‘TRANSACTIONS

ROYAL ‘SOCLETY OF VICTORIA.

=

VOL. I, PART 1

ican D. Sie. PL fe S., Fusnow OF Gon 5 Congr and Demonstrator

AND ~ Agsistax? Lecronmn ix Brotogy’ IN THE UNIVERSITY OF

eth)

i ee Da ORGANI8ATION AND Caassirreation OF THE Cacixmes

oh

nh wits Dnscnrenions or tine Vicrortan Species. (With

Mie MELBOURNE.
ross FOR ee 2 BOYAL: socrnTY

JULY, 1891

- PUBLISHING COMPANY, uaMITED, 270 post pErioe PLACE.

DOV SOC hh ry OF VICTORIA.

Patron:
HIS EXCELLENCY THE EARL OF HOPETOUN, G.C.M.G.

[President :
PROFESSOR W. C. KERNOT, M.A., C.E.

VicezPresidents :
K. J. WHITH, F.B.A.S. | J. COSMO NEWBERY, B.Sc., C.M.G.

Ibon, Treasurer :
C2) BEACKHED, EC.s.

Ibon. Secretaries :
H. K. RUSDEN, F.RB.G:S.

PROFESSOR W. BALDWIN SPENCER, M.A.

bon. Librarian :
JAMES E. NEILD, M.D.

Council :

hk. L. J, HLLERY, F.R.S. | H. MOORS.

G. 8. GRIFFITHS, F.R.G.S. | PROF. H. LAURIE, LL.D.

By We OW TOT ARGS. » PROBS Riots LYLE, MA«

JAMES JAMIESON, M.D. ALEXANDER SUTHERLAND, M.A.
A. H: 8: LUCAS, M.A;,. B.Sc. ©: A. TOPP, M.A., LL.B.

PROF. ORME MASSON, M.A., D.Sc. Boi WAY, MA:

TRANSACTIONS

OF THE

ROYAL SOCIETY OF VICTORIA.
WO. JUL PARI oe

1891.

CONAN ES:

A MONOGRAPH OF THE VICTORIAN SPONGES, sy Artur Denpy,
D.Sc., F.L.8., Frtnow or Qurzn’s CoL~LEce and DrMoNSTRATOR
AND Assistant LecturER IN BioLoGy IN THE UNIVERSITY OF

MELBOURNE.

PART I. ‘Tue ORGANISATION AND OLASSIFICATION OF THE CALCAREA
Homocana, with Derscriprions oF THE VicTorIAN Species. (With

Plates I.—-XI.)

MELBOURNE,
PUBLISHED FOR THE ROYAL SOCIETY
BY THE SPECTATOR PUBLISHING COMPANY LIMITED, 270 POST OFFICE PLACE.

JULY, 1891

It has been decided by the Council to devote Volume III.
Of the Transactions to the publication of Dr. Arthur Dendys
Monograph of Victorian Sponges. This forms the first part, and

succeeding parts will be issued from time to time.

—

A MonoaraPH oF THE VICTORIAN SponaEs, BY ArtTHUR Dernpy, D.Sc., F.L.S.,
FELLOW oF QUEEN’s CoLLEGE AND DrMonsTRATOR AND AssISTANT LECTURER IN

Biotogy IN THE University or MELBOURNE.

PART I.—THE OrGANISATION AND CLASSIFICATION OF THE CatcarREA Homoce.a,

witH DEscRIPTIONS oF THE VicTorIAN Species. (With Plates I.—XI.)

(Read December 11, 1890).

TABLE OF CONTENTS.

PAGE.

I.—INTRODUCTION Bs a0 ne mae Be ae ne 2
II—THE ORGANISATION OF THE CALCAREA HOMOC(GLA a me 5
A.—TuHeE OtyntHus TyPE no 56 Sa ae ne ae 5
B.—Tue Histonocy or THE CancarEA Homoca@ta Bd a ae 6

C.—Tue Canau SystEM OF THE CaLcAREA Homoca@na AS ILLUSTRATED BY THE

VIcTORIAN SPECIES ae ote oe ae An se 23
D.—Tuer Canau SystEM oF THE CALCAREA Homoca@:La IN GENERAL .. ts 37
IlIl.—_THE CLASSIFICATION OF THE CALCAREA HOMOCGLA ay A 39

IV.—DESCRIPTIONS OF THE VICTORIAN SPECIES OF CALCAREA HOMOCGLA 45
Homoca:na SIMPLICIA
Homocana REeticuLaTa
Homoca:ta Rapiata

DovstruL SPrcIES

V.—DESCRIPTIONS OF PLATES Oc a O0 50 o0 De 72

2 A MONOGRAPH OF THE VICTORIAN SPONGES.

I.—_INTRODUCTION.

It is now nearly three years since the Port Phillip Biological Survey Committee
of the Royal Society of Victoria placed in my hands for the purposes of investigation
the splendid collection of sponges dredged by Mr. J. Bracebridge Wilson in the neigh-
bourhood of Port Phillip Heads. Since that time Mr. Wilson has yearly added to the
collection, which now numbers nearly two thousand specimens, and is certainly one
of the finest collections of sponges in the world.

In dealing with so large a mass of material the first thing necessary was to
roughly sort out the specimens into their main groups. In order to accomplish even
so much it was necessary to make many hundreds of microscopical preparations,
and this preliminary work alone has already occupied a very considerable amount
of time.

Meanwhile I wrote to the Natural History Department of the British Museum
and asked for fragments of all their Australian sponges for purposes of comparison
and identification. Dr. Giinther, F.R.S., the Keeper of the Zoological Department
of the Museum, responded to my request in the most liberal manner, and I have to
offer him my most sincere thanks for sending me several hundred named specimens
of Australian sponges, which are of the greatest value in determining the identity or
otherwise of the various species in the Port Phillip collection with those previously
described by other writers.

I am also greatly indebted to Sir Frederick M‘Coy, F.R.S., who has most kindly
placed at my disposal all the sponges in the National Museum at Melbourne ; to Mr.
J. Gabriel for a number of valuable specimens dredged by him chiefly at Westernport
(Victoria) ; to Professor W. Baldwin Spencer, M.A., for the valuable advice which he
has always readily given on difficult points ; to the Rey. Walter Fielder, for assistance
in making microscoyical preparations, and aboye all to Mr. J. Bracebridge Wilson,
M.A., Head Master of the Church of England Grammar School, Geelong, for the great
majority of the specimens and for several opportunities of accompanying him on
his dredging trips and seeing the sponges in the living condition as they first come
from the water. .

I must also not omit to mention my indebtedness to the Council of the
Royal Society of Victoria, who readily agreed to my suggestion that a special
volume should be devoted to the monograph on the Victorian sponges, thus
enabling the work to appear ultimately in the form of a separate publication instead
of being scattered in a series of isolated papers through the Transactions of the

A MONOGRAPH OF THE VICTORIAN SPONGES. 3

Society. Itis hoped that this plan of publication will render the work much more
easy of reference to students. It will probably be a long time before the volume is
completed, as the work can only be undertaken in the intervals between other duties.

It is proposed to complete the examination of the collection systematically,
sroup by group, and to publish the account of each larger or smaller group as soon as
it is completed.

The present contribution contains the account of the simplest and most lowly
organised group of sponges, the Calcarea Homoceela. Considerable difficulty was
experienced in working out this group owing to the very serious imperfection of our
knowledge of the anatomy of these sponges. There appears to be a general idea that
the Calearea Homoceela are all very much alike one another in organisation, and that
such small differences as may exist are scarcely worthy of investigation. Heckel,
indeed, in his famous Monograph of the Calcareous Sponges, recognises several distinct
types of canal system, but he seems to regard them more as “ sports’ than anything
else, and illustrates them all from what he considers to be a single species (A scandra
reticulum).* I do not consider that there is any sufficient evidence in favour of such
an extraordinary variation in canal system within the species as Heckel imagines.
Of course the whole difficulty turns upon the question, ‘‘ What is a species ?”’ and
this is a question which no man can presume to settle off-hand. Whatever else a
species may be, however, there can be no doubt that such a thing has no existence in
Nature, the term being used by zoologists and botanists merely as a matter of
convenience to indicate a group of individuals which resemble one another to a certain
extent.+ If, however, we are going to place in one and the same species forms so
widely divergent in anatomical characters (canal system) as those which Heckel places
in his Ascandra reticulum we might as well do away with the term altogether.

Von Lendenfeld has added two types of canal system to the Calearea Homoceela,
for which, while accepting Heckel’s views as to the intra-specific variation of the
canal system in other forms, he creates two new families, Homodermide and Leucop-
side. The former are supposed to be a transition form between the Asconidx and
Syconid, and the latter between the Asconidie and Leuconide. The mania for dis-
covering (or inventing) ‘‘ connecting links” has probably done an immense amount
of harm to the progress of biological science. If zoologists would be content to
describe what they see in a straightforward manner, instead of going out of their way
to discover connecting links, the study of zoology in general and of spongology in
particular would be greatly facilitated.

*«Die Kalkschwiimme.’’ Vol. 8, Plate 20.

| This question is further discussed later on, in the section dealing with the classification of the Calearea
Homoceela.

4 A MONOGRAPH OF THE VICTORIAN SPONGES.

The descriptions* given by Dr. von Lendenfeld both of Homoderma and Leucopsis
are so inadequate and the specimens upon which the two new ‘‘ families” are based
avowedly so minute, that I cannot help feeling considerable hesitation in accepting
them without further evidence. In thus refusing to aceept Dr. von Lendenteld’s
Homodermide and Leucopside without further evidence I am only following the
example of so eminent a spongologist as Dr. Vosmaer.t

These considerations have induced me to pay particular attention to the minute
anatomy of the Calcarea Homoccela, and my researches have shown that in the more
highly organised forms the canal system becomes far more complex than has been hitherto
believed, leading to a complexity of organisation amongst the Homocela perhaps equal
in degree, although different in kind, to any found at any rate amongst the Sycon
Heterocela. Moreover my observations support those of Dr. von Lendenfeld in that
Thave found a type of canal system closely agreeing with that of his Homoderma, but
in a totally distinct species (Leucosolenia tripodifera, Carter sp.) of such a size and
represented by so many specimens that mistake is impossible.

In working out the Calearea Homoccela I have been much aided by the kindness
of Mr. H. J. Carter, F.R.S., the well-known English writer on sponges. Some years
ago Mr. Bracebridge Wilson forwarded large collections of sponges to Mr. Carter, by
whom they were described in a series of papers in the “Annals and Magazine of
Natural History,” the types being afterwards lodged in the British Museum. Thus
many of the species with which I shall have to deal in this monograph have already
been described by Mr. Carter, but the descriptions are brief, without illustrations,
and the anatomy is scarcely touched upon. Thus it becomes a matter of great
difficulty to recognise many of them, but this difficulty in the case of the Calcareous
sponges has been to a large extent obviated by the kindness of Mr. Carter, who sent
me a copy of his work with numerous manuscript illustrations and unpublished notes.
For this and numerous other acts of kindness I must express my hearty thanks to
Mr. Carter.

Type specimens of the sponges described in the present work will be deposited
in the Biological School of the Melbourne University, and it is proposed also to send
a set to the National Museum in Melbourne and to the British Museum in London.

* Proceedings of the Linnean Society of New South Wales, Vol. IX., Part 4., pp. 1088, 1089.
t Bronn’s ‘“‘ Klassen und Ordnungen des Thierreichs,” ‘‘ Porifera,” p. 387.

A MONOGRAPH OF THE VICTORIAN SPONGES.

Ou

II.—THE ORGANISATION OF THE CALCAREA HOMOCGILA.

A.—Tue Otyntuus Typr.

In the hope that the present work may be of use to naturalists, and especially
Australian naturalists, who are not professed spongologists, I propose to deal more
in detail with the anatomical side of the question than would otherwise be necessary,
and to commence by describing briefly what may be regarded as the simplest and
most typical form of organisation met with in the group. We may then investigate
the more readily in what manner and to how great an extent the other members of
the group differ from this typical form.

The simplest type of organisation met with in the group is known by Heckel’s
term ‘‘ Olynthus”’,* and this name may conveniently be retaimed for the type of
organisation in question, though not as indicating a ‘‘genus” in the ordinarily
accepted meaning of that term.

In the Olynthus type, then, the sponge consists of a small, thin-walled tube or
sac, the lower end of which is closed and attached to some foreign object, while at
the upper end is a large opening known as the osculuim. The wall of the tube,
moreover, is perforated by numerous very minute apertures known as the inhalant
pores or prosopyles.+ The prosopyles, the cavity of the tube (gastral cavity), and the
osculum together constitute what is known as the canal system of the sponge, here
met with in its simplest condition.

The wall of the tube is made up of three distinct layers. On the outside is a
single layer of flattened, plate-like, nucleated epithelial cells, polygonal in shape and
in contact with one another at their edges; this layer is known as the ectoderm. On
the inside is a single layer of ‘ collared cells” known as the endoderm, and between
the endoderm and ectoderm is the mesoderm, consisting of a layer of gelatinous
material in which are embedded the skeleton of the sponge (consisting of calcareous
spicules) and various kinds of cells.

The histological characters of these three fundamental layers will be described in
detail later on, meanwhile it is important to remember that these three layers are
found in all sponges and that they always exhibit the same essential characteristics,
although numerous differences in detail are met with.

* “Die Kalkschwamme,’’ Vol. 1, p. 219 etc. Vol. 3, Plate 1, Fig. 1.

| The term ‘prosopyle” is used by Sollas, in his article on Sponges in the ‘ Encyclopadia Britannica,” to
designate the openings of the inhalant canals into the flagellated chambers and to distinguish them from the inhalant
pores on the surface of the sponge. Ina thin-walled Olynthus the inhalant pores and prosopyles are so close together
that they may for all practical purposes be regarded as identical. In higher forms, however, it is very convenient to have
distinct names for the two structures.

6 A MONOGRAPH OF THE VICTORIAN SPONGES.

Throughout the life of the sponge a stream of water may be observed issuing
through the osculum from the gastral cavity, the water which thus flows out being
constantly replaced by smaller streams which flow in through the prosopyles. This
stream of water is maintained by the activity of the flagella of the collared cells com-
posing the endoderm.

Such then, in brief, is the structure of the Olynthus, which is to be regarded as
a single individual or Ascon-person.* No examples of the Olynthus type occur in the
collection of Victorian sponges, but for an illustration I would refer students to
Heeckel’s Ascetta (Leucosolenia) primordialis, admirably illustrated in his Atlas of the
Calcareous Sponges.t It is probable that all Calcareous sponges pass through an
Olynthus stage at an early period of their life history.

B.—Tue Histotocy or THE CaLcaAREA Homoca.a.

The Ectoderm.

The true nature of the ectoderm of Calcareous sponges was first elucidated by
Schulze in his well-known memoir on the structure and development of Sycandra
vaphanus.' Heckel, in his Monograph on the Calcareous Sponges, had altogether
failed to comprehend the true condition of things, and had confused the two layers,
ectoderm and mesoderm, under the one name Exoderm (used as a synonym for
Ectoderm). Useckel states§ that the ‘‘ exoderm”’ consists firstly of a ‘‘ syncytium,”
and secondly of the spicules. ‘“* Syucytiuim nenne ich bei den Kalkschwammen die
ganze Gewebsmasse, welche durch die Verschmelzung der Geisselzellen des Exoderms
der Flimmerlarve entstanden ist, mit Ausschluss der darin gebildeten Kalknadeln.
Dieses Syncytium ist aus folgenden Bestandtheilen zusammengesetzt: (1) der
Sarcodine, einer hyalinen, structurlosen, contractilen Grundsubstanz, dem modificirten
Protoplasma der verschmolzenen Zellen ; (2) den bleibenden und sich vermebrenden
Kernen dieser Zellen, und (3) den Sficula-Scheiden, welche durch Verdichtung der
Grundsubstanz rings um die Oberflaiche der Spicula entstanden sind.” Heckel
himself, in his latest work on sponges, recognises his error, and acknowledges the
correctness of Schulze’s view ; to quote his own words: ‘“‘ Three years later (in 1875)
this conception was corrected by the accurate observations of Franz Hilhard Schulze,
the excellent spongiologist, who has advanced in so many important directions the

* We may conveniently retain this term although Heckel’s group name “‘ Ascones’’ has -been abandoned.
} “Die Kalkschwimme,” Vol. 3, Plate 1.

ae Balos ot und die Entwicklung von Sycandra raphanus Heckel.” Zeitschrift f. wissensch. Zoologie, Vol.

§ “Die Kalkschwamme,” Vol. I., p. 160.

A MONOGRAPH OF THE VICTORIAN SPONGES. i

knowledge of this class of Cwlenterata. Employing new methods of histological
examination, he discovered on the surface of many sponges a delicate external
pavement-epithelium not hitherto observed, and deduced from this observation the
following important conclusions :—

“The body of the sponges is originally composed not of two, but of three
primitive cell-layers, corresponding to those which in the higher organised Metazoa
are called exoderm, mesoderm and entoderm. The exoderm (or outer layer covering
the external faces) and the entoderm (or inner layer lining the canal-system internally)
are two simple epithelial plates, and between them is enclosed the mesoderm (or the
middle layer); this latter isa kind of connective tissue, and produces not only the
skeleton, but also the sexual cells (eggs and sperm).

‘“« The conception of the sponge-tissues given by F. EK. Schulze is now generally
accepted, and it is very probable that it has general value, though it was not possible
to demonstrate clearly in all sponges the delicate exodermal epithelium.”*

The Calcisponge in which Schulze discovered the ectodermal epithelium was
Sycandra vaphanus, one of the Heterocela. It was of course tolerably safe to assume
that a similar ectodermal layer exists also in the Homocela, but I am not aware that
this has hitherto been definitely proved. Poléjaeff,} indeed, gives a figure of a section
of Leucosolenia poterium in which what are evidently intended for ectodermal cells
appear, but the figure is very diagrammatic, and there appears to be no mention of
the ectodermal cells anywhere in the letterpress.

As a matter of fact I find that the ectoderm of the Homocela agrees precisely
with what Schulze has described for Sycandra raphanus, and what I myself found and
described in Grantia labyrinthica.| xcept in the case of very well preserved speci-
mens it is a matter of great difficulty to make out satisfactorily the structure of the
ectodermal epithelium. In cases, however, where the specimen has been at once
immersed in a sufficient quantity of strong spirit and the sections carefully prepared
by means of the paraffin method, care being taken to avoid overheating, the ectoderm
generally appears in section as a delicate but sharp outline with a moniliform or
beaded appearance due to the swelling caused by the presence of the nucleus in the
centre of each cell. In my sections of Leucosolenia wilsoni, n. sp., for example, I
found the ectoderm unusually well preserved. In sections vertical to the plane of its
extension it presented the appearance described above (PI. VII., Figs. 2, 3), but
where it happened to be cut tangentially, or sliced off, its true character could
readily be seen (Pl. VII., Fig. 4). It consists of thin, flattened, plate-like cells,

* Report on the Deep-Sea Keratosa of the ‘Challenger ’’ Expedition, p. 13.

| Report on the Calcarea of the ‘ Challenger”? Expedition, Plate 3, Fig. 1.

} ‘Studies on the Comparative Anatomy of Sponges, III.—On the Anatomy of Grantia labyrinthica, Carter, and the
so-called I’amily Teichonidw.” Quarterly Journal of Microscopical Science, January, 1891.

8 A MONOGRAPH OF THE VICTORIAN SPONGES.

polygonal in outline, and each with a swelling in the centre where the nucleus is
situate. The cell itself averages about 0-0136 mm. in diameter, and the nucleus,
which is very distinctly outlined and more or less spherical (or perhaps somewhat
flattened in the same manner as the cell), has a diameter of about 0°0034 mm.
Within the nucleus appear a few small, deeply staining granules. Around the nucleus
the protoplasm is highly granular, exactly as described by Schulze, while towards the
periphery of the cell it becomes gradually hyaline. Adjacent cells are in contact at
the edges, and all together form a single-layered, continuous epithelium over the out-
side of the Ascon-tube. As a rule, in ordinary preparations, although the nuclei and
granules of the ectoderm cells may be clearly enough visible, it is very difficult to
distinguish the outlines.

Von Lendenfeld* observes: ‘‘The whole outer free surface of the Sponge is
covered with a low Epithelium, consisting of flat covering cells, each of which may
possess one swinging cilia.” ‘This author habitually figures cilia on the ectodermal
cells both of calcareous and non-calcareous sponges, and it would be interesting to
know if he has actually seen them or merely assumed their presence, perhaps from
observing currents in the surrounding water. It does not seem to me probable
that the ectodermal cells of the Calcarea Homoccela are ciliated.

The Endoderm.

According to Heckel} the cells (termed by Sollas ‘‘choanocytes” and by Carter
‘‘spongozoa,’” but generally known as ‘‘collared cells’) of which the endoderm in the
Calearea Homoccela is entirely composed are so uniform in structure, not only
amongst the Caleareous sponges but also in other groups, that the description of
one species will serve for all. Until recently this has been the general view and it
has become customary with some spongologists to figure the collared cells in their
drawings like so many bricks in the plan of a building, apparently without taking the
least trouble to investigate for themselves the true state of the case. No onecansuppose,
for stance, that Poléjaeff ever saw the collared cells in the not too well preserved
material at his disposal anything like so plainly or so regularly arranged as he figures
them in his Reports on the Calcarea and Keratosa of the ‘‘Challenger’ Expedition. The
structure of the collared cells having been made out by James-Clark,{ Carter,§ Heckel||
and Schulze{] in a few cases, it was assumed that all collared cells were the same, and

* Proceedings of the Linnean Society of New South Wales, Vol. 1X., Part 2, p. 318.

“Die Kalkschwamme.” Vol. I. p. 137.

“ Spongie ciliatz, as Infusoria flagellata.” Memoirs of the Boston Society of Natural History, Vol. i. Part 3, 1867.

: “Notes Introductory to the Study and Classification of the Spongida.’ Annals and Magazine of Natural
History. Ser. 4, Vol. xvi., 1875.

|| “ Die Kalkschwémme,”’ loc. cit.
“| Loc. cit.

Or ++

A MONOGRAPH OF THE VICTORIAN SPONGES. 9

spongologists seem to have thought themselves justified in figuring them in what was
considered to be the orthodox manner whether they had been able to make out the
structure or not. Certainly they were tolerably correct as to what may be
called the typical form of the collared cell, and before passing on to describe the
various modifications which have been shown to exist by more independent investi-
gation we may with advantage describe the structure of this typical form.

Typically, then, the collared cell consists of a rounded or sometimes cylindrical
body produced above into a neck or collum. The neck is surmounted by a long,
vibratile, whip-like flagellum, and the flagellum is surrounded by a very delicate,
transparent, membranous collar which is usually more or less funnel-shaped and is
inserted on the collum around the base of the flagellum. In the body of the cell is
a large nucleus, provided sometimes, if not always, with a distinct nucleolus. ‘The
protoplasm around the nucleus is more or less granular and sometimes contains,
according to James-Clark, Carter and Heckel, one or more contractile vacuoles. Both
collar and flagellum appear to be capable of complete retraction and in this retracted
condition the collared cells are generally met with in sections. According to
Carter* the collared cell (spongozoon), after removal from the body of the
sponge, may be seen moving about the field of the microscope in the form
of an Ameba, with or without the flagellum. It is not improbable that the amoeboid
phase may also occur normally within the sponge. For further details and excellent
figures I may refer the reader to the works of James-Clark, Carter, Heckel, and
Schulze last quoted.

In his article on Sponges in the “Encyclopedia Britannica” (lid. [X.), Sollas first
showed that in some Tetractinellid forms a peculiar modification of the collared cells
occurs, and he gives further details in his Report on the ‘‘ Challenger” Tetractinellida.
This modification consists in the fusion of the margins of the collars of adjacent
collared cells in such a manner as to form a distinct membrane, stretching from
collar to collar and perforated by eireular apertures through which the flagella
project. Soon after this I discovered the same membrane very clearly shown in a
horny sponge (Carter’s Stelospongus flabelliformis), and proposed} for it the name
‘ Sollas’s membrane.” Bidder also mentioned} its existence in Caleareous sponges,
a discovery confirmed by myself in the case of Grantia labyrinthica.\ So far,
however, no one had been able to see both Sollas’s membrane and the flagella of
the collared cells co-existent, but since then I have very clearly seen both at once in

* Loc. cit.

+ “Studies on the Comparative Anatomy of Sponges, II.—On the Anatomy and Histology of Stelospongus
Jlabelliformis, Carter; with Notes on the Development.” Quarterly Journal of Microscopical Science, December, 1888.

| Proceedings of the Cambridge Philosophical Society, Vol. VI., Part 4, p. 183.
§ “Studies on the Comparative Anatomy of Sponges, III.,”’ loc. cit.

10 A MONOGRAPH OF THE VICTORIAN SPONGES.

preparations of the siliceous sponge Halichondria panicea.* Yor more detailed
descriptions and illustrations of this remarkable modification of the collared cells in
certain sponges I must refer the reader to the writings of Sollas and myself just
quoted.

