rays. The unpaired ray is straight and slender, varying in
length from about 1 1 0 /* to as much as 1 85 /i. The.paired rays are generally
slightly thicker than the unpaired, sometimes distinctly so ; they curve
symmetrically, first in a backward direction for the proximal two-thirds of
their length, then forwards for the distal third ; in length they vary from
100/t to 145 M, the average length l>eing about mid-way between these two
extremes. The anterior angle is an open one, but the lateral angles are
always greater than right angles, the spicules never l>eing T-shaped as in

YOU IT. ^ I

C. F. JENKIN.

botryoides. In young specimens the triradiates are generally smaller and
more slender than in the larger colonies.

Quadriradiates.

(d) These spicules are generally much scarcer than the simple triradiates, but
can always be found. Their facial rays are similar to the triradiates, with
perhaps slightly less tendency to elongation of the unpaired rays. The
gastral ray is short or of moderate length, laterally compressed, and set
well back on the unpaired facial ray ; its proximal end runs straight
upwards for about two-thirds of its length, then the shaft curves forwards,
the distal extremity, however, being again nearly straight.

Speculation of the Oscular Rim.

The skeletal spicules of the oscular rim appear to be quadriradiates alone.
They are of Relatively small size and slender dimensions ; the lateral angles
are very nearly or even quite right angles, so that the spicule is more or
less T-shaped ; the unpaired rays are always distinctly longer than the paired,
the latter being strongly curved, and the gastral rays are relatively long.

The monaxons of the oscular rim are, with rare exceptions, of the smallest size.
Both ordinary and refringent monaxons are found ; the former differ from
those in other parts in the slenderness of the shaft, the proximal end of
which is scarcely or not at all swollen. The refringent monaxons in this
region are also very slender and sharp.

LEUCOSOLENIA MINCHINI.
(Plate XXVIIL, Figs. 14-15.)

There are about eleven specimens of this new species in the collection, most of
them being oscular tubes only, with very little root tube. Their appearance and habit
of growth are well shown in Fig. 14, drawn by Professor Minchin. The oscular
tubes, which are 3 mm. to 4 mm. long, are usually largest towards the middle of their
length. Diameter, flattened, about 5 mm. The bulk of the skeleton spicules are
triradiates, regularly arranged and often in rather open order, so that the walls have
a transparent appearance. The oxea, which are small, usually about 90 M long and
never exceeding 160/u, are usually scarce, and in some specimens almost, if not entirely,
wanting. In the specimens with plentiful oxea there is a fringe round the oscule,
see Fig. 15, but in the specimens with few oxea there is little or no fringe.

Spicules (Fig. 15).

Oxea. — There are two sorts of oxea, ordinary and refringent :—
(a) Ordinary oxea, straight or slightly curved, with a lance head which is
usually bent slightly to one side, 60 M to IGO/x long x 3/x to 6/x thick.

• \ir\i:r.\

I 'MMKineflt sfoe is 90/u long x 4/u to SM thick. In some npocuncns

tin- maximum size is 90 M long.

Am. •ii- the unliiiary oxea there are a very few r type oxea. It is uncertain
whether these are always present It seems probable that in those
.•ijMviim ii- in which the oxca arc very scarce there may lie none of P type.

(/») Rcfringent oxea, similar to the smaller and thinner sizes of (a) and usually

nearly straight

'l'h»« Trir<t<liates ate of one sort : —

•

(r) Alate triradiates, with the basal ray longer than the paired rays. Basal ray
straight, tapering, fairly sharply pointed, 90 M to 140 M long x 5/» to 10 M
thick. Paired rays equal, bent very slightly downwards, tapering slightly
for about two-thirds of their length, then fairly sharply pointed, 50 M to
90 M long x 5ft to 7 A* thick ; oral angle 130° to 140°.

The Qtuidriniiliate* are of one sort :—

(</) Alate quadriradiates, facial rays the same as the triradiates (c). Apical
ray sharply pointed.

Oscular spicules : —

(«*) Oxea from the fringe, like (a) but shorter, with bluntly pointed lance heads,

55 M to 65 M long x 5 M to 6 M thick.
(/) Quadriradiates from the oscular e<lgc like (»/), but of the smallest size, the

paired rays rather more bent Oral angle 160°.

GRADE HETEROCOELA POL.

FAMILY 8YCETTIDAE DKNDY.

TENTHRKNODW.

A Sycettid in which the radial chambers, with freely projecting distal cones, are
" linked " so as to form a reticulated pattern round the large inter-canals.

This genus contains two species. In both these species the sponge is in the form
of a tube, usually more or less bent, the cross-section being probably circular when
alive, though it is considerably flattened in some of the preserved specimens. The
diameter of the tube is greatest near the middle of its length. It tapers to the base,
which is rounded off, and also to the top, which terminates in the oscule. The interior
of the tube forms the gastral cavity. The inner or gastral layer is supported by a
special skeleton of tri- and quadriradiates, the apical rays projecting into ^thc gastral
cavity. Outside the gastral layer stand the flagellated chambers, pointing more or

2 I 2

10 C. F. JENKIN.

less radially, and " linked " into a meshwork pattern. They communicate with the
gastral cavity either in groups through excurrent chambers (T. scotti) or
independently (T. antarcticus).

TENTHRENODES SCOTTI.*
(Plate XXVIL, Fig. 9; and Plates XXVIII. and XXIX., Figs. 16-27.)

There are two specimens of this new species in the collection, the larger one perfect,
the smaller one a broken fragment. Both were fixed in osmic acid and preserved in
alcohol. The larger and perfect specimen is referred to in the following description.

The sponge as preserved is flattened so that the opposite walls are in contact. Its
shape in this flattened condition is shown in Fig. 16. The length is 115 mm. and
the diameter (flattened) 32 mm. The colour is brown, owing to the osmic acid, and
was probably white in life.

The whole surface is covered with a conspicuous honeycomb pattern due to the
linking of the flagellated chambers. When slightly magnified the pattern is seen to
consist of a principal network of larger and deeper meshes divided by a finer network
of shallower meshes (see Fig. 9). To the naked eye the meshes appear to be smooth.
The pattern extends up to the oscular edge, the size of the meshes diminishing
gradually from the middle of the sponge to the oscule.

Canal System. — The flagellated chambers communicate through contractile
apopyles with excurrent chambers, which open through irregular ports into the gastral
cavity. The canal system is extremely difficult to make out in this species, owing to
the very irregular shapes of the flagellated chambers near their bases and the erratic
way in which they are crowded together. The presence of the .excurrent chambers can
be ascertained most easily by examining the gastral surface under a low power. The
ports in the gastral wall are largo enough to allow the interiors of the excurrent
chambers to be inspected and two or three of the apopyles to be seen in each.
Sections cut either perpendicular or parallel to the axis of the sponge are almost
useless for demonstrating the structure, but serial tangential sections enable the
actual connections to be followed satisfactorily and bring out the peculiar " linked "
arrangement of the chambers admirably. Figs. 17-20 represent corresponding parts
of four tangential sections chosen for illustration from a complete series ; Fig. 17 shows
the ports in the gastral layer ; Fig. 1 8 the grouping of the chambers in the sub-gastral
layer; Fig. 19 the grouping higher up, where both large and small meshes occur;
Fig. 20 the grouping on the surface, where only large meshes occur, the smaller
ones not* reaching so high; the section includes the tops of most of the chambers,
but a few project further and therefore appear cut through. Several groups of cells
are numbered, and may be followed from one section to another.

In order to show the structure more clearly, the group of flagellated chambers com-

* This species, represented by the finest specimen in the collection, is named after the distinguished leader
of the Expedition,

cM.i'ARBA. II

tl>'' Urgent excurrent c-haniU-r in Kip. 18 in shown in Figs. Zl-'2-\ <>n a
I II n-preaente the subgastrnl structure, and shown the evils gnmpe«1
r.. mi. I nul partly over the excurrent chamber: the section includes the bottom
.li:i|ihragms and apopyles of five of the flagellated chambers The excurrent chamber

.U 1« l<.w these apopyh-, l.ut exactly how far cannot be seen. The position of the
]x>rt in the gastral layer, which is much smaller than the full extent of the excurrent
. h mil »T. is shown by the dotted line. Fig. 22 represents the next section higher, and
in. -lades the top of the excurrent chamber with one apopyle ; the second apopyle
1 ..•!,, n-nii; to flagellated rhamber K cannot U> clearly made out and has not been
shown, thou-_'h it mast l>e present Fig. 23 shows the same group of rhaml>ers higlu-r
up and Fig. '_' 4 higher still.

Skeleton. —The gastral skeleton is a dense felt of large quadriradiates, with the
basal rays pointing in all directions, though the majority point more or less towards
tin- l«se of the sponge. The apical rays, which arc very sharp, project into the gastrnl
. u\it\. The arrangement of the spioules is shown in Fig. 26, which represents the
same port that is shown in Fig. 21 ; in this drawing only a certain number of the
>]>i. -nl«'s a. tiiiillv present are drawn. The large number of the distorted r spiculet is
rather irnmrkaMe. The articulated tubar skeleton is formed of triradiates which arc
bent over the distal end of the chamber, forming a thick rounded end and not a cone.
The tops of contiguous flagellated chambers are joined together by the fusion of their
skeletons, which thus form a sort of dermal cortex, cf. Fig. 20, similar to that in genus
Dtrmatreton. In the present species, however, the flagellated chambers are of different
lengths, and consequently the dermal cortex is very irregular and not clearly defined,
so that it seems advisable to include it in Tenthrenwle* rather than Dermatreton.
Roun-1 the distal ends of the chamtars there are rings of oxea which project slightly.

The oscular skeleton differs very little from the normal gastral and dermal
.-k.-h-t.iii8. The quadriradiates are rather smaller than the ordinary gastral
quadriradiates. There is a thick fringe of hair oxea. The meshwork of flagellated
chambers extends right up to the oscular edge, terminating in a scolloped edge.
Fig. 25 represents a longitudinal section through the oscule at a point where the
mesh reaches the edge.

Spicules (Fig. 27).

The Oxea are of one sort :

(a) Projecting oxea, 160/i to 24 OM long x 10/t to 15 M thick, slightly and

irregularly bent, thickest towards the outer end, which is bluntly pointed ;
the inner end is rather more sharply pointed

The Triradiate* are of one sort :

(b) Alate triradiates from the body-wall and dermal cortex. Basal rays

straight, tapering to a sharp point, 90 p to 330 M long x 9 M thick. Paired
rays, unequal (appearing the more so owing to folding), bent slightly

12 C. F. JENKIN.

upwards, 80 /A to 150 M long x 8 to 10 M thick. Oral angle 110° to 125°.
Occasionally a T shaped variety is found.

The Quadriradiates are of one sort :

(c) Large alate quadriradiates from the gastral layer. Basal ray straight,

tapering uniformly to a very sharp point, 420 M to 680 M long x 10 /A
thick. Paired rays equal, bent slightly downwards near the centre, then
straight, 330 AI to 380 n long x 10 p. to 11 n thick. Oral angle 140° to
150°. The T shaped variety is fairly common. Apical ray slightly bent
orally, sharply pointed, 50 n long x 8 p, to 12 M thick. Occasionally the
apical ray appears to be wanting.

Oscular spicules.

(d) The oscular fringe consists of hair spicules over 400 M long.

The triradiates are the same as (b) ; the quadriradiates are similar to (c), only
much smaller. The projecting oxea are the same as the smaller sizes of (a).

TENTHRENODES ANTARCTICUS.
(Plate XXIX., Figs. 28-32.)

There are three specimens of this new species in the collection ; they are easily
distinguished by the naked eye from T. scotti, owing to their having bare basal tubes
and bare oscular collars, so that the sponge presents the appearance of an oval ball
pierced by a tube. The dimensions of the three specimens are : —

11 mm. long x 3 mm. diameter.
4 mm. „ x 1 '2 mm. ,,
10 mm. „ x 1 • 6 mm. ,,

In colour they are white as preserved in spirits.

The flagellated chambers are branched near their proximal ends, and each opens
directly into the gastral cavity through a contractile apopyle. They have articulated
skeletons which terminate at the distal end in freely projecting cones of triradiates,
amongst which are a few small oxea (see Fig. 28).

The gastral skeleton (see Fig. 29), consisting of large tri- and quadriradiates,
extends the whole length of the tube.

The oscular edge has a very regular skeleton of small quadriradiates closely
packed together, with a fringe of hair oxea (see Fig. 30). Unfortunately the
specimens are not in a sufficiently good state of preservation to allow the exact
structure of the basal tube to be made out. There are indications of a lining of
flagellated cells in the basal tube, continuous with the lining of the lowest radial
flagellated chambers. These chambers appear to arise as bulgings of the gastral layer.

OALCARKA. 13

It i- miliiitimatr that this interesting structure cannot be made out with certainty.
Tlic origin of (he lla^cllatcd chambers no»r the osculc can lie men clearly ; an isolated
group of flagellated cells forms outaide the gastral layer, and over this a few triradiatett
tiinl oxea grow (nee Fig. 31«i). As the Hagcllated chamber grows it tips up the
ii.itcs and oxea, which thus begin to point outward* while they continue to grow
(see Fit: -U6). At the oscular end of the sponge therefore the flagellated chambers
<1» in it grow as divcrticula of the gastral layer, nor is there any lining of flagellated
"•IN in this part of the gastral cavity.

Spicules (Fig. 32).
The Oxfa are of one sort :

(a) Small projecting oxea, slightly irregularly bent, faintly hastate at the
thinner end, bluntly pointed at the thicker end. 70 M to 210 M long x CM
I- in a thick.

radiates. The triradiates are of two sorts :—

(c) Alate triradiates from the gastral layer. Basal ray straight, tapering
slightly, then pointed ; maximum size 480 M long and 12 M thick. Paired
rays unequal, slightly irregularly bent, maximum size 160 M long x 12 M
thick. Oral angle 130° to 135°.

((/) Alate triradiates from the tubar skeleton. Boaal ray straight, tapering
slightly, then pointed, 100/* to 140ft long x CM to 7 n thick. Paired
rays nearly equal, straight, 70 M long x CM to SM thick. Oral angle 120°
to 130°.

The Quadrirad iales are of one sort :—

('•) Alate quadriradiutes from the gastral layer. Banal ray straight, tapering
slightly, then bluntly pointed, 320 M to 480 M long x 10 M to 12 M thick.
Paired rays unequal, nearly straight, 140 M to 240 M long x 10 M to 12 M
thick. Oral angle 130° to 140°. Apical ray bluntly pointed, slightly lient
orally, 1:50 M to 180 M long, oval in section, SM to liV thick x 1'2 n to
16/u deep. In side view the apical rays sometimes appear swollen
near the point, the depth being greater than at the junction with the
facial rays.

Oscular spicules :—

(/) Quadriradiates from the oscular edge, Imaal ray straight or slightly bent
Inckwards, HO/* to 140 M long x SM thick. Paired rays about equal,
bending downwards, bluntly pointed, 80 M to 110M long x 10 M thick.
Oral angle 145° to 155°. Apical rays short and sharply pointed.

14 C. F. JENKIN.

The triradiates in the oscular collar are similar to (c), but have longer
paired rays.

The fringe consists of :—

(g) Minute oxea, 120 /u long x 1 M to 2M thick.

FAMILY GRANTIIDAE DENDY.
LEUCANDRA Dendy.

The collection contains five new species and one new variety belonging to
this genus.

The genus Leucandra, as defined by Dendy, is too comprehensive and needs
subdivision. The minute structure of the species belonging to the genus has not been
sufficiently examined ; the wide variations in the development of the mesoderm, and
also the arrangement of the incurrent and excurrent canals, require investigation.
Some notes on these points are contained in Part III.

Want of accurate knowledge about the structure makes the classification of the
species in this genus difficult and uncertain. The species Leucandra frigida, Leucandra
brumalis, and Leucandra gelatinosa can only be considered as provisional ; Leucandra
cirrata and Leucandra hiberna are more definitely differentiated.

LEUCANDRA PRIMIGENIA, H. var. leptoraphis.

(Plate XXIX., Figs. 33-34.)
Lemetta primujenia Haeckel (3), Vol. II., p. 118.

There are two specimens of this new variety in the collection. The larger consists
of an irregular mass of anastomosing branches (Fig. 33), dirty white in colour as
preserved in spirits. The surface is mostly smooth, but in places is more open or
spongy in structure. The branches have a very small gastral cavity running down
them, ending in inconspicuous closed oseules. It agrees with Haeckel's description of
Leucetta primigenia (var. isoraphis), except that the spicules are very much slenderer.
Haeckel gives the ratio of length of arm to thickness as 10 to 12. Polejaeff, for the
' Challenger ' specimens of Leucetta primigenia, which he calls Leucetta fruticosa, gives
the ratio 12 to 15. The new variety has the ratio 25 to 40. The spicules are
shown in Fig. 34, with one of the ' Challenger ' specimens of Leucetta primigenia for
comparison beside them. Haeckel does not describe the alate oscular spicules, which,
in the new variety, differ widely from the normal spicules, as will be seen from the
figure. These spicules lie with the basal rays circumfereritially round the oscule, and
not, as might have been expected, symmetrically.

PolejaefFs species, Leucetta dura, cannot now remain, since it is only based on the
existence of an oscule which was supposed not to be always present in Leucetta
mici'oraphis.

« M.OARRA. 15

Spicules (Pig.
Tin- //••./'/ ipicuU* are of one sort :--

i It. .jul.u triradiataL . Rays straight, very slender, parallel, bluntly pointed,
140M to 200 n long x 4* to ?M thick. Ratio of length to tliirknew,
to 40.

The Oscular spicults are of one sort :—

(k) Mate triradiates, strongly folded. Basal ray straight, nearly parallel,
bluntly pointed, 55 p to 90 M long x 1 it. to SM tliick. Paired rays nearly
straight in facial view, and twice bent when viewed parallel to the basal
ray. Tapering, moderately sharply pointed, 90 M to 140/* long x 8 M to
10 M thick. Oral angle 105° to 115°. Angle of fold 110°.

LEUCANDRA KRIGIDA.
(Plate XXVII., Fig. 6, and Plate* XXIX. and XXX., Figs. 35-40.)

Thi-re are eleven specimens of this new species in the collection. They are all fig-
shu|H-.l, \\itli the oscule at the thick end (see Fig. 6). The surface is smooth, pure white
and rather chalky in appearance. They vary in size from 10 mm. x 3 mm. to
30 mm. x 10 mm. (nee Fig. .15). They are all solid to the touch, though they vary

much in texture when cut open. All but one of the specimens have the osculc
completely closed, and its presence is only indicated externally by a slight
protuberance. One specimen, 21 mm. long x 7 mm. diameter, has the osculc open ; it
is 2 mm. in diameter,* and is surrounded by a short spiculated membranous collar (see
Fig. 39).

Tlu- canal .system varies considerably in the different Kpecimens ; in some there it*
.1 wide gaatral cavity with large branching canals opening into it, while in others there
is no gastral cavity, the excurrent canals all converging to the oscule. In cross-section
tlu- >JM-. iiiH-as differ less than might Iw expected, for the gastral cavity appears to lie
divided by delicate septa, which give the whole section a mesh'work appearance even
when longitudinal sections show a large cavity.

The excurrent canals are usually surrounded with a fairly distinct skeleton of
Iri radiates, among which are a few quadriradiates.

The incurrent canal system is made up of a series of very irregular spaces leading
from the surface into the rather more regular longitudinal canals ; these run parallel
to the excurrent canals, from which they may Ite distinguished by their lest*
pronounced skeleton, which never contains quadriradiates.

The bulk of the skeleton consists of regular triradiates ; the quadriradiates round
the gastral cavities and excurrent canals are very constant in size. In some specimens
tin -y arc so rare that they can only be found with great difficulty. The apical rays,
which are slender and crooked, project into the canals.

VOUIT. 2 K

16 C. F. JENKIN.

Spicules (Fig. 40).

The Triradiates are of one sort :—

(«) Approximately regular triradiates. Rays straight, tapering, rather bluntly
pointed, 130 yu to 310 M long x 10 M to 20 M thick. One of the rays is
occasionally rather longer than the other two, the maximum ratio observed
being 5/4. Very rarely one of the rays is shorter than the other two, the
maximum ratio observed being 3/2. The angles in all cases are almost
exactly equal ; r forms are occasionally found.

The rare Quadriradiates are of two sorts :—

(l>) Approximately regular quadriradiatcs from the lining of the excurrent
canals and gastral cavities. Facial rays straight, slender, slightly
tapering, bluntly pointed, IGOjU to 200 M long x 9 /u to 12 /x thick. Angles
equal, 120°. Apical rays slender, wavy, sharply pointed, 280 M long x 8/u
to 9 M thick.

(c) Abnormal quadriradiates apparently due to the growth of a very short
apical ray on one of type (a) triradiates.

LEUCANDRA BRUMALIS.
(Plate XXX., Figs. 41-43.)

There are five specimens of this new species in the collection. They have no very
definite shape (see Figs. 41 and 42), but all are considerably longer than they are
broad, and each has a single small gastral cavity, surrounded at the upper end by a
small membranous collar about 2 '5 mm. long. They are smooth and white as
preserved in alcohol. The consistency of the sponge is fairly solid.

Canal System. — There are small incurrent chambers under the dermis,
communicating with irregular, more or less radial, incurrent canals. The excurrent
canals are also more or less radial, opening into the gastral cavity.

The Skeleton consists of slender regular triradiates with a few regular
quadriradiates round the gastral cavity. The apical rays, which project into the
gastral cavity, are very slender and crooked, the point being usually bent rather
sharply.

This species resembles Leucaltis pumila, var. Bleekii (H), but the triradiates are
much smaller and slenderer, while the quadriradiates are regular instead of irregular.

Spicules (Fig. 43).

The Triradiates are of one sort :—

(a) Regular triradiates, rays straight, tapering uniformly to a fairly sharp point,
170 p. to 180 p. long x 6/t to 10/t thick.

CALCARKA. 17

Tin- Qiuttlrinuliate* are of one sort :—

(A) Regular quadriradiates of varying size from the lining of the cxcurrcut
canak Facial rays almost straight, tapering to a point 60 M to 180/a
long x 4 M to 8 M thick. Apical ray very slightly wavy, sharply pointed.
The point usually bent a little to one side, 50 M to 120/u long x SM to
4 M thick.

LEDCAXDRA CKLATINOSA.
(Plate XXX., Figs. 44-53.)

There are two specimens of this new species in the collection ; one fixed in osmic
arid, which is therefore brown, and the other in spirits, which in quite white. The

istency is firm and the surface smooth. The general shape of the larger specimen
shown in Fig. 44 is irregular and contorted, suggesting a tul>crou8 root ; the oscule,
which is entirely closed, is at the top of the dome-shaped end. The diameter of this
part of the specimen is 1 1 mm. ; the smaller specimen is in the form of a nxl 24 mm.
long, nearly straight, rounded off at the oscular end, which is 4 mm. diameter, and
tapering gradually to the lower end, which is 2 mm. diameter.

The mass of the sponge is formed of the solid gelatinous mcsoderm, which is
ivmarkably developed in this species, and the incurrent and excurrcnt canals are reduced
to definite pipes lined with epithelium in this jelly. There is no gastral cavity, the
numerous excurrent canals converge to the oscule in the top of the domed end These
canals are regular circular pipes of small section lined with an extremely elastic
epithelium, which has contracted to such an extent as to almost close the lumens of the
canals (see Figs. 45 and 46).

The flagellated chamlwrs arc scattered throughout the Ixidy of the sponge and
communicate with the excurrent canals by means of narrow outlet pipes. The outlet
pipes from several flagellated chamliers usually join together before they open into the
main excurrent canals. These small outlet pipes are lined with elastic epithelium,
which is continuous with the lining of the main excurrent canals, so that when the
main canals close the small pipes are stretched out lengthwise in order to maintain their
connection with the central lumen of the main canal (see Fig. 46).

There are a large number of small ostia (see Fig. 48), close together all over
tin- dermal surface (see Fig. 47), which communicate through narrow nmMgM (see
Fig. 49) with irregular branching chamtars under the dermal cortex ; these in
turn lead into the incurrent canals which run as irregular pipes radially inwards, and
communicate with more regular circular canals running parallel to the axis of the
sponge. These canals have not contracted. The flagellated chamlxjrs are often situated
close against the incurrent canals, so that the communication is direct, presumably
through a pore cell which, in the contracted condition of the sponge, is closed. When
there is no incurrent canal near the chaml>cr a narrow branch leads from the incunvnt
canal to the flagellated chamber (see Figs. 51, 52 and 53).

2 K 2

18 C. F. JENKIN.

The Skeleton is built up almost entirely of regular triradiates, but among the
triradiates lining the excurrent canals there are a very few regular quadriradiatcs with
slender, wavy, apical rays projecting inwards.

The oscule is closed by the folding in of a thin oscular collar which, when open,
probably stands erect as a short tube. The collar has a thin skeleton of triradiates.

Spicules (Fig. 50).

The Triradiates are of one sort :—

(a) Regular triradiates of widely differing sizes. Rays almost perfectly straight,

but not quite, tapering moderately to near their ends, then sharply pointed,
140 M to 320 M long X 10 n to 20 M thick.

The rare Quadriradiates are of one sort :—

(b) Regular quadriradiates from the lining of the excurrent canals. Facial rays

equal, straight, tapering moderately to near their ends, then sharply pointed,
110 M long x 8 M thick. Apical rays wavy, sharply pointed, 80 M to 100 M
long X 3 M to 8 M thick.

LEUCANDRA CIRRATA.
(Plate XXXI., Figs. 54-56.)

There are two specimens of this new species in the collection, one complete, the
other much distorted, and possibly only a portion of a complete sponge. The follow-
ing description applies to the complete specimen. The sponge is ovoid in shape,
8*5 mm. x 3 mm., and is white as preserved in spirits. It is completely covered by
long, angularly bent projecting oxea, which to the naked eye give the surface the
appearance of being covered with curling hair.

The structure of the body wall which surrounds a large gastral cavity is shown in
Fig. 54. The gastral skeleton consists of triradiates and quadriradiates lying
tangentially with their basal ray downwards ; the body skeleton consists of sub-
gastral triradiates and the inner ends of the projecting oxea ; the dermal skeleton is
strong, consisting of about eight layers of triradiates, irregularly placed. The oxea
and sub-gastral triradiates are arranged in groups up which the body substance
(flagellated chambers) creeps to the dermis, forming as it were columns supporting the
cortex. Between these columns under the cortex is a large space forming an incurrent
chamber which is continuous over the whole sponge.

There are only two or three layers of the flagellated chambers, which are roughly
spherical, between the gastral and dermal membranes. The excurrent canals are short
and simple. .

The oscule has a fringe of stumpy oxea (Fig. 56) ; there is no diaphragm.

CALCAREA. 19

Spicules (Figs. 55 and 56).
The Oxea an off on* sort :—

(</) Very large bent projecting oxea, of all size* from 400 M to 1,830/t long, x
20 M to 33 M thick, pointed at both ends, usually straight for alwut two-
thirds of their length, then l>cnt at two or three points angularly, the end
being often bent through a total angle of 125° from the axis of the
straightcr portion of the spicnle. The [joints where the angular In-nds
occur are often marked on the convex Hide by flattened areas, a* if the
spicule had Iwen deflected by contact with a plane surface.

The Tri radiates are of one sort :—

(li) Alate tri radiates of very variable Kha{x>. There is no distinction lictwecu
the gastral, Ixxly wall and dermal triradiatea. Basal ray straight, tapering,
sharply pointed, 190^ to 360 M long x 10 M to 14 M thick. Paired rays
straight or bent slightly up or down, 160^ to 220 M long x 10 M to 14 M
thick. Tapering slightly and rather bluntly pointed. Oral angle 110° to
145°.

"idriradiate*. — There is one sort of quudri radiate :—

(?) Alatc quadriradiates from the gastral cortex, Iwsal ray straight, tajtoring to
a sharp point, 280 M long x 12 M thick. Paired rays equal, slightly bont
upwards, 140 to 220^ long x 10 to 12 M thick. Oral angle alnmt 130°
folded backwards, i.e. away from the face taaring the apical ray. Angle of
fold about 170°. Apical ray 40 M long x 8 to 10 M thick, sharply pointed,
l>ent orally.

Oscular spicules.

(/) The oscular fringe consists of short club-shaped oxea tapering to a point at
the inner end and bent near the outer end, which is rounded, I .'><>/« to
230 M long x 12 M to 45 M thick. The thicker spicules are the commonest.
The tri radiates at the edge have an oral angle of nearly 180°.