Sollas, again,t has pointed out that in the calcareous sponges the collared cells
are usually, if not always, larger than in other sponges, a statement which is
supported by my own observations. According to Heckel, the collared cells in the
Calearea range from 0-005 to 0:009 mm. in diameter, while in the non-calcareous
sponges Sollas gives the diameter of the collared cells as only 0-003 mm.

Enough has been said to show that the collared cells are not so uniform in
structure in all sponges as has generally been supposed, and I must now make a few
observations on the condition of these elements in the Calcarea Homoccela.

We have seen above that the collared cell is a more or less polymorphic structure,
and this polymorphism is exhibited mainly in a great readiness to withdraw the
collar and flagellum into the protoplasmic body of the cell. The mere act of
removing a sponge from the water and placing it in alcohol seems, as a rule, to be
sufficient to cause the total retraction of the collars and flagella, and it is in this
condition of retraction that the collared cells are generally found in microscopical
preparations. Even in this state, however, they are readily recognisable. Exactly
as the ectodermal cells form an epithelial membrane over the outside of the Ascon-
tube, so the endodermal, collared cells form a continuous epithelium lining the
inside of the tube. For the sake of clearness this liming of collared cells is coloured
red in all the figures. When looked at from above or below this epithelium presents
the appearance shown in the red-coloured portion of Figure 5, PI. VI. It consists of a
number of retracted collared cells lying side by side. The individual cells are small, in
Leucosolenia cavata, for example, measuring about 0:0085mm. in diameter. They
usually appear more or less polygonal from mutual pressure, although at the same
time, owing, doubtless, to shrinking caused by the action of reagents, they generally
stand alittle apart from one another in the sections, so that the outline of each is distinctly
visible. The nucleus of each cell is generally clearly visible ; in Leucosolenia cavata
its diameter is only about half that of the cell itself. The collared cells, and
especially the nucleus, stain deeply with the ordinary staining fluids, such as Borax
Carmine and Kleinenberg’s Hematoxylin, and are thus rendered very conspicuous in
stained preparations. The usual appearance of the collared cells, when seen in sections
taken at right angles to the surface which they cover, is shown in Figure 4, Pl. VI.

* «Studies on the Comparative Anatomy of Sponges, IV.—On the Flagellated Chambers and Ova of Halichondria
panicea.” Quarterly Journal of Microscopical Science, January, 1891.
} Article “Sponges.” Encyclopedia Britannica, Ed. 1X.

A MONOGRAPH OF THE VICTORIAN SPONGES. ils

In many cases the collared cells appear to be but very feebly attached to the
underlying layer of mesoderm, and frequently they come away over large areas
in the course of preparation. Hence, sometimes, they cannot be found at all.

Occasionally, however, the collared cells are exceptionally well preserved and
only partially retracted, even in ordinary sections prepared by the paraffin method.
This is the case in one of my preparations of Leucosolenia proxima. Figure 3, Pl.VILI.
shows three collared cells placed side by side in a row, and each one exhibits the
characteristic collar still expanded. The collars are funnel-shaped or tubular, and
come into contact at their margins, without, however, forming a distinct Sollas’s
membrane. Of course in the case of collars which thus come into contact at their
margins, a temporary fusion may take place as between any two naked masses of
protoplasm, and in this manner Sollas’s membrane in all probability first arose.
Between the bodies of the cells, filing up the interstices, a transparent substance is
seen, which is probably part of the gelatinous ground substance of the mesoderm in
which the bodies of the collared cells are partially embedded. Figure 4, Pl. VIII. shows
another collared cell from the same preparation which has become entirely separated
from the rest of the sponge and yet distinctly exhibits its tubular collar. None of
these cells show the flagellum, which always seems to be retracted before the collar.
The transverse diameter of the body of the collared cell is, in this species, scarcely
0:005mm., while the length of the body and collar taken together is 0°013mm.

Even in the small group of the Calcarea Homoccela at least one interesting
deviation from what may be regarded as the typical condition of the collared cells is
found, and probably more will subsequently be shown to occur when the subject
receives the attention which it deserves. ‘The peculiarity to which I now refer occurs
in the very remarkable species Leucosolenia tripodifera, Carter sp., and is illustrated by
Vigures 5and 6, PI.VIII. In the first place there is a well-developed Sollas’s membrane
in the ordinary position. ‘The flagella of the cells are retracted, but the collars are in
some cases well preserved and funnel-shaped. Their margins do not come into direct
contact, but are united by the membrane, which runs from one to another at right angles
to the long axes of the collars. In vertical sections this membrane appears as a thin,
sharp line running parallel to the row of collared cells, and at a little distance (about
the length of a collared cell body) from it (Fig. 6). Not only, however, is Sollas’s
membrane present, but it presents in itself a very peculiar modification. Instead of
being perfectly smooth the outer surface is thickly studded with delicate, rod-like
processes of uniform length and projecting at right angles from it into the gastral
cavity. ‘These processes are about 0-007 mm. in length and 0-0012 mm. in diameter ;
they stain lightly and have a distinctly granular appearance ; they are arranged with
great regularity and occur in specimens from different localities, so that they would
appear to be of constant occurrence in the species. From their general appearance one

12 A MONOGRAPH OF THE VICTORIAN SPONGES.

is at first tempted to imagine that they are bacilli, but their constancy and peculiar
and regular arrangement argue strongly against this view. If, on the other hand, we
suppose that Sollas’s membrane serves to catch or filter food particles from the
streams of water flowing through the sponge, as maintained by Bidder and myself,
then it seems possible that the curious rod-like processes in Leucosolenia tripodifera
may be a further device for the same purpose. Such, at any rate, is the only
explanation which I can offer of their presence, although I freely admit that it is far
from being proved to be the correct one.

The Mesoderm.

The mesoderm consists of a gelatinous ground-substance in which are embedded
the spicular skeleton and various cell-elements.

The Ground-Substance.*—This is usually but feebly developed in the Calearea
Homocela. It is doubtless of an inter-cellular nature and formed probably as a
secretion of certain of the cells embedded in it, though to which of these cells it owes
its origin we cannot with certainty say. Heckel, indeed, sayst that “it is secreted
by the connective tissue cells of the mesoderm, which are derived originally from the
primary exoderm cells,” and this view is probably correct. In the Homoceela, and,
indeed, so far as we at present know, in all the Calcavea, the ground-substance is clear
and transparent and destitute of the numerous granules so characteristic of many
horny sponges (¢.g., Stelospongos flabelliformis). Upon the extent to which the
sround-substance is developed depends the thickness of the mesodermal layer. In
some species, é.g., Leucosolenia lucas: (P1. IV., Fig. 1) the mesoderm forms a very
thin and inconspicuous layer. In others, however, such as L. stolonifer (PI. VL.,
Fig. 1), L. cavata (PI.VI., Fig. 4) and L. wilson: (P1. VII., Figs. 2, 3), it is much more
strongly developed and forms a layer of considerable thickness. Immediately around
each spicule the ground substance is concentrated in the form of a delicate “ spicule
sheath,” visible as a distinct structureless membrane when the spicule itself is
dissolved out by means of weak acids.

The Mesodermal Cell-Elements—The different kinds of mesodermal cell-elements
as yet recognised in the Calcarea Homoccela are as follows :—

(1.) The ordinary multipolar or Stellate Connective Tissue Cells ave the most
abundant.

In Leucosolenia stolonifer, for example (P1.VI., Figs. 1, 3), these cells, though very
small, are easily visible and plentiful. Each consists of a small, deeply staining,

*“*Maltha” of Heckel. Report on the Deep-Sea Keratosa of the ‘‘ Challenger ’’ Expedition, p. 15.
| Loe. cit.

A MONOGRAPH OF THE VICTORIAN SPONGES. 13

spherical nucleus, about 0:0025 mm. in diameter, around which is aggregated a
small body of granular protoplasm running out in various directions into long, thread-
like processes. Adjacent cells are united together by these processes so as to form a
network, as in higher sponges. Similar cells are shown in the case of Leucosolenta
cavata on Pl. VI., Figs. 4, 5.

We may mention here a very remarkable development of the stellate connec-
tive tissue cells which takes place in Leucosolenia proxima. In this sponge the cells
in question, instead of all remaining embedded in the gelatinous ground-substance
between the ectoderm and endoderm in the normal manner, have grown out between
the collared cells of the endoderm into the gastral cavity, where they form a delicate
network, as shown on PI. VIIT., Fig. 2. The cells composing this network are of very
large size, but otherwise do not differ from normal multipolar mesoderm cells, They
are nucleated and around the nucleus are a number of granules. Some of the proto-
plasmic processes of the cells retain connection with the mesoderm of. the tube-wall,
thus supporting the network in the gastral cavity. Such an extraordinary develop-
ment of the mesoderm seems almost incredible, but it is far too distinct and obvious
in my numerous sections of the sponge in question to admit of any doubt; moreover
it leads on to a still further development in the same direction in an allied species

(Leucosolenia wilsont), which will be fully described in dealing with the canal
system.

(2.) The Ameboid Cells (‘‘Wanderzellen’’ of German authors) are difficult to
distinguish from the ordinary stellate ones, closely resembling the latter with the
long processes cut off. They are somewhat more massive and rounded in outline.

(3.) Lam not able to state with certainty whether or not sub-dermal Gland Cells
occur in the Homocela. In Leucosolenia ventricosa, however, I have found certain
elements resembling to a great extent the sub-dermal gland cells described by me*

in Grantia labyrinthica, but I do not venture to express a definite opinion as to their
nature.

(4.) In some cases certain of the mesodermal elements become specially
modified to form more or less plate-like Endothelial Cells. These occur in two distinct
situations and differ in character accordingly. ‘The two kinds of mesodermal
pavement cells (endothelial cells) which Thave as yet met with in the group are
(a) the lining cells of the embryo-containing cavities and (b) the cells which
ensheath the spicule-rays projecting into the gastral cavity.

I have shown elsewhere} that in Sfelospongos the developing embryos are lodged in
special cavities, excavated, as it were, in the substance of the mesoderm, and that

* “Studies on the Comparative Anatomy of Sponges, IIT.,”’ &e. loc. cit.
t ‘Studies on the Comparative Anatomy of Sponges, II.,”’ &e. loc. cit.

14 A MONOGRAPH OF THE VICTORIAN SPONGES.

these cavities are lined by a single layer of gigantic pavement cells whose function
appears to be to supply the growing embryo with nourishment. Poléjaeff* had
previously called attention to ‘the extraordinary growth of the endothelial cells
surrounding the growing embryos” in various horny sponges. Schulzet long since
showed that the embryos of Sycandra raphanus lie in special cavities in the
mesoderm lined by flattened pavement cells, and I have found the same thing in °
Grantia labyrinthica.t Butin neither of these two cases do the pavement cells of
the embryo capsule exhibit anything remarkable in size or form, as in the horny
sponges referred to above. Similar endothelial capsules occur around the ova of
Leucosolenia pelliculata, but do not require further description.

I have now, however, to record the occurrence, in Leucosolenia wilsom, of
embryo capsules composed of large, thick, plate-like cells resembling, though on a
smaller scale, those found in Stelospongos. The peculiar position of the embryos in
Leucosolenia wilsont will be described later on, when we come to speak of the canal
system. Suffice it now to observe that they le in spherical cavities excavated in the
mesoderm. ‘The cavity itself is about 0°14 mm. in diameter, and is lined by a single
layer of large, polygonal cells of unusual thickness (Pl. VII., Fig.3). Hach cell is about
0:028 mm. in diameter, and 0:009 mm. in thickness. The protoplasm composing it
is highly granular and stains well. The nucleus is small and spherical, measuring
only about 0:005 mm. in diameter; owing doubtless to the granular nature of the
protoplasm surrounding it, it is not easy to make out. There can be no doubt that
the function of these cells is to supply the developing embryo with nutriment, but I
haye not seen any connection between them and the ectodermal cells of the embryo
such as I have described in Stelospongos.

We have next to deal with the endothelial cells which ensheath the spicule-rays
projecting into the gastral cavity. In many Homoccela quadriradiate spicules are
present, and the apical ray usually, if not always, projects through the layer of
collared cells into the cavity of the Ascon-tube (gastral cavity). Probably in all cases
this projecting ray is not naked, but clothed by an investing sheath of flattened,
plate-like, nucleated cells. This cellular sheath is very distinctly shown in my
preparations of Leucosolenia stolonifer (Pl. VI., Fig. 2), and again in Leucosolenia
ivipodifera. If the spicule is dissolved out by means of weak acid, as happens in
staining the preparations by the borax-carmine method, the cellular sheath is left as an
empty, elongated, conical bag, hanging on to the wall of the gastral cavity. The cells
composing the sheath somewhat resemble the cells of the ectoderm, and are arranged
like them in a single layer. Their nuclei are very distinct, about 0°0034 mm. in

* Report on the Keratosa of the “‘ Challenger ” Expedition, p. 52.

wa t ue den Bau und die Entwicklung yon Sycandra raphanus Heckel.” Zeitschrift f. wissensch. Zoologie,
ol. xxy. Suppl.

{ “Studies on the Comparative Anatomy of Sponges, III.,”’ &e. loc. cit.

A MONOGRAPH OF THE VICTORIAN SPONGES. 15

diameter, and stain deeply, but the actual outlines of the cells I have not been able to
distinguish. The collared cells of the endoderm are not continued over this sheath,
but, as shown in PI. VI., Figs. 1, 2, cease abruptly around its base. These structures
must not be corfounded with the ordinary ‘ spicule sheath” always found investing
Calcareous sponge-spicules (vide supra). The latter has been shown by Heckel* to
‘be simply a concentration of the gelatinous ground-substance immediately
surrounding the spicule, and has no cellular structure.

The origin of the cellular spicule sheaths is open to debate. I prefer at present
to regard them as mesodermal structures, but, on the other hand, there is nothing to
prove that they are not endodermal. If, however, they are endodermal then Lenden-
feld’s emended diagnosis of the Homoccela as ‘‘Calcispongie the entoderm of which
consists throughout of frilled flagellate cells,’ + will no longer hold good. ‘The spicules
themselves are certainly mesodermal in origin, and if they can penetrate through the
endoderm what is to prevent them from carrying with them an investment of
mesodermal cells? Probably these cells are really calcoblasts, which secrete constant
additions of carbonate of lime whereby the spicule is enabled to maintain its growth.
It is not certain which of the mesodermal cells, as a general rule, function as
calcoblasts (7.e., secrete the spicules), but I have elsewhere given it as my opinion
that the ordinary stellate connective tissue cells do the greater part of the work.

(5.) The next mesodermal elements of which we haye to speak are the
Reproductive Cells. Schulze,§ Polejaeff,|| and others have shown that the ova and
spermatozoa in Calcareous (and, indeed, other) sponges develop in the mesoderm, and
are formed directly from the amceboid cells (** Wanderzellen”’). The ovum is formed
simply by increase in yolume of the amceboid cell and the spermatozoon by a process
of fission; for further details of these processes, I must refer to the works of
Schulze and Poléjaeff just quoted.

The spermatozoa I have not seen, at any rate in the Calearea Homoceela, but we
may assume that they resemble more or less closely those described by Poléjaeff in
Sycandra raphanus, consisting of a very minute spherical head and a very long and
slender tail (0°03 mm. long).

* Die Kalkschwiimme.”’
{| Proceedings of the Linnean Society of New South Wales, Vol. IX., Part 4, p. 1083.

| In my memoir on the Anatomy of Grantia labyrinthica (loc. cit. supra) I have proposed to distinguish between two
kinds of spicule-secreting cells, or calcoblasts, in the Caleareous sponges, viz., primary calcoblasts, within which the spicules
take their origin as in mother cells, and secondary calcoblasts, which add successive layers of carbonate of lime to the
spicule after it has once been formed. I am not aware that anyone has ever yet seen a primary calcoblast, but their
existence is rendered probable by the analogy of siliceous sponges, in which the smaller spicules are often seen within
mother cells (primary silicoblasts). ‘The ordinary stellate connective tissue cells, or some of them, probably function as
secondary calcoblasts, as also do the endothelial cells ensheathing the rays of spicules which project into the gastral
cavity.

§ ‘Ueber den Bau und die Entwicklung yon Sycandra raphanus Heckel.” Zeitschrift f. wissensch, Zoologie,
Vol. XXV. Suppl.

|| ‘Ober das Sperma und die Spermatogenese hei Sycanda raphanus Heekel.” (* Aus. dem LXXXVI. Bande der
Sitzb. der k, Akad, der Wissensch., 1 Abth. Noy. — Helt, Jahrg., 1882.)

16 A MONOGRAPH OF THE VICTORIAN SPONGES.

The ova, on the other hand, are easily recognisable, and frequently met with in
preparations. Figure 4, Plate VI., for example, shows an ovum of Lezcosolenia cavata,
lying in the gelatinous ground-substance of the mesoderm beneath the layer of
collared cells. This ovum is simply a very large ameceboid cell, about 0.025 mm. im
its longer diameter, with an oval form and avery distinct nucleus about 0-006 mm. in
diameter. The nucleus is granular, and stains (in this case) more deeply than the
surrounding protoplasm, which is also granular. Around this ovum are congregated
a number of other mesodermal cells (probably amceboid) of much smaller size. These
are probably the commencement of the characteristic capsule which seems generally
to surround the developing embryo of calcareous sponges. Possibly the particular
ovum here described is not adult, for it does not present quite the characteristic
appearance of a ripe calcisponge ovum.

In a specimen of Leucosolenia pelliculata, however, I found very numerous
mature ova of much more characteristic structure, and these may be taken as a type
for the whole group. In this species the wall of the Ascon-tube is very thin, and the
mesoderm is but feebly developed. The mature ova, on the other hand, are very
large, their shorter diameter measuring two or three times the thickness of the wall
of the tube. Hence, wherever an ovum occurs, the tube-wall is swollen out, and the
swelling always projects inwards, into the gastral cavity, and not outwards, for the
ovum hes inside the layer of rigid spicules composing the skeleton of the sponge.
Hence the layer of collared cells lining the gastral cavity is lifted up into a mound-like
swelling over each ovum, exactly as I have described in the case of the developing
embryos of Grantia labyrinthica.* Moreover, the ovum already lies in a distinct
cavity lined by a delicate endothelial membrane just like the embryo capsule of
Sycandra raphanus or Grantia labyrinthica.

~

The ovum itself (PI. VIIL., Fig. 7) is usually irregularly oval in longitudinal section,
flattened on the side which lies next to the spicules and swollen out on the other. It
measures about 0-085 mm. in longest diameter. The nucleus is very large, and its shape
appears to follow that of the ovum to which it belongs; its longer diameter measures
about 0-042 mm. in length ; it lies pretty nearly in the centre of the ovum. There
is always one very large spherical nucleolus whose position in the nucleus varies. It
lies sometimes towards one end, sometimes to the side, and in one case I distinctly
saw the nucleolus lying half inside the nucleus and half in the surrounding proto-
plasm. The nucleolus is about 0-01 mm. in diameter. With regard to the more
minute structure of these different parts of the ovum we must notice that the body of
the ovum (7.e., the part outside the nucleus) is a naked mass of coarsely granular
protoplasm which stains very deeply. The-nucleus is enclosed in a very distinct

* «On the Pseudogastrula Stage in the De al : ” : i i
Victoria, 1889, p. 93, g velopment of Calcareous Sponges.” Proceedings of the Royal Society of

A MONOGRAPH OF THE VICTORIAN SPONGES. 17
nuclear membrane, which stains deeply, and it is itself composed of a much more
finely granular material which stains less deeply than the outside protoplasm. The
nucleolus stains more deeply than any other part and appears to be homogeneous and
devoid of structure. In addition to the large nucleolus there are generally visible in
the nucleus a number of very much smaller bodies resembling the nucleolus in
appearance, but it is doubtful whether they have really the same composition as the
latter, especially when we compare them with similar bodies in the ovum of Lewcoso-
lenia depressa, which I propose to describe next.

In my only specimen of Leucosolenia depressa adult ova are extremely abundant.
Like the embryos in the closely allied species, Lewcosolenia wilsoni, the ova
are imbedded in the mesodermal tissue which in part blocks up the cavities of the
Ascon-tubes (gastral cavities), but there is no definite layer of large, plate-like, nutritive
cells around each. Such may, however, be formed later on as the embryo develops.
The ova are of great size, measuring 0°17 mm. in longer diameter, and, as is
usually the case with the ova of the same sponge, are all in very much the same
state of development and resemble one another closely. Figure 8, Plate VIII., repre-
sents a section of one of these ova. ‘The outline of the section is indicated by the
dotted line and, to save labour, only a portion of the protoplasm outside the nucleus
is filled in. ‘The ova vary somewhat in shape and may be described as irregularly
spherical. Owing to the absence of spicules from the surrounding tissues they are
not definitely flattened on one side as in Leucosolenia pelliculata. The nucleus is
very large indeed, 0:076 mm. in diameter, and approximately spherical. The
nucleolus is distinct and also spherical, 0°0085 mm. in diameter. Within the
nucleolus several small, highly refringent granules are visible.

The body of the ovum, outside the nucleus, 1s composed (principally at any rate)
of distinct, closely packed, more or less spherical granules, each about 0-003 mm. in
diameter ; the whole forming a deeply-staining mass. The nucleus is composed as
usual of a very finely granular and less deeply-staining substance, and is invested in
a very distinct, deeply-staining membrane. Just inside this membrane, and for the
most part touching it, lies an irregular row of spherical granules closely resembling
in size and other characters the granules of the outer protoplasm. ‘These granules
occur in the some position with great constancy in all the ova, and form a character-
istic and conspicuous feature. They look as if they were granules belonging to the
outer protoplasm which have actually passed through the nuclear membrane.
Whether this passage is normal or pathological (due to reagents) I know not. In
Leucosolema pelliculata we have already seen similar granules, but they are not
arranged with anything like the same regularity around the inside of the nuclear
membrane, nevertheless a tendency towards such arrangement may sometimes be
observed, as is shown in Figure 7, Plate VIII. In the latter species also I noticed, as

18 A MONOGRAPH OF THE VICTORIAN SPONGES.

described above, the nucleolus itself half in and half out of the nucleus, and this
fact supports the supposition that the granules of the outer protoplasm may under
certain circumstances pass through the nuciear membrane (in the opposite direction
to that in which the nucleolus appeared to be travelling). Another feature of
considerable interest in the ova of Leucosolenia depressa is the presence im the
nucleus of distinct traces of a nuclear network, composed of very delicate, branching
and anastomosing threads, which seem to stain more darkly than the surrounding
protoplasm, and appear somewhat as shown in the figure.

These examples may suffice to illustrate the structure of the ovum of the
Calearea Homoccela, and indeed, of calcareous sponges in general.* Although the
adult sponge is generally admitted to be amongst the simplest and most lowly of
organisms, the ovum appears to be possessed of a complexity of structure far beyond
what might be expected, and this complexity appears especially to concern the
nucleus.

Such then are the various cell-elements which take part in the formation of the
-mesoderm of the Calcarea Homocela, but in addition to these there occur in the
mesoderm of Leucosolenia cavata a great number of bodies whose nature is still a
matter of conjecture, but which may be conveniently described in this place.

Yellow Granules —The bodies in question lie embedded in large numbers
and grouped at fairly regular intervals in the mesoderm beneath the layer
of collared cells (Pl. VI., Figs. 4 and 5); they consist of spherical granules (perhaps
small cells), mostly aggregated in very definite clusters and each with a dark
spot in its centre (nucleus ?). The arrangement of the granules in each aggregation
is peculiar and characteristic. First they may be arranged in a solid heap
of comparatively small size, about 0:025 mm. in diameter. This appears
to be an immature condition (Pl. VI., Fig. 5). Sometimes there appears in such
a heap a dark area resembling a nucleus, which seems to indicate that the whole mass
of granules may be originally a single nucleated cell. Then the granules, each
about 0°0023 mm. in diameter, appear to spread themselves out, so that in the next
stage, which I regard as the more advanced condition, they form a ring surrounding
an empty, discoid or sometimes crescentic space (Pl. VI., Fig. 5). One side of the
ring appears to be always thicker (7.¢., composed of more granules) than the other,
which may be composed of only a single row of granules. The diameter of this
annular ageregation of small cells or granules is about 0-034 mm.

* For further particulars as to the ova of sponges vide Dendy, ‘‘ Studies on the Comparative Anatomy of Sponges,”
III. and IV., loc. cit.

A MONOGRAPH OF THE VICTORIAN SPONGES. 19

The nature of these “yellow granules,” as I propose to call them, is extremely
enigmatical and the first question which presents itself is, do they really belong to
the sponge in which they are found or are they parasitic or symbiotic organisms ?