LKITANDRA HIBKRNA.*
(Plate XXXI., Figs. 57-58.)

There are two specimens of this new species in the collection, one straight, vase-
sliajied. 7 mm. x l£ mm., the other bent through at right angle, 6 mm. x 1 mm. In
colour they are dull yellow as preserved in spirits, and appear" minutely hispid to the
naked eye.

Tin- structure of the l>ody wall, which surrounds a large gastral cavity, is shown in

• Most of the specimens were OflBtlrtH at the ' Diacovcrj '» ' Winter Quarter*.

20 C. F. JEN KIN.

Fig. 57. The gastral skeleton consists of quadriradiates with the basal ray pointing
downwards : the body skeleton consists of sub-gastral triradiates and the inner
ends of the projecting oxea. Among the sub-gastral triradiates there are a very few
quadriradiates exactly like the triradiates, except for the development of the apical ray,
which is roughly perpendicular to the plane of the other rays, and does not project into
the gastral cavity.

The dermal skeleton consists of a thin layer of triradiates with the basal ray
pointing downwards.

The space between gastral and dermal membranes is filled up with three or
four layers of spherical flagellated chambers between which are small incurrent
and excurrent canals. There are no large incurrent chambers as in Leucandra
cirrata.

The oscular skeleton consists of quadriradiates with an oral angle of nearly 180° and
oxea. The dermal triradiates do not reach quite up to the edge. The thin oxea which
form the fringe are straight, with hastate ends. Just below the oscule there are a few
very short stout projecting oxea, similar in form to the longer ones lower down.

Spicules (Fig. 58).

The Oxed are of one sort : —

(a) Large projecting oxea, bluntly pointed at the inner end, hastate and
sharply pointed at the outer end, nearly straight for about three-quarters
of their length, then bent at the outer end, 260 n to 670 M long x 12 M to
24 A* thick at the thickest point. The thickness varies irregularly along
the length.

Triradiates. — There are two sorts of triradiates :—

(6) Sub-gastral alate triradiates. Basal ray straight, tapering uniformly to a
sharp point, 160 M to 200 M long x 6/« thick. Paired rays equal, more or
less bent downwards, 80 p. long x 6 M to 8 /x thick, nearly uniform in
thickness all along, then bluntly pointed. Oral angle 135° to 150°.

(c) Dermal alate triradiates. Basal ray straight, occasionally shorter than the

paired rays, 60 M to 200 M long x 8 p to 1 1 p thick, tapering slightly, then
bluntly pointed. Paired rays usually differing slightly in shape and size ;
bent upwards in a wide sweep at the centre and then usually slightly
downwards near the points. 100 M to 140 n long x 9 M to 12 /u, thick.
Oral angle 105° to 115°.

The Quadriradiates are of one sort :—

(d) Gastral alate quadriradiates. Basal ray straight, tapering uniformly to a

sharp point. 100 p. to 300 /x long x 8 /u, to 12 /A thick. Paired rays equal,
bending upwards in a wide sweep at the centre, outer half straight, 80/u. to

CAI/URRA. 21

180ft long x 9/t to 13/i thick. Oral nngle 120°. Apical ray curved
slightly orally, sharply pointed, 80 p. long x 8 p. tliirk.

fipicidff.

< Fringe oxea, slender, hastate, straight except for a slight bend at the
junction of the spear-head. IGO/i to '240 p. long x 7 p. to $ p. thick. The
maximum thickness is just l»clow the spear-head.
) Stumpy, stout, hastate oxea, from just lx?lo\v the edge, similar to («) 140/i

to 160 ft long x 14 ft to 16 p. thick.

(//) Small quadriradiates from the oscular edge. Basal my 180/i or leas x 8 p.
to 10 fu Paired rays equal, sharply pointed, bending slightly downwards,
70 p. to 100 p. long x 8/1 to 10 p. thick. Ural angle 155° to 100°. Apical
ray short, sharply pointed, bent slightly orally.

DKRMATRETON.

Grantiids in which the radial chambers are " linked " so as to form a reticulated
pattern round the large inter-canals, and are covered with a continuous reticulated

c«irt«-\.

The definition of this new genus in the family Grantiulae corresponds to that of
Tenthreiwdes in the family Sycettidaf.

The remarkable reticulated cortices of the two species in this genus are shown
in Kigs. 61 and 71. The importance of the " linking " of the flagellated chambers
as a generic characteristic is emphasised by the fact that it produces a corresponding
tv|x> of cortex. The cortical spicules have developed from the distal spicules of
tlu' tubar skeleton. In Grantia the inter-canals are small, and the skeleton bridges
i iv. r them. In Dermatreton the inter-canals are too large to be bridged over, and
a reticulated cortex results.

In Wh the species in this genus the sponge is in the form of a tube more or less
I M-nt, of circular cross-section, the maximum diameter being near the middle. The
tube tapers gradually to the base, which is very small, and also to the upper end,
which terminates in the oscule. The interior of the tube forms the gastral cavity.
The inner <>r gastral layer is supported by a special skeleton of Urge quadriradiates
regularly arranged, with the basal ray pointing downwards and the apical ray
projecting into the gastral cavity. Outside the gastral layer stand the flagellated
cliaml>ers, pointing radially outwards and linked into a meshwork pattern. They
communicate with the gastral cavity either independently (D. chartaceum) or in
groups through excurrent chambers (D. hodgsoni). The distal ends of the flagellated
chambers are covered by a reticulated cortex, which follows the " linked " pattern of
the chambers. Many of the dermal triradiates lie in inclined positions, more or less
eml>raciiig the rounded tops of the flagellated chaml>ers. It thus happens not

22 0. F. JEXKIN.

infrequently that spicules may be found with one of the paired rays pointing almost
radially inwards, and at first sight suggesting that sub-dermal triradiates are present
(cf. Fig. 59). The peculiar shape of the dermal triradiates, however, makes it easy
to prove that it is always a paired ray, and not the basal ray, which is centripetally
directed.

DERMATRETON CHARTACEUM.
(Plate XXVIL, Fig. 5, and Plates XXXI. and XXXII., Figs. 59-64.)

There is only one specimen of this new species in the collection. It was found at
a depth of 180 fathoms. It is nearly straight, 23 mm. long x 3 mm. maximum
diameter, quite white as preserved in alcohol, and has the appearance of a delicate paper
tube (see Fig. 5). The walls are only 0 • 32 mm. thick at their thickest point. The
reticulated pattern on the surface is too small to be^readily seen with the naked eye.

The structure of the body wall is shown in Fig. 59. The flagellated
chambers open directly into the gastral cavity through contractile apopylcs (see
Fig. 60). The skeleton of the flagellated chambers consists normally of a single
row of sub-gastral triradiates, but there is occasionally a smaller triradiate half-way
between the gastral and dermal layers. There are also fine hair oxea, not shown in
the figure, surrounding the flagellated chambers. The distal ends of the flagellated
chamber skeletons are strengthened by oxea, which project a short distance beyond the
dermis. The distal ends of the chambers are covered by a reticulated cortex with a
skeleton formed of curved triradiates (see Fig. 63).

The gastral skeleton (Fig. 62) is formed of large quadriradiates, among which
there are a very few which appear to have no apical ray.

The oscular skeleton is shown in Fig. 61. The edge, which is slightly
everted, is formed of small quadriradiates. There is a very short fringe of hair oxea,
amongst which are a few stout oxea similar to those in the body wall. The dermal
triradiates extend up to the edge.

Spicules (Fig. 64).

Oxea. — There are two sorts of oxea : —

(a) Projecting oxea, slightly crooked, slightly hastate, 130/u to 270 n long

x 11/u to 16/x thick.
(ft) Hair oxea about 1 /j. thick.

Triradiates. — The triradiates are of two sorts : —

(c) Alate sub-gastral triradiates forming the tubar skeleton. Basal ray straight,
tapering to a moderately sharp point, 320 p. to 700 /« long x 9/* to 10/u
thick. Paired rays nearly equal, slightly crooked, 100,u to 160/« long
X 9 M to 12 M thick, bluntly pointed. Oral angle variable from 170° to
120° ; strongly folded ; angle of fold 120° to 130°.

CALCAREA.

(d) Alate trinulintcs from the dermal corUsx. Basal ray nearly straight, bluntly
pointed, 80 /» to 140 M long x DM to 10 M thick. Paired rays unequal,
slightly curved upwards, often longer than the basal ray, 65 it to ISO/*
long x SM to 10 M thick Oral angle 100° to 110°.

The Qiinilrmulitites are of one sort :—

(f) A late quadriradiates from the gastral layer. Basal ray straight, tapering
uniformly to a sharp point, 500 M to 700 M long x 10 M thick. Paired rays
usually equal, slightly bent in either direction, 100 M to 230 M long x SM
to 11 M thick. Oral angle 130°. Apical ray nearly straight, sharply
pointed, 70 M to 100 ft long x 6/u thick.

Oscular spicules. — The fringe is formed of oxea of types (a) and (b).
The oscular edge is formed of small quadriradiates similar to (f), but smaller.
Basal ray lOOyu long x 8/1 thick. Paired rays about 100/x lotig.

DKRMATRETON HODOSONI.*
(Plate XXVII., Fig. 1, and Plate XXXII., Figs. 65-74).

There is only one specimen of this new species in the collection. It is lient and
irregularly swollen (see Fig. I), the length being 60 mm. and the maximum diameter
14 mm., tapering at both ends to about 3 mm. It is quite white as preserved in
alcohol. To the naked eye the surface appears to be smooth and minutely reticulated
The walls are delicate, only 1 mm. thick at their thickest point

The structure of the body wall is partly shown in Fig. 73. The flagellated
chambers open through contractile apopyles into excurrent chambers, three or four into
each, and these in turn communicate with the gastral cavity through large irregular
ports. The excurrent chambers have no proper skeleton, but are merely spaces left
between the proximal ends of the chaml>ers and the gastral layer. Four of the
flagellated chambers opening into an excurrent chamber arc .shown in Fig. 66.
The skeletons are drawn and the interior linings indicated by dotted lines. A cross
section of the same excurrent chamber is shown diagrammatically in Fig. 67. The
form of the excurrent chambers varies widely, they are often much deeper than the
one illustrated. The flagellated chaml>ers have ordinary articulated skeletons of many
joints, strengthened at the distal ends by oxea, which project a short distance beyond
the dermis, and also by long hair oxea, which project with the thicker oxea and extend
inwards nearly to the gastral layer. The " linked " arrangement of the flagellated
chambers is shown in Figs. 68, 69 and 70. Fig. 68 shows the ports in the gastral layer,
Fig. 69 the flagellated chambers just above the gastral layer, and Fig. 70 the same
chambers higher up, arranged in a "linked" pattern. All three figures are drawn

• The author venture* to name this specie* after Mr. T. V. Hodgson, the biologist of the Expedition, to
whose untiring industry and ingenuity the magnitude of the collection is due.

vou iv. 2 L

24 C. F. JENKIN.

from corresponding portions of a series of tangential sections. Several of the chambers
are numbered, and may be traced from one section to another. The distal ends of the
flagellated chambers are covered by a reticulated cortex with a skeleton formed of
curved triradiates (see Fig. 71).

The gastral skeleton, formed of large quadriracliates, is shown in Fig. 72.

The oscular skeleton is shown in Fig. 65. The quadriradiates at the edge
are smaller than the gastral quadriradiates. There are a few stout oxea and hair oxea
projecting. The dermal spicules extend to the edge.

Spicules (Fig. 74).

Oxea, — There are two sorts of oxea :—

(a) Projecting hastate oxea, slightly crooked, 150 /A to 240 p. long x 12 p. to

16 /A thick, bluntly pointed at each end.
(i>) Hair oxea of considerable length.

Triradiates. — The triradiates are of two sorts : —

(c) Alate triradiates from the tubar skeleton, of very variable size. The largest

are in the proximal joint of the skeleton, and they diminish towards the
distal end, where they change into type (d). Basal ray straight, sharply
pointed, 330 /A long x 9 /A thick. Paired rays equal, straight, 120 /A long
x 10 p thick. Oral angle of sub-gastral spicules 160°. Oral angle of
spicules from the middle of the tubar skeleton 125°.

(d) Alate triradiates from dermal cortex. Basal ray straight, bluntly pointed,

70 p. to 130 p. long x 10 p. thick, often shorter than the paired rays.
Paired rays equal, straight or slightly curved upwards, bluntly pointed,
forming a short curve in the centre where they join the basal ray, 130 /A
to 140 /A long x 10 LI thick. Oral angle 105° to 110°, strongly folded;
angle of fold 116° to 130°.

The Quadriradiates are of one sort :—

(e) Alate quadriradiates from the gastral layer. Basal ray straight, tapering

uniformly to a sharp point, 500 p. to 850 p. long x 10 p. to llti thick.
Paired rays sometimes very unequal in length, straight, or slightly curved
downwards, 150 p. to 400 p, long x 9 M to 12 p. thick, bluntly pointed.
Oral angle, 135° to 145°. Apical ray sharply pointed, nearly perpendicular
to the facial rays, slightly bent orally near the point, 7 p. or 8 p. thick.

FAMILY CHIPHORID^E. nov.

Definition. — Flagellated chambers elongated, arranged radially around a central
gastral cavity, their ends projecting more or less on the dermal surface and not

CALCAREA. 25

covered over by a continuous dermal cortex. The tubar skeleton is articulate, the first
Joint Itfinij fortntd of chiactines.

The above only differs from Dendy's definition of Syctttidse by the addition of the
words in italics.

STRITTOOONUS.

Chiphorids in which the radial chamlwrs are usually more or less united at places
where they come in contact with one another, and are always crowned at the distal
extremity with tufts of oxeote spicules. The tubar skeleton is articulate, the first
joint being formed of chiactines.

The definition of this new genus in the Chiphoridse corresponds to that of Sycon
in the Sycettidee.

STREPTOCONUS AUSTRALIA
(Plate XXVII., Fig. 3, and Plates XXXII. and XXXIII., Figs. 75-80.)

There are three specimens of this new species in the collection ; all are small,
shaped like a Florence flask and covered with long projecting oxea, sec Fig. 3. They
arc white as preserved in spirits.

Their dimensions are between 7 mm. x 3 mm. and 10 mm. x 4 mm.

The gastrnl cavity terminates in a long oscular collar 2 mm. to 3 mm. long,
with slightly everted edge crowned with a fringe of slender hastate oxea.

The Inxly wall is made up of the radial flagellated chambers (sec Fig. 75 ), which
touch each other for most of their length and have large projecting distal cones
crowned by tufts of long slightly l)ent hastate oxea ; these oxea converge in cones over
the chambers, and then, crossing spirally, spread out and interlace. Each chamber
opens separately into the gas t ml cavity. In tangential section the chambers appear
more or less regularly arranged as hexagons with triangular intercanals.

The Skeleton. — The whole gastral cavity and oscular tul>e is lined with a sparse
layer of large quadriradiates (see Figs. 77 and 79), regularly arranged with the *Ttiiml
ray pointing downwards and the apical ray projecting into the gastral cavity ; the apical
ray is slightly bent orally. A few of these large radiates appear to lack the apical ray.

The radial chambers have articulated skeletons (see Fig. 75), the proximal joint
consisting of chiactines, all of which have their apical rays turned towards the
flagellated chamber, so that when looking at the gastral surface the apical rays appear
to form a protection to the apopyle (see Fig. 79). The upper joints of the articulated
skeleton of the flagellated chamber consist of triradiates of ordinary form. The
unpaired rays of the outermost triradiates project, with the oxea, forming the bases
of the large distal tufts.

The oscular collar has a very fragile skeleton consisting of large qu ad ri radiates
widely spaced (see Fig. 77). On the outside it is thinly covered with long
oxea which project downwards and outwards at an angle of about 45° from the axis.

2 L 2

26 C. F. JENKIN.

The rim of the tube is formed by a ring of small quadriradiates regularly and closely
packed, with the unpaired ray downwards and apical ray inwards. The rim is
crowned by a fringe of long nearly straight hastate oxea. (See Fig. 78.) The sharply
defined line of the small quadriradiates forming the rim of the collar, standing
above the widely spaced quadriradiates which form the lower part of the skeleton,
produces the effect of a vacant space with no spicules between the two. The space
is however no wider than between other rows of the collar skeleton, and occasionally
one of the big quadriradiates is found in the rim itself. No diaphragm can be seen.

Fig. 80 represents a longitudinal section through the junction between the
oscular collar and body. The gastral skeleton is continuous and the flagellated
chambers grow outside it. The first is very minute, with no skeleton ; the second is
larger, with a primitive skeleton. The length of the chambers continues to increase
down to the base of the sponge.

Spicules (Fig. 76).

Oxea. There are three sorts of oxea :—

(a) Stout projecting hastate oxea, usually straight, but occasionally bent

sharply through an angle of about 120°, varying irregularly in thickness ;
300 n to 1,250 /x long x 9 A* to 15 M.

(b) Thin projecting oxea, quite straight, faintly hastate, maximum size

1,020 ju, x 6yu.

(c) Small hastate oxea 140/x x 8 A*.

Triradiates. There are two sorts of triradiates :—

(d) Large alate triradiates from the gastral layer. Basal rays straight,

tapering uniformly to a sharp point from 190/u to 300 /x long x 6/u to 8/t
thick. The paired rays are nearly equal, nearly straight, from 80/t*tol50jw
long X 6 M to 8 /X thick. Oral angle 130° to 140°.

(/) Small alate triradiates from the tubar skeleton. Basal ray straight,
100/u x 6M, tapering to a sharp point. Paired rays irregularly bent,
70/u x 6/x. Oral angle 120° to 135°. Considerably folded.

Quadriradiates. There are two sorts of quadriradiates :•—

(g) Large alate quadriradiates from the gastral layer. Basal ray straight,
tapering uniformly to a sharp point, 230 A* to 430 M long x 6/* to 12 /x
thick ; the larger sizes occur in the oscular collar. Paired rays bracket-
shaped, nearly equal in length, 140/u to 215/x long x 8 M to 10/x thick.
Apical ray thin and sharply pointed, bent orally. Apparent size seen
facially, 32 M x 4ju. Oral angle varying from 110/u in body to 150° in
oscular collar.

(A) Small chiactines from the tubar skeleton. Basal ray, 140/« to 220 n
long x 4 it to 6 tt thick, tapering uniformly to a sharp point, appearing

r )

CAIA'AREA. 27

straight in facial view, but often slightly U'lit in side view at a point a
short distance from the junction with the paired rays. Tliis ln-inl i>
eonnected with the close manner in which jours of these sjiicules lie
together. Paired rays lx;nt slightly downwards, nearly equal in length,
sharply pointed, 70 M to 110 p. long x 3ft to 5 ft thick. Oral angle 140°
to 150°. Folding considerable, angle of fold 115° to 125°. Apical ray
thin, 50 p. to 65 ft x 3ft, considerably set over out of line with the Imsal
ray. Angle between apical and basal ray alx>ut 170°.

Oscular ftpicult-H. The fringe flomhtii of :—
(i) Hastate oxea, nearly straight, 240ft to 700ft long x 5 ft to 8 ft thick.

The edge wnthtff of :—

(k) Small quadriradiates. Basal ray straight, uniformly tapering to a sharp
point, 110ft to 170 p. long x 5 p. to 8 ft thick. Paired rays equal,
considerably curved downwards, sharply pointed, GO ft to 100 ft long x SM
thick. Oral angle about 160°.

HYPODICTYON.

Chiphorids in which the radial chamtars with freely projecting distal cones are
linked so as to form a reticulated pattern round the large intcrcauals. The tulmr
skeleton is articulate, the first joint being wholly or partially formed of chinctines.

This new genus in the family Chiphoridee corresponds to the genus Terithrenodes in
the family Sycettidx.

There is only one species in this genus.

HYPODICTYON LONGSTAFFI.
(Plate XXVIL, Fig. 10, and Plates XXXIII. and XXXIV., Figs. 81-97.)

There are four specimens of this new species in the collection. All are in the form
of longish tubes more or less bent, stouter in the middle and tapering towards each
end ; the upper ends terminate in short, smooth, oscular collars, and the lower ends
are rounded off. The whole sponge, except the oscular collar, is covered by a
meshwork of projecting spicules which give the surface a bristly honeycombed
appearance (see Figs. 10 and 81).

The dimensions of the four specimens are as follows :—

Length. Maximum diameter.

60 nun. .... 15 nun.
37 mm. .... 7 mm.

50 mm. .... 5 mm.

37 mm 9 mm.

All the specimens are white in colour as preserved in spirits.

C. F. JENKIN.

The body walls are formed of the long branching flagellated chambers
(see Fig. 89), which are roughly oval at their proximal ends and are packed
closely together on the gastral cortex ; each chamber opens directly into the gastral
cavity through an irregular apopyle. Following the flagellated chambers outwards
by means of serial tangential sections (Figs. 83 to 87 and 92-97), it will be seen that
the chambers usually divide into two branches close above the gastral layer, and that
these branches often divide a second time near their distal ends. At the same time it
will be seen that the irregular arrangement of the flagellated chambers on the gastral
layer gradually changes till at the outer surface it assumes the regular linked pattern
which is typical of the genus.

The Skeleton. — The gastral skeleton consists almost entirely of large tri- and
quadriradiates irregularly arranged, with the basal ray pointing more or less down-
wards (see Fig. 91). In addition to these there are the paired rays of the chiactines
which occur occasionally round the apopyles.

The gastral skeleton continues into the oscular collar and is modified near the
oscular rim, as shown in Fig. 88. The junction of the oscular collar and
body wall is shown in Fig. 82, in which the gradual growth of the flagellated
chambers may be seen. Near the rim the triradiates almost or entirely disappear,
and the quadriradiates are much smaller. There is a fringe of hair oxea (these are
absent in one specimen) amongst which are a few stout oxea, apparently of the same
sort as those on the distal ends of the flagellated chambers. There are a few thin
bent oxea scattered irregularly near the edge.

The skeleton of the flagellated chambers is of the ordinary articulated type,
consisting of small triradiates, which converge at the top of the chamber to form a
point which is strengthened by a bundle of oxea of two types which project freely.
Every here and there in the proximal joint a chiactine may be found replacing one
of the ordinary triradiates ; its apical ray projects diagonally into the gastral cavity
and across the apopyle. It is the presence of these chiactines which has made it
necessary to form the genus Hypodictyon for this species instead of putting it in the
genus Tenthrenodes.

Spicules (Fig. 90).

Oxea. — There are two sorts of oxea : —

(a) Projecting oxea from the distal cones, slightly irregularly bent, some
slightly hastate, from 100 M long x 12 M thick to 900 M long x 26 M thick.
(V) Hair oxea projecting with (a).

Triradiates. — There are two forms of triradiates :—

(c) Large alate triradiates from the gastral layer. Basal ray straight, tapering
uniformly to a sharp point, maximum size, 800 /x X 10 M- Paired rays,
nearly straight, of uniform thickness all along, bluntly pointed, usually

CALCAREA. 29

of unequal lengths, ISO/* to 430 M long x 10 M to 16 M thick. Oral ungle
about 135°.

(</) Smaller alate triradiates from the tubar skeleton. Basal ray straight,
tapering to a sharp point, 140 M to 260 M long x 8 p. to 10 p, thick.
Paired rays nearly equal and straight, 80 /i to 150 /x long x 8/t to 10 p.
thick, slightly folded. Oral angle 130°.

Quadriradiatea. — There are two sorts of quudriradiates :—

(f) Large quadriradiatcs from the gastral layer. Basal ray straight, tapering
uniformly to a sharp point, maximum length 1,000/i x 9/i to 12/x.
Paired rays of unequal length, often slightly crooked, of irregular thick-
ness, bluntly pointed, 220 p. to 420 p. long x 1 1 p. to 13 /A thick. Apical
ray sharply pointed, bent orally, about 80 ft x 12 p. Oral angle 128°
to 138°.

(f) Chiactines. Similar to (</) with the addition of an apical ray 90 p. long x
G p. thick, sharply pointed. These are not shown in the drawing of
spicules.

Oscular spicules : —

(y) Fine straight oxea from the fringe, 3 p. to 4 p. thick. All are broken off, so

their length is uncertain. Among them are a few stout oxea of the same

sort as (a).
(A) Thin zigzag oxea lying tangentially near the oscule. 170 p. to 300 p. long x

3 p. to 4 p. thick, sharply pointed at both ends,
(i) The edge is formed of small quadriradiates. Basal ray straight, tapering

to a sharp point, 130 p. to 200 p. long x 9/t to 12 p. thick. Paired rays

nearly equal, considerably bent downwards, 90 p. to 150 p. long x 9 p. to 12 p.

thick. Oral angle 150° to 160°. Apical ray short and sharp, about

40 p. long.

FAMILY STAURORRHAPHID^.

Definition. — There is a distinct and continuous dermal cortex, completely covering
over the chamber layer and pierced by inhalent pores. There are no subdermal sagittal
triradiates nor conspicuous subdermal quadriradiates. The flagellated chambers vary
from elongated and radially arranged to spherical and irregularly scattered ones, while
the skeleton of the chamber layer varies from regularly articulate to irregularly
scattered, but there are always regularly disposed subgastral chiactines.

The above only differs from Dendy's definition of Grantiidse by the addition of
the words in italics.

30 C. F. JENKIN.

ACHRAMORPHA.

Staurorrhaphids in which the elongated flagellated chambers are arranged radially
round the central gastral cavity ; they are covered over by a dermal cortex, composed
principally of triradiate spicules, and without longitudinally disposed oxea. The
tubar skeleton is articulate, the ifirst (and sometimes only) joint being formed of
chiactines.

The definition of this new genus in the family Staurorrhaphidse corresponds to
that of Grantia in the Grantiidse, but omits the limitation introduced by Dendy, which
transfers all Grantise with tufts of spicules, including Grantia compressa, to the
family Sycettidse. This limitation appears to be artificial and unsatisfactory.

The genus contains three species, Achramorpha nivalis* Achramorpha glacialis*
and Achramorpha grandinis*

In all the species the flagellated chamber skeleton consists of the basal rays of the
chiactines. There is no gastral skeleton proper ; the gastral layer is supported by the
paired rays of the chiactines, except in the oscular tube, where there are quadriradiates
lying tangentially in the ordinary position, i.e., in the gastral layer, with the basal ray
pointing downwards. At the junction between the oscular tube and the body, the
quadriradiates lie at all angles intermediate between the centrifugal position and the
tangential position. This remarkable change in position of the quadriradiates
in passing from the oscular tube to the body, which is conspicuous in this
genus, is discussed in Part III.

The skeleton of the dermal cortex consists of triradiates lying tangentially with
the basal ray, pointing more or less downwards ; there are also large projecting oxea.
In one species there are also projecting hair oxea, and in the other two there are
minute prickly hastate oxea projecting round the ostia.

In two of the species the skeleton of the oscular edge is well differentiated. It
consists of a few rows of special quadrira'diates closely and regularly packed, and
partially hidden in the dense fringe of oxea. In the top row these quadriradiates have
remarkably snub apical rays, which usually hardly penetrate through the fringe.
From these apical rays springs the diaphragm which closes the oscule (Figs. 100 & 102).
This diaphragm can be made out in some specimens, but not in others, probably owing
to defective preservation.

In the third species, A. grandinis, the oscular skeleton is hardly differentiated
from the body skeleton and there is no fringe.

In all the species the flagellated chambers open one or more together into small
shallow excurrent chambers, which open into the gastral cavity through ports in the
gastral membrane.

* Ignis, grando, nix, glades spiritus procellarum quir faciunt verbum ejus.

CAI.C.\I:I:A 31

ACHKAMORI'HA OLACIALIK.
(Plate XXXIV., Figs. 98-102.)

There are six specimens of this new species in the collection. All are small,
slender, vase-shaped Sponges, which when examined with the naked eye might easily
lie mistaken for Ascons. In colour they are white tinged with orange, as preserved in
alcohol. The dimensions of the four perfect specimens are as follows : 9 mm. x 2 mm. ;
4 mm. x '75 mm. ; 8 mm. x 1'5 mm. (flattened) ; 9 mm. x .1$ mm.

The structure of the body wall is shown in Fig. 101. The flagellated chamliers
often taper considerably at their distal ends, thus leaving rather large inhalent
chambers under the dermal cortex. The hair oxea lie in the very thin mcsoderm
tatween the chambers and project beyond the dermal cortex. Neither they nor the
large projecting oxea are arranged in regular tufts or cones, but appear to be
irregularly scattered.