There are two arguments in favour of the view that they really belong to the
sponge. (1) They occur in great numbers, arranged in the same manner and having
the same characteristic structure in all specimens of Leucosolema cavata that I have
examined, so much so as to form one of the most striking characteristics of the
species. (2) They bear a rather striking resemblance to the developing flagellated
chambers described by me in the embryo of Stelospongos flabelliformis.*

On the other hand there are very serious arguments against this view. (1)
They are only known to occur in this species and, perhaps, in Lewcosolenia
coriaceat and L.. osculuim (vide infra) (I find them also in L. dubia but it is
not impossible that that ‘“‘species’” may be a young form of L. cavata.) (2)
With the exception of the structures already mentioned in the embryos of Sfe/ospongos
flabelliformis they are unlike any other structures found in sponges, and it is scarcely
conceivable that they can be developing flagellated chambers, or masses of collared
cells, because the true collared cells in L. cavata are plainly visible, altogether
different in appearance and distributed in a totally different manner. The idea that
they might form a kind of accessory chambers at one time commended itself to me
but I can see nothing to warrant such a supposition ; moreover the crescentic or discoid
space in the centre of the annular group of granules shows no connection with the canal
system of the sponge. (8) The behaviour of the granules towards reagents is not
altogether what one would expect if they were really sponge-cells. Their colour,
when mounted in balsam without staiming, is distinctly yellowish. They stain well,
however, with hematoxylin. When treated with solution of iodine alone, or
with iodine and sulphuric acid, no blue colour is discernible, but when
fragments of the sponge are boiled in caustic potash solution for the purpose of
isolating the spicules the yellow granules withstand the action of the potash and
appear in the preparations of the spicules. This last observation leads me to

* «Studies on the Comparative Anatomy of Sponges, II. &c.” Quart. Journ. Micro. Sei., Dee. 1888 (vide
especially Pl. XXXIII. Fig. 23.)

+ In a paper in the Annals and Magazine of Natural History for July 1884 (pp. 21, 22), Mr. Carter describes certain
granular bodies in Bowerbank’s Leucosolenia coriacea which must I think be of a similar nature to the yellow granules of
Leucosolenia cavata. He deseribes them fully, but unfortunately without figures; the description runs as follows :—
“Taking the granule singly, it is spherical, translucent, and glairy, glistening from refraction of light, of a faint yellow
tinge, and varying under 1-6000th of an inch in diameter, although rarely attaining this size in this state. They are,
when in sitw, congregated round a nucleated cell (the ‘ Kern’) which is often so indistinct here as to be very difficult
to see, owing to its delicate (? polymorphic) structure and the opaque mass which the granules form when closely
applied to it in juxtaposition; or they are scattered throughout the syncytium in the same way as in the Foraminifera

0 Todine does not turn them purple, nor does liquor potasse dissolve them; but strong nitric acid appears
to destroy their sphericity, which may be brought back again by the addition of liquor potassw. ‘This glairy refractive
appearance gives them the aspect of fat or albumen; while, like the green granules in Spongilla, they appear in the

sulphur-yellow and scarlet varieties of Grantia clathrus (—=Le ucosolenia coriacea] to be the seat of these colours
respectively, when they might be termed ‘pigmental.’ It is possible that they grow into the larger cells of the
protoplasm (the ‘ Kerne’), from which they appear to be derived, when they may fulfil other offices, . . » But

whatever the oflice of the granules may be no one as yet has demonstrated beyond conjecture what they fare or what
purpose they may subserve either in the sponges or in the Rhizopoda,—so they are still called ‘ the granules.’

20 A MONOGRAPH OF THE VICTORIAN SPONGES.

suppose them to be of vegetable nature. The iodine-sulphuric acid test is to my
mind by no means a conclusive one, and in spite of my failure to detect starch or
cellulose by its means I am inclined to favour the view that these very characteristic
and remarkable bodies are symbiotic alge.

To judge from a passage in Mr. Carter’s description of Leucosolenia (Clathrina)
cavata* and from the fact that the yellow granules occur abundantly in a
specimen of that sponge sent by him to the British Museum and thence to
me, I think that Mr. Carter must have seen them. The passage referred to
runs as follows :—‘ Structure that above mentioned, whose staple is the ‘ tubulated
thread,’ of which the wall is very thin and skeletally composed of a single
layer of radiate spicules held together by sarcode supporting the softer parts,
which here appear to consist chiefly of a layer of spongozoa in juxtaposition,
and not gathered into the form of ampullaceous sacs, together with a remarkable
quantity of those organs which consist of nucleated cells surrounded by an abundance
of glistening spherical granules, which Heckel has figured and named ‘nucle’
(Kerne) of his ‘ Syncytium.’” The portion of the passage italicised leaves no doubt
in my mind that Mr. Carter refers to the yellow granules, but he is probably mistaken
in supposing that they have anything to do with the nuclei of Heckel’s “‘ Syncytium.”
The latter are simply the nuclei of the various cells which compose the ectoderm
and mesoderm, and these also are clearly visible in my preparations of Leucosolenia
cavata (PI. VI., Fig. 5).

The Skeleton—Having now discussed all the cell-elements which occur embedded
in the ground-substance of the mesoderm, we come next to the skeleton.

The skeleton of all caleareous sponges consists of isolated spicules never united
together into fibres or networks as in the siliceous forms. We will first speak of the
spicules themselves and then of their arrangement.

The excellent researches of von Ebnert+ have recently thrown much light upon the
physical nature of the calcareous sponge-spicule. According to this author the
spicule contains no organic basis, but is a crystalline formation, composed chiefly of
calespar, but with an admixture of other salts :—‘‘ Die Nadeln der Kalkschwaémme
sind hauptsichlich aus Kalkspath bestehende, keine organische Substanz enthaltende
Individuen von Mischkrystallen, deren diussere Form—ohne Begrenzung durch wahre
Krystallflachen—yon der specifischen Thatigkeit eines lebenden Organismus bedingt
ist und deren innere Structur, obwohl vollstindig krystallinisch, durch eine

* “Annals and Magazine of Natural History,” June, 1886, p. 502.

t “Uber den feineren Bau der Skelettheile der Kalkschwimme nebst Bemerkungen iiber Kaikskelete iiberhaupt.”
(‘‘ Aus dem XCY. Bande der Sitzb. der kais. Akad. der Wissensch. I. Abth. Mirz-Heft., Jahrg. 1887.’’)

A MONOGRAPH OF THE VICTORIAN SPONGES. 21

eigenthiimliche Vertheilung der Gemengtheile mit der ausseren Form in Beziehung
steht.” The spicules, as already stated, are formed as secretions of certain
mesodermal cells, and each is invested by a very delicate organic sheath which is,
however, no part of the spicuie itself, but simply a concentration of the gelatinous
sround-substance surrounding it.

Three, and only three, main forms of spicules occur amongst the Calcarea, and
all three occur amongst the Homocela. They are—(1) The triradiate spicules, which
are by far the commonest, and which seem to be the primitive form amongst the
Calearea. (2) The quadriradiate spicules. (3) The oxeote spicules.

Trivadiate Spicules—Wach consists of three rays or arms radiating from a
common centre. Of these spicules Heckel distinguishes* three groups—(1) Regular
trivadiates, in which the three angles and the three rays are alike. (2) Sagittal
trivadiates, in which two of the angles, or two of the rays, form a pair, while the third
is in some way different. (3) Irregular trivadiates, in which the three rays, or the
three angles, or both, are al] unlike. ‘This classification is convenient ; but it must be
carefully borne in mind that the three types are not sharply defined, but merge
insensibly into one another, so that it is often impossible to say, without very exact
measurement, to which of the three groups a particular spicule should be referred.

In the simplest cases the rays of the spicule are straight, and generally sharp-
pointed at the extremity. Frequently, however, they become more or less curved.
The curvature may take place either in the plane of the three rays, as, for example, in
the case of many sagittal spicules, where the two paired rays often curve either
towards or away from one another; or the rays may curve in such a manner that if
the spicule be laid on a level surface it will rest on the apices of the rays only, with
the centre of the spicule elevated to a greater or less extent. An excellent example
of this latter kind of curvature is shown in the so-called ‘ tripod-spicules’’ of
Leucosolenia tripodifera (Pl. XI., Fig. 5). Amongst the Calcarea Homoccela I have
frequently noticed that the spicules on the outside of a sponge colony are more
robust, and also exhibit a greater degree of curvature than those in the interior
(c.g., Leucosolenia tripodifera).

In the sagittal triradiates the two paired rays are known as the /aéeral or oral rays,
the angle between them as the oval angle and the unpaired ray as the basal ray.

Quadriradiate Spicules.—With Heckel, I regard the quadriradiate spicule as a
derivative of the triradiate, and the fourth ray, which projects from the centre of the
spicule in a plane at right angles to that of the other three, as a secondary develop-

* «Die Kalkschwiimme.’”’ Vol. I., p. 187.

22, A MONOGRAPH OF THE VICTORIAN SPONGES.

ment. The three primary rays are spoken of by Heckel as the facial rays, and the
fourth as the apical ray. As in the case of the triradiates the quadriradiate spicules
may be classified as recular, sagittal and irregular.

In some species triradiate spicules pass almost insensibly into quadriradiates, the
apical ray appearing first as a tiny wart in the centre of the spicule, and only a very
few quadriradiates occurring amongst a vast majority of triradiates. Sometimes only
one or two quadriradiates can be found in a preparation, and it is quite impossible to
be certain, from the mere fact that no quadriradiates at all can be found, that none
exist in the sponge. Hence the hopelessness of attempting to classify the Calcarea
by their spicules alone or even principally.

Oxeole Spicules (Oxea).—I use this term for the spicules named by Heckel
“¢ Stabnadeln,” because they resemble in form the siliceous spicules to which the term
‘“‘oxeote’’ was originally applied by Mr. Ridley and myself in our Report on the
Monaxonida of the ‘Challenger’ Expedition. The oxeote spicule has only a single
axis, and thus consists of a simple rod which, in its most typical condition, is spindle-
shaped and pointed at the two ends, which are precisely similar. ‘This typical
condition, however, is seldom found, ‘The spicule is nearly always curved, and one
end is nearly always of different shape to the other. It appears to me, however,
useless to attempt to give special names to all the varieties of oxeote spicules.

Further details as to the forms assumed by the spicules in different species are
unnecessary in this place, the nomenclature having been explained I must refer the
student for further particulars to Heckel’s monograph and to the systematic portion
and illustrations of the present memoir (Plates IX., X., XI.).

With regard to the arrangement of the spicules in the Calecarea Homoccela there
are one or two points of general interest which offer themselves for consideration.
In the simplest case the skeleton consists of triradiates only, placed side by side with
considerable regularity in a single layer in the thickness of the mesoderm. If the
spicule be sagittal it is so placed that the lateral rays and oral angle are directed
towards the osculum and the basal ray towards the base of the sponge. If quadri-
radiates be present they appear nearly always to be so placed that the apical ray
projects through the endoderm into the gastral cavity, generally curving upwards
towards the osculum, and thus affording a protection against the entrance of
parasites. If, on the other hand, oxeote spicules be present, they seem always
to lie nearer to the ectoderm than either the triradiates or quadriradiates, and very
commonly one end,which is usually slenderer and more sharply pointed than the
other, projects outwards and upwards through the ectoderm, forming a protection to

the outside of the sponge. The position and arrangement of these three kinds of
spicules is well illustrated in Leucosolenia lucasi (Pl. IV., Fig. 1).

A MONOGRAPH OF THE VICTORIAN SPONGES. 23

C.—Taer Canat System oF THE CaLtcAREA Homocaia As ILLUSTRATED BY THE
VICLORIAN SPECIES.

The portion of our subject with which we have now to deal is perhaps the most
difficult. We have already described the Olynthus type, and have seen that it
consists essentially of a narrow, thin-walled tube, open at one end only. Now the
question before us is—Given a narrow tube capable of indefinite elongation and
capable also of branching and anastomosing to an indefinite extent, to determine how
mary different forms the sponge colony resulting therefrom may assume.

It will be seen that under the term ‘“‘canal system” I include not only the canal
system of the individual Ascon-person, but also that of the sponge colony as a whole.
This 1s necessary, for in some cases, as we shall see (e.g., Leucosolenia cavata), the
individuals of which the colony is composed are so intimately united together that we
can no longer disentangle them, and we cannot treat of the canal system of one
Ascon-tube without at the same time treating of that of its neighbours; indeed, the
individuals no longer have each its own canal system, but the colony has a canal
system common to them all. The canal system of the colony is of course determined
by the manner in which the individual Ascon-tubes are united together, and hence in
describing the canal system we must at the same time describe the form of the colony,
and vice versd. In the more complex forms of Calcarea Homoccela it is almost as
impossible to recognise individuals as it is in the most highly developed sponges
such as the Keratosa, and the term canal system is equally applicable, with the same
significance, to both.

The Olynthus is the individual unit of the sponge, and must be regarded as a
single animal developed directly from: a smele ovum. It may in fact be compared to
a Hydra. But the Olynthus has generally, like the Hydra, a strong tendency to form
buds or branches, but, unlike in Hydra, these branches never separate from the
parent. ‘They form with the parent permanent colonies, as in such hydroids as
Obelia. Still we have no difficulty as yet in distinguishing the individuals (Ascon-
persons) which compose the sponge colony from one another, each is only connected
with its fellows at the base, and each has its own gastral cavity and its own osculwm.
But now the different members of each colony begin to branch more and more, and
to fuse with one another wherever the branches come in contact, communications
being established at these points between adjacent gastral cavities. Thus more or less
complex networks of tubes are formed, and every chance of distinguishing between the
individual ‘persons’? which compose the colony is lost. We only know, from
analogy, that the whole sponge is a colony, and not a single individual. In this
anastomosis of the different members of the colony to form complex networks, and
their consequent loss of individuality, the sponge colony differs widely from the

24 A MONOGRAPH OF THE VICTORIAN SPONGES.

hydroid colony. The connection between the individuals which compose the colony
is a far more intimate one than in any other colonial metazoa. The entire colony
now behaves as one individual, it has its own characteristic size and shape, its own
exhalant openings (oscula* or pseudoscula) and its own inhalant pores (psewdopores)
and external skin (pseudoderm).

We must now describe in detail the principal modifications which the sponge
colony, and as a direct consequence the canal system of the colony, present amongst
the Victorian Calcarea Homoccela.

According to the modifications in the form of the colony and of the canal system
I propose to divide the genus Lewcosolenia, which with Poléjaeff and Vosmaer I
regard as the sole genus of Homoccela, into three sections, Simplicia, Reticulata, and
Radiata. We will deal with these divisions in order, beginning with the least
complex.

Section |. smMPLicra.

The Simplicia include such simple Olynthus types as never form colonies, and also
those colonial forms in which the whole colony consists of individuals (Ascon-persons)
which may branch, but which never form complex anastomoses nor give off radial
tubes, so that the individuality of the different members of the colony is always
recognisable.

As already stated, no species which permanently retain the simple Olynthus
form have been met with amongst the Victorian Sponges, The simplest form which
we have appears to be Leucosolenia lucasi.

This sponge is very minute, the individuals (Ascon-persons) of which the colony
is composed measuring only about 2 or 3 mm. in length and O°7 mm. in diameter.
The colony increases by budding, the buds appearing on the parent tube first as blind
outgrowths, which subsequently develop each an osculum at its free end. Hence the
whole colony forms a loose, branching mass of indefinite size and shape, attached to
some foreign object. A portion of such a colony is represented on Plate I.,
Figure 1.

Each individual tube resembles an Olynthus except that instead of being closed
at its lower extremity it remains in open communication with its neighbours. ‘The
minute structure of a single individual is shown on Plate IV., Figure 1. It consists
of a thin-walled, sub-cylindrical or sometimes nearly conical tube with a wide osculum

*The osculum of a compound sponge is not strictly homologous with the osculum of an Olynthus, but corresponds
to a number of Olynthus oscula run together.

A MONOGRAPH OF THE VICTORIAN SPONGES. 25

at its upper extremity. ‘The thin wall of the tube is perforated by numerous circular
inhalant pores (prosopyles) placed in the intervals between the spicules and leading
directly from the exterior into the gastral cavity. The diameter of the prosopyles
in the spirit-preserved specimen is about 0°01 mm. but they may have been a good deal
larger in life, (According to Heckel* the inhalant pores |Twbi porales| are not
permanent structures but appear and disappear again in the living sponge with no
regularity. They appear in any part of the tube wall simply by a process of shrinking
away of the surrounding soft tissues. The ectoderm appears to be the first part to
give way and the endoderm the last. Then they disappear again by a reversal of the
process, the collared cells being the last to close in and resume their original position.
When they have closed completely no trace is left of their existence. Hence it
happens that in spirit-preserved specimens the prosopyles are often not visible ; they
are completely closed. In other cases their position is indicated only by gaps in the
layer of collared cells, the ectoderm and the mesoderm only having closed in. The
diameter of the fully open prosopyles is usually, according to Heckel, from 0:01-0:02

mm.) The number of the prosopyles is very great, as will be seen by reference to
Figure 1, Plate IV.

A slight advance in complexity is exhibited by Leucosolenia stolonifer. Here
the individual Ascon-persons arise from a creeping, tubular, stolon-like structure
(the spongorhiza) which spreads over the substratum on which the colony is growing
in every direction. The spongorhiza may perhaps be regarded as the parent of the
colony, it grows horizontally itself but its buds (the younger Ascon-persons) grow
upwards and become much wider in diameter than the parent tube. The parent
spongorhiza also produces short downward outgrowths which serve to attach the colony
to the substratum on which it grows (Pl. L., Fig. 2). A strong tendency to form a
spongorhiza is also exhibited in Leucosolenia lucasi but is not so well marked as
in L. stolonifer. ‘The spongorhiza is comparable to the similar structure (iydrorhiza)
found in many hydroids.

The upright Ascon-persons of Leucosolenia stolonifer resemble those of L. lucast
except in their very much greater size and inthe greater thickness of the tube wall. They
rarely branch or anastomose, and, except when very young, each one is provided with
a wide terminal osculum ; the youngest individuals end blindly. When fully grown
they attain a height of about 35 mm., and a diameter of about 3°5 mm.; they
narrow slightly towards the osculum. Owing to the unusual thickness of the Ascon-
tube wall the inhalant pores are somewhat different from those of L. ducasi. They
are no longer simply circular apertures perforating a thin membrane. ‘The wall of the
tube averages about 0:14 mm. in thickness, and its outer surface is extremely uneven.
Numerous wide apertures, or depressions, lead into irregular canals in the thickness

* “Die Kalkschwimme” Vol. I., p. 220, et seq.

26 A MONOGRAPH OF THE VICTORIAN SPONGES.

of the wall. These canals, after branching slightly and diminishing greatly in
diameter, terminate in small openings (the true prosopyles), only about 0-02 mm. in
diameter, which perforate the layer of collared cells (Pl. VI., Fig. 1). The
elongation of the inhalant pores into distinct canals is due to the unusual thickness
of the mesoderm. Since these canals appear to be lined by an ingrowth of the
ectodermal epithelium they are probably more permanent structures than is usually
the case with the inhalant pores of the Calcarea Homoccela. Possibly they are
capable of opening and closing only at the gastral end (prosopyle) where they perforate
the layer of collared cells.

Section I]. ReTIcULATA.

In this section of the genus the sponge colony forms a more or less complex
network of branching and anastomosing tubes, and it is no longer possible to
distinguish the individual Ascon-persons of which the colony is composed.

The canal system presents great variations within the section, and before
proceeding to describe individual examples it will be as well to give an outline scheme
of the chief types, of which I recognise six, and which I propose to distinguish as_
types A, B, C, D, HE, F of the reticulate form of canal system. These six types of
canal system may be arranged in two groups, which I propose to regard as subsections
of the Reticulata, and to term Indivisa and Subdivisa respectively.

Subsection 1. IJndivisa.

The gastral cavities of the Ascon-tubes composing the colony retain the
primitive hollow condition, there being no ingrowth of the mesoderm or endoderm.

Division 1.—The colony forms a loose network of tubes (Ascon-tubes) still
well separated from one another, and without a common investing skin (pseudoderm).
Hence there are no definite pseudopores.

Type A., with the characters of the division.

Example.—Leucosolenia dubia, n. sp.

Dwision 2.—The colony is formed of tubes (Ascon-tubes) branching and
anastomosing in a very complex manner. The whole colony forms a compact mass, and
the outer walls of the outer tubes generally become specially thickened and contain
specially large spicules, forming a more or less definite investing skin
(pseudoderm). This skin is perforated by numerous small apertures, the
pseudopores, which communicate with the narrow interspaces between the Ascon-

A MONOGRAPH OF THE VICTORIAN SPONGES. 21

tubes in the interior of the colony. The pseudoderm is also perforated by
larger exhalant openings (oscula or pseudoscula).

Subdivision 1.—The exhalant openings, through which the water leaves
the sponge, are true oscula, z.e., they lead directly into a space lined
by collared cells and formed by the union of a number of Ascon-
tubes.

Type B., canal system normal.

E-xanples.—Leucosolenta stipitata, n. sp. Leucosolenia
pulcherrima, ni. sp.

Type C., canal system reversed.
Example.—Leucosolenia cavata, Carter sp.

Subdivision 2.—The exhalant openings through which the water leaves
the sponge are pseudoscula, z.c., they lead at first into a space not lined
by collared cells, but, presumably, by ectoderm. This space is a
pseudogaster, it really lies outside the colony and is formed probably
by the upgrowth of the colony around it. The Ascon-tubes open
into the pseudogaster.

Type D., with the characters of the subdivision.

Example.—Leucosolenia ventricosa, Carter sp.

Subsection 2. Subdivisa.

The gastral cavities of the Ascon-tubes are more or less subdivided into
incomplete chambers by ingrowths of mesoderm, or of both mesoderm and
endoderm.

Type -.—The collared cells have not yet spread in over the
ingrowths of mesoderm.
Example.—Leucosolenia proxima, n. sp.
Type .—The collared cells have spread in over the ingrowths of
mesoderm.
Examples.—Leucosolenia wilsont, nu. sp. Leucosolenia
depressa, lL. Sp.

Having thus briefly reviewed the six types of canal system which I have found
amongst the Victorian reticulate Homocela we may pass on to describe somewhat
more in detail an example of each.

28 A MONOGRAPH OF THE VICTORIAN SPONGES.

Type A.—-Leucosolenia dubia consists of a network of tubes spreading in
many planes and varying in the closeness of its meshes not only in different specimens
but even in different parts of the same colony. Figure 3, Plate I., represents a small
portion of a colony enlarged four times. We cannot of course distinguish the
Ascon-persons of which the colony is composed, we can only distinguish between
what I propose to call Ascon-tubes and interspaces. The Ascon-tubes are of course
lined by collared cells (endoderm) and the interspaces by ectoderm. Although the
interspaces in the outer portions of the colony (and indeed also in the inner portions)
are often no wider than the Ascon-tubes themselves yet no definite external skin
(pseudoderm) is formed, and hence no definite pseudopores. The structure of the
Ascon-tube is the same asin the Homoccela Simplicia. It is cylindrical and about
0-5 to 1 mm. in diameter, but varies a good deal in different parts of the same colony.
True oscula are occasionally visible on the upper surface of the colony. ‘Two or three
Ascon-tubes open by the same osculum and the diameter of the latter is rather less
than that of an Ascon-tube. The prosopyles are not visible in my specimens, they
are probably closed. As the thickness of the tube-wall is unusually great (about
0-1 mm.) the inhalant pores must be elongated, probably they resemble those of
Leucosolema stolontfer described above.

Type B.—The simplest example of this type is afforded by Leucosolenta stipitata.
The entire colony is only about 8 mm. in height and is perched on the
summit of a short stalk. Figures 4, 5, 6, Plate I., show the external form. The
colony consists of a network of Ascon-tubes so disposed that there is one main
tube running vertically through the centre of the colony, into which the remaining
tubes debouch. The main tube is probably not to be regarded as the parent stock
from which the others have been budded out but as formed by the fusion of a number
of tubes to form a single axial cavity. Itis lined throughout by collared cells and
ends at the summit of the colony in a single, wide, true osculum. Figure 2, Plate IV.,
shows a vertical section through the colony, in which the relations of the different
Ascon-tubes is made clear. Irom this it will be seen that tie interspaces are necessarily
confined to the peripheral portions of the colony. The outer tubes of the colony are
very wide and the interspaces between them correspondingly small, hence there is a
common pseudoderm formed by the outer walls of the outer tubes, in which the
interspaces appear as definite pseudopores (Pl. I., Fig. 6). The pseudoderm is
strengthened by spicules of somewhat greater size than those found in the interior of
the colony. ‘The walls of the Ascon-tubes are very thin and the prosopyles are very
clearly visible in my stained preparations both in the inner and outer tubes of the
colony ; they are circular apertures about 0-02 mmm. in diameter.

Another good illustration of this type is found in Leucosolenia pulcherrima,
the only point of difference to notice as regards the canal system being the presence

A MONOGRAPH OF THE VICTORIAN SPONGES. 29

of several true oscula instead of a single one, but this character is probably not
constant. Figure 7, Plate I., shows the external form of the colony and Figure 3,
Plate IV., the arrangement of the canal system.