The structure of the oscule is shown in Figs. 100 and 102. The section shows
the transition from normal to centrifugal position of the quadriradiates. It also
shows the diaphragm.

Fig. 98 represents a view of the gastral layer. The section includes some of
the gastral ports and some of the apopyles which lie immediately above the ports.
There are a very few minute oxea lying radially in the gastral layer. One of these is
shown in Fig. 102. In some specimens these minute oxea are scattered throughout
the body wall. Several of the specimens contain ova of the usual form.

Spicules (Fig. 99).
Oxea. — The Oxea are of three sorts : —

(a) Stout projecting oxea, irregularly bent, sharply pointed at the inner end,
and bluntly pointed at the outer end, 280 /* to 440 ft long x 12 ft to 24 /*
thick.

(A) Minute, hastate, slightly bent oxea, some smooth, some spined, from the
dermal layer, 35 /* to 40 ft long x 2 ft, to 3 /* thick.

(c) Hair oxea, 400 /* or more long.

The Triradiales are of one sort : —

('/) Alate triradiates from the dermal cortex. Basal rays straight, tapering and
sharply pointed, 200 /x to 380 ft long x 12ft thick. Paired rays equal,
nearly straight, 130ft to 180/x long x 10ft to 14 ft thick. Oral angle
varying widely, 110° to 160°. Folding very variable.

(e) ' Chiactines. Basal ray straight, tapering, sharply pointed, 340ft to 400 p.
long x 15ft thick. Paired rays equal, 130 /A to 180 ft long x 16 ft thick,
bent slightly irregularly. Oral angle, 145° to 165°. Angle of fold, 150°

YOU. IV. g M

32 C. F. JENKIN.

to 160°. Apical rays sharply pointed, slightly irregularly bent, 70 ft to

WO p. long X 12 ft thick; set-off slight; angle between apical ray and

basal ray 165°.
Oscular Spicules. — The fringe consists of hair spicules, and stout oxea of types

(c) and (a) :—
(/) Small quadriradiates from the oscular edge. Basal ray straight, tapering,

and rather bluntly pointed, 100 ft to 120ft long ftx9fitolOft thick.

Paired rays about equal, bent downwards, 50 ft to 100ft long x 7 ft to

8 ft thick. Oral angle about 150°.
(cf) The quadriradiates from the oscular collar are similar to (/), but much

larger. Basal rays 220 p- or more long ; paired rays up to 200 /A long ;

apical rays up to 120 p. long.

ACHRAMORPHA GRAKDINIS.

(Plate XXVII., Fig. 4, and Plates XXXIV. and XXXV., Figs. 103-104.)

There is only one broken specimen of this new species in the collection ; it consists
of the upper part and oscular collar of a sponge which probably was of considerable
size, judging by the size of the oscule (see Fig. 4). The specimen is straight, 18 mm.
long, and circular in section, tapering from 4£ to 3 mm. in diameter at the oscular
end. The lower part is covered by long oxea, which lie along the surface pointing
downwards ; the upper part is smooth and transparent. There is no fringe.

The structure of the body wall is the same as in the other species of this genus.
There are no hair oxea, but there are a few small prickly hastate oxea in the dermis.

The oscule (see Fig. 104) differs considerably from the other two species. There
is no fringe of hair oxea and no special ring of quadriradiates at the edge, the skeleton
being made up of both tri- and quadriradiates lying tangentially. The transition from
tangential quadriradiates to chiactines occurs in the same way as in the other species.
The basal rays of the chiactines project a long way outside the dermis.

Spicules (Fig. 103).
Oxea. There are two sorts of Oxea : —

(a) Very large straight projecting oxea, size of longest fragments 3' 5 mm. x

23 ft. Very sharply pointed at the inner end.

(b) Small hastate oxea, from the dermis. Slightly irregularly bent, some

spined, some smooth, 65 ft to 120 ft long, 3 ft to 6 p, thick.

The Triradiates are of one sort : —

(c) Large, alate triradiates from the dermal cortex. Basal rays straight,

tapering uniformly to a sharp point, 420 ft to 500 /* long x 12ft to 15 ft
thick. Paired rays usually equal, straight or bent slightly downwards,

CALCAREA. 33

bluntly pointed, 220 /t to 260 /i long, x 12/i to 14 ft thick. Oral angle
130° to 160°.

(</) Chiactines. Basal rays straight, tapering to a more or leas sharp point,
450/i to 550 /t long x 12 ft to 16 /i thick. Paired rays equal, straight,
240/i to 270/i long x 12/i to 14/i thick. Oral angle, 155° to 160°.
Apical ray sharply pointed, maximum length IGO/i long x 14/i to 16 /i
thick. Angle between apical ray and basal ray, 165° to 180°.
Oscular gpiculcs.

(e) Quadriradiates from the oscular tube. Basal rays straight, tapering
uniformly to a sharp point, 650 /i long x 12/i thick. Paired rays often
unequal, maximum size 560 /i long x 14 /i thick, nearly parallel, bluntly
pointed. Oral angle about 160°. Apical ray sharply pointed, curved
orally, 1 20 /i long x 9 /i thick.

*

ACHRAMORPHA NIVALIS.

(Plate XXVII., Figs. 7 and 8. Plates XXXV. and XXJCVI., Figa 105-112.)

There are fourteen specimens of this new species in the collection. They vary
considerably in shape, but are mostly more or less pear-shaped, with the oscule at the
narrow end (see Figs. 7 and 8). In some specimens the neck of the pear is considerably
extended, but the actual oscular collar (without flagellated chambers) is always quite
short There is a dense fringe of long shining silver-white hair spicules round the
oscule, which sometimes attain a length of 2*5 mm. The whole sponge is covered
with long projecting oxea which lie in all directions, giving it a very untidy appear-
ance. The base of the sponge is rounded.

The dimensions of the perfect specimens are as follows : — 12 mm. x 3 mm.
17 mm. x 3 mm. ; 18 mm. x 9 mm. ; 20 mm. x 8 mm. ; 12 mm. x 4 mm. ; 8 mm.
X 4 ram. ; 11 mm. x 5 mm. ; 17 mm. x 8 mm. ; 8 mm. x 2£ mm. ; 26 mm. x
12 inia. ; 29 mm. x 10 mm.

The structure of the body-wall is shown in Fig. 106, where the subdermal
cavities and exhalent chambers may be seen. The small projecting oxea are mostly
arranged round and over the ostia. This may be seen in Figs. 107 and 108 repre-
senting the dermal cortex in plan and section. The very large projecting oxea are
scattered quite irregularly, and often project through the body-wall into the gastral
cavity. The large oxea are all broken ; the longest fragment found is 2 • 7 mm. long.
They lie at all angles. Fig. 109 represents the gastral membrane as seen from inside.

Figs. Ill and 112 represent the oscular structure. The highest flagellated
chambers are shown in Fig. 112, and the short oscular collar projecting about
1*3 i ii in. beyond them. Fig. 112 shows the junction of the oscular tube and
the body to a larger scale. Above the last chamber the skeleton consists of quadri-
radiates and oxea only ; the triradiates forming the dermal skeleton do not extend

2 M 2

34 C. F. JEKKIN.

beyond the last flagellated chamber. At the junction there are specially large quadri-
radiates lying diagonally, apparently strengthening the junction between the dermal
and gastral skeletons. The basal rays of the chiactines are much longer than the
thickness of the body in the upper part of the sponge and project, often half their
length, outside the cortex. The regular arrangement of the oscular quadriradiates at
the oscular edge is shown in Fig. 111. Side views of these spicules are drawn,
showing the diminution in the length of the apical ray. These spicules lie in the
middle of the fringe, so that they are almost hidden from both inside and outside. In
the same figure is shown one of the large quadriradiates from the junction between
tube and body.

There are numerous ova in some of the specimens.

Spicules (Fig. 105).

O.rea. There are three sorts of Oxea :—
(a) Large projecting oxea, straight, and pointed at both ends. Largest fragment

found, 2'7 mm. long x 14 p, thick.
(/>) Minute, spined, slightly crooked, hastate oxea from the dermal layer, 75 p.

te 85 fj. long x 5 p, to 6 p. thick.

(c) Rather longer, small, straight hastate oxea, 120 p. to 140 p. long x 4/i thick.

The Triradiates are of one sort : —

(d) Alate triradiates from the dermal layer (these are probably the same as

those occasionally found in the body- wall). Basal ray straight, tapering to
a point, 200 /* to 380ft long x 8/u, to 10 p. thick. Paired rays nearly equal
in length, curving slightly upwards, forming a rounded bend at the centre
140 p. to 210 p. long x 8 ft thick. Oral angle 102° to 106°. There are
a very few triradiates with an oral angle of 160°. It is doubtful where
these come from.

(e) Chiactines. Basal ray straight in facial view, but bent slightly in side view,

tapering to a sharp point, 400 p, to 600 p. long x 8/u, to 10 p. thick. Paired
rays about equal, sharply pointed, nearly straight, 160 p, to 200 p, long x 8 p.
to 12 p. thick. Oral angle 160°. One of the paired rays is sometimes
deflected from its usual position into line with the opposite paired ray.
(T type of triradiate system.) Apical ray straight, or slightly bent,
tapering uniformly to a sharp point, 110/u. to 130 p, x 8/u to 12 //, thick.
Set-over about equal to the thickness of the ray. Angle between basal
. ray and apical ray 150° to 155°.

Oscular spicules.

(/) Long, straight slender oxea forming the fringe, 2' 5 mm. long x G p. or less
thick, minutely hastate at the outer end.

CAI.C\I:K\ 35

•

(</) Ijirgc alato quadriradiates, lining the oscular tulx?. These arc largest at
the junction with the Ixxly-wall. liasal ray straight, tapering uniformly
to a very sharp point. Maximum dimensions 850 ti long x 12 /x thick.
Paired rays bracket-shaped, equal in length, 450 /i long, oval in section,
16 /i deep x 6/1 thick. Oral angle, 155°; folded angle, 150°. Apical
ray curved orally, 100/i long x 10 it thick.

These get smaller up to the oscular edge, where their dimensions are as follows :
l>asal ray, 150/t long. Paired rays, which are bent downwards, 120 /A long
X 10/* thick. Apical ray reduced to a blunt cone, 20 /x long x IG/i thick.

SUBORNUS GRANT10PSIS Dendy.

The sponge has the form of a greatly elongated hollow tultc whose wall is
composed of two distinct layers of about equal thickness. The outer (cortical) layer
is provided with a very strongly developed skeleton of triradiate spicules, and is
penetrated by narrow ramifying incurrent canals. The inner layer is formed by
elongated radial chambers arranged very regularly side by side. The skeleton of the
inner layer is very feebly developed. The tul>ar skeleton is articulate and composed
of very abnormal sagittal triradiates, whose paired rays are greatly reduced ; the inner
jirint of the tubar skeleton consists of chiactines.

The above is slightly modified from Dendy 's definition (7, p. 73).

The sub-genus contains only one specie*.

GRANTIOPSI.S CYLIXDRICA Dendy.

(Plate XXXVI., Fig. 113.)
Grantioptu rylituirira IK-ndr (7), p. 90.

A single specimen of this species found in Australia was described by Dendy
(7, p. 90), who called special attention to the chiartiues in its skeleton. The structure
of this species differs considerably from that of the other species in the genus
Achramorpha, so that it seems advisable to retain the sub-genus, Graiitiopxis, only
transferring it from the Grantiulte into the Slatavrrkepkidtf.

The spicules, which have not hitherto been illustrated, are shown in Fig. 113;
they differ widely from any found in the new species. The sponge is fully dcscril»ed
in (7) and figured in (2).

MEOAPOGON.

Staurorrhaphids in which the flagellated chambers are spherical or sac-shaped,
never arranged radially around the central gastral cavity, with which (or with the
main excurrent canals derived therefrom) they communicate by a more or less
complicated excurrent canal system. The skeleton of the chamber layer is largely

36 C. F. JENKIN.

composed of irregularly scattered radiate spicules, but it always lias regularly placed
subgastral chiactines.

The above only differs from Dendy's definition of Leucandra by an alteration in
the last sentence.

This genus contains five species, four new, and PolejaefFs " Leuconia cruel/era "
(Megapogon cruciferus), which is now transferred to it.

All the species have well-marked gastral cavities, but no gastral skeleton
proper. The spiculation in all the species is very similar. The size and shape
of the flagellated chambers is very variable ; they are sometimes so long as to
resemble the radial chambers typical of Grantia or Achramorpha, and at other times
they are spherical.

The close similarity between Megapogon cruciferus found at the Azores and the
new species from the Antarctic is remarkable. It is noteworthy that Megapogon
cruciferus was dredged from a great depth, 450 fathoms. The temperature of the sea
was not measured at the station where it was found, but judging from the
temperatures measured at neighbouring stations it must have been about 47° F. at
the bottom. It is possible that successive larvae of this sponge may have travelled in
cool water at great depths all the way from the Antarctic to the Azores.

MEGAPOGON CRUCIFERUS Pol.
(Plate XXXVI, Fig. 114.)
Leuconia crucifera Poldjaeff (4), p. 60.

The following description is abstracted from PolejaefFs (4). Only a fragment
of the inferior part of the sponge was found. The outer surface is bristly, the inner
slightly roughened by the apical rays of the gastric quadriradiate spicules [chiactines].
These are all more or less cruciform, all the rays lying in the same or almost the same
plane. By this characteristic the species can be very easily distinguished from all
other Leuconidse. Leuconia crucifera and Leuconia blanca are of peculiar interest, as
forms inhabiting the greatest depth (450 fathoms) from which Calcarea have been
hitherto obtained.

Skeleton.— The skeleton consists of gastral quadriradiates [chiactines], of
parenchymal triradiates [of minute spined oxea *], of dermal triradiates and of stout
acerate spicules, piercing the parenchyma obliquely and projecting with their free end
from the outer surface, and of slender acerate spicules scattered here and there on the
outer surface in small bundles. The spicules (omitting the hair spicules) are shown in
Fig. 114 drawn from spicule preparations made by the author. For description and
dimensions see (4).

* These spicules were overlooked by PoWjoeff. They are included in the new drawing of the spicules.

' ALCARBA. 37

MfCUAPOGON VILLOHUS.

(Plate XXXVI., Figs. 115-119.)

There are four specimens of this uew species in the collection, but only one ia
perfect, the other three being broken pieces only. All are white as preserved in
alcohol, and are covered with a dense mat of very long oxea. The following
description applies to the perfect specimen (see Fig. 117).

The sponge is in the form of a nearly straight circular tube, a good deal distorted
at the base, where it has its maximum diameter. It tapers gradually to the short
oscular collar which terminates the large gastral cavity.

The structure of the body-wall is shown in Fig. 115 ; it gets gradually thinner
towards the oscule.

Canal System. — The dermis is pierced by numerous ostia which lead into
luge incurrent canals which extend radially inwards nearly through the body-wall.
The excurrent system consists of large radial chambers or canals which do
not appear to be branched, each opening through a large port into the gastral
cavity.

The flagellated chambers are very variable in size and shape ; they are mostly
more or less thimble-shaped, opening by apopyles directly into the excurrent chambers
round which they are grouped. Some of the flagellated chambers are nearly spherical,
some are so long that they reach almost across the whole thickness of the wall ; these
long chambers are often branched.

Skeleton. — The skeleton of the body-wall is made up of the chiactines
which reach completely through it and often project, and a few scattered triradiates
whose basal rays also often project It is pierced at variable angles by the very long
projecting oxea. All these oxea are broken in the specimens ; the longest fragment
measured is 15 mm. long. Chiactines occasionally occur at the top of the excurrent
chambers, and may thus appear to be in the middle of the thickness of the wall. There
is no gastral skeleton proper; the only supports to the gastral wall (see Fig. 118) are
the paired rays of the chiactines, amongst which are a few minute oxea, some smooth
and some spined. The dermal skeleton is formed of a thick layer of triradiates
(Fig. 116) lying in all directions, with a few minute oxea standing semi-erect
round the ostia.

Oscule. — The oscule is at the end of an oscular collar about 1 mm. long which
has the appearance of a thin, almost transparent extension of the body. The skeleton
of the oscular collar consists of quadriradiates lying tangentially on the inside, with
the basal ray downwards and the apical ray projecting into the gastral cavity. Outside
are triradiates, also regularly arranged, with the basal ray downwards. The edge is
made up of both types of spicule, with a few special oxea which lie horizontally, but
do not project beyond the edge, which has no fringe. The junction between the
oscular collar and the body-wall is very similar to that shown in Fig. 110.

38 C. F. JENKIN.

Spicules (Fig. 119).
Oxea. — There are two sorts of Oxea :—

(a) Enormous straight projecting oxea more than 15 mm. long x 36 p. to 43 p.

thick. The inner end tapers gradually to a very sharp point, the outer
end is hastate.

(b) Minute irregularly bent hastate oxea, some spined, some smooth, from the

gastral and dermal layers, 80ft to 160 p. long X 6 /A thick.

The Triradiates are of one sort :—

(c) Alate triradiates from the body-wall and dermal layer. Basal rays straight,

tapering uniformly to a sharp point, 220 p. to 700 p. long x 10 p. to 16 p.
thick. Paired rays equal, nearly straight, 170 p, to 300 p, long X 12 p. to
14 p. thick. Oral angle 135° to 155°.

(d) Chiactines. — Basal ray straight, tapering uniformly to a sharp point from
600 p, to 1120j«, long X 10 p. to 16 p. thick. Paired rays equal, straight,
tapering to a sharp point, 240 p. long x 13 p. thick. Oral angle 155° to
160°. Slightly folded, angle of fold 175° to 150°. Apical ray straight,
slender, sharply pointed, 140/i to 220 p. long X 8 p. to 16 p, thick, set-off
small ; angle between apical ray and basal ray 170° to 180°.

MEGAPOGON RARIPILUS.
(Plate XXXVI., Figs. 120-124.)

There are three specimens of this new species in the collection. The natural shape
appears to be that of a straight tube, enclosing the gastral cavity, slightly thicker in
the middle and circular in section. The lower end is closed very squarely, and the
upper terminates in an oscule without a collar. The largest specimen, which is much
distorted, is 30 mm. x 11 mm. ; the next, which has the shape described above, is
19 mm. x 4 mm., and the smallest specimen 2£ mm. x 1 mm. They are white as
preserved in alcohol, their surface is hispid.

The structure of the thick body- walls is shown in Fig. 122.

Canal System. — The stout dermal cortex (Fig. 124) is pierced by small
ostia leading into small incurrent chambers, from these run incurrent canals branching
and getting smaller as they approach the gastral cavity. The excurrent canals are
connected in groups of three or four to excurrent chambers which open into the gastral
cavity through large ports.

Skeleton. — The body skeleton consists of the chiactines, amongst which are a few
large subgastral triradiates and a few scattered small triradiates ; the basal rays of the
latter often project beyond the dermis. There are also dense tufts of hair spicules with
a few stout bent oxea among them, which project and make the surface hispid. There
are minute spined hastate oxea scattered all through the body. There is no gastral

CALCABEA.

skeleton proper; the gastral cortex is supported by the paired rays of the chiactines
and suhgastral triradiates (see Pig. 120). The dermal cortex is formed of a thick layer
of irregularly scattered triradiates (Fig. 124).

The osc.ule is shown in Fig. 121. There is a ring of small quadri radiates lying
tangcntially round the inner edge, with short pointed apical rays from which springs
the diaphragm. There is a thick fringe of hair oxea, outside which there is an open
fringe of large oxea. The oscule is thick and the Hagellated chambers reach almost to
the edge. Near the oscule they are only in a single layer, but in the middle of the
sponge they are ten or twelve layers thick.

Spicules (Fig. 123).
Oxen. — There are four sorts of Oxea : —

(«) Large projecting oxea, pointed sharply at the inner end, and very bluntly at
the outer end, wluch is thicker. Curved all along, more sharply near the
thicker projecting end. Length fairly constant about 700 p., thickness 30 p.
to 35 ft.

(/<) Hair oxea, projecting in dense tufts, quite straight, length over 500ft,
thickness about 1 ft.

(c) Minute spined hastate oxea, scattered all through the body-wall. Slightly
bent, with the largest spines in a ring round the " set-over " of the spear-
head. Usual length 60 ft, occasionally 150ft, thickness 4 ft to 5 p..

(</) Small irregular oxea from the body-wall al>out 350 p. X 20 ft, often set-over

in the middle of their length.
Tri radiates.— There are two sorts of Triradiates : —

(?) Large alate subgastral triradiates. Basal ray straight, tapering uniformly
to a sharp point, about 700 p long x 16 p. thick. Paired rays equal, nearly
straight, 200 p, to 320ft long X 20ft thick ; oral angle 160°.

(/) Smaller alate triradiates from the dermal cortex (probably the same as the
rare ones in the body-wall). These vary widely in size. Basal ray
straight, tapering uniformly to a sharp point, 170 p. to 700 ft long x 10 ft
to 18 ft thick. Paired rays equal, forming a rounded bend at the centre,
but straight for the rest of their length, 120 ft to 270ft long x 10 ft to 16ft
thick. Oral angle very constant, 108° to 112°.

(y) Chiactines. Basal ray straight, sharply pointed, 600ft to 750 ft long x 14 ft
thick. Paired rays bent more or leas forwards, 200 ft to 280 ft long x 1 6 ft to
20 ft thick. Oral angle variable, 1 35° to 1 60°. Apical ray about 80 ft x 1 2 ft.

Oscular spiculfs.

(h) Large oxea like (a), but shorter ; 380 ft to 480 ft long x 18 ft thick. These
form an external fringe. The inner fringe is formed of hair oxea of
unknown length,

TOL. IV. 2 N

40 C. F. JENKIN.

(i) Small alate quadriradiates forming the edge. Basal ray

straight, tapering uniformly to a sharp point. Paired rays bent slightly
downwards, about 70 ^ x 10 /A. Apical ray short and conical, 20 /x long x
8 /A thick at base, perpendicular to facial plane. Oral angle nearly 180°.

MEGAPOGON POLLICARIS.*
(Plates XXXVII. and XXXVIII. , Figs. 125-130.)

There are two specimens of this new species in the collection ; the larger one is fixed
in osmic acid and is consequently brown ; the smaller one, which is fixed in alcohol, is
white ; both are hispid. The larger specimen is in the form of a flattened tube 15 mm.
long. The diameter at the widest part near the middle is 3 mm. The smaller
specimen is ovoid, 2£ mm. x l£ mm. There is a large gastral cavity which terminates
in the oscule, which in both specimens is bent to one side. The larger specimen has an
oscular collar about 1 mm. long.

The structure of the body -wall is shown in Fig. 126. The space between
gastral and dermal layers is filled up with three or four layers of flagellated chambers.
Each flagellated chamber is surrounded by a slender skeleton of minute prickly hastate
oxea (see Figs. 129 and 127) ; these oxea often lie together in small bunches of
three or four, and as they may lie partly on one and partly on another chamber, they
build up serpentine lines of minute spicules threading about between the chambers.
Some of these minute oxea also lie in the gastral and dermal layers, and also in the
columns supporting the dermis.

Canal System. — There is an incurrent chamber under the dermal layer which is
probably continuous round the whole body of the sponge. The dermal layer is
supported over this chamber by columns of spicules consisting partly of the basal rays
of the chiactines and partly of oxea which project outside the dermis. The incurrent
canals lead inwards from the incurrent chambers. The excurrent canals are connected
in groups to excurrent chambers which open into the gastral cavity through large
ports ; these are more or less surrounded by the apical rays of the chiactines (see
Fig. 125).

The Skeleton. — The dermal skeleton is formed by a dense layer, five or six
spicules thick, of triradiates, pierced at intervals by the projecting bunches of oxea.
The body skeleton consists of chiactines and a very few scattered triradiates. The
basal rays of the chiactines are bunched more or less together at their distal ends and
occasionally project through the dermal layer ; with each bunch are grouped a number
of the oxea, which project with their thicker bent ends about one-third of their length
beyond the dermis. There is no special gastral skeleton ; the gastral cortex (see Fig.
130) is supported by the paired rays of the chiactines, and a few minute oxea
irregularly scattered.

* Pollicaris = "like a thumb," uamed from the shape of the oxea,

CALCAREA. 4 1

The oscular rollar in thick and densely packed with spicules. On the inside arc
(juadriradiates lying tangentially with the Iwisnl rayg downwards, and outside are
triradiates. The rim, which has no fringe, is formed of tri- and <juadriradiatcs, with
oral angles of nearly 180°. There is a diaphragm in one specimen.

Spicules (Fig. 128).

Oxea. — There are three sorts of Oxea :—

(a) Large projecting oxea, straight for about two-thirds of their length, then

bending smoothly to one side. The inner (straight) end tapers gradually
to a point The outer end is blunt, shaped like the end of a thumb.
Most of the oxea are alxmt the same size, 460ft long x 24 p. thick.
Maximum size, 640ft x 25 p. There arc a few smaller, imperfectly formed,
possibly young, oxea.

The minute spined oxea scattered throughout the body and forming the flagellated
chaml»er skeletons are of two forms :—

(b) Straight, refringent, thin, slightly spined, 50ft x 2ft.

(c) Curved, hastate, thicker and well spined, 50ft to 00 ft x 2*3 ft to 3'lft.

The junction of the spear-head is marked by a ring of four or more large
spines.

The Triradiates are of one aort :—

(rf) Alate triradiates from the dermal cortex and body-wall. These vary very
much in size and proportions. Most have the paired rays only slightly
shorter than the basal ray, but occasionally there is great disparity. Hasal
ray straight, tapering uniformly to a sharp point, 130ft to 330 ft long x
8 ft to 15 ft thick. Paired rays equal, nearly straight, but slightly
irregular, 90 ft to 320ft long x 8ft to 14 ft thick. Oral angle 108° to 115°.

(e) C/tiacthieft. — Basal ray straight, tapering uniformly to a sharp point, 400 ft
to 580/x long x 14ft to 18ft thick. Paired rays straight, bluntly pointed,
160ft to 280ft long x 14ft to 16 ft thick. Oral angle 160°. Angle of fold
155° to 165°. Apical ray nearly straight, pointed, 100 ft long x 12 ft
thick. Set-over small. Angle Instwecn apical ray and Itasal ray 175°.

MEOATOGON CRISPATUS.*
(Plat* XXVII., Fig. 2, and Plate XXXVIIL, Figs. 131-136.)

There are four specimens of this new species in the collection, all found together.
The sizes are 9 mm. x 3 mm. ; 6 mm. x 2 mm. ; 6 mm. x 3 mm. ; 4 mm. x 2 mm.
They are vase-shaped, white or pale orange as preserved in spirits, with a curly surface

, early-haired.

2 N 2

42 C. F. JENKIN.

due to the bent projecting oxea (see Fig. 2). The gastral cavity, which is nearly
constant in diameter throughout its whole length, terminates at the oscule, which is at
the top of an oscular collar about l£ mm. long.

The structure of the body-wall is shown in Fig. 131.

Canal System. — There is an incurrent chamber under the dermal layer which is
probably continuous round the whole body of the sponge. The dermal layer is
supported over this chamber by columns of spicules, consisting partly of the basal rays
of the chiactines, and partly of oxea which project in spreading tufts outside the
dermis. The incurrent canals, which are small and irregularly branched, lead inwards
from the incurrent chamber. The excurrent canals are large and extend radially from
the gastral cavity nearly through the body-wall, often branching once. The flagellated
chambers are grouped round these large excurrent canals, the larger number opening
directly into them, the remainder being connected by short branches.

The Skeleton. — The skeleton of the body-wall consists of chiactines. There are
also a few triradiates scattered irregularly, and numerous minute spined oxea. There
is no gastral skeleton proper; the gastral cortex (see Fig. 132) is only supported by
the paired rays of the chiactines. The dermal skeleton is made up of triradiates, which
are approximately equiangular, and lie tangentially to the surface. There are also tufts
of oxea which project for most of their length and give the sponge its characteristic
" curly " appearance.