Type C.—The only example which I know of this very remarkable, reversed
type of canal system is afforded by Lewcosolenia cavata.

This sponge forms large, massive colonies (Pl. II., Fig. 7) which may attain
a height of four or five inches (100 or 125 mm.) The colony is characteristically
flattened from side to side and also on the upper surface. Onthe upper surface are a large
number of true oscula, each surrounded by a membranous collar (Pl. V., Fig. 1).
Between the oscula are a few small pseudopores but the majority of the pseudopores
occur on the sides of the colony as little oval openings closely scattered all over the
well-developed pseudoderm (PI. II., Fig. 7).

The lowermost portion of the colony, which is also the oldest portion and the
part attached to the substratum on which the sponge grows, consists of a number of
well separated, although branching and anastomosing Ascon-tubes, In short, the
first-formed part of the colony retains the comparatively simple form of type A (e.g.,
Leucosolenia dubia), except that the Ascon-tubes for the most part grow vertically
upwards, being connected together by cross-branches. In this first-formed portion
of the colony there is no difficulty in distinguishing between Ascon-tubes and
interspaces, the latter are wide, and there is no investing skin (pseudoderm) around
the outer portion of the colony, and hence no definite pseudopores. But now as the
colony grows older and larger a remarkable change im the mutual relations of the
Ascon-tubes and interspaces takes place. ‘This change affects only the newly formed
portions of the colony, and leaves the old portion at the base in its original condition.
The change takes place somewhat as follows:—As the young Ascon-tubes grow
upwards the outer members of the colony fuse together so that their outer walls form
a definite pseudoderm, leaving small apertures, the pseudopores, which still place the
interspaces inside the colony in communication with the outside world. At the same
time all the Ascon-tubes, outer and inner, increase in diameter and become extremely
irregular in form, branching and anastomosing very freely. In proportion as the
Ascon-tubes become expanded and irregular, so do the interspaces between them
become constricted and well defined, until at last the tubes and interspaces appear
to have changed places entirely, the tubes having become irregular interspaces, and
the interspaces definite tubes. For this extraordinary condition I propose to use the
term ‘‘ reversal of the canal system.” The pseudopores in the external skin of the
colony now lead into narrow tubular inhalant canals, while the oscula on the top of
the sponge communicate with the irregular spaces between these inhalant canals.
These irregular spaces really represent the Ascon-tubes, and are lined on the inside
of their walls by collared cells, while the narrow inhalant canals, which resemble in

30 A MONOGRAPH OF THE VICTORIAN SPONGES.

appearance the original Ascon-tubes, are lined on the owtszde of their walls by collared
cells, so that a section across one of these inhalant canals looks like a section of an
Ascon-tube turned inside out. The tubular oscula are also lined inside by collared cells.
It is almost impossible to imagine a more complete loss of individuality on the part
of the Ascon-persons which make up the colony than is exhibited in this sponge. It
is difficult to convey in words a clear idea of the real state of the case, but I hope
that reference to the figures will enable the reader to follow my description.
Figure 1, Plate V., is a slightly diagrammatic view of a young colony cut in half
vertically. Figure 2, Plate V., 1s a diagram, simplified as far as possible, and
intended merely to show how the well-defined Ascon-tubes of the lower part of the
colony become transformed into irregular interspaces in the upper portion, and vice
versd. The layer of collared cells is, as usual, coloured red. As already stated, this
diagram is very much simplified; thus all the tubes are made vertical instead of
branching and ramifying in every direction, and no inhalant canals are represented
as entering from the sides of the colony as they mostly do in nature.

As, in the upward growth of the colony, the true Ascon-tubes gradually become
transformed into irregular interspaces, and the true interspaces become gradually
converted into definite tubes, a transition stage is reached at a certain point in
which there are no definite tubes at all, and both Ascon-tubes and interspaces are
represented by a series of irregular cavities. This condition usually prevails over a
considerable portion of the centre of the colony (the portion marked x in Fig. 2,
Pl. Y.), so that here we have instead of tubes and interspaces two systems of irreeular
cavities. All the cavities of each system communicate freely with one another, but
the one system never communicates with the other except through the minute
inhalant pores (prosopyles) in the tube walls, which latter now form a very complex and
irregular series of undulating membranes dividing the two systems of cavities. It
is still possible to determine whether a given cavity in the sponge is gastral cavity
or interspace by examining carefully its bounding membrane, and observing on which
side the collared cells le.* The undulating membrane (= Ascon-tube walls fused
together) exhibits the usual characteristics of the Ascon-tube wall except for the
presence in it of the yellow granules described above (p. 18). The mesoderm is,
however, well developed, so that the wall averages about 0-035 mm. in thickness. I
have not succeeded in detecting the prosopyles, apparently they are all closed.

Type D.—This type of canal system is exemplified in Leucosolenia ventricosa.
In this species the colony forms large, compact, irregular masses, with
uneven, undulating outer surface rising into mounds and ridges. In the typical

* As the collared cells very readily become detached and float away in spirit specimens, the task is rendered still
more difficult, unless the material is very well preserved. If the collared cells have gone we can still tell on which side
they lay by the position of other structures in the wall, as, for example, the position of the spicules and yellow granules.
In well-preserved specimens, however, collared cells can easily be found. j

A MONOGRAPH OF THE VICTORIAN SPONGES. 31

forms there is in the centre of the colony a wide space or hollow in the form of a
tube, penetrating almost to the base of the colony and opening above by a wide
pseudosculum surrounded by a membranous frill or collar. This cavity is the
pseudogaster. Its irregularity in form, the fact that it often shows subsidiary
openings through the wall of the colony iv addition to the terminal pseudosculum,
and, above all, the fact that it is not lined by collared cells, all point to the conclusion
that it is not a true gastral cavity but really les outside the colony and is formed by
the latter erowine up in the form of a thick wall around a central space. Figure 8,
Plate I, shows a very small specimen drawn three times the natural size. Figure 9,
Plate I. shows the external appearance of a larger specimen, and Figure 10, Plate I.
shows the same specimen cut in half longitudinally so as to exhibit the pseudogaster.

The colony (7.e. the thick wall of the pseudogaster) is made up of a dense plexus
of branching and anastomosing Ascon-tubes separated from one another by
well-developed interspaces and opening by true oscula into the pseudogaster,

All over the outer surface of the colony are numerous pseudopores, leading into
the interspaces between the Ascon-tubes. They are small rounded or oval openings,
ranging in diameter from about 0.3 to nearly 1 mm., and they are separated from
one another by intervals of about the same breadth as themselves.

The investing skin or pseudoderm of the colony is in this species remarkably
well developed and it is clearly differentiated into two parts. (1) The part
covering the outside of the sponge and perforated by the pseudopores, and (2)
The part linmg the pseudogaster, not perforated (as a rule at any rate) by any
pseudopores but by numerous true oscula, arranged in groups, as shown in Fieure 10,
Plate I.

Now the first of these two parts consists of a great number of Ascon-tubes fused
together side by side so as to leave small but frequent interspaces, the pseudopores.
These outer Ascon-tubes are much smaller than those in the interior of the colony,
but their outer walls are very thick and strengthened by very large spicules. Hence
the skin on the outer surface of the sponge is a thick double membrane in which the
sastral cavities of its component Ascon-tubes are still present as small spaces lined
by collared cells (Pl. IV., Fig. 4).

The portion of the pseudoderm which lines the pseudogaster constitutes, on the
other hand, a very distinct and easily separable membrane, about 0-1 mim. in thickness,
perforated by the exhalant openings of the Ascon-tubes (true oscula). This membrane,
however, instead of showing a number of spaces lined by collared cells in the thickness
between its two surfaces, is solid, consisting of the usual gelatinous mesodermal
eround-substance with cells and spicules embedded in it, and lined on each side by

32 A MONOGRAPH OF THE VICTORIAN SPONGES.

ectoderm. The walls of the fused Ascon-tubes of which it is probably, like the outer
portion of the pseudoderm, composed, have grown together and obliterated the gastral
cavities, and the collared cells have disappeared. ‘The openings of the Ascon-tubes
(v.e., oscula) appear to be formed secondarily; an Ascon-tube approaches the inner
side of the membrane and the wall of the tube fuses with it. Then a perforation
is established through the fused portion, leading from the gastral cavity of the
Ascon-tube into the pseudogaster.

Around the pseudosculum the lining membrane of the pseudogaster is continued
into a frill or collar (Pl. I., Fig. 10).

This is the only way in which I can explain the formation of the lining
membrane of the pseudogaster, which differs essentially from the wall of an Ascon-
tube in that it is not lined on either surface by collared cells, but on both by an
epithelium of ordinary plate-like cells. We find an analogous process going on in the
stalks of stipitate Calcisponges, which, originally formed out of one or more Ascon-
tubes (or flagellated chambers in the higher forms), become solid by the obliteration
of the cavities of the tubes and the disappearance of the collared cells.*

We come now to speak of the structure of the Ascon-tubes themselves. These,
as already stated, are much narrower on the outside of the colony than elsewhere,
and their outer walls are very thick (0.14 mm.) and contain numbers of very large
spicules, so as to form a hard, resistent skin over the whole outer surface, penetrated
by the pseudopores. As the Ascon-tubes approach the pseudogaster, towards which,
notwithstanding their complex branching and anastomosing, the course of all the
principal tubes is directed, their diameter increases and their walls become much
thinner, so that while the Ascon-tubes themselves may have a diameter of 0.5 mm.,
their walls measure only about 0.05 mm. in thickness. Still these walls are very
strong and firm, for they are supported by a very well developed skeleton of good-sized
spicules, though much smaller than those on the outer surface.

The inhalant pores (prosopyles) in the walls of the Ascon-tubes are clearly visible
in my preparations. ‘They are simply circular apertures about 0-085 mm. in
diameter.

The interspaces between the Ascon-tubes are wide and irregular, and continue
right up to the lining membrane of the pseudogaster without diminishing in size.
They communicate with the exterior through the small pseudopores on the outer
surface of the colony.

Type E.—This very remarkable type of canal system, in which is shown the

* Compare my account of the development of the stalk in Grantia labyrinthica. (Studies on the Comparative
Anatomy of Sponges III. Quarterly Journal of Microscopical Science, January, 1891.)

2
v

co

A MONOGRAPH OF THE VICTORIAN SPONGES,

first indication of the subdivision of the gastral cavities of the Ascon-tubes by
mesodermal inegrowths, is exemplified in Leucosolemia proxima.

The external appearance of this sponge is represented in Figures 1 and 2, Plate II.
The colonies are small, only about 7 or 8 mm. in diameter, and attached to foreign
objects, such as the stems of Algz, by means of root-like processes. Over the upper
surface of the colony are numerous small true oscula, each raised on a conical
projection. The surface of the colony is pitted by numerous pseudopores. Except
for the ingrowth of the mesoderm into the gastral cavities of the Ascon-tubes the
canal system agrees with type B, as will be seen at once by reference to Ficure 1,
Plate VIII., representing a vertical section through the colony. The pseudopores on
the outer surface of the colony are oval and measure about 0:27 mm. in longer
diameter (in the specimen selected). They lead into perfectly irregular inter-
spaces between the Ascon-tubes. The Ascon-tubes have thin walls measuring
about 0°028 mm. in thickness. The inhalant pores (prosopyles) which perforate the
tube-walls are, as usual, small circular apertures, easily visible in my preparations as
areas devoid of collared cells and about 0-014 mm. in diameter; they may very likely
be larger when fully opened. The Ascon-tubes of which the colony is composed
collect together into several larger tubes, each of which leads up to and opens through
an osculum.

In many of the Ascon-tubes, and especially in the larger branches which lead up
to the oscula, there is plainly visible a network (Pi. VIII., Figs. 1, 2) of large,
nucleated, stellate cells, sometimes appearing only near the wall of the tube, leaving the
central portion of the gastral cavity quite empty, but at others stretching right across
from wall to wall in the section. These stellate cells resemble ordinary mesodermal
connective tissue cells except in their large size. The network which they form
inside the gastral cavity remains connected with the mesoderm in the tube-wall by
means of certain of the long slender processes of the cells, which penetrate between
the collared cells of the endoderm. It appears to me highly probable that this
network of cells has a nutritive function and serves to entangle food particles
floating in the stream of water which flows through the tubes. There is no trace
of a similar network in the interspaces between the Ascon-tubes.

It seems very strange that the mesoderm should thus break through the
endoderm and come to lie on both sides of it. A somewhat analogous occurrence
takes place in the case of the ovum of Grantia labyrinthica, which, as I have
elsewhere* shown, breaks through the ectoderm and hangs suspended in the inhalant
canals awaiting fertilisation;} but this is only a temporary condition and more easily
understood, for the ovum at an early stage is known to possess the power of
amoeboid movement.

* « Studies on the Comparative Anatomy of Sponges, III. On the Anatomy of Grantia labyrinthica, Carter, and the
so-called family Teichonidm.”’ Quarterly Journal of Microscopical Science, January, 1891.

{ A careful re-examination of my preparations of Grantia labyrinthica, since this was written, has made me doubt

the correctness of this observation, it being difficult to determine the true nature of the cavities in which the ova are
suspended,

34 A MONOGRAPH OF THE VICTORIAN SPONGES.

Type F.—This, the last type of reticulate canal system, is obviously a further
development of type E, and hence especially interesting. It is best exemplified in
Leucosolenia wilsont. The canal system exactly resembles that of Leucosolenia
proxima except that the mesodermal ingrowths into the gastral cavities of the
Ascon-tubes are more strongly developed and the collared cells have spread in over
them, so that the mesoderm is again covered all over by a layer of endoderm. ‘The
result of this proceeding is to divide the Ascon-tubes into a series of irregular and
incomplete chambers, by means of what Heckel terms ‘ endogastric septa.”

The entire sponge forms a low, irregular, encrusting growth (Pl. II., Figs. 3, 4)
attached to the substratum by short, root-like processes. On the upper surface are the
minute oscula, each on the summit of a small and almost solid, conical projection, and
also numerous pseudopores which (on the upper surface of the sponge) are elongated
and slit-like. Owing to the partial blocking up of the Ascon-tubes by the endogastric
septa the whole sponge acquires an unusual degree of solidity, which is a very
characteristic feature of the species.

The pseudopores lead direct into the irregular and perfectly empty interspaces,
about which there is nothing to attract attention. ‘These lead through the inhalant pores
(prosopyles), which are easily visible in some parts of my preparations and measure
about 0.02 mm. in diameter, into the Ascon-tubes. The Ascon-tubes, after collecting
several together, open through the minute oscula at the apices of the conical
projections (Pl. VII., Fig. 1). The appearance of the Ascon-tubes in section
varies a good deal in different parts of the same colony and would seem
to depend upon their age. In parts of the colony which seem to have been
newly formed they are comparatively thin-walled, the wall of the tube being
only about 0.04 mm. in thickness, while the inhalant pores (prosopyles) are clearly
visible. The mesodermal ingrowths, also, are not nearly so strongly developed
as in older tubes, consisting simply of a loose network of stellate cells, over which,
however, the collared cells have already spread themselves. In the older parts of the
colony the walls of the Ascon-tubes are thicker, the mesodermal ingrowths
are more solid and the network of stellate ceils of which they are at first composed is
hardly visible through the dense layer of collared cells which invests it. This
condition is represented in Figures 2 and 8, Plate VII. In the thick walls of the
Ascon-tubes small stellate cells can be seen imbedded in the gelatinous ground-

substance. The prosopyles are not visible in my sections of the older parts of the
colony, being apparently closed.

No spicules are developed in the mesodermal ingrowths in the gastral cavities
of the Ascon-tubes, so that if any doubt existed as to the position and extent of the

originally simple and perfectly hollow Ascon-tubes the arrangement of the skeleton
would be sufficient at once to determine it.

A MONOGRAPH OF THE VICTORIAN SPONGES. 85

In many of the older tubes developing embryos* were found suspended in
the middle of the gastral cavity, embedded in the mesoderm and surrounded by a
peculiar capsule ; but these have already been described.

A similar, if not identical, condition of the Ascon-tubes, characterised by the
formation of endogastric septa, was long since described by Heckel in his Ascetta
primordialis var. loculosa and A. clathrus vay. clathrina.+

Srotion II]. raprara.

This section of the genus, as its name implies, includes such species as exhibit a
radiate structure—the sponge consisting of a single central Ascon-tube from which
smaller tubes are budded off radially.

The only example of the section with which I am as yet acquainted is
Leucosolema tripodifera, a very interesting sponge originally described by Mr.
Carter under the name Clathrina tripodifera. Fortunately, a considerable number
of specimens of this sponge have come into my hands, including a piece of one of Mr.
Carter’s type specimens in the British Museum, so that I have been able fully to
elucidate the anatomy of the species, an undertaking greatly facilitated by the
large size to which the sponge grows. One specimen in my possession is nearly
four inches (100 mm.) in height, and two inches (50 mm.) in diameter, it is
represented of the natural size in Figure 6, Plate II. The remainder of my
specimens, however, are none of them more than about one inch and a half (88 mm.)
high, and three quarters of an inch (19 mm.) in diameter (Pl. IL., Fig. 5).

The whole sponge resembles a very thick walled sack, ‘The cavity of the sack is
the gastral cavity of the central Ascon-tube, which is extraordinarily large, its
transverse diameter being as great as, or greater than, the thickness of the wall of the
sack. This cavity terminates in a wide osculum at the summit of the sponge (PI. V.,
Fig. 3). The thick wall of the sack is made up of a great number of radial
tubes given off from the central tube, into which they open by means of
irregular groups of openings (Pl. V., Fig. 3). The way in which these radial
tubes are formed is very plamly shown in one of my specimens, represented in
a slightly diagrammatic manner in Figure 3, Plate V. At the summit of the sponge,
immediately below the osculum, the wall of the large central Ascon-tube is very
thin and membranous and there are no radial tubes developed as yet, so that it looks
as if the osculum were surrounded by a membranous oscular frmge. ‘The

* In the closely-allied species Leucosolenia depressa, ova only were found, in the same position as the embryos

in L. wilsoni.
{| “ Die Kalkschwiimme,”’ Vol. 2, pp. 23, 85. Vol. 3, Pl. IV., Vig. 5.

36 A MONOGRAPH OF THE VICTORIAN SPONGES.

uppermost part of the sponge, in fact, is still in the condition of a simple Ascon-
person, with an unusually wide gastral cavity, whose thin walls (about 0-025 mm.
thick) exhibit the usual structure, being lined by a layer of collared cells on the inside,
and perforated by numerous inhalant pores (prosopyles).

At a short distance below the osculum, however, small, hollow buds make their
appearance as outgrowths of the thin tube-wall. These buds are at first simple and
unbranched. As they grow older they elongate and branch freely towards their distal
ends, which terminate blindly; this condition prevails lower down in the colony.
The branches of these numerous radial tubes he close to one another, frequently
touching, and form all together the thick wall of the sponge, which, of course,
oradually increases in thickness as the colony grows older and is hence much thicker
in the lower than in the upper portion of the sponge. Occasionally the cavities of
the radial tubes communicate directly one with another but this does not seem to be
at all general ; an instance is shown in Figure 4, Plate V. As they branch towards
their distal extremities the radial tubes become gradually narrower (PI. V., Fig. 4)
and their blind ends, for the most part in contact with one another, and protected
by special spicules, form the outer surface of the sponge. Abundant small interstices
are left between the ends of the radial tubes, which serve for the admission of
water into the irregular interspaces between the radial tubes.

The structure of the individual radial tubes is shown in Figure 4,
Plate V., and conforms to the usual Ascon type. The inhalant pores (prosopyles) are

very distinctly shown in my preparations, they are simply circular apertures about
0.014 mm. in diameter.

The interspaces between the radial tubes are wide and irregular, widening as
they approach the wall of the central Ascon-tube.

By cutting off portions of the wall of the central Ascon-tube and staining them
separately, as well as by means of sections, I have been able to prove beyond doubt
that the central cavity of the sponge is really lined by collared cells; its wall in fact
presents the same structure as those of the radial tubes and is, like the latter,
perforated by inhalant pores (prosopyles).

According to Mr. Carter this species is sometimes lipostomous, or devoid of
osculum. He says,* describing the general form of the sponge, “ conical, rather
compressed, sessile, fixed, with cloaca and wide mouth; or ovoid and free, with
cloacal cavity but o mouth, that is Heckel’s ‘Auloplegma’-form.” If Mr. Carter’s
account be correct it is a very remarkable thing, and I cannot at all understand how
the stream of water is kept up through the sponge when there is no osculum.

*Annals and Magazine of Natural History, June, 1886, p. 505.

A MONOGRAPH OF THE VICTORIAN SPONGES. 37

Certainly if the sponge is to continue to exist there must be some exhalant opening ;
moreover, if there be no osculum, why should there be a “cloacal cavity,’ as Mr.
Carter calls it? I can understand the sponge having been perhaps damaged, and its
growth consequently disturbed, and the osculum perhaps concealed by overgrowth of
the neighbouring parts, but I cannot believe that there exists a sponge with no
exhalant opening (vide infra).

Abnormalities in the form of the sponge may certainly occur, for in-one of my
specimens a kind of secondary osculum, if one may use the term, has been
established near the base of the sponge, the radial tubes being absent over a small
area and the thin wall of the central Ascon-tube perforated by a relatively large
opening. Whether this is the result of injury or not it is impossible to say, but it
seems not unlikely.

D.—Tse Cana SyYsteM oF THE CaLCAREA Homoca@LA IN GENERAL.

After what has been said in the previous section there remains but little to add
concerning the canal system of the Homoccela. The Victorian species of Leucosolenia
so completely represent the genus that a knowledge of their anatomy is nearly all that
is required to understand the anatomy of the Homoccla in general. The different
forms of canal system enumerated and described above include with a few doubtful
exceptions all the known types, and even add to those already described. The most
important exceptions to this statement are what are known as the lipostomous
forms, 7.e., forms without any osculum, including Heckel’s “ artificial genera,”
Clistolynthus and Auloplegma, the former being a simple Olynthus with no osculum,
and the latter a reticulate colony with no osculum. According to Heckel the
osculum may be entirely wanting, and then part of the prosopyles (Tubi porales,
Heckel) serve for the admission of water and part for its expulsion from the gastral
cavity. Itis a significant fact that amongst the numerous forms of Homoccla which
I have had the opportunity of examining, I have never met with an example of what
can be considered as true lipostomy. Certainly, as I have pointed out previously (in
the case of Leucosolenta lucasi and L. stolonifer) the young Ascon-persons, formed as
buds from the parent tube, at first end blindly and acquire an osculum only when
they reach a certain age; but this is only an immature condition, and it seems to me
not impossible that Heeckel’s Clistolynthus may simply be an immature form in which
the osculum is not yet developed. As to the Auloplegma form, I cannot help regarding
this as simply a reticulate colony in which the oscula, opening cither into a
pseudogaster or directly on the surface of the colony, have been overlooked owing to
their small size. In those forms provided with a pseudogaster which L have
examined (e.g., Leucosolenia ventricosa) I have always been able to find direct

38 A MONOGRAPH OF THE VICTORIAN SPONGES.

communication between the gastral cavities of the Ascon-tubes and the pseudogaster.
Then again we must not forget that in many sponges the osculum has the power of
closing up entirely, and many forms described as lipostomous are probably only forms
in which the osculum is temporarily closed, possibly by the action of the spirit in which
the specimen has been preserved. I am unwilling to admit without further evidence
that species exist in which the prosopyles serve both as incurrent and excurrent
openings, there being no special excurrent opening or openings present in the adult
sponge. Even, however, if we admit the existence of truly lipostomous Homocela,
there is no reason to alter or add to the classification of the canal system proposed
above, for the lipostomous forms might stand side by side in the various groups with

the ordinary osculate types.

Of the fifteen ‘artificial genera” described by Heckel and diagnosed in the
next section of the present memoir, the first seven, Olynthus, Olynthella, Olynthium,
Clistolynthus, Soleniscus, Solenula and Solenidium fall under my section Homocela
simplicia ; Nardorus, Nardopsis, Nardoma, Tarrus, Tarropsis and Tarroma are
reticulate forms, while the last genus Ascometra is, according to Heckel, a compound
of several of the preceding genera in a single colony, but I prefer to regard it as a
colony composed of individuals in various stages of development.

Concerning the types of canal system represented by von Lendenfeld’s genera
Homoderma and Leucopsis* there is little to say; Homoderma, if it really be a
Homoccelous sponge, comes under my section Radiata, while Leucopsis is too
imperfectly described to be taken into serious consideration, though as far as we can
judge from the description and figure it appears to be simply a reticulate Leucosolenia
with well developed pseudogaster.