Oscule. — The thin oscular collar is about l£ mm. long and the same diameter
as the gastral cavity (see Figs. 133 and 134). On the inside are quadriradiates lying
tangentially with the basal ray downwards, and outside are triradiates and oxea.
There are numerous minute spined oxea scattered amongst the radiates. At the
rim (see Fig. 131) the collar is slightly thickened. There is a ring of small
quadriradiates forming the edge on the inside ; they do not appear to be very closely
or regularty packed. The diaphragm extends from the apical rays of these spicules.
There is a scanty fringe of prickly hastate oxea of small size. On the outside the
oscular collar is thickly covered by large bent oxea of the same type as those forming
the tufts on the body, but shorter. They stick in the collar, often projecting through
it into the interior, and are directed upwards usually at about 45° with the axis of the
sponge ; some lie tangentially along the surface.

At the junction of the oscular collar and the body the collar thickens, a few small
flagellated chambers appear, the quadriradiates begin to turn their basal rays outwards,
and the oxea project more radially, a few even pointing downwards.

Spicules (Figs. 135 and 136).
Oxea. — There are three sorts of Oxea :—

(a) Large projecting oxea, irregular but nearly straight for about three-quarters
of their length, then bending (at the outer end) considerably to one side
by two or three angular deflections. The inner end is sharply pointed and

CALTAREA. 43

usually slightly Ix-nt to the opposite side from the outer end, giving the
spieule an 8 Hhapo. The uutor end in slightly swollen uud roughened.
The thickness varies very irregularly throughout the length, 550 ft to
950 /t long X 20 ft to 28 ft thick. Many of those spicules have long
Hattened areas just below the outermost licnd, looking as if the spicule
had been diverted by contact with a. plane surface.

There are minute spiued oxea of two types :—

(b) Straight, refringent, and thin, 50ft long x 1 '2 ft thick.

(c) Curved, hastate, and thicker, 55 ft long x 4 ft to 5 ft thick.

The triradinte* are of two sorts :—

(il) Approximately regular tri radiates from the dermal cortex and body-wall
with straight rays, tapering slightly to near their extremities, then sharply
pointed. The rays and angles are nearly but not exactly equal. Kays
UO/i to 220 ft long x 8/i to 12 ft thick.

(«) Large alate triradiates from the Ixxiy-wall. Basal rays .straight, tapering
moderately, then fairly sharply pointed, 400 p. to 500 /x long X 12ft to
14 ft thick. Paired rays nearly straight, tapering, then sharply pointed,
160 /x long x 12 ft to 14 p. thick. Oral angle 120° to 145°.

(/) Chiactines. Basal rays straight, tapering slightly to near the end, then
sharply pointed, 300 p. to 500 p. long x 1 1 ft to 1 2 p. thick. Paired rays
straight, 120 p. to IGO/x long x 10 ft to 12 ft thick. Oral angle 115° to
150°. Angle of fold very variable, 115° t« 100°. Apical ray nearly
straight, sharply pointed, 100 p. long x 8 p. thick. Set-off moderate.
Angle between apical my and Iwisal ray 165°.

Oscular spicules. — Near the osculc type (</) spicules get smaller, 330 p. long x
20 ft thick.

(y) Straight, strongly hastate, thin, spined oxea forming the fringe, 100 ft to
200 ft long x 1 ft to 3 ft thick. The spear-head is swollen at the junction
and very sharply pointed. These spicules are also found in the oscular
collar a short distance below the edge lying tangentially and occasionally
projecting.

(h) Small quadriradiates from the oscular edge. Basal ray straight, pointed,
50 ft to 90 ft long x 6ft thick. Paired rays equal and l»ent slightly down,
30ft to GO ft long x 5ft thick. Apical ray straight, pointed, 20 ft long x
5ft thick. Oral angle 160°.

(i) Large quadriradiates lining the oscular collar. Basal ray straight, tapering
uniformly to a fine point, sometimes bent backwards in side view. Length
near oscular edge 200 ft, increasing considerably lower down. Thickness
7 p to 9 fu Paired rays equal, straight or slightly bracket-shaped, 110 ft

44 C. F. JENKIN.

to 130 /n long x 9 /* thick. Oral angle 155° to 1GO°. Apical ray sharply
pointed, 50 p. to 70 ft long x 10 p to 13 p. thick, pointing considerably
upwards. Angle with basal ray 130°.

PART III.

'HISTOLOGY.

The specimens are not sufficiently well fixed to enable many histological details to
be made out satisfactorily, but as most of them are unique, it seems advisable to record
all that can be determined with reasonable certainty and accuracy.

PORES, OSCULES, ETC.

Several different methods are found among calcareous sponges for regulating the
circulation of water and preventing the entry of dirt or enemies.

Ostia. — The admission of water to the incurrent canals is regulated in many
species by the opening and closing of the ostia. These pores are inter-cellular, and are
closed by the action of the surrounding cells. Fig. 48 shows a closed ostium at
the end of the short passage leading through the dermal cortex into the incurrent
chamber. The same passage is shown in tangential section in Fig. 49.

Incurrent Canals. — So far as the author is aware, there is no species in which
the incurrent canals close.

PrOSOpyles. — It is probable that the prosopyles can be closed in all species.
Considerable discussion has taken place as to whether prosopyles are inter- or intra-
cellular. The explanation appears to be simple. The prosopyle is an opening in a
pore cell * (intra-cellular) which leads into the flagellated chamber between the
flagellated cells (inter-cellular). As, however, the flagellated cells are always much
more conspicuous in a surface view than the epithelial cells, the pore when seen in this
view presents the misleading appearance of being an inter-cellular opening.

In most of the Heterocoela (excepting some species of Lcucandra) the gelatinous
mesoderm is very little developed, so that the walls of the flagellated chambers appear
to consist of epithelium lined with flagellated cells. The pore cells are special cells in
this epithelium. In the Homocoela, where there is a certain thickness of mesoderm
containing the spicules, Minchin has shown (1) that the pore cells reach through this
to the level of the flagellated cells, thus forming pipes through the wall. In the
very thin walls of the flagellated chambers of most species of LeucanJm the pore cells
have no appreciable length, the opening appearing as a hole through a thin membrane,
see Figs. 36, 37 and 38.

In Leucandra gelatinosa, the gelatinous mesoderm is so much developed that the
flagellated chambers appear as spherical spaces lined with flagellated cells, and situated

* Vide Minchin (1).

CALCAREA. 45

in the middle of a solid jelly. There is no epithelial Hack, l>ut the epithelium
appears as the lining of the iiicurrent canals, which are more or less circular pipes
through the jelly. The pore cells have not been made out with certainty in this
species, hut the cells lettered p.c. in Figs. 51 and 53 arc prol>ably the pore cells,
which of course are closed.

Apopyles. — The apopylc usually has the appearance of an oval opening in a
thin elastic membrane stretching across the end of the flagellated chamber. It is
probable that the apopylcs can be closed in all species.

Excurrent Canals. — In Leucandra <jeltitint>m the excurrent canals close by the
contraction of the epithelial lining. When the canal contracts, the spiculcs arc left
in their normal position, and the lining cells stretch the gelatinous mesoderm behind
them, so that it is drawn into the canal. The apical rays do not appear to interfere
with this action. In cross-section the contracted canal, therefore, appears as an
irregular ring of spicules filled up with jelly, in the centre of which lies the contracted
group of lining cells (see Figs. 45 and 46). It is probable that similar contractility
may be possessed by other species which have a strongly developed mesodcrm ;
as a rule, however, the excurrent canals do not close.

Gastral Cavity. — In some Homocwla the gastral cavity <«n lie closed solid.*
The author does not know of any instance in which this occurs among the Ifi'trrocorla.

Osculum. — The osculum may be closed in a variety of ways :—

(1) By simple contraction of the surrounding mass (»v/., Lnu'andra primiijcnm).

(2) By the folding inwards of the oscular collar (>'.<)., h-iuxindra f/flatiiwsa), etc.

(3) By means of a diaphragm across the mouth (>:</., Megapogon raripilm).

Diaphragms have been found in several of the new species, but not in good
enough preservation to enable their structure or method of closing to be investigated.

SPICULES.

The dimensions given for the spicules in the detailed descriptions of the several
species in Part II. have generally l>een taken from the drawings, which were traced by
camera lucida accurately to scale, and show as far as possible a representative selection ;
they must not !>«• understood to !>•• cither limiting (maximum and minimum) dimen-
sions or accurate average dimensions. To ascertain the true limiting and average
dimensions would require a detailed examination of a large numl>er of specimens and
great care in the preparation of the spiculc slides, in order to insure that all the
spicules from each specimen were mounted. There appeared to l>e no advantage to
be gained in the present instance by attempting such accuracy, which, indeed, would
only have been possible in the few cases where a sufficient number of specimens
existed.

When the facial rays of a triradiatc or quadriradiate spicule arc " folded," /.»•., do

Minchin (1).

46 C. F. JENKIN.

not lie in one plane, foreshortening always produces an apparent distortion which
affects both the relative length of the rays and the angles between them. No attempt
has in general been made to correct the measurements for this ; the figures given are
the apparent lengths and angles as the spicules lie in various positions on the slide.
The distortion is large in such cases as Streptoconus australis, where the folding
angle is 120°.

The position of the spicules relatively to the other parts of the sponge body is
of some interest. It may be stated as a general rule, and probably as a universal rule,
that spicules do not pass through flagellated chambers. They usually lie entirely
within the mesoderm, except such parts as project beyond the dermal or gastral
layers. They occasionally project into, or cross, the incurrent and excurrent passages.
They appear in* these cases often to be surrounded by a considerable amount of body
substance ; if this is covered with an epithelial layer of cells the spicules, strictly
speaking, still remain within the mesoderm. The minute spicules (" Mortar spicules,"
Haeckel), which in Leucandra often appear to be scattered irregularly through the
whole body-wall, all lie in the mesoderm surrounding the flagellated chambers, and
thus build up what may be described as a flagellated chamber skeleton (see Figs. 127
and 129). The very thin hair spicules, which extend in straight lines through the
body-wall of some species, all lie within the mesoderm. The enormous projecting oxea,
which pierce the body-wall at all angles in some species and appear to be quite
independent of the flagellated chamber structure, all lie in the mesoderm between
these chambers. Similarly the regularly arranged tubar or body skeleton lies in the
mesoderm.

The complex adjustment between the different spicules and flagellated chambers
must be arrived at during the growth of the sponge. It therefore seems probable
that the longer spicules are formed before the flagellated chambers and are pushed
about by the latter as they grow. The mortar spicules, on the other hand, are
probably formed after or during the growth of the flagellated chambers.

r spicules. — The Greek letter T has been used to designate malformed triradiatc
systems in which the paired rays are in line.

T spicules. — The Greek letter T has been used to designate a type of oxeote
spicule in which the end is bent over sharply through about a right angle.

The hastate ends of the oxea in Leucosolenia discoveryi* amongst which T
spicules are found are formed by two angular bends close together in the axis of the
spicule. The r spicules may arise owing to the absence of the second bend.

Chiactines. — -The facial rays of these spicules are very similar to those of
ordinary quadriradiates. The basal ray, which is the longest, is usually straight, but
sometimes bent slightly at a point not far from its junction with the paired rays (see
Fig. 76) and is round in section. The paired rays are usually slightly curved and are
often oval in section, being flattened in the facial plane. Viewed along the axis of the

* r spicules also occur in Leucotolenia minchini and Leucosolenia lieberkuehni.

CALCAREA. 47

basal ray, the paired rays arc .seen to Iw folded to one side ; the fold appears in some
cases to occur sharply ncnr the IMI.SC of the rays, which arc otherwise straight ; in other
OMBn the fold is gnulual, the rays being curved for the greater part of their length.
The apical ray springs from the junction of the facial rays on the side to which they
arc folded, but is immediately bent down nearly into line with the baaa! ray. Viewed
from the side the apical ray appears as a continuation of the Imsal ray, with a sharp
"set-over" at the junction. The amount of the set-over varies widely (see Streptocomut
australta and Meyapogon crucif(rus). The axis of the apical ray is usually slightly
inclined to the axis of the basal ray, sloping towards the side to which the paired rays
are folded. In facial view the chiactine does not differ much from the ordinary
quadriradiate, since the apical rays of the latter are usually curved orally, but in side
view the difference is wide.

The ordinary position of quadriradiates in calcareous sponges is tangential to the
gastral surface, with the l»sal ray downwards and the apical ray projecting into the
gastral cavity, so that the paired rays lie circumfcrcntially embracing the gastral
cavity. In Amphoriscu* and Syculmis there are quadriradiates which lie tangent ially
to the gastral surface with the apical ray pointing radially outwards, and in all the
Amphori»cidat there are quadriradiates which lie tangentially to the dermal surface with
the apical ray pointing radially inwards. The chiactines lie in a plane perpendicular
to all these positions. Their l>asal ray is directed radially outwards, the paired rays
lie in the gastral cortex and the apical ray poiutu radially inwards and projects into
the gastral cavity. The plane of the facial rays is therefore perpendicular to the
gastral and dermal surfaces. The position of the paired rays is no longer fixed ; they
may lie circumferentially as l>efore, but they may turn round the axes of the Itasal ray
to any extent without coming out of the gastral cortex, and in fact they are found
lying in all positions between circumferential and longitudinal.

The relation between the chiactines and the ordinary quadriradiates is an
interesting problem. All the nine new species containing chiactincs have ordinary
quadriradiates lying taugcntially round the oscule, and also lining the whole oscular
collar, if any. In six * of the new species the ordinary quadriradiates cease
entirely at the base of the collar and are replaced by the chiactines which form
the body skeleton. At the junction between the oscular collar and the body-wall
there is a short space in which quadriradiate spicules are found in every intermediate
position between tangential and centrifugal. This fact suggests most forcibly that the
spicules turn round. The top of the body appears to be the most recently developed,
and the spicules formed in the oscular collar might !»• supposed to !•«• turned round
by the development of flagellated chambers under their basal rays. A very similar
tipping up of dermal triradiates, due to the growth under them of the flagellated
chambers, occurs at the base of the oscular collar in Tenthrenodet antarcticu*. There is

• In three of the new specie* the ordinary qnadriradiates in the oscular collar continue throughout the whole
gutral cavity in conjunction with the chiactines in the body.

VOL. IT. 2 O

48 C. F. JENKIN.

one serious objection to this theory. The apical rays of the ordinary quadriradiates
would have to be bent down and straightened while the spicule turned round. The
shape of the apical ray in the chiactines suggests that this does occur, but it is difficult
to realise how an actually formed spicule can be altered in shape.