As to Heckel’s Ascaltis canariensis and Ascaltis lamarckit, which von Lendenfeld
nowt includes in his family Homodermide, I cannot see any reason for placing them
amongst the radiate Homocela. The endodermal papilla which project into the
gastral cavity around the apical rays of the quadriradiates appear, to judge from
Heeckel’s description and figures,} to be merely exaggerations of the slight projections
of the endoderm which frequently surround the bases of projecting spicule-rays
(cf. Leucosolenia stolonifer, Pl. VI., Fig. 1) and they certainly in no way indicate
the formation of radial tubes, which originate as outgrowths and not as ingrowths.
A moment's reflection will show that such ingrowths of the endoderm as those
figured by Heckel could never give rise to the formation of radial tubes.

* Proceedings of the Linnean Society of New South Wales. Vol. IX., Part 4, pp. 1088, 1089.

Tt ‘Das System der Spongien.” (Separatabdruck aus den Abh ; gi
Gegelisshatty Peete tea (Sep us den andlungen der Senckenbergischen naturforschenden

} “Die Kalkschwimme,” Vol. 2, pp. 52, 60. Vol. 3, Plate IX, Figs. 1, 2, 3, 5.

A MONOGRAPH OF THE VICTORIAN SPONGES. 39

UI.—THE CLASSIFICATION OF THE CALCAREA HOMOCCLA.

In classifying the Calcarea we have two primary sets of characters to guide us ;
these are (1) the structure of the imdividual (person) and (2) the structure of the
sponge-colony as a whole.

Heckel apparently considers the structure of the individual to be the most
umportant. In his great monograph of the group, he distinguishes three ‘ families”
of calcareous sponges :—

1. Ascones,* = Calcisponges ‘‘ mit Loch-Canalen.”’
2. Leucones, = Calcisponges ‘‘ mit Ast-Candlen.”

3. Sycones, = Calcisponges ‘‘ mit Strahl-Candalen.”

But here already we are landed in confusion, for the ‘‘ Loch-Canalen,”’ ‘“ Ast-
Canalen,” and ‘Strahl-Canalen ” are not homologous or even comparable structures.
The Loch-Canalen of the Ascones (Homoccela) are simply the inhalant pores (prosopyles)
in the thin wall of the Ascon-person, while the “ Strahl-Canalen” of the Sycones are
totally distinct structures, being really, according to Heckel, the homologues of entire
Ascon-persons, and the Sycon-person as a whole equivalent to a colony of Ascon-
persons formed by gemmation. Each “ Strahl-Canal” is, according to this view,t
nothing but an Ascon-person, and each certainly has its own ‘ Loch-Canalen”
exactly as ina typical Homoccelous (Ascon) form.

If Heckel’s opinion that the Sycon-person is equivalent to a colony of Ascon-
persons is correct, he really makes use of the form of the colony and not of the
structure of the individual as the family characteristic of the Sycones. When,
however, he comes to subdivide the Ascones (—= Homoccela) into genera he denies
the value of the form of the colony as a whole as a natural character, relying
entirely for the distinction of what he terms his ‘natural’? genera upon the
structure of the skeleton, and using the form of the colony only in what he terms his
“artificial” system.

Accordingly we find two schemes of classification in Heckel’s monograph placed
side by side, the artificial and natural, as he calls them. We will consider at present

* Heeckel’s Ascones are equivalent to our Homocala, his Leucones and Sycones to the Heterocala.

| Ido not wish to commit myself at present to any definite view on this question, which will be discussed in a later
portion of the present work.

40

A MONOGRAPH OF THE VICTORIAN SPONGES.

only that part of each system which deals with the Calcarea Homoccela (Ascones), of
which the following is an epitome* :—

Genus
Genus
Genus
Genus
Genus
Genus

Genus

Genus
Genus

Genus

Genus

Genus
Genus

Genus

(a.) NaTURAL SYSTEM.
Family Ascones.

Ascetta.—Spicules all triradiate.

Ascilla.—Spicules all quadriradiate.

Ascyssa.—Spicules all oxeote.

Ascaltis.—Spicules partly trivadiate, partly quadrivadiate.

Ascortis—Spicules partly trivadiate, partly oxeote.

Asculmis.—Spicules partly quadriradiate, partly oxeote.

Ascandra.—Spicules partly trivadiate, partly quadriradiate and
partly oxeote.

(p.) ARTIFICIAL SYSTEM.
Family 1. Olynthida.
Ascones consisting of a single person with a single osculum.

Olynthus.—A single person with naked osculum.
Olynthella.—A single person with probosciform osculum.

Olynthinm.—A single person with the osculum provided with a
fringe.
Family 2, Clystolinthida.
Ascones consisting of a single person without any osculum.

Clystolynthus.—A single person without an osculum.

Family 8. Soleniscida.

Ascones consisting of a colony of simple persons provided with
oscula.

Soleniscus.—A colony of simple persons with naked oscula.

Solenula.—A colony of simple persons with probosciform oscula.

Solenidium.—A colony of simple persons with fringed oscula.

* “Die Kalkschwimme,” Vol. I., pp. 84, 85.

A MONOGRAPH OF THE VICTORIAN SPONGES. Al

Family 4. Nardopsida.
Ascones forming a colony with a single common osculum.

Genus 8. Nardorus.—A colony with a single naked osculum.
Genus 9. Nardopsis.—A colony with a single probosciform osculum.

Genus 10. Nardoma.—A colony with a single fringed osculum.

Family 5. Tarromida.

Ascones in which the sponge 1s composed of several colonies, each
with a single osculum.

Genus 11. Tarrus.—A colony composed of several Nardorus colonies.
Genus 12. Tarropsis.—A colony composed of several Nardopsis colonies.

Genus 18. Tarroma.—A colony composed of several Nardonia colonies.

Family 6. Auloplegmida.
Ascones consisting of a colony with no osculum.

Genus 14. Auloplegnia.—A colony with no osculum.

Family 7. Ascometrida.

Ascones consisting of a colony in which several generic forms are
united. ;

Genus 15. Ascometra.—A colony consisting of several generic forms united.

The second of these two extremely ingenious schemes of classification is well
termed by its author artificial, but this term, as it seems to me, applies equally well
to the so-called “natural” system. Neither system is in itself sufficient, and the
use of the two side by side by Heckel has led to an almost hopeless confusion in the
nomenclature of genera and species, as every spongologist finds to his cost when
endeavouring to identify a calcareous sponge.

The fault of each system lies in the fact that it is based upon a single group of
characters, and the imadequateness of the ‘natural system,’ which Heckel
principally uses, is shown by the host of ‘ generic,” ‘ specific,’ and ‘* connective ”
varieties which he has been obliged to create, The first species described in Heckel’s
monograph is Ascetla primordialis and this embraces seyen generic varieties, four
specific varities and three connective varieties, each with a different name. ‘The only
‘ Species-character ” given is a short description of the spicules, and no type of the
species is described ; and so on for the other species.

49, A MONOGRAPH OF THE VICTORIAN SPONGES.

Allforms of Homoccela, no matter what their anatomical peculiarities, which happen
to have spicules of the same shape, are brought by Heckel under a single “species.”
Now there are only three fundamental forms of spicules in the entire group of calcareous
sponges, viz., the trivadiate, quadriradiate and oxeote, and each of these may vary
greatly in shape within a single specimen. Moreover some portions of a colony may
contain say oxeote or quadriradiate spicules, while other portions taken from the
same specimen may contain none. Heckel was well aware of this variation in the
spiculation and hence his “‘ connective varieties,’ which often bring the same species
under two or three distinct genera.

I grant that there are certain forms of spicules which may be regarded as typical
for each species, but the presence of similar spicules is not alone sufficient to justify
us in including under one specific name forms which otherwise differ widely in
organisation.

Heckel’s natural genera are particularly artificial, and this might be expected
when we see that the same series of generic diagnoses, depending solely upon the
combinations of the three types of spicules, is made to serve for all his three families
of calcareous sponges.

Since neither of Heckel’s alternative systems will satisfy the requirements of
the modern zoologist we must find a new method, and in this new method we must
be content to follow the example of workers in other branches of zoology and give
up all hope of finding a royal road to classification. In other words, we must make
use of as many characters as possible and not of as few, and in this way—classifying
by an assemblage of characters*—we may fairly hope to make satisfactory progress.
At the same time we must not forget that Heckel’s classification of the Calcarea was
the first serious attempt at anything of the kind; that he had an extremely difficult
task to deal with, and that immense credit is due to him for the energy and
perseverance with which he led the way in this department of biology. All schemes
of classification are but tentative and each must give place to later modifications as
our knowledge advances, for evolution plays as important a part in our study of
organisms as 1t does in the history of the organisms themselves.

Of course we shall find that different characters often contradict one another,
forms which agree in spiculation may differ widely in canal-system, and vice-versa.
So great is the variation in both these characters amongst the Calcarea Homoccela
and so difficult is it to determine which is the most important that Poléjaeff has, in
his work on the Calcarea of the ‘Challenger’ Expedition, abandoned all Heckel’s
seyen natural genera of Homocela and fallen back upon the old genus Leucosolenia

* “No doubt organic beings, like all other objects, can be classed in many ways, either artificially by single
characters or more naturally by a number of characters.”—Darwin, “ Origin of Species,” Ed. VL., p. 364.

A MONOGRAPH OF THE VICTORIAN SPONGES. 43

of Bowerbank as the sole genus of the group. In this I agree with Poléjaeff, but at
the same time, as a mere matter of convenience and not necessarily as indicating
genetic relationship, I prefer to subdivide the genus into sections and subsections
according to the nature of the canal system of the colony. I prefer to make use of
the canal system for this purpose rather than the spiculation, for I have frequently
found the spiculation varying greatly in different parts of the same colony but never
so the canal system (except m cases where a later-formed portion of the colony
shows an advance in organisation as compared with an earlier-formed portion, as, for
example, in Lewcosolenia cavata; but this is simply a case of different ontogenetic
stages being present simultaneously. May not a similar state of affairs have given
rise to Miklucho-Macleay’s* and Heckel’s statements as to the co-existence of
different types of canal system in the same colony, as in Heckel’s genus
Ascometra ?). On the other hand I have excellent series of specimens of obviously
the same species in which the canal system is remarkably constant (e.g., Leucosolenia
cavata, L. tripodifera, L. pelliculata). Moreover the canal system offers a wider
range of characters to choose from than do the combinations of spicules, of which
seven only are possible.

The classification which I therefore propose to make use of for the Calcarea
Homoceela is as follows :—
Order Homocela (Poléjaeff).
Calcareous sponges in which the endoderm consists throughout of collared cells.
Genus, Leucosolenia (Bowerbank).

With the characters of the order.

Section 1. Homocela Sinplicia.

Homoccela in which the Ascon-persons either remain solitary and do not form
colonies or they form simple colonies in which the component Ascon-persons may
branch but never form complex anastomoses nor give off radial tubes, so that the
individuality of the different members of the colony is easily recognisable.

Section 2. Homocela Reticulata.t

Homocoela in which the sponge-colony forms a more or less complex network of
branching and anastomosing tubes, so that it is no longer possible to distinguish the
individual Ascon-persons of which the colony is composed.

* Jenaische Zeitschrift, Vol. 1V., 1868.

{ This section is practically synonymous with Gray’s genus Clathrina, used by Carter in his descriptions of the
Victorian sponges (loc. cit.).

44 A MONOGRAPH OF THE VICTORIAN SPONGES.
Subsection 1. IJndivisa.

The gastral cavities of the Ascon-tubes composing the colony retain their
primitive hollow condition, there being no ingrowths of mesoderm or endoderm.

Subsection 2. Subdivisa.

The gastral cavities of the Ascon-tubes are more or less completely subdivided
into chambers by ingrowths of mesoderm or of both mesoderm and endoderm.

Section 8. Homocela Radiata.

Homocela in which the sponge consists of a single, central Ascon-tube from
which secondary tubes are budded off radially.

This classification will, so far as I am aware, serve to include all the known forms
of Homoceela with the doubtful exceptions of Heeckel’s A scometra and von Lendenfeld’s
Leucopsis,* concerning both of which we require further details. Ascometra may, as I
have already suggested, simply represent a series of ontogenetic stages present at the
same time, for there can be little doubt that the most complex adult forms pass
through several stages, beginning with the Olynthus, in their life history ; while the
extremely doubtful Leucopsis is probably only one of the Homocela reticulata, and
whatever it is it is so insufficiently described that we cannot possibly recognise it.

In distinguishing species all characters are of use, and a well-marked difference
in any cne character is, in my opinion, a sufficient justification for a distinct specific
name. ‘This, of course, necessitates a good many specific names, but it is better to
have too many than too few, and so long as each form is properly described increase
of species only adds to our knowledge, while the merging of many forms under one
name makes hopeless confusion, for the author who does so seldom thinks it
necessary to give an adequate description of each variety and it then becomes
impossible to sort them out and to determine which is really the type of the
species.

The term species is, as I understand it, a purely arbitrary one, meaning simply
an assemblage of individuals more or less resembling one another and presumably
descended from a common parent, but whether a particular individual belongs

to a particular species or not is a question which each observer must decide for
himself. +

* Proceedings of the Linnean Society of New South Wales, Vol. IX., part 4, p. 1089.

| ‘From these remarks it will be seen that I look at the term species as one arbitrarily given, for the sake of
convenience, to a set of individuals closely resembling each other, and that it does not essentially differ from the term
variety, which is given to less distinct and more fluctuating forms. The term variety, again, in comparison with mere
individual differences, is also applied arbitrarily, for convenience’ sake.”—Darwin, ‘ Origin of Species,” Ed. VI., p. 42.

A MONOGRAPH OF THE VICTORIAN SPONGES. 45

IV.—DESCRIPTIONS OF THE VICTORIAN SPECIES OF CALCARKA HOMOCGLA.

I follow Poléjaeff* in dividing the calcareous sponges into two orders, Homocela
and Heterocela. The term Homoceela, as already stated, is synonymous with Heeckel’s
term <Ascones and the diagnosis of the order is the same as that of the sole genus
(Leucosolenia) comprised therein.

Genus Lencosolenia, Bowerbank. |

Diagnosis. —Caleareous sponges in which the endoderm consists throughout of
collared cells.

In the following arrangement of the Victorian species of Leucosolenia I include
only such as I have been able to examine myself or such as are sufficiently fully
described as to make their indentity tolerably certain. I give at the end a list,
together with the original descriptions, of what must at present be considered as
doubtful species stated to have been found in Victorian seas.

Section 1. Szmplicia.

The Ascon-persons either remain solitary and do not form colonies, or they form
simple colonies in which the component Ascon-persons may branch but never form
complex anastomoses nor give off radial tubes, so that the individuality of the
different members of the colony is easily recognisable.

1. Leucosolenia lucasi, n. sp.
CE Eicwe Birney ice Ws Pl xX... Wie...)

(a.) General Appearance and Canal System.—The sponge (PI. I., Fig. 1) forms
loose colonies, the Ascon-persons being connected at their bases by a hollow
spongorhiza which creeps over foreign objects. The Ascon-persons are small,
cylindrical, thin-walled tubes, 2 or 8 mm. in height and about 0-7 nun. in diameter.
When fully grown there is a wide osculum at the summit of each individual. The
outer surface of the tubes is very minutely hispid and the colour in spirit is white.

* Report on the Calcarea of the ‘Challenger’ Expedition, p. 22.
+ Philosophical Transactions of the Royal Society of London, 1862, p. 1094,

46 A MONOGRAPH OF THE VICTORIAN SPONGES.

Particulars of the canal system are given on pp. 24, 25.

(b.) Arrangement of the Skeleton.—The skeleton consists of quadriradiate, triradiate
and oxeote spicules. The quadriradiates and trivadiates are arranged in a single
layer in the thickness of the mesoderm, the apical rays of the quadriradiates projecting
into the gastral cavity. The oxeotes have their broader ends embedded in the
mesoderm, while their narrow ends project outwards and upwards through the
ectoderm and give to the outer surfaces of the tubes their hispid appearance. The
arrangement of the spicules is shown in Fig. 1, Pl. IY.

(eeike Spicules-| (Pl SUX shiossl):

(1.) Trivadiates.—These are sagittal, but the three angles are about equal.
The basal ray is long and gradually sharp-poimted, measuring 0-1 mm.
in length by 0:005 mm. in diameter at the base. The oral rays are
slightly curved away from one another, gradually sharp-pointed,
0.07 mm. long.

(2.) Quadriradiates.—These are of about the same size and shape as the
triradiates with the addition of the apical ray, which is shorter than
the others, gradually sharp-pointed, curving slightly upwards and
projecting freely into the gastral cavity.

(8.) Oxeotes——These are irregularly fusiform, sharply pointed at both ends
but broader at one end than at the other; usually bent suddenly at a
slight angle near the broader end; often slightly and irregularly
curved; tapering gradually to a fine point at the narrow end but with a
very slight annular swelling ata short distance below the apex. They
measure up to 0-16 mm. in length by 0:005 mm. in diameter at the
broadest part.

(d.) Affinities.—In external appearance this species resembles Heckel’s Soleniscus
variabilis* but it differs markedly in the form of the spicules.

(e.) Locality.—Outside Port Phillip Heads, (Coll. J. B. Wilson.)

2. Leucosolenia stolonifer, n. sp.
(bias 26) P)e Vi lies. 1525 3) SPL Pxe Uiion oe)

(a.) General Appearance and Canal System.—The sponge (Pl... Te; Era: 79)
consists of a colony of Ascon-persons springing vertically from a slender, tubular

* «Die Kalkschwimme,” Vol. III., Pl. 18, Fig. 6.

A MONOGRAPH OF THE VICTORIAN SPONGES. AT

spongorhiza which spreads horizontally in every direction over the substratum on
which the colony is growing. Occasionally, but only rarely, the Ascon-persons
anastomose with one another. The spongorhiza also produces short downward
outgrowths whereby the colony is attached to the substratum. The Ascon-persons
are cylindrical tubes much wider than the spongorhiza, each provided, when adult,
with a terminal osculum. They attain a height of about 35 mm. and a diameter of
about 3°5 mm., narrowing slightly towards the osculum. The wall of the tube is
about 0:14 mm. thick and distinctly hispid on its outer surface. ‘The colour in spirit
is white.

Particulars of the canal system are given on pp. 25, 26.

(b.) Arrangement of the Skeleton.—The skeleton consists of quadriradiate and
oxeote spicules. The quadriradiates are very closely packed in the mesoderm,
forming a rather confused layer, with their facial rays freely overlapping and their
apical rays projecting into the gastral cavity. The oxeotes have their broader ends
embedded in the mesoderm and their narrow ends projecting outwards and obliquely
upwards through the ectoderm and giving to the surface of the tube its hispid
character.

(Ger Spiculess (Pl xe ice) 2).

(1.) Quadrivadiates.—These vary a good deal both in size and form. They
are generally more or less markedly sagittal but not very strongly so,
and may be almost or quite regular. The rays may be all straight
or the oral rays may be slightly curved. ‘The apical ray is usually
short, slender, sharp-pointed, and curved towards the oral angle ;
much shorter and slenderer than the facial rays; in a small, but by
no means insignificant proportion of the spicules, however, it becomes
enormously exaggerated, measuring as much as 0°6 mm. in length
and 0-035 mm. in diameter at the base, so that its dimensions greatly
exceed those of the facial rays; when developed in this way it is
often irregularly bent. The facial rays are usually conical, gradually
sharp-pointed, and measure about 0-2 by 0:015 mm.; sometimes they
are slightly constricted at the base.

(2.) Oxeotes.—These are large, asymmetrically fusiform spicules, the widest
part being much nearer to one end than to the other. They may be
quite straight but are more generally somewhat curved like a sickle,
the curvature may be even or more or less irregular. They are sharply
pointed at each end. When fully developed they measure about 0-7 by
0-03 mm,

48 A MONOGRAPH OF THE VICTORIAN SPONGES.

The variability and frequent irregularity of the spicules in this species render
them difficult to describe accurately, but it is hoped that the figures will serve to give

a correct idea of them.

(d.) Affinities—In external form this species bears a marked resemblance
to Heckel’s Ascometra variabilis* while in spiculation it approaches the Asculnis
form of the same species.t Heeckel’s species, however, differs from Leucosolenia
stolonifer in the presence of an annular swelling near one extremity of the oxeote
spicule and in the wavy curvature of the rays of the quadrivadiates, moreover
triradiate spicules are almost always present. Carter’s <Aplroceras asconoides,
from near Port Phillip Heads, perhaps makes a nearer approach to our species and
I at first believed the two to be identical. In Leucosolenia asconoides, however, judging
from Carter’s description! the oxeote spicules are much larger and symmetrically
fusiform, while the quadriradiates have much slenderer rays and none of them
appear to have the greatly enlarged apical ray characteristic of Leucosolenia stolonifer.
I have five distinct colonies of this species and in all five the enlarged apical rays
projecting into the gastral cavity form a conspicuous feature.

(e.) Locality—Near Port Philip Heads. (Coll. J. B. Wilson.)

3. Leucosolenia asconoides, Carter, sp.

A phroceras asconoides, Carter, Annals and Magazine of Natural History, August
1886, p. 134.

(As I have not had the opportunity of examining this species myself I must rely
entirely on Mr. Carter’s description, which, in conjunction with the manuscript
illustrations of the species which he has kindly sent me, appears to me to be
sufficiently detailed for recognition.)

(a.) General Appearance and Canal System.—The sponge consists of a colony
of Ascon-persons growing from a contracted base. The individuals are long,
narrow, tubular, sessile, somewhat compressed, diminishing in size towards the free
end, which is truncate, and contracted towards the other, which is fixed. They attain a
height of about 87 mm. and a diameter of nearly 8 mm. ‘They often braneh
slightly. The osculum, at the upper end of the tube, is not surrounded by any
fringe. In dry specimens the outer surface is even, glistening, composed of a layer

* “Die Kalkschwiimme,” Vol. III., Pl. 18, Fig. 9.
+ Op. cit. Vol. II., p. 108; Vol. III., Pl. 16, Fig. 4.
} Annals and Magazine of Natural History, August, 1886, p. 134.

A MONOGRAPH OF THE VICTORIAN SPONGES. AY

of very large oxeote spicules arranged longitudinally and very near together, between
which are the inhalant pores. The wall of the tube is 0°2 mm. in thickness. The
colour (dry or in spirit ?) is yellowish white.

(b.) Arrangement of the Skeleton.—The skeleton consists of quadriradiate and
oxeote spicules, arranged in the normal manner, the quadriradiates forming an
internal layer with the apical rays projecting into the gastral cavity and the oxea
forming an external layer. Judging from the description the oxea do not project
beyond the outer surface of the tube but the surface is minutely hispid from the
projection of some of the facial rays of the quadriradiates.

(c.) The Spicules.

(1) Ouadrivadiates—These are comparatively small and delicate spicules,
with very slender rays. They are more or less sagittal in form,
with a long basal ray and two very much shorter oral rays extended
almost at right angles to the basal ray. The basal ray is perfectly
straight and tapers gradually to a fine point. The oral rays are also
finely and gradually pointed, but curve slightly backwards towards the
apex of the basal ray. The basal ray measures about 0°375 by 0-004
mm. and the oral rays about 0°18 by 0-004 mm.

(2.) Oxeotes—These are very long and stout, symmetrically fusiform,
shehtly curved and gradually sharp-pointed at each end, averaging
about 2°0 mm. in length by 0°1 mm. in diameter.

(d.) Affinities.—Although Mr. Carter places this species amongst the Leucons,
yet his description makes it perfectly clear that it is really one of the Homocela
simplicia and nearly related to Leucosolenia stolonifer. If the description be correct,
however, as there is no reason to doubt it is, it 1s readily distinguishable from that
species by the size and shape of the spicules, and possibly also by the external form,
for there appears to be no spongorhiza and the individuals are stated to grow from
a contracted base. (Vide also the remarks on the affinities of Leucosolenta stolonifer,

supra.)
(e.) Locality,x—Near Port Phillip Heads. (Coll. J. B. Wilson.)

Srorion 2. Reticulata.

The sponge-colony forms a more or less complex network of branching and
anastomosing tubes, and it is no longer possible to distinguish the individual Ascon-
persons of which the colony is composed,

50 A MONOGRAPH OF THE VICTORIAN SPONGES.
Subsection 1. /ndivisa.

The gastral cavities of the Ascon-tubes composing the colony retain the
primitive hollow condition, there being no ingrowth of the mesoderm or endoderm.

4, Leucosolenia dubia, n, sp.
(GEG Moy oie, aye tele IDK Ihe, G3.)

(a.) General Appearance and Canal System.—The sponge-colony forms irregular,
low-growing masses, composed of a network of branching and anastomosing, more or
less cylindrical Ascon-tubes. ‘The size of the interspaces between the tubes varies
ereatly in different specimens and, indeed, in different parts of the same specimen.
No definite external skin (pseudoderm) is formed, the interspaces between the surface
tubes being as wide as in the interior of the colony. In compact colonies the surface
has a curious vermiculated appearance, in loose ones (PI. I., Fig. 3) it is a lattice-work
with oval or rounded meshes. On the under surface of the colony are sometimes
present little root-like processes, which serve for attachment. The oscula, when
discoverable, are visible as very small, round apertures, situate on the apices of small
papillz formed by the anastomosis of several Ascon-tubes.

The diameter of the cylindrical Ascon-tubes varies a good deal in different parts
of the same colony, say from 0.5 to 1mm. ‘To the naked eye some of the tubes
appear smooth on the outer surface, others minutely but distinctly hispid. The
colour im spirit is white or nearly so. Further details as to the canal system, which
conforms to type A, will be found on p. 28.

(b.) Arrangement of the Skeleton—The skeleton consists of triradiate and
oxeote spicules. The former are arranged in a confused layer in the mesoderm, the
spicules being placed close together with rays overlapping. The latter occur only in
certain parts of the colony, namely, in some, and apparently only some, of the
outer Ascon-tubes. Hence a boiled out preparation from one part of a specimen
may show no oxeote spicules at all, while another from a different part of the same
specimen contains plenty of them. When present the oxeotes project more or less
from the outer surface of the Ascon-tubes, whence the hispid appearance of the latter.

(c.) The Spicules (Pl. 1X., Fig. 3.)
(1.) Trivadiates—These are approximately regular or slightly sagittal.
The rays are fairly stout, conical or subfusiform in shape, fairly

sharply but rather abruptly pointed, often a little irregular in diameter
and mode of pomting ; measuring about 0-18 by 0-014 mm.

A MONOGRAPH OF THE VICTORIAN SPONGES. 51

(2.) Oxeotes.—These are stout, club-shaped, often irregularly and strongly
curved, very broad near one end and tapering gradually to a fine point
at the other; abruptly but sharply pointed at the broad end. The
size of these spicules is very variable, up to 0°48 by 0:02 mm.

Tn addition to the spicules above-mentioned I have met with a few quadriradiates,
which do not seem, however, to be characteristic but only occasional.

(d.) Affinities—There are probably several distinct species which closely resemble
Leucosolenia dubia in general form and canal system, as, for example, certain forms of
Heckel’s Ascetta coriacea and A. clathrus, which differ from it, however, in spiculation.
Tn one character Leucosolenia dubia strikingly resembles L. cavata, to be described
later on, namely, in the presence of great numbers of ‘‘ yellow granules”? embedded in
the mesoderm. The structure and meaning of these bodies has already been discussed.
They appear to be of constant occurrence in the species in which they are found and
remind one in this respect of the curious dumb-bell-shaped filaments, also of unknown
significance, present in the Hircinide amongst horny sponges. Indeed, it is by no
means improbable that the specimens which I distinguish as Leucosolenia dubia may
be young forms of L.cavata, for the general organisation (including § spiculation)
resembles closely that of the first-formed part of a colony of L.cavata. L. dubia,
however, attains a considerable size (the largest specimen being about 62 mm. in
greatest breadth and 20 mm. thick) without acquiring the very peculiar canal system
of L. cavata.

(e.) Locality.—Near Port Phillip Heads. (Coll. J. B. Wilson.)

5. Leucosolenia stipitata, n. sp.
(ER Bigs 4ios Ge Pili Vshie. 2): Pl. DX. Bigs 5.)

(a.) General Appearance and Canal System.—This is a very minute sponge,
consisting of a more or less oval body perched on the summit of a short, slender
stem attached at its lower extremity to some foreign body. ‘The total height of the
sponge is only about 12 mm. Figures 4 and 5, Plate I., show two specimens of the
natural size, and Figure 6, Plate I., one of them greatly enlarged. ‘The stem is solid
and cylindrical. The body has the reticulate surface characteristic of this section of
the genus, the pseudopores being rather large in proportion to the size of the sponge.
On the summit of the body is a relatively large osculum. Only one osculum is present
in the specimens figured, but I should imagine that there might sometimes be more
in older specimens. The colour in spirit is white.

The canal system conforms to type B., and is described on p. 28 and illustrated
in Figure 2, Plate IV.

52 A MONOGRAPH OF THE VICTORIAN SPONGES.

(b.) Arrangement of the Skeleton—The skeleton consists of triradiate spicules
only, arranged in a single layer in the mesoderm. In the outer skin (pseudoderm)
the spicules are arranged fairly regularly, in a single layer, with all the basal rays
pointing towards the stem. In the stem the spicules are very densely and rather
irregularly arranged, but still with a large proportion of the basal rays pointing
downwards.

(c.) The Spicules (Pl. TX., Fig. 5).—As already stated these are all triradiate ;
they are more or less sagittal, and the rays are all straight, conical or subfusiform,
and gradually sharp-pointed. In the spicules of the pseudoderm the basal ray
measures up to 0'1 by 0°01 mm., and the lateral rays about 0-07 by 0:0085 mm.
The spicules of the deeper parts of the body are more nearly regular and somewhat
slenderer than those at the surface. The rays of the dermal spicules are slightly
curved, in such a manner that the centre of the spicule is uplifted above the plane
in which the apices of the three rays lie.

(d.) Affinities—This species very closely resembles von Lendenfeld’s A scetta
macleayi* in external form and spiculation; indeed for a long time I regarded
the two as identical, but if von Lendenfeld’s description of the canal system in
Leucosolemia macleayi be correct the two species differ very markedly in this respect.
According to von Lendenteld Ascetta macleayi is one of Heckel’s ‘‘Auloplegma ”’
forms, with a pseudogaster and pseudosculum, while in Leucosolenia stipitata there is
no pseudogaster and the osculum is a true osculum. It appears to me very desirable
to distinguish by different names forms which differ so widely in anatomical
characters, but this question has been already discussed in the section of the
present work dealing with classification.

The species closely resembles one form of Miklucho-Macleay’s Guancha blanca,+
but considering the great variety of forms included under that name and the fact that
the spicules appear to be much stouter and more sharply pointed in Leucosolenia
stipitata it seems best to give a distinct name to the latter.

(¢.) Locality.—Near Port Phillip Heads. (Coll. J. B. Wilson.)

6. Leucosolenia pulcherrima, n. sp.
CAME pI BVewari pe tele Wi/es Tivos, Guat Teil, OX. Tove 3)

(a.) General Appearance and Canal System.—The sponge consists of a relatively
large body perched on the end of a short stalk. Figure 7, Plate I., represents a
typical example multiplied seven diameters. The body is irregularly ovoid in shape,

* Proceedings of the Linnean Society of New South Wales, Vol. IX., Part 4, p. 1086.
+ Jenaische Zeitschriff, Band IV., 1868.

A MONOGRAPH OF THE VICTORIAN SPONGES. ats)

compressed laterally, and with a somewhat flattened upper surface on which are
several true oscula each surrounded by a membranous oscular fringe. The sides of the
body exhibit a number of narrow ridges, of which the principal ones run vertically,
branching and anastomosing more or less and connected by less pronounced, more or
less transverse ridges ; between the ridges are numerous interspaces of irregularly
rounded shape, these are the pseudopores. ‘The height of the whole colony is only
about 10 mm. and the greatest breadth about 6 mm. (in the specimen figured). In
some specimens the ridges are much less regularly arranged, and the oscular fringes
not so well developed. The ridges are, of course, the outer surfaces of the outer
Ascon-tubes of the colony. ‘The colour in spirit is white.

The canal system conforms to type B (described on p. 28) and 1s illustrated
in Figure 3, Plate IV.

(b.) Arrangement of the Skeleton.—The skeleton consists of triradiate spicules
only, arranged in the usual manner. In the outer Ascon-tubes the spicules are often
sagittal and the basal ray then points downward as usual.

(c.) The Spicules (Pl. X., Fig. 8).—The triradiate spicules vary considerably in
form according to their position in the sponge, so that we may roughly distinguish
between deep and dermal spicules. The dermal spicules are sometimes nearly
equiangular and equiradiate, but with the rays strongly curved so that the centre of
the spicule is uplifted considerably above the level of the plane in which the apices
of the three rays le. The rays are stout, conical and not very sharply pointed,
measuring about 0°056 by 0:008 mm. Very frequently, however, the dermal spicules
become markedly sagittal, the lateral (or oral) rays embracing a much wider angle
(nearly 180° in some cases) and curving somewhat towards one another, while the
basal ray may be elongated to about O0'l mm. ‘The deep spicules, lying in the walls
of the internal Ascon-tubes, are approximately equiangular and equiradiate, with the
rays lying pretty much in one plane, so that the spicule would le almost or quite
flat on a plane surface. The rays are straight and slender, of nearly the same
diameter all along until close to the apex and then rather abruptly and not very
sharply pointed. They measure about 0°084 mm. in length by 0°0042 mm. in
diameter.

(d.) Affinities—This species is obviously closely related to L. stipitata and I
should not be at all surprised to find a perfect series of intermediate forms. Typical
examples of each are, however, easily distinguishable by their external appearance,
and they may conveniently receive separate names.

(e.) Locality.—Neaw Port Phillip Heads. (Coll. J. B. Wilson.)

54 A MONOGRAPH OF THE VICTORIAN SPONGES.

7. Leucosolenia pelliculata, n. sp.
(DE TMU, Iie Ae IB WON. Tiiey, 7/3 IP Oo, IDS, Ip 2s)

(a.) General Appearance and Canal System.—The sponge forms rather small,
low-growing, lobular masses of irregular shape (Pl. III., Fig. 2,) and with a very
well marked outer skin or pseudoderm. The largest specimen in the collection is
scarcely 87 mm. in greatest diameter. The skin is smooth and somewhat glabrous.
Here and there over the surface of the lobes are scattered small, true oscula,
generally surrounded by a very slightly developed oscular fringe. The general
external appearance of the sponge varies a good deal on account of the variation in
the size of the pseudopores. In some specimens these are very minute and very
closely arranged, so small that to the naked eye the surface has merely a minutely
punctate appearance, while in others they are a little over 0°5 mm. in diameter.
Even in different parts of the same specimen they vary much in size. The small
size of the pseudopores causes the skin to appear unusually well developed, whence
the specific name pelliculata.

The colour of the sponge in spiit may be pure white, pale brown or even
pinkish.* The canal system belongs to type B and agrees precisely with that of
Leucosolema pulcherrima, figured on Plate IV., Figure 3. The walls of the Ascon-tubes
are characteristically thin and delicate, and, as is usual in such cases, the prosopyles
are very clearly shown. Even the outermost Ascon-tubes generally retain their wide
gastral cavity, so that the skin (pseudoderm) is formed out of the outer walls only of
the outer tubes, strengthened by specially large spicules.

(b.) Arrangement of the Skeleton.—The skeleton consists of quadriradiate and
triradiate spicules, arranged in a perfectly normal manner ; the spicules in the pseudo-
derm are markedly larger than those inside the colony and the spicules in the oscular
membrane have a strong tendency to become sagittal. The apical rays of the
quadriradiates, of course, project freely into the gastral cavities of the Ascon-tubes.

(c.) The Spicules.—As these vary somewhat in different specimens, especially in
size, I propose to give separate accounts of the spiculation of four specimens, which
we may call A, B, C and D, and in this way to give a fair idea of the extent of
variation. The spicules of the two extreme forms (A and D) are figured (Pl. X.,
Figs. 1, 2.)

Specomen A. (Pl. X., Hig. 1.)

(1.) Drivadiates—Dermal. Regular; rays conical, fairly gradually sharp-
pointed, measuring about 0:15 by 0-014 mm. Deep. Of the same
form as the dermal spicules but with smaller and slenderer rays,
measuring 0°11 by 0:0085 mm.

* Possibly stained by other specimens in the same jar with the sponge.

A MONOGRAPH OF THE VICTORIAN SPONGES. 55

(2.) Quadriradiates—Abundant. Of the same size and shape as the deep

triradiates except for the usually perfectly straight apical ray projecting
from the centre at right angles to the facial rays. The apical ray is
well developed and gradually and very sharply pointed, but not quite
so long as the facial rays and a trifle slenderer.

In the oscular membrane many of the spicules are markedly sagittal.

Specimen B.

(1.) Trivadiates—Dermal. Regular; rays conical or slightly fusiform,

gradually sharp-pointed, measuring about 0°24 by 0°03 mm. Deep.
Regular; rays slender, conical, fairly sharp-pointed, measuring about
0-14 by 0:01 mim.

Quadrivadiates.—Oft about the same size and shape as the deep
trivadiates, but with an apical ray projecting at right angles from the
centre. The apical ray is usually straight, slightly fusiform, not quite
so long as the facials, very slender and finely pointed.

In the oscular membrane the spicules often become beautifully sagittal, both

quadriradiates and triradiates having the oral rays curved backwards towards the
basal ray, which is rather longer than the orals. (Here, and in the other specimens
also, the spicules of the oscular fringe are of about the same size as the deep
triradiates.)

Specimen C.

(1.) Trivadiates—Dermal. (a) Regular, large and stout; rays conical or

subfusiform, gradually sharp-poimted, measuring about 0°25 by 0-035
mm. (b) Sagittal, occurring in small numbers around some of the
pseudopores and probably also in the oscular membrane. ‘These are
approximately equiangular and the rays are about equal in length,
but the oral rays are curved symmetrically, while the basal ray 1s
straight. Rays conical, sharp-pomted, measuring about O-1 by
0-‘Ol mm. Deep. Regular; rays slender, conical or slightly fusiform,
sharp-pointed, measuring about 0°15 by 0:014 mim.

(2.) Quadriradiates.—These are scarce and also of two forms. (a) Regular,

like the deep triradiates in shape and size but with a straight, slender
apical ray projecting from the centre. (0) Sagittal, found in the
oscular membrane and (?)around some of the pseudopores ; much
like the sagittal triradiates in size and shape but with a long and well-
developed, though slender apical ray, curving gradually upwards,
sickle-like, from the centre of the spicule and gradually and_ finely
pointed.

56 A MONOGRAPH OF THE VICTORIAN SPONGES.

Specimen D. (PI. X., Fig. 2.)
(1) Triradiates—Dermal. Regular, large and stout. Rays conical or
slightly fusiform, gradually and sharply pointed, measuring 0°3 by
0-035 mm., or a little less. Deep. Regular, rays slender, conical,
fairly sharp-pointed, measuring about 0°14 by 00085 mim.

(2.) Quadrivadiates—Of about the same size and shape as the deep
triradiates, but with an apical ray projecting at right angles from the
centre. The apical ray is usually straight, not quite so long as the
others, very slender and finely poimted, slightly fusiform. There are
also a few larger quadriradiates amongst the dermal spicules.

The spicules of the oscular membrane are, as usual, distinctly sagittal, with
backwardly curved oral rays.

From the above details it will be seen that the variation in spiculation concerns
chiefly the dermal triradiates, the rays of which range from 0:15 by 0-014 to 0:3 by
0-035 mm., and the quadriradiates, which are sometimes scarce and sometimes
abundant, with the apical ray variously developed. In all specimens the spicules of
the oscular membrane tend to become markedly sagittal and a similar state of things
is sometimes observable around the pseudopores.

(d.) Affinities.—This species would come, of course, under the genus Ascaltis of
Heeckel’s system, but I know of no other species which resembles it sufficiently
closely to demand special notice. The sagittal triradiates and quadriradiates in the
neighbourhood of the oscula appear to be very characteristic.

(¢.) Locality.—Near Port Phillip Heads. (Coll. J. B. Wilson.)

8. Leucosolenia cavata, Carter, sp.
(HUES ae Se Welt, he IP WL, lies, 4 Be FP, 1X. Fig. 4.)

Clathrina cavata, Carter. Annals and Magazine of Natural History, June,
1886, p. 502.

(a.) General Appearance and Canal System.—The sponge (PI. II., Fig. 7) forms
massive colonies of considerable size and with a fairly even surface. The colony
grows vertically upwards, and is conspicuously flattened on the top and generally
also more or less flattened from side to side, sometimes very much so. The lowest
portion of the colony, attached to the substratum, consists of a number of separate,
slender Ascon-tubes ; the greater part is, however, enclosed in a common external

A MONOGRAPH OF THE VICTORIAN SPONGES. 57

skin (pseudoderm) pierced by the numerous, closely placed pseudopores. The
pseudopores are nearly circular and rather small, ranging up to about 0-564 mm. in
diameter, they are abundant on the sides of the sponge but rare on the upper surface.
On the upper, flattened surface of the colony are numerous good-sized oscula, each
surrounded by a very well developed oscular fringe, and averaging about 3 mm.
in diameter. The largest specimen in the collection is represented in Figure 7,
Plate II. It is about 110 mm. in height, and over 50 mm. in its lesser horizontal
diameter at the top. Unfortunately the specimen had to be cut to make it go into
the bottle, so its greater horizontal diameter cannot be determined. The colour in
spirit is white.

Full details of the canal-system are given on pp. 29, 80. It belongs to the
‘reversed ”’ type (type C) and is illustrated in Figures 1, 2, Plate V. The peculiar
‘yellow granules,’ which seem to be characteristic of the species, are described and
their nature discussed on p. 18 ef seg. (Pl. VI., Figs. 4, 5.)

(b.) Arrangement of the Skeletonn—The skeleton consists mainly of triradiate
spicules lying in a rather confused layer in the mesoderm (Pl. VI., Fig. 5). A
good many quadriradiates are also present, with the apical rays as usual projecting
into the gastral cavities of the Ascon-tubes. A very few oxeotes also occur in boiled
out preparations.

(c.) The Spicules (Pl. 1X., Fig. 4).

(1) Trivadiates.—Dermal. Regular. Rays conical, tapering a little more
suddenly towards the apex, which is fairly sharp, measuring about
0°16 by 0°014 mm. Deep. The same as the dermal, but with the
rays perhaps a little stouter.

(2.) Quadrivadiates.—Like the triradiates but with a small apical ray,
which is slender, gradually sharply pointed and slightly curved.

(3.) Oxeotes—These are rare, small, asymmetrically fusiform, generally
slightly and irregularly curved, measuring about 0°2 by 0.0083 mm.

(d.) Affimties—My. Carter’s description of this species is not very satisfactory,
as he mistook the true nature of the arrangement of the canal system, stating that
the oscula are “only in communication with the dilated parts of the interspaces,”
whereas, as we have seen, they are really true oscula and what Mr. Carter considered
as interspaces are gastral cavities. The very peculiar reversal of the canal system is
extremely misleading, and this doubtless gave rise to Mr. Carter’s error. Comparison
with a piece of one of Mr. Carter’s type specimens kindly sent to me from the
British Museum has convinced me that my specimens are identical with Mr. Carter’s
Clathrina cavata.

58 A MONOGRAPH OF THE VICTORIAN SPONGES.

I have already indicated the possible identity of Leucosolenia cavata with
L. dubia in the description of the latter. The species is common near Port Phillip
Heads.

(c.) Locality.—Near Port Phillip Heads. (Coll. J. B. Wilson.)

9. Leucosolenia protogenes, Heckel sp.
(Pian Bice 1; Be) ie et)

Ascetta primordialis, var. protogenes, Heckel. ‘‘Die Kalkschwamme,” Vol.
los TOs Lute

Ascetta procumbens, von Lendenfeld. Proceedings of the Linnean Society of New
South Wales, Vol. IX., Part 4, p. 1086.

Clathrina primordialis, Carter, Annals and Magazine of Natural History, June,
1886, p. 510.

(a.) General Appearance and Canal System.—The colony (Pl. I1., Fig. 1) forms
large, irregularly rounded, lobose masses, sometimes attaining a diameter of as much
as 125 mm. and attached by the lower surface to the substratum. The texture of the
colony is characteristically soft, spongy and friable. The surface of the colony is thrown
into strong, irregular, rounded ridges and depressions. ‘The pseudopores are much
more abundant in the depressed portions of the surface than on the ridges, where
the outer skin (pseudoderm) is better developed ; they are irregularly rounded or oval
in shape and measure about 1 mm. in maximum diameter. On the upper surface of
the colony are numerous large circular pseudoscula, arranged along the tops of the
ridges and each with a rather feebly developed membranous collar surrounding it. The
pseudoscula measure about 6 mm. in diameter and lead each into a distinct, tubular

pseudogaster which penetrates deeply into the substance of the colony. The colour
in spirit is nearly white.

The canal system conforms to type D, and hence agrees with that of Leucosolenia

ventricosa, which will be found described on pp. 80, 31, and illustrated in Figure 4,
Plate IV.

(b.) Arrangement of the Skeleton.—The skeleton consists of trivadiate spicules
only, and these are arranged in the normal manner ; in a dense, confused layer, with

freely overlapping rays in the pseudoderm, and less densely packed in the walls of the
internal Ascon-tubes.

A MONOGRAPH OF THE VICTORIAN SPONGES. 59

(c.) The Spicules (Pl. XI., Fig. 1).—Triradiates only, of slightly different size and
shape according to whether they le in the pseudoderm or in the deeper parts of the
colony. Dermal. These are equiradiate and equiangular (regular), fairly sharply
pointed ; the rays are almost cylindrical till pretty near the apex when they begin to
taper off gradually, they measure about 0°14 by 0:0186 mm. Deep. Like the
preceding but with slenderer rays measuring about 0:14 by 0.009 mm. Numerous very
small trivadiates occur amongst these as in other species and as figured by Heckel
for Ascetta primordialis; these are doubtless young forms. ‘The spicules of the
pseudoscular fringe and of the membrane which lines the pseudogaster resemble
the ordinary deep-seated trivadiates.

(d.) Affinitics.—The identification of the four large specimens of this species in
the collection with Carter’s Clathrina primordialis is a safe one, because not only
have I his long, published description and manuscript figures to go by, but also a
piece of one of his own types from the British Museum. Mr. Carter’s identification of
his specimens with Heckel’s Ascetta primordialis is more open to doubt but appears
to me to hold good if restricted to Heckel’s variety Ascetta primordialis vay.
protogenes, although in our specimens the dermal spicules are slightly thicker than
the deep ones. I do not however place any great importance on so slight a
distinction, although the specimens thereby make an approach to Heckel’s A scetta
primordialis vax. poterium with its stout dermal spicules.

Poléjaeff records in the case of the specimens in the ‘ Challenger” collection
identified by him as Leucosolenia poterium (2) the presence of sagittal spicules in
the pseudoscular membrane. I have found none such in my specimens of
Leucosolenia protogenes.

Heeckel’s Ascetta primordialis vax. protogenes includes of course specimens with
various types of canal system, but I propose to admit the specific, or at any rate
varietal, identity only of forms with the same type of canal system, and hence the
nomenclature becomes a matter of considerable difficulty. Inthe absence of definite
information as to the canal system of the original type of protogenes I propose at
present to restrict the name to specimens which conform to my type D of the
Reticulata.

My reason for considering that von Lendenfeld’s Ascetta procumbens is
synonymous with Leucosolenia frotogenes is that I have examined a fragment of one of
his type specimens from Port Jackson, sent to me from the British Museum, and can
find no poimts of distinction. Von Lendenfeld says, loc. cit.—‘* Our species is
distinguished from the allied species by the rays of the spicules being neither
cylindrical as in Ascetta coriacea nor pointed as in the numerous varieties of Ascetta
primordialis. Besides that, the spicules are shorter than in the latter and thicker

60 A MONOGRAPH OF THE VICTORIAN SPONGES.

than in the former.’ The fragment (named by Dr. von Lendenfeld) examined by me
shows nothing distinctive in the spiculation, and I should certainly say that the
spicules, at any rate many of them, are pointed. In his description of Ascetta
procumbens, von Lendenfeld says—‘‘ The sponge consists of numerous slightly curved
cylindrical tubes, extending in one plane in one or more layers. The sponge has the
appearance of a perforated plate, and attaims a diameter of 25 and a thickness of
2°5 mm.” One of his figures (Pl. LXI., Fig. la) agrees with this description, but
three others, said to be of the same species (Pl. LXI., Figs, 10, 1c, 1d), represent a
massive, lobose sponge of very different appearance. The fragment also, sent to
me from the British Museum, is evidently a portion of a massive sponge. Hence it
seems not impossible that von Lendenfeld has confused two species, the description of
the external form and spicules being taken from specimens which may perhaps differ
from Leucosolenia protogenes. The canal system is not properly described, nor is the
fragment in my possession sufficiently perfect to allow me to completely elucidate it,
but this fragment so closely resembles, in all the characteristics which it exhibits,
the Victorian specimens of Leucosolenia protogenes, that I have little hesitation in
identifying it. Moreover, von Lendenfeld gives Port Phillip as a locality for his
Ascetta procumbens.

(e.) Locality.—Near Port Philip Heads. (Coll. J. B. Wilson.)

10. Leucosolenta ventricosa, Carter, sp.
(Cele Uc, Tere tes INOS JEG Wvos lives, ALS TRI, OX Iie, 4b)

Clathrina ventricosa, Carter, Annals and Magazine of Natural History, June,
1886, p. 512.