Hair Spicilles. — In many of the new species there occur long straight spicules
of extreme thinneSs, varying from about • 7 /A to about 2 /u, in diameter ; they are
called hair spicules. Their length is always difficult to determine. When they
occur in the oscular fringe the outer ends are usually broken ; when they occur in the
body-wall they are usually so crowded together that it is impossible to distinguish the
ends of individual spicules. It is probable that they often attain to lengths of 500 fj.
or 1,000/i. They often project from the body-wall with the ordinary thick oxea. The
hair spicules are usually omitted in the drawings of spicules, since their thickness is
too small to draw to scale and their length is uncertain.

Flattened Spicules.— In two of the new species (Leucamlra cirrata and
Megapogon crispatus) the large projecting oxea are marked near their distal ends by
curious flattened areas situated on the convex sides of the curved ends. Their
appearance might suggest that the spicule had come in contact with some flat
obstruction which had diverted its growth and caused the mark. This, however, does
not seem to be possible, and no explanation of the origin of the marks has
been found.

REPRODUCTION.

Larvae similar to those of Sycon ciliatum were found in Tentkrenodes acotti and
Megapogon pollicaris.

Ova of the ordinary type were found in Ilypodictyon and in Achramorpha
glacialis.

Duplicate ova of a remarkable type were found in Achramorpha nivalis and
Megapogon raripilus; no satisfactory explanation of their structure has been found.
They were in large numbers in three specimens of Achramorpha nivalis ; the best
preparations were made from one of the specimens which had been fixed in osmic acid.
None of the specimens of Megapogon raripilus were as well fixed, nor could the
structure of the ova be made out so satisfactorily ; it appears to differ slightly from
that of Achramorpha nivalis.

The following description applies to the ova of Achramorpha nivalis: —

Each ovum (Fig. 110) appears to be made up of two unequal parts; the
larger part (a) is very similar to the ordinary large ovum cell and contains a large
transparent nucleus (b) and small strongly staining nucleolus (c) ; the smaller part
appears to be a multicellular structure, consisting of a large inner cell .surrounded by a
sheath of small cells (d) ; (but it is possible that it may be a single cell, the central
portion (a) being the nucleus). The inner cell (</) contains two structures ; one (/«)
strongly staining like the nucleolus of the larger part, the other a hyaline sphere (i)

49

packed with alxmt a dozen grains (X-) of one colour and one odd one (b) (nucleus,
nuclcoUis ?), which stains a different shade. The outer sheath of rolls each has a small
nucleus (• ) and minute nucleolus (/). This sheath appears in some cases to surround
the inner cell entirely, and in other cases only to surround the outer part, and not to
exist between the inner cell (</) and the other half (<i) of the ovum.

It is possible that the smaller half of the ovum may be a feeding cell, supplying
nourishment to the larger half. The hyaline sphere of grains may I PC food material.

In the ovum of Meynpoytm mripihis, the smaller half has a somewhat different
appearance, which suggests that it is made up of several concentric sheaths of cells
round a central one. These ova, however, are not well enough fixed to repay detailed
examination.

REFERENCE LIST OF MEMOIRS CITED.

1. Mix. HIV. E. A.— A Treatise on Zoology. Edited by E. Rny Lankeatcr. Part II. The Porifcra and

(VlmUr.i. Chapter III. Sponges, by E. A. Mint-bin, M.A. 1'JIK).

2. DKHDY, A..— Studies on the Comparative Anatomy of Sjionyts. V. Obwrvatioiw on the Structure and

Classification of the Calcarea Heterococla. Q. J. M. 8., vol. 85 (1893), pp. 169-257, PI. X.-XIV.
New Sum*

3. HABCKBU— Die KaUuchtcamme. Berlin. 1872.

4. PmMAKrr.—Challfnffer Rep. VIII. 1883.

5. BKMTruss.— Zool. Jahrbuch, Syst. XI. (1898.)

6. LKNDKJJKKLU, R. VON. — Die Spongim tier Adria. I. Die Kalkachwamme. Zeit. f. wisg. ZooL LI 1 1.

(1891.) pp. 1-212.

7. DuiDT, A. — Synoptit of Australian Calearea Heterototla .- irith a proponetl clattification of the group and

dfMriptiont of tome nfir genera and sj)ff iff. Proc. Roy. Soc. of Victoria. Vol. V. (1892), pp. 69-116.

8. SIiHCHlH, E. A. — The Char otter » and Synonomy of Uu British ^jierifs of Sponget oftheyentu Leucotolenia

Proc. ZooL Soc. Lond. 1904, Vol. ii., p. 349.

2 o 2

DESCRIPTION OF PLATES XXVII. TO XXXVIII.

The figures on Plate XXVII. are reproduced from photographs made by the Author ; the rest of the
figures are reproductions of drawings made by the Author with the Abbe drawing apparatus. All the
figures are reduced from these drawings, most of them to two-fifths of the size of the originals ; the
final magnification is marked on each figure.

The letters on the Spicule drawings correspond with those used in the text to indicate the different
sorts of spicules.

N.B. — Hair spicules, which occur in many of the sponges, are not shown in the figures.

The significance of the letters on the other drawings is given in the following list : —

ost. Ostia.

pros. Prosopyles.

ap. Apopyles.

p. Ports.

». ch. Incurrent chambers.

i. ca. „ canals.

e. ch. Excurrent chambers.

e. ca. „ canals.

osc. Oscule.

osc. e. Oscular collar.

osc. e. „ edge.

g. c. Gastral cavity.

d. Dennis.

ft. ch. Flagellated chamber.

ov. Ovum.

sp. Spicule.

a. r. Apical ray.

LIST OF FIGURES.

1. Dermatreton hodgsoni.

2. Megapogon crispatus.

3. Streptoconus australis.

4. Achramorpha grandinis.

5. Dermatreton chartaceum.

6. Leucandra frigida.
7-8. Achramorpha nivalis.

9. Tenthrenodes scotti.

10-11. Hypodictyon longstaffi.

12-18. Leucosolenia discovery i.
14-15. „ mimhini.

16-27. Tenthrenodes scotti.
28-82. „ antarcticus.

83-34. Leucandra primigenia var. lep-

t or aphis.

85-40. Lewandra frigida.
41-48. brumalis.

44-58.

54-66.

57-58.

59-64.

65-74.

75-80.

81-97.

98-102.

103-104.

105-112.

118.

114.

115-119.
120-124.
125-130.
181-186.
137.

Leucandra gelatinosa.
„ cirrata.
„ hiberna.
Dermatreton chartaceum.

„ hodgsoni.
Streptoconus australis.
Hypodictyon longstaffi.
Achramorpha glacialis.

„ grandinis.

„ nivalis.
Orantiopsis cylindrica.
Megapogon cruciferus.

„ villosus.

„ raripilus.

„ pollicaris.

„ crispatus.
Sycon tenellum.

Of -

\^c

nat size

Antarctic 'Discovery) Exp.

Sponges PI. 27

J«nk?n photo Hignley hth

Ml

< /'B- »

i, 15 .100

Antarclic(Discovery) Exp.

Sponges PI 28.

Jenkin del Hiyhley hth

• •

UNIVERSITY

o e

< 4BO

AntarcUc(Discovery)Exp

Sponges PI. 2 9

Jenlon <U1 HujhUv

or THE
[ UNIVERSITY )

OF

51 ««•«

I . .1

f

f 1 ch

..V-'H-

« J F

^ .

ca

Antarcuc (Discovery; Exp

52

4M

Sponges PI. 30

e* f oa

53 >?4«

Antarctic (Discovery ) Exp

Sponges PI. 31

JtnkmtUl Highley hth

I

100

72 . «.

Antarctic (Discovery) Exp

Sponges PI 32

Jenkm <kl ILqhlwy lith

Or rH£ * '

UNIVERSITY

(». .•

AmarcUc (Discovery) Exp

Sponges PI .33

Jenkuvdsl.

•'IV

103 » 100

Antarctic (Discovery) Exp.

Spor.yes PL 34

Jenkindel Highley

of
I UNIVERSITY

!>-.. (

104 x so

\
//

\rT'

-

' V AQ

109 '' x 68

105 x 100

i .

ci.st

I \

107 x aoo

T

108 x 100

osc.e .

Large quidnradi&te
from junction of tube

andbooy

111 X 100

no

720

Antarctic (Discovery) Exp

Sponges Pi 35

Jenkindal Highl«/lithi

-' •

117 inat

114 x 100

112» ' x 23

112* x 65

115 x 2*

119 X 50

Antarctic (Discovery) Exp

Sponges PI 3 6

Jenkin del Highley hth

127 x zse

128 (part) x 100

Antarctic (Discovery) Exp

Sponges PI 37

Jentan del. Highley ktH

OF THE

f UNIVERSITY

c b c

Antarctic (Discovery) Exp.

Sponges Pi 38

Jenlnr del Highlev luh

OF THE '

UNIVERSITY

OF

INI) i: \.

Afhrniiitir/tfiii, 80.

ylarialu, 81.
„ ijranilinin, 82.

„ nicaJif, 88.

table, 2.

antarrtifii*, TtnUurmod»»t 12.
Apopy]«, defined, 5.

•„ histology, 45.
Atffila roriiifta, 6.

„ pnmordialif, 6.
aiuiralu, Strrptotoniu, 25.
Maura, Leveonia, 86.
brumalu, Lttttandra, 16.
Chiactincs, 8.

„ histology, 46.
Chiphoriiiaf, chiactines in, 25.
„ definition, 24.
„ new family, 8.
cirrata, Ltutandra, 18.
Clathrina, 2, 6.

„ foriatxa, 6.
„ primordial™, 6.
Clathnnfiiif, 2, 6.
rompliftita, Ltitevtolftiia, 6.
foriafta, Clathrina (A»c«tta\ 6.
critpafu*, Mrgapogon, 41.
ertififrrut, Mtgapogon (Ltiifonia), 86.
cylindrica, Orantioptit, 85.
Drrmalrtion thartactum, 22.
„ hodgtoni, 23.

Uble, 2.

ducortryi, Lnitosoltnia, 6.
Drawing of gpicules, 45.
Duplicate ova, 6.
rfuni, Ltwttta, 14.
Excnrrent chamben, 5.
„ canals, 5, 45.
friyida, Ltwandrn, 15.
>Wb>M, Lntrttta, 14.
Gamma spicoleB, 7, 9, 46.
Gastral cavity, histology, 45.
ytlattnota, Lfiifandra, 17.
glariali*, Arhramorpha, 81.
gnmdinit, Achramorpka, 32.
Orantiidae, 2, 14.

Graittiop*i» rylitulrica, 3, 8A.

table, 2.
Hfltrofotla, 2, '.».
hilvrna, Lnu-andra, 19.
lli-i'iliiL'v. 44.
hodgwni, Dfrmatrrton, 28.
Homoetela, 2, C>.
Honeycomb patU-rn, 10.
Hyixxliflyon, definition, 2.

„ longitaffi, 27.

Incurrent canals defined, 5.

„ canals, huttology, 44.

.. cbambcra defined, 5.
Jelly in L.grlatinoM, 17.
Irptoraphi* var. Lrtitamlra primigrnia, 14.

nilii, 16.

i, 2, 14.
„ l>r a inn! i*, 16.
„ firmta, 18.
,. friyida, 15.
,. »/'/"/"'""'. 17.
„ hibfrna, 19.
„ primiyrnia, 14.
Lfitffttn dura, 14.

„ f rut if ota, 14.
Ltwonia blawa, 86.

„ rrtitifrra, 86.
Lfitcoxolrnia, 2, 6.

., rompHrata, 6.
„ ilinfovrryi, 6.

„ liebtrkuhnii, 46.
„ minrhini, 8.
LfuroMlfiiiuhif, 2, 6.
litberkUhni, Ltueotoltnia, 46.
" Linked " chambers, definition, 4.

„ „ in Tmthrmodm, 9.

„ „ in Dtrmatrtton, 21.

Mtgapogon, table, 2.

„ rri*patu*, 41.

„ frveiftrun, 85.
„ poUieari», 40.
raripilut, 3M.
„ rillnnu, 5, 87.

mirroraphu var., 14.
minrhini, Lfueotoltnia, 8.

52

INDEX.

New families, 2, 3.
New genera described, 4.

„ table, 2.
Oscular collar defined, 5.
Oscule defined, 5.

„ closed in different ways, 45.
Oscnles, histology, 44.
Ostia defined, 5.

„ histology, 44.
Ova, 48.

„ duplicate, 6, 48.
pollkaris, Megapogon, 40.
Ports defined, 5.
Pore defined, 5.
Pores, histology, 44.
primigenia, Leucarulra (Leucetlti), 14.
primordialis, Clathrina (Ascettci), 6.
Prosopyles defined, 5.

„ histology, 44.
pumila, Leucaltis, ref., 14.
Quadriradiates, turning round, 30, 47.
raripilus, Megapogon, 38.
Kefringent spicules, 7, 9.
Reproduction, 48.

Reticulated cortices in Dermatreton, 21.
scotti, Tenthrenodes, 10.
Spicules, histology, 45.
„ hair, 48.
,, folding, 45.
„ flattened, 48.

gamma,- 7, 0, 46.
„ how drawn, 45
„ position in sponge, 46.
„ tau, 46.
Staurorrhaphidse, definition, 29.

„ new family, 2, 3.

Streptoconus australis, 25.

„ table, 2.

Sycant/ta tenella, described, 4.
Sycettidse, 2, 9.
Tau spicules, 46.
Terminology, definitions, 5.
Tenthrenodes antarcticus, 9, 12.
„ scotti, 10.
table, 2.

Thumb-shaped spicules, 40.
villosus, Megapogon, 37.

MMBI

raiimai IT «n I.UK ruiww AXD -•»«. Liimtn.
«r>En. xTAMrum (run. .« r. **» o«*»r «I*I>»ILL «rmur. w.

FOURTEEN DAY USE

RETURN TO DESK FROM WHICH BORROWED

LOAN DEPT.

This book is due on the last date stamped below, or

on the date to which renewed.
Renewed books are subject to immediate recall.

General Library

University of California

Berkeley

-- - -

• ' ,; • ;
',',-'•'. m '••

n • I m • jB "";:'"^

• • • /• i "'

• I

A

r s

L/^I

-- 1 B -'-" ' -
.',.-•• •

mm

•

•

••• • .

H •

•.

H m s -, • : g

... „ ,

I P i • •••..• i ' ;

' j ":, • : £ I

- . '"''".'

, - -

• ._