(a.) General Appearance and Canal System.—The sponge (Pl. I., Figs. 8, 9, 10)
forms large, irregularly lobose masses with nearly even or very uneven surface,
each colony forming a thick wall around a very large pseudogaster which penetrates
to the base of the sponge. Several such colonies may occur more or less fused
together side by side, each with its large pseudogaster. Well-grown specimens
measure about 75 mm. in height and the same in diameter. Each pseudogaster
opens at the top of the colony through a wide pseudosculum surrounded by an
extension of the lining membrane of the pseudogaster in the form of a pseudoscular
fringe. The pseudopores are thickly and pretty evenly distributed in the pseudoderm,
varying in size up to about 1 mm. in diameter, and mostly of oval shape.

From the lower surface of the colony short, root-like processes are given off

A MONOGRAPH OF THE VICTORIAN SPONGES. 61

which attach the sponge to the substratum on which it grows. The colour in spirit

is white and the texture of the sponge is characteristically harsh and unyielding to
the touch.

The canal system conforms to type D, and is described in detail on pp. 30, 31,
and illustrated in Fig. 4, Plate IV.

(b.) Arrangement of the Skeleton.—The skeleton consists of triradiate, quadri-
radiate and oxeote spicules. The triradiates form nearly the entire skeleton and are
arranged in an irregular layer in the walls of the Ascon-tubes and in an especially
dense layer of especially large spicules mixed with smaller ones in the pseudoderm.
The quadriradiates are relatively few but easily discoverable and even abundant in
some specimens, they occur in the walls of the Ascon-tubes, with the apical ray
projecting into the gastral cavity, and in the lining membrane of the pseudogaster
with the apical ray projecting into the latter. The oxeote spicules vary greatly as to
the numbers in which they occur in different specimens. ‘Their position, however,
appears to be constant. They have one end embedded in the margin of the
pseudopore, while the greater part of the spicule projects freely, sometimes singly
and sometimes in dense tufts, towards the centre of the pseudopore. In some
specimens the pseudopores literally bristle with these tufts of oxeotes converging

towards but not reaching the centre of the space, in others only a few oxeotes can be
found,

(caine Sprciles (Pl. xX.) Big. A).

(1.) Trivadiates.—Dermal. Regular ; rays conical or subfusiform, gradually
sharp-pointed, very stout, measuring about 0°35 by 0:056 mim.
(Numerous smaller spicules occur mixed with these.) Deep. Regular ;
rays conical or subfusiform, gradually sharp-pointed, measuring
0-17 by 0:014 mm. Numerous intermediate sizes occur.

(2.) Quadriradiates.—Ot the same shape and size as the smaller triradiates,
but with a short, slender, curved and sharply pointed apical ray.

Sometimes the smaller spicules in the pseudosecular membrane
become strongly sagittal, the oral rays being curved backwards
towards the basal.

(3.) Oxeotes.—Short, slender, straight, gradually sharply pointed at both
ends, measuring about 0°16 by 0-003 mm.

(d.) Affinities.—This species, which Mr. Carter characterises as ‘by far the
largest and most abundant in specimens of all Mr. Wilson’s calcareous sponges "’ is
easily recognisable by both general appearance and spiculation. My. Carter makes no

62 A MONOGRAPH OF THE VICTORIAN SPONGES.

mention of the oxeote spicules, but as I find them present in one of his type
specimens sent to me from the British Museum this must be an oversight on
his part.

(c.) Locality.—Near Port Phillip Heads. (Coll. J. B. Wilson.)

10a. Leucosolenia ventricosa var. solida, nov.
(Plate III., Fig. 3.)

This variety, represented in the collection by two good specimens, differs from
the typical form in the flattened, cake-like shape and the suppression of the pseudogaster.
The exhalant canal system is reduced toa shallow pit varying up to about a quarter
of an inch in diameter, in the bottom of which are visible several small cireular
openings (exhalant openings of the Ascon-tubes). Around this pit there is a slghtly-
developed fringe. Numerous pits of this kind occur on the upper surface of each
specimen. It is doubtful whether one is justified any longer in speaking of a
pseudogaster and pseudosculum in this variety, for the collared cells appear to
approach very closely to the edge of the pit, at any rate in some cases. Perhaps it
would be better to regard the pits as groups of true oscula or even each as a single
true osculum formed by the union of many Ascon-tubes. Hence the canal system is
referable rather to type B than to type D.

In spiculation this form is identical with Lewcosolenia ventricosa as described
above, and as the spiculation is very characteristic I therefore regard it as a well-
marked variety of that species.

Locality.—Near Port Phillip Heads. (Coll. J. B. Wilson.)

Subsection 2. Subdivisa.

The gastral cavities of the Ascon-tubes are more or less subdivided into
chambers by ingrowths of mesoderm or of both mesoderm and endoderm.

11. Leucosolenia proxima, n. sp.

(Pl. IL, Figs. 1,2; Pl. VIIL, Figs. 1, 2, 8, 4; Pl. XI, Fig. 2).

(a.) General Appearance and Canal System.—The sponge forms small colonies
attached by little, solid, root-like processes to foreign objects (e.g. Polyzoa). The form

A MONOGRAPH OF THE VICTORIAN SPONGES. 63

of the colony is irregular and somewhat variable and will be best understood by
reference to Figures 1, 2, Plate II., representing two slight varieties in one of which
the oval pseudopores are conspicuously larger than in the other. On the upper
surface of the colony are numerous conical projections, each having a small osculum
at its summit. The greatest diameter of a colony is usually only about 6 or 7 mm.
The colour in spirit is nearly white.

The canal system belongs to type HK. and is fully described on pp. 32, 33
and illustrated in Figures 1, 2, Plate VIII.

(b.) Arrangement of the Skeleton—The skeleton consists of triradiate spicules
only, arranged as usual in the thickness of the mesoderm, the dermal spicules being
larger than the deep ones.

(c.) The Spicules (Pl. XI., Fig. 2).—Dermal, Regular; rays stout, conical,
gradually and sharply pomted, measuring about 0°16 by 0.021 mm. Deep.
Similar to the dermal spicules but with the rays shorter and scarcely half as stout.

(d.) A ffinities—Were it not for the peculiar ingrowth of the mesoderm into the
gastral cavities this species would present no peculiarity and would come near to
Leucosolenia pulcherrima; as it is it forms a very interesting connecting link between
the species with simple gastral cavities (Imdivisa) and such forms as Lezcosolenia
wilsont.

(e.) Locality.x—Near Port Phillip Heads. (Coll, J. B. Wilson.)

12. Leucosolenta witlsonz, n. sp.
(Ebi icin ote bl, Vil: Pl, Xf. Kiev 3.)

(a.) General Appearance and Canal System.—The sponge forms a low-growing,
much flattened, spreading mass (PI. IL., Figs. 8, 4), sometimes growing out into
rounded lobes (as atin Figure 4). From the lower surface arise a few short,
root-like processes which serve to attach the sponge to the substratum. On the
upper surface, at irregular intervals, occur small conical papille, each with a minute
osculum at its apex. On the upper surface the pseudopores are mostly elongated,
narrow and slit-like, on the lower surface they are more rounded. The greatest
horizontal diameter of the single adult specimen in the collection was about 385 min.
and its thickness very variable. In spirit the specimen has a characteristic dead
white, solid appearance, due to the partial filling up of the Ascon-tubes.

The canal system (Pl. VII., Figs. 1, 2, 3) belongs to type F and is fully

64 A MONOGRAPH OF THE VICTORIAN SPONGES.

described on p. 34. The developing embryos are contained in separate spherical
cavities lined by very large, polygonal, plate-like cells (PI. VIL., Fig. 3).

(6.) Arrangement of the Skeleton—The skeleton consists of triradiate spicules
only, arranged as usual in the thickness of the mesoderm and often, especially in the

thick pseudoderm, in several layers.

(c.) The Spicules (Pl. X1., Fig. 8).—Dermal. Regular triradiates ; rays stout,
conical, gradually sharp-pointed, measuring about 0.14 by 0.014 mm., or a
trifle stouter. Deep. Like the dermal spicules but with the rays a little shorter

and slenderer, still fairly stout.

(d.) Affinities.—This species obviously comes very near to Heeckel’s Ascetta
primordialis var. loculosa* ut it is no easy matter to determine what <Ascetta
primordialis var. loculosa veally is, as the name is evidently applied by Heeckel to a
number of forms which differ widely in canal system and must therefore, according

to our view, be considered as distinct species.

Ascetta loculosa, H. (Ascetta primordial vay. loculosa ) is diagnosed by Heckel as
follows :—‘ Spicula alle von gleicher Grosse, einer doppelte oder mehrfache Schicht
im Exoderm bildend, ihre Strahlen 8-16 mal so lang als dick. | Kntoderm verdickt,
ein geschichtetes Epithelium bildend, welches durch blattformige Fortsatze oder
Septa die Darmhohle in Fiicher abtheilt. (Azs¢vaten).” As it is placed in the genus
Ascetta the spicules are, of course, all triradiate. ‘The only further description given
by Heckel of the remarkable species or variety /oca/osa 1s as follows :—‘‘ Hine andere
sehr bemerkenswerthe Varietiit, die ich anfangs fiir eine ganz verschiedene Art hielt,
bekam ich aus Siidaustralien (Ascetta Joculosa). ier war sowohl Hxoderm als
Entoderm verdickt. Das Exoderm enthielt mehrfache Nadelschichten und bildete
diinne membranoese Vorspriinge und Scheidewinde imneshalb der Darmrohren,
durch welche dieselben in Facher abgetheilt wurden. Diese Scheidewande enthielten
keine Spicula und trugen ein aus mehreren Schichten von Hpithelzellen gebildetes
Entoderm. Bei einigen Stocken fand sich in jedem Fach ein reifer Embryo, ganz
iihnlich wie bei Ascetta clathrus, wo dies Verhialtniss genauer beschrieben werden wird
(vel. Taf. 4, Fig. 4-7). Die Stocke von Ascetta loculosa, welche solche Facher in der
Darmhohle enthielten, wiirden im _ kiinstlichen System zu vier verschiedenen
Gattungen gehort haben, namlich Soleniscus, Tarrus, Auloplegma und Ascometra.”

Probably our Lewcosolenia wrlsont is identical with one or other of the forms
included by Heckel under the name Ascetta /oculosa, although without more exact
information as to the latter it is difficult to be certain of this. But, granting that it
is so, the question of nomenclature is still by no means settled, as we have to decide

* «Die Kalkschwiimme,’’ Vol, II., pp. 17, 23.

A MONOGRAPH OF THE VICTORIAN SPONGES. 65

to which of the various forms included by Heckel under the name Zocz/osa that name
should be restricted. None of these forms are figured and none are specially
described, but the first mentioned is the So/ezzscus, hence I propose to restrict the name
loculosa to a Soleniscus form. It must be remembered, in considermg Heeckel’s
account, that his term ‘“‘/voderm” includes both ectoderm and mesoderm. Although
Heckel found embryos he does not appear to have noticed the curious embryo-
capsules which are so conspicuous in Leucosolenia wilsonz, perhaps they do not occur
in any of Heckel’s forms, if so they might form a distinctive character for Leucosolenia
welsont.

Heeckel’s Ascetta clathrina* also exhibits the same peculiarity in regard to the
subdivision of the Ascon-tubes by ingrowths of mesoderm and endoderm
(‘‘endogastriche Septa,’ Hk.), but in that species, which Heckel regards as a variety
of Ascetta clathrus, the rays of the spicules are knobbed at the extremities.

The affinities of Lezcosolenta wilsoni to L. proxima have already been indicated
in describing the latter species.

(e.) Locahty.—Near Port Philip Heads. (Coll. J. B. Wilson.)

13. Leucosolenia depressa, Nn. 8).
(Pl ics, 4) 4ar abl Vill ic 8); Pl Xd, Hig. 4.)

(a.) General Appearance and Canal System.—The single specimen (Pl. ILI., Figs.
4, 4a) forms a flattened, spreading, irregular crust, attached to the substratum (if at
all) at a few points only. The greatest diameter of the specimen is a little over
100 mm, and the average thickness about 10 mm. or less. The surface is smooth
but irregular and undulating. On the upper surface are scattered numerous very
small, conical papille, each with a very minute osculum at its summit. On the
upper surface also the pseudoderm is very strongly developed, with comparatively few,
small pseudopores scattered in irregular patches here and there. On the under
surface, on the other hand, the pseudopores are larger and thickly scattered all over,
giving to it a sieve-like appearance. In spirit the colour of the sponge is greyish
white,

The canal system is like that of Lewcosolenta wrlsont (type F), but with the
ingrowths into the gastral cavities of the Ascon-tubes more irregular and less strongly
developed, consisting principally of irregular proliferations of the endoderm.

* «Die Kalkschwiimme,” Vol. 2, p. 31.

66 A MONOGRAPH OF THE VICTORIAN SPONGES.

(6.) Arrangement of the Skeleton—The skeleton consists of triradiate and
quadriradiate spicules arranged irregularly in one or more layers in the thickness of

the mesoderm.
(c.) The Spicules (Pl. X1., Fig. 4).

(1) Zyiradiates—Regular. Dermal. Large and stout; rays conical or
slightly fusiform, gradually sharp-pointed, measuring about 0-2 by 0-028
mm. Deep. Rays long and rather slender, fairly gradually and sharply
pointed, measuring about 0:14 by 0-008 mm.

(2) Ouadrivadiates.—These spicules are abundant, they resemble the deep
triradiates but have a feebly developed apical ray.

(d.) Affinities—I at first regarded this sponge as a variety of Leucosolenia
w2/sonz, which it closely resembles in external appearance and in canal system, but
considering the differences in spiculation it is perhaps better for the present to regard
it as a distinct species. The specimen contains numerous unsegmented ova*, in the
same position as the embryos in Z. wz/sonz, but there is no special epithelial capsule
visible around them, perhaps this is developed only after segmentation.

(e.) Locahty.—Near Port Phillip Heads. (Coll. J. B. Wilson.)

Section 8. Radiata.

The sponge consists of a single central Ascon-tube from which smaller tubes
are budded off radially.

14. Leucosolema tripodifera, Carter, sp.
(RT. Bigs: 0.6; bl V2, Bigs. 3,4; Ble Vie Micsw 5.6. aexelee Fig. 5.)

Leucetta clathrata, Carter, Annals and Magazine of Natural History, 1883,
Vol. XI., p. 83.

Clathrina tripodifera, Carter, Annals and Magazine of Natural History, June,
1886, p. 505.

Clathrina tripodifera var. gravida, Carter, Loc. cit., p-. 507.
(a.) General Appearance and Canal System.—As 1 have already given a full

* Described on pp. 17, 18.

A MONOGRAPH OF THE VICTORIAN SPONGES. 67

account of the external form and canal system of this remarkable sponge on pp. 35,
36, I need only refer the reader to that description and to the illustrations (PI. II.,
Figs. 5, 6; Pl. V., Figs. 3,4; Pl. VIIL., Figs. 5, 6) and pass on to describe the
arrangement of the skeleton and the form of the spicules.*

(6.) Arrangement of the Skeleton.—The skeleton consists of rather slender sagittal
triradiates and quadriradiates and of the “tripod” spicules (triradiate). The
triradiates and quadriradiates are arranged in a single layer, though of course with
frequently overlapping rays, in the mesoderm of the walls of the tubes, the
quadriradiates with the apical ray projecting freely into the gastral cavity (Pl. V.,
Fig. 4). As already stated the spicules are sagittal, and in the central Ascon-tube
the basal ray points towards the base of the sponge, while in the radiating tubes the
basal ray is directed towards the outer, blind end of the tube just as in the articular
tubar skeleton of the Sycons. The ‘ tripod” spicules form a protective layer on the
outside of the colony, being placed at the blind ends of the radiating tubes with all
three rays pointing inwards (PI1.V., Fig. 4).

(tie *Spiculgs (Pl. XN., Fie: 5).

(1.) Zrexadtates.—Dermal. These are the ‘ tripod” spicules and the
extent to which they are developed varies in different specimens.
When fully developed the rays are very stout and curved towards one
another in such a manner that if the spicule were placed with the
apices of the rays resting on a level surface the centre of the spicule
would lie very much above that surface; towards their apices the
rays curve outwards again. Rays stout, conical, gradually but not
very sharply poimted, measuring about 0°12 by 0.02 mm. Deep.
Sagittal, rays long and slender; angles about equal; rays straight,
or nearly so, fairly gradually and fairly sharply pointed. Basal ray
longer than the orals, averaging, say, about 0°16 by 0:0095 mm. when
the orals measure about 0-1 by 0:0075 mm., but the size of these
spicules is subject to a good deal of variation.

(2.) Quadrivadiates.—Like the deep trivadiates but with longer and
slenderer facial rays and with the addition of a shorter apical ray
generally rather strongly and often suddenly curved upwards. In a well
developed example the basal ray measures about 0:24 by 0°005 mm.
and the orals about 0°21 by 0:005 mm., but here agai there is a good
deal of variation, especially in the development of the apical ray.

(d.) Affinities.—I Nave a good many specimens of this species, which is very

* T may add that the colour of the sponge in spirit is grey, or nearly white, and the texture rather soft and delicate.
The peculiar structure of the endoderm is deseribed on p, 11.

68 A MONOGRAPH OF THE VICTORIAN SPONGES.

easily to be recognised. Curiously enough Mr. Carter, in his original description,
makes no mention of the sagittal character exhibited by most of the spicules, but
this character is very clearly shown in a portion of one of his types of the species
sent to me from the British Museum ; to such an extent, indeed, that the oral rays
are sometimes slightly recurved towards the basal, in addition to being markedly
shorter.

The only other species which could possibly be included amongst the Homocela
radiata is von Lendenfeld’s Homoderma sycandra*, indeed his diagnosis of the
Flomodermide, Viz., “Homocela with radial tubes”? would serve very well for the
radiate section of the genus Lewcosolenza. His figures of Homoderma, however, and
especially of the collared cells, are so exceedingly diagrammatic and his description of
the histology so scanty that, as stated previously, I must agree with Vosmaert in
regarding the matter as still sab judice. Hyen if Homoderma sycandra be a
Homocelous sponge still it is totally different from Lezcosolenia tripodifera,
conforming as it does in every particular except the problematical extension of
the collared cells into the gastral cavity, to the normal Sycon type with unbranched
radial tubes (flagellated chambers).

Unfortunately I have not had the opportunity of examining a specimen
of Carter’s Clathrina tripodifera var. gravida,t but there seems no doubt
that it really is a slight variety of Lewcosolenia tripodifera; the only doubt seems to be
whether it deserves a special varietal name. The variety was represented by a single
very small specimen with several oscula ; according to Mr. Carter’s description there
are no quadriradiate spicules present, but in an unpublished illustration of the
variety which he has kindly sent me a quadriradiate spicule of the usual form is figured ;
another difference lies in the more slender character of the rays of the internal
triradiates in the variety evaveda. An unfortunate complication in nomenclature
arises from Mr. Carter’s statement after the description of this variety (loc. czt.):—
“There can be no doubt that this is the same sponge as, only in a larger form than,
that which I described and illustrated under the name of ‘ Leucetta clathrata’
(‘ Annals’, 1883, Vol. XI., p. 33; Pl. L, Figs. 13—17).” As the name clathrus has
already been used for a species of Lewcosolenia§ and as the original description of
Leucetia clathrata was so imperfect we may perhaps with advantage retain
Mr. Carter’s later name ¢77fodifera for the species under consideration, in preference
to reverting to the name which seems to have a prior claim.

(¢.) Localztzes.—Near Port Phillip Heads. (Coll. J. B. Wilson); Westernport
(Victoria), Kent Islands (Bass Straits). (Coll. J. Gabriel.)

* Proceedings of the Linnean Society of New South Wales, Vol. IX., Part 4, p. 1088. V7
+ Bronn’s ‘“ Klassen und Ordnungen des Thierreichs.” « Porifera,’”’ p. 387.

{ Annals and Magazine of Natural History, June, 1886, p. 507, :

§ Vide Heckel, “Die Kalkschwimme,” Vol. Il., p. 30. (dscetta clathrus).

ide also p. 70 of this memoir.

A MONOGRAPH OF THE VICTORIAN SPONGES. 69

DovustTFUL SPECIES.

For the sake of completeness it is proposed in this place to quote the original
descriptions of those Victorian sponges which have been described as Homoccela (or
Ascones), but concerning which we are not at present in possession of sufficient
information to make identification and classification, without examination of the
original specimens, which is at presentimpossible, a matter of anything like certainty.

I may mention here that I have carefully examined a fragment of Mr. Carter's
“Clathrina latitubulata (provisional, incerte sedis)”* sent by Mr. Carter to the
British Museum and thence to me, and that I find it to be an undoubted Sycon with
small, but branched, flagellated chambers (radial tubes).

1. Leucosolenia osculum, Carter, sp.

Clathrina osculum, Carter, Annals and Magazine of Natural History. June, 1886,
p-. 908.

‘“‘ Tndividualised, social. Globular, stipitate, presenting on the summit a short, cylindrical, hollow process, and
ending below in one or more filiform stems fixed to the object on which it has grown, composed throughout of a
mass of tubulated thread-like filament growing by almost infinite and irregular branching and anastomosis into the
form above mentioned. Colour sponge—brown when fresh, when dry dark grey. Surface even, uniformly reticulate,
interstices about 1-120thin. in diameter. Pores numerous, passing through the wall of the hollow thread. Vent
single, tubulated, at the summit of the specimen, composed of a thin, cylindrical extension of the walls of the tubulation,
which at this part opens into it by a plurality of holes, and thus enters into its composition. No definite cloacal dilatation.
Structure already stated, composed of the same kind of staple thread as C. cavata, but smaller and more compact in its
reticulation ; wall of the tubulated thread very thin and skeletally composed of a single layer of triradiate spicules held together
by sarcode, and lined by the softer parts, which here also appear to consist chiefly of a layer of spongozoa in juxtaposition,
that is without being gathered into the form of ampullaceous sacs, together with a remarkable quantity of those organs
which consist of nucleated cells respectively surrounded by an abundance of glistening spherical granules, which Hiickel
has figured and named ‘nuclei’ (Kerne) of the syncytium, as before stated. Stem apparently an extension of the
tubulated thread, but more solid. Spicules of one kind only, viz., triradiates of different sizes, but for the most part equiarmed
and equiangulated, intercrossing each other on the surface so as to give the interstices of the reticulation here a polygonal
border ; spicules more plentiful and larger than in C. cavata, ray of the larger ones averaging 42 by 5-6000ths in. in its
greatest dimensions. Size of individual, of which there are two joined together, about 5-24ths in. in diameter; stem
about 1-24th in. long and 1-48thin. in diameter.

“Obs.—To what size this species might ultimately grow I am ignorant, but that above described appears to be very
small. It is, however, amply large enough to show in the section that the tubular vent is the outlet of the tuwbulated
structure, and that, although there is no absolutely cloacal dilatation, this is indicated by the reticulated structure in the
centre immediately under the vent being more open than towards the circumference. In these two particulars, then, it
differs from C. cavata, not more so, perhaps, than in the size and abundance of its spicules, especially on the surface,
whereby the thickness of the wall of the tubulation here appears to consist of a plurality of layers instead of one only as
in @. cavata. The tubulation is charged internally with ova in the unsegmented state, large, and presenting the
germinal vesicle.

“Upon the authority of Hickel I have stated that the ‘nuclei,’ mentioned in the two last species, are in his
‘syneytium ;’ but, entertaining a different view of their nature, I must refer the student for my explanation of this

assumption to the ‘Annals’ of 1884, Vol. XIV., pp. 20 and 21. ‘The species is very like Schmidt's Nardoa reticulum
(Spong. Kiiste vy. Algier, p. 28, Taf. V., Figs. 7 and 8).”

This species seems to me to be based on an insufficient number of specimens
which may not impossibly be simply young forms, as, indeed, Mr. Carter's
description suggests. The canal system, to judge from the description and from an
unpublished illustration sent to me by Mr. Carter, belongs to my type Bb. of
the Homoccela reticulata, and in this respect the sponge closely resembles

* Annals and Magazine of Natural History, June, 1886, p. 515.

70 A MONOGRAPH OF 'THE VICTORIAN SPONGES.

Leucosolenia pulcherrima, with which species I should be inclined to identify 1t were 1t
not for the great difference in the size of the spicules and the absence of any
information as to the shape of the spicule-rays in Leacosolenta osculum.

The specimens upon which Mr. Carter’s description is based formed part of the
collection sent to him by Mr. J. Bracebridge Wilson from the neighbourhood of Port
Phillip Heads.

2. Leucosolenia (?) laminoclathrata, Carter, sp.

Clathrina laminoclathrata, Carter. Annals and Magazine of Natural History.
June, 1886, p. 509.

‘Specimen a subcireular patch about 3in. in diameter and 74 in. thick, which has grown over a rocky substance.
Clathrous, massive, sessile, spreading, lamino-reticulate. Colour now (that is in its dry state) steel-grey. Surface even,
smooth, reticulated by the clathrous holing of the structure generally, which here makes its appearance in the form of
circular interstices of different sizes up to »:in. in diameter. Pores in the lamina. No appearance of a vent or vents
of any kind, i.c. spurious or real. Structure lamino-clathrous; lamina solid, composed of a thin layer of radiate spicules
supporting the sarcode and other soft parts. Spicules of one form only, viz. triradiate, equiarmed and equiangled, varying
in size under 75-6000ths in. in diameter, ray alone about 45 by 5-6000ths in. Size above given.

“Obs. In this instance, which is unique among the calcareous sponges so far as I know, the tubulated staple thread
of Clathrina, which so generally characterizes this genus, is replaced by a flat, solid, ‘ tape-like’ form or staple, whose
edge when cut presents no appear ance of mesodermal structure or parenchyma whatever, although towards the angles of
union, where of course the lamina branches off to form the clathrous structure of the mass, there is a small angular
space left which bears a faint trace of parenchyma, and this seems to introduce us to what in this way will become so
much more evident hereafter. It is represented among the non-calcareous sponges by ‘ Hchinoclathria favus.’ (‘ Annals,’
1885, vol. XVI., p. 292.)

“In the next species that will be described, viz., Clathrina primordialis, the reticulated flat lamina of C.
laminoclathrata appears to be replaced by a vermiculated tube, in which the walls are just as thin as the lamina of this
species, but which tubulation by repeated branching, contortion, and anastomosis, all more or less in apposition, assumes
the form of a solid mass of this kind of structure in which the intervals between the tubulation afford a much larger
space for parenchymatous structure than in C. laminoclathrata; in short, wherein the quantity of parenchymatous structure ~

is much greater.”

The specimen here described was sent by Mr. Wilson from the neighbourhood
of Port Phillip Heads to Mr. Carter. In considering the description we must
remember that it is based upon a small, dy specimen, in which probably the walls of
the Ascon-tubes had collapsed together so as to form a solid thread, so that very
likely Mr. Carter was misled in this case precisely as he had been previously misled
in the case of his Leuwcetta clathrata (vide supra p. 68). Unfortunately Ihave neither
specimens nor figures of the species to assist me in arriving at a true conclusion as

to its real nature, but I think there can be little doubt that the species must be
abandoned.

3. Leucosolenia (?) sycandra, von Lendenfeld, sp.

Flomoderma sycandra, von Lendenfeld. Proceedings of the Linnean Society of
New South Wales. Vol. IX., Part 4, p. 1088.

A MONOGRAPH OF THE VICTORIAN SPONGES. 71

** Quadriradiate, triradiate and aceratespicules. The radial tubes in regular strobiloid circles around the cylindrical
Gastral cavity which is clothed up to the margin of the Osculum with frilled flagellate cells. Gastral quadriradiates,
centripetal radial ray 0:02—0-04 x 0:0024 conic, pointed and straight ; lateral tangental ray slightly curved, convex outside
0:05 x 0:0038 ; longitudinal, aboral tangental ray 0-04 x 0:0038. Parenchymal triradiates; internal triradiates with unequal
rays, radial. centrifugal ray 0:048 x 0:0032 conic, sometimes protruding into the Gastral cavity. Tangental basal rays
curved 0:0074—0:011 x 0:0048, convex towards the outer side often equatorially situated.

“Median triradiates regular, rays conic 0:048 x 0:003. Dermal rays similar in size and shape to the former on
the summits of the radial tubes some triradiates are situated, the outer rays of which protrude beyond the surface.
Dermal acerates protruding and leaning towards the Osculum under an angle of 45° 0-71 x 0:0071 mm., cylindrical,
pointed, the centrifugal end abruptly pointed to a sharp point. Situated in groups of 10 to 12 on the summits of the
radial tubes. Oscular acerates a longitudinal cylinder forming a kind of Oesophagus with a frill of horizontal acerates.
The former slightly curved, convex on the inner side 0°57 x 0:0016, the latter 0:21 x 0:003 slightly concave to the front.

‘ Persons attaining a height of 14 mm., and a breadth of 5 mm.

‘‘Homoderma Sycandra is connected with the Asconide by forms such as Ascaltis canariensis,* and Ascaltis
Gegenbauri.+

“ Locality : East coast of Australia, Port Jackson, south coast of Australia, Port Phillip, Von Lendenfeld.”

In view of the meagreness of the above description and especially of the entire
absence of histological details, and in spite of the elaborately diagrammatic figures
which accompany it, I fear we must at present place Homoderma sycandra amongst
the doubtful species. It remains to be proved that the collared cells really do extend
beyond the limits of the radial tubes, if they do not then the sponge is an ordinary
Sycon. It will be observed that the sponge, whatever it may be, differs widely
in structure from Lezcosolenta tripodifera.

* Heckel. ‘Die Kalkschwimme,” Vol. IJ., p. 52; Vol. III., Pl. 1X., Vig. 1—3, Pl. X., Figs. la—le.
+ Loc. cit. Vol. II.,p.62; Vol. I1I., Pl. IX. Figs. 6—8; Pl. X., Figs. 5a—5d.

72 A MONOGRAPH OF THE VICTORIAN SPONGES.

V.—DESCRIPTIONS OF PLATES.

PLATE I.

Figure 1.—Leucosolenia lucast. Part of a colony; x 6,

Figure 2.—Leucosolenia stolonifer. Part of a colony ; x 2.

Figure 3.—Leucosolenia dubia. Part of a colony; x 4.

Figures 4, 5.—Leucosolenia stipitata. Two entire colonies; natural size.

Figure 6.—Leucosolenia stipitata. The same specimen as represented in Figure
Ae xall:

Figure 7.—Leucosolenia pulcherrima. An entire colony ; x 7.
Figure 8.—Leucosolenia ventricosa. A small colony; x 3.
Figure 9.—Leucosolenia ventricosa. A good-sized colony ; natural size.

Figure 10.—Leucosolenia ventricosa. The same specimen as represented in
Figure 9, but cut im half longitudinally to show the pseudogaster ; natural size.

PLATE II.

Figure 1, 2.—Leucosolenia proxima. ‘Two entire colonies ; x 8.
Figure 38.—Leucosolenia wilsoni. Upper surface of the colony; x 2.
Figure 3a.—Leucosolenia wilsoni. A very young colony ; x 2.
Figure 4.—Leucosolenia wilsoni. Half of the specimen represented in Figure 3 _

divided transversely ; showing the cut surface and, at x, a lobular outgrowth with an
osculum; x 2.

Figure 5.—Leucosolenia tripodifera. Group of specimens attached to Algae, &e.;
from a photograph ; natural size.

Figure 6.—Leucosolenia tripodifera. The largest specimen in the collection; from
a photograph ; natural size.

Nateaan s . > .
Figure 7.—Leucosolenia cavata. Part of a large colony, seen from the side. (The

specimen has been cut, and one of the cut surfaces is seen in perspective on the
right.) From a photograph ; natural size.

A MONOGRAPH OF THE VICTORIAN SPONGES. 13

PLATE III.
Figure 1,—Leucosolenia protogenes. Part of a large specimen; natural size.
Figure 2.—Leucosolena pelliculata. A good-sized specimen; x 2.

Figure 3.—Leucosolenia ventricosa, var. solida. Part of a large colony; natural
size.

Figure 4.—Leucosolenia depressa. Upper surface of the specimen; natural size.

Figure 4a.—Leucosolenia depressa. Lower surface of the specimen; natural size.

PLATE IV.

Figure 1.—Leucosolenia lucasi. The upper portion of a single Ascon-person,
highly magnified. At the summit of the tube is shown the wide osculum (o0.) and
in the right hand upper portion of the figure a piece of the thin wall of the tube is
supposed to be cut away so as to show the gastral cavity. The collared cells are
coloured red and the spicules blue.

a.v. Apical ray of quadriradiate spicule projecting into the gastral
cavity.
o. Osculum.

ox. Oxeote spicule projecting from the outer surface of the tube
wall.

pr. Prosopyle.
va, 'Trivadiate and quadriradiate spicules lying in the thickness of
the mesoderm.

Figure 2.—Leucosolenia stipitata. Vertical section, passing through the
osculum, of the specimen represented in Figure 6, Plate I.; showing the arrangement
of the canal system. For the sake of clearness the thickness of the walls of the
Ascon-tubes is slightly exaggerated, ‘The endoderm is coloured red. (Zeiss ag, Oc. 2,
Camera.)

g.c. Gastral cavity.

it. Interspace between Ascon-tubes.
o. Osculum.

fp. Pseudopore.

fd. Pseudoderm.

Figure 3.—Leucosolenia pulcherrima. Vertical section, passing through an
osculum, of the specimen represented in Figure 7, Plate I.; showing the arrangement
of the canal system. The endoderm is coloured red. (Zeiss ay, Oc. 2, Camera.)

st. Stalk.

Other lettering as in Figure 2.

74 A MONOGRAPH OF THE VICTORIAN SPONGES.

Figure 4.—Leucosolenia ventricosa. Vertical section, passing through the
pseudosculum, of the small specimen represented in Figure 8, Plate I.; showing the
arrangement of the canal system. The endoderm is coloured red. (Zeiss a, Oc. 2,
Camera.)

o. Opening of Ascon-tube into pseudogaster (—osculum in
Figures 2 and 8).

pd! The portion of the pseudoderm which forms the lining
membrane of the pseudogaster.

p.g. Pseudogaster.

p.o. Pseudosculum.

Other lettering as in Figure 2.

PLATE VY.

Figure 1.—Leucosolenia cavata. A small colony cut in half vertically to show
the arrangement of the canal system. Hnlarged and slightly diagrammatic. In the
lower portion of the figure are seen the narrow Ascon-tubes separated by wide inter-
spaces: in the upper portion are seen the narrow, tubular, inhalant canals (—inter-
spaces of the lower portion) separated by wide, irregular cavities (=merged gastral
cavities of the Ascon-tubes) out of which the oscula open.

a.t. Ascon-tubes in their primitive condition, forming the lower-
most portion of the colony.

Gastral cavities of the Ascon-tubes merged into irregular
spaces.

int. Interspaces between Ascon-tubes.

it! Interspaces transformed into narrow inhalant canals.

o. Osculum.

f. Pseudopore leading into wide irregular interspace.

p. — Pseudopore leading into narrow, tubular interspace.

pd. Pseudoderm.

Figure 2.—Leucosolenia cavata. Simplified diagram, based upon Figure 1, to
illustrate the reversal of the canal system. The endoderm is coloured red and the
space occupied by the merged gastral cavities is coloured black (except where the
endodermal lining is seen en face). For the sake of clearness all the tubes are
represented as unbranched and running vertically instead of in every direction as

A MONOGRAPH OF THE VICTORIAN SPONGES. 75

they really do (cf. Figure 1). The transition between the normal and the reversed
portions of the canal system takes place in the area marked x, where it will be seen
that there are really no tubes at all but two systems of spaces separated by an
undulating membrane.

Lettering as in Figure 1.

Figure 3.—Leucosolenia tripodifera. A small specimen the upper portion of
which has been divided longitudinally so as to exhibit the arrangement of the canal
system. The red colour indicates the distribution of the collared cells, which are
visible wherever the central cavity or the radial tubes have been cut into. The
figure also shows how the radial tubes originate as little hollow diverticula (r't') of
the wall of the central cavity a short way below the osculum, at first unbranched
but as they grow older (r.t.) branching freely.

g.c. Gastral cavity.

int. Interspaces between radial tubes,

o. Osculum.

op. Openings of radial tubes into gastral cavity.

y.t. Nadial tubes.

rt’. Young radial tubes.

Figure 4.—Leucosolenia tripodifera. Small portion of a thin section along the

radial tubes seen under a low power of the microscope. In the upper right-hand
part of the section the tubes are not cut into and their walls are seen en face.

Hilsewhere the tubes are cut right through so as to show their walls only in section.
The endoderm is coloured red and the spicules blue.

a. Anastomosis between two neighbourig radial tubes.
pr. Prosopyle.

i.s. Large “tripod” spicules protecting the blind outer ends of
the radial tubes.

Other letterig as in Figure 3.

PLATE VI.

Figure 1.—Leucosolenia stolonifer. Small portion of a transverse section of the
wall of an Ascon-person. ‘The endoderm is coloured red and the spicules blue.
(Zeiss C, Oc. 2.)

76 A MONOGRAPH OF THE VICTORIAN SPONGES.
ap.v. Apical ray of quadviradiate spicule, projecting into the
gastral cavity.
ect. EKctoderm.
end. Kndoderm.
i.c. Inhalant canal.
mes. Mesoderm.
pr. Prosopyle.
st. c. Stellate cell of mesoderm.
Figure 2.—Leucosolenia stolonifer. The projecting apical ray of a quadriradiate
spicule, enclosed in its cellular sheath. (Zeiss I’, Oc. 2.)
ap. v. Apical ray.
c.c. Collared cells around the base of the spicule ray.
sp. s. Sheath of flattened endothelial cells around the spicule ray.
Figure 3.—Leucosolenia stolonifer. A multipolar connective tissue cell of the
mesoderm (Zeiss. F', Oc. 2).
Figure 4.—Leucosolenia cavata. Vertical section of tube-wall (Zeiss F, Oc. 2).
c.c. Collared cells.
ect. Nucleus of ectoderm cell.
mes. Mesodermal ground-substance.
ov. Ovuin, surrounded by other mesodermal cells.
sp. Spicule.
st. c. Stellate cell of mesoderm.
y. gr. Group of yellow granules.
Figure 5.—Leucosolema cavata. Portion of tube-wall laid out flat and examined

as a transparent object; the entire thickness is supposed to be in focus. (Zeiss F,
Ocr2s)

y. §”. 1. Young (solid) group of yellow granules.

y. §7. 2, Fully developed group of yellow granules with central
space.

Other lettering as in Figure 4.

A MONOGRAPH OF THE VICTORIAN SPONGES. TCE

PLATE VII.

Anatomy and histology of Leucosolema wilson. (Red=endoderm; blue=
spicules).

Figure 1.—Part of a vertical transverse section of the specimen represented in
Figure 3, Plate II.; showing the arrangement of the canal system. (Zeiss ay, Oc. 2,
Camera.)

em. HKmbryo.

g.c. Gastral cavities of Ascon-tubes, more or less subdivided by
endogastric septa.

mt. Interspaces between Ascon-tubes.
o. Osculum.

fp. Pseudopore.

fd. Pseudoderm.

r. Root-like process which attaches the sponge to the substratum.

Figure 2,—Part of a section of a single Ascon-tube more highly magnified
(Zeiss C, Oc. 2, Camera); showing the subdivision of the gastral cavity by endogastric
septa. The collared cells appear to form several layers in some places, but this is
probably due chiefly, if not entirely, to the irregularity of the surface which the
endoderm covers causing the latter to be often cut tangentially.

ect. Jctoderm.
e. Ss. Hndogastric septum.
g.c. Gastral cavity.

st. c. Stellate cells in mesoderm.

Figure 3.—Section similar to that represented in Figure 2, but with the endo-
gastric septa more developed and containing embryos lying in special cavities lined
by large endothelial cells. (Zeiss C, Oc. 2, Camera.)

em. Himbryo.
n.c. Nutriant endothelial cells forming the embryo-capsule.

n.c. An embryo capsule cut tangentially so as to show the
polygonal form of the endothelial cells.

Other lettering as in Figure 2.

Figure 4. Ketodermal epithelium seen in a tangential section of an Ascon-tube.
(Zeiss I’, Oc. 2, Camera.)

78 A MONOGRAPH OF THE VICTORIAN SPONGES.

Figure 5.—Small portion of a thin section of an Ascon-tube, showing how the
collared cells spread in from the wall of the tube over the mesodermal ingrowths,
thus forming with the latter the endogastric septa. (Zeiss F, Oc. 2.)

c.c. Collared ceils.

mes. Mesoderm of the tube-wall.

mes’. Inerowth of mesoderm into the gastral cavity.
PLATE YIII.

Figure 1.—Leucosolema proxima. Vertical section of the specimen represented
in Figure 1, Plate H.; showing the arrangement of the canal system. The
endoderm is coloured red. (Zeiss a,, Oc. 2, Camera.)

g.c. Gastral cavity of Ascon-tube.
int. Interspace between Ascon-tubes.
0. Osculum.

p. Pseudopore.

pd. Pseudoderm.

iP Root-like attaching process.

st. c. Stellate mesoderm cells forming a network in the gastral
cavity.

Figure 2.—Leucosolenia proxima. A small portion of a section more highly
magnified (Zeiss F', Oc. 2); showing part of the wall of an Ascon-tube and part
of the network of stellate mesodermal cells in the gastral cavity.

c.c. Collared cells (collars and flagella completely retracted).
ect. HKetoderm.

mes. Mesodermal ground-substance.

nu. Nucleus of stellate cell.

sp. Spicule.

st. c. Network of stellate cells in the gastral cavity.

Figures 3, 4.—Leucosolenia proxima. Group of three collared cells and isolated
collared cell. The collars are still extended but the flagella are withdrawn. (Zeiss
F, Oc. 4.)

col. Collar.

nu. Nucleus.

A MONOGRAPH OF THE VICTORIAN SPONGES. 79

Figure 5.—Leucosolenia tripodifera.. Small portion of a section of the tube-
wall, showing the arrangement of the collared cells and Sollas’s membrane, with the

rod-like bodies. °
c.c. Collared eells.

v. b. Rod-like bodies.
s. m. Sollas’s membrane.

Figure 6.—Leucosolenia tripodifera. Portion of a section similar to that repre-
sented in Figure 5, but more highly magnified.

col. Collar of collared cell.
Other lettering as in Figure 5. .
Figure 7.—Leucosolenia pellieulata. Section of an ovum (Zeiss F’, Oc. 2).

gr. b. Granular body of the ovum.
no. Nucleolus.
nu. Nucleus.

Figure 8.—ZLeucosolenia depressa. Section of an ovum (Zeiss F, Oc. 2). The
outline of the section is indicated by a dotted line, the granular body of the ovum
being filled in only at one place.

gr. Granules lying just within the nuclear membrane.
n.m. Nuclear membrane.

n.n. Nuclear network.

Other lettering as in Figure 7.

PLATE IX.

Spicules.
Fieure 1.—Leucosolenia lucasi.

a. ‘'Triradiate.

b. Quadriradiates; facial view, the apical ray represented as
seen in optical section at its base.

b.' Quadriradiate; side view, showing the apical ray (a. 7.) in
full length, the three facial rays being broken off short.

c. Oxeotes.

Figure 2.—Leucosolenia stolonifer.

b. Quadriradiates; facial view, the apical ray represented as
seen in optical section at its base.

6.’ Quadriradiates; side view, showing normal apical ray (a. .)
in full length.

b.” Quadriradiates; side view, showing immensely hypertrophied
apical ray (a. 7.) in full length.

c. Oxeotes.

80 A MONOGRAPH OF THE VICTORIAN SPONGES.

Figure 8.—Leucosolenia dubia.
a. ‘Triradiates.
a.' Strongly sagittal triradiate (abnormal).

b. Quadriradiate; facial view, the apical ray represented as seen
in optical section at its base.

c. Oxeotes.

Fieure 4.—Leucosolenia cavata.
a. ‘'Triradiates.

b. Quadriradiate; facial view, the apical ray represented as seen
in optical section at its base.

c. Oxeotes.

Figure 5.—Leucosolenia stipitata.

Triradiates from the body of the sponge.

Nore.—The figures are all drawn to the same scale, with Zeiss’s camera lucida, and
multiplied by about 260 diameters. After the drawings were completed it was
found that the camera employed caused a certain amount of distortion; this,
however, only affects the largest spicules to any perceptible degree, causing
some of the rays to be unduly elongated. As full descriptions and measure-
ments of the spicules are given in the text it is hoped that the error thus
introduced will not cause any serious inconvenience.

PLATE X.
Spicules.
Figure 1.— Leucosolenia pelliculata (Specimen A).
a. ‘Triradiates.
a.' Sagittal oscular triradiate.

Quadriradiates; facial view, the apical ray represented as
seen in optical section at its base.

b.' Quadriradiate; side view, showing the full leneth of the
apical ray (a. 7.), the facial rays being broken off short.

6." Sagittal oscular quadriradiate; facial view, the apical ray
represented as seen in optical section at its base.
Figure 2.—Leucosolenia pelliculata. (Specimen D.)

Lettering as in Figure 1.

A MONOGRAPH OF THE VICTORIAN SPONGES. 51

Figure 3.—Leucosolenia pulcherruna.

a. Ordinary deep-lying triradiates.

a.’ Stout, curved dermal triradiates.
Fieure 4.—Leucosolenia ventricosa.

a. ‘Triradiates of various sizes.

b. (uadriradiates; facial view, the apical ray represented as

seen in optical section at its base.

c. Oxeote.

Nore. —The note after the description of Plate IX. applies equally to this plate.

PLATE XI.
Spicules.
Figure 1.—Leucosolenia protogenes.
Trivadiates of different sizes.
Figure 2.—Leucosolenia proxima.
Triradiates of different sizes.
Fieure 3.—Leucosolema wilson.
Trivadiates of different sizes.
Fieure 4.—Leucosolenia depressa.
a. ‘Triradiates.

b. Quadriradiates; facial view, the apical ray represented a

op.

seen in optical section at its base.
Figure 5.—Leucosolenia tripodifera.
a. ‘T'riradiates.

t

a.' Dermal trivadiate (tripod spicule), side view.

b. Quadriradiates; facial view, the apical ray represented as
seen in optical section at its base.

b.' Quadriradiate; side view, showing the apical ray (a. 7.) at
full leneth; one of the facial rays is broken off and one
represented as seen in optical section at its base.

b." Quadrivadiate; facial view, showing the whole of the much
curved apical ray (a. /.).

Norn.—The note after the description of Plate IX. applies equally to this plate.

Trens LS. Victoria Vol3 Flere/

ii
Sag

e4

am DA
Tr 2

wee
oa

f

drthur Dendy de

(I. Leucosolenia lucasi2L.stolonifer 3L.dubia, 45,6,Lstipitata 7Lpulcherrima,
89,10, L. ventricosa)

A hur Dendy del, e€ proto

(l2,Leucosolenia proxima, 3,3, 4, Lwilsoni, 9, 6,L. tripodifera, Z L. cavata-)

RC:

2

ln,

a

: ty
=
*
=<.
ote F
~. ,
oe
Be -_ i
‘

a

aT

I=) CY

Trans B.S Victor,

DIST Wid ad nat.del, ed. tidh

7 (| Leucosolenia protogenes, 2 L pelliculata3.L ventricosa, var solida,4.4.a. |. depressa)

--——

ANATOMY Fig
| Leucosolenia lucasi, 2.L.stipitata, 3.L.pulcherrima, 4.L ventricoss |
NT LL <—<<—

@

y
as
vee

as AYY ary
Cpa ad

eS

A AX
I>

2 456 «

ANATOMY

Art} > De, Ly del . 5 .
iil alae ima (1.2. Leucosolenia cavata, 3.4. L.tripodifera.)}

, ee —— eo

A
i
|
Vy
rt \) S }
ma]
KN) > 2)
)
} ~
\
, | ;
\C)
2) =< — 2 *
: = i 3 a> ae
5 _—- esse a \ * er
~ = Kini © 2
S ee re =
K™ ~ ie =o

*
|
l j
4 wy oe
- y =H
8 r
&
SS ‘
= S
S @ See
8 2 e
sees. Huse
s ©,
OS a Es
a
—_ — _ i} — a am

ANATOMY & HISTOLOGY

(1.2.3 Leucoso

r) ,
JERAY LE:

Ath /

45 | cavata.,)

enia stolonifer

,
¥
‘ @.
. 1
a
‘
‘ ‘ itt
‘
.
« .
é
.
’
‘
4
‘
aan
.
4
7
.
} '
¥ t
\ .
at’
i
‘
' ‘’
,
}

a

ANATOMY & HISTOLOGY

(Leucosolenia wilsoni|

2)
2
©6088 6068668808 OGE O80

y a
pre

: fs
TO} as

@ tT) ee ¢¢®
Ceeee® @ eee@ eee 2
.

EZ

1

60806

x
S
Ody

<<

4 4 7 : yy
ALCL? Lk nay Aéel

ANATOMY AND HISTOLOGY
(123.4. Leucosolenia proxima 5.6.L.tripodifera 7L.pelliculata, 8..depresse

i: 5D

Sy

N

Arthur Denadly del Ao al (Oy Oe oS

(| Leucosolenia lucasi,2 L.stolonifer 3 Ldubia, 4 Lcavata

Trans. FUS. Victoria Vol3 Plate /O

- SHICULES
(12. Leucosolenia pelliculata,3 L. pulcherrima, 4, L ventricosa.)

ts Ls,

iss

=o

— as

Ciara 4 it Md

Fh aha. ek ‘ aA ba - . 7 > es 2 %) : f
3.L.wilsoni,4.Ldepressa, 5.L.tripodifera)

»

MITHS! SONIAN INSTITUTION LIBRARIES

wii tn

3 9088 01308 8