Part IT. of Volume 11. will be published on’
November 15, 1903.

The Fauna and Geography

of the
Maldive and Laccadive Archipelagoes

Being the Account of the Work carried on and
of the Collections made by an Expedition
during the years 1899 and 1900

Edited by

J. STANLEY GarDINER, M.A.

Fellow of Gonville and Caius College and late Balfour Student
of the University of Cambridge.

VOLUME ll. PART |.
With Plates XXVI—XXXIV

CAMBRIDGE:
at the University Press.

Lonpon: C. J. Cuay anp Sons,
Cambridge University Press Warehouse,

Ave Maria Lane.

Price Fifteen Shillings net.

The Fauna and Geography
of the
Maldive and Laccadive Archipelagoes

VOLUME li. PART I.

Dondon: C. J. CLAY anp SONS,
CAMBRIDGE UNIVERSITY PRESS WAREHOUSE,
AVE MARIA LANE,

AND
H. K. LEWIS,

136, GOWER STREET, W.C.

Glasgow: 50, WELLINGTON STREET.
Leipzig: F. A. BROCKHAUS.
few Work: THE MACMILLAN COMPANY.
Bombay and Calcutta: MACMILLAN AND CO., Lrp.

[All Rights reserved.]

The Fauna and Geography
of the
Maldive and Laccadive Archipelagoes

Being the Account of the Work carried on and
of the Collections made by an Expedition

during the years 1899 and 1go0o0

Edited by
J. STANLEY GaRDINER, M.A.

Fellow of Gonville and Caius College and late Balfour Student
of the University of Cambridge.

VOLUME Il. PART I.

With Plates XXVI—XXXIV

=—=—
=
—
===
—=e (Oy. 3:
— AMBRIDGE :
—= rn mg :
at the University Press.
n923

it

i

i

i

0 030)

UNTNT

Wut
aut! ty
SO 8 Taig

ae Ny,
meer oO My,
ui“O%

CAMBRIDGE: _
PRINTED BY J. AND C. F. CLAY,
AT THE UNIVERSITY PRESS.

CONTENTS OF VOR. ih BARE I

Reports.

I,

The Alcyonaria of the Maldives. Part I. The Genera Xenia,
Telesto, Spongodes, Nephthya, Paraspongodes, Chironephthya,
Siphonogorgia, Solenocaulon, and Melitodes. With Plates
XXVI and XXVII . : : : : . :

By Prof. Sypney J. Hickson, M.A., D.Sc., F.R.S.

The Alcyonaria of the Maldives. Part II. The Genera Sarco-
phytum, Lobophytum, Sclerophytum, and Alcyonium. With
Plates XX VITI—XXXI ‘ : : : : : :

By Epira M. Pratt, M.Sc.

Nudibranchiata, with some Remarks on the Families and Genera
and Description of a New Genus, Doridomorpha. With
Plate XXXII : 2 :

By Sir Cuarues Extor, M.A., K.C.M.G.

Marine Crustaceans. IX. The Sponge-crabs (Dromiacea). With
Plate XXXIII . : : : ‘ é
By L. A. BorrapalLe, M.A.

On a Land Planarian from Hulule, Male Atoll, with a note on
Leptoplana pardalis Laidlaw
By F. F. Larpnaw, B.A.

Lagoon Deposits. With Plate XXXIV.
I. General Account .

By J. Srantey GARDINER, M.A.

II. Report on Certain Deposits
By Sir Jonn Murray, K.C.B., F.R.S.

PAGE

473

503

540

574

579

<1 AP A AT PS Muay We

4 ; a © <- ty
os =

2 1s:
.

civie he winaie =afY Poked -eyitdast aly font j
Re ea sapenars auitignt” nine
© Tata al M4 oriole bi Gwokiinactebare . eee ee
ie ee ee hush EM
“ee wt panes meni .% conn ¥

Soa . ‘ ‘

Seyi deer) oll Abi hab evils ahd, Wee pages rs
SHAE — vasitionenl ty Fata ew eas si eal ee
ww pu oo - > . Wee on LV Oe s
: wie vera veo som ee:

eet b. ‘Ma eh Aidt ott ie afin nl TAS ae a fs
iW ahe sonnnltenit aust) “a ew chp eee

-
=

’ ‘ * * ee

; : « Saha wl wud os escteae wi mt
iW sino ata) Aleresynde all vA site ieeGnteree apwall
. ‘ . . . . ™ . Vi
PO Ce ic

ae sian vale Hara ahs) styl HAG: uh ry — en
i ie oe lb a eat,
! a4 Pioww =~ F 3!

my

vider ag ethqncal ll ae

: i cece ied os »
Aes) ea 77
el "

aD

THE ALCYONARIA OF THE MALDIVES.
PARE

THE GENERA XENIA, TELESTO, SPONGODES, NEPHTHYA,
PARASPONGODES, CHIRONEPHTHYA, SIPHONOGORGIA,
SOLENOCAULON, AND MELITODES.

By Sypney J. Hickson, M.A., D.Sc. F.R.S., Beyer Professor of Zoology in the
Owens College, Manchester.

(With Plates XXVI and XXVIL.)

INTRODUCTION.

Tus collection of Alcyonaria from the Maldive archipelago is of more than ordinary
interest. In most cases the species of Alcyonaria have been founded on a single specimen
obtained from a single dredging or on two or three specimens obtained from a single
locality. The interest of the present collection lies in the fact that specimens were obtained
from a considerable number of dredgings made in several of the atolls or off the reefs
of a very extensive archipelago, and consequently an opportunity is afforded for the study
of the variation of form, colour, and other features within the limits of a wide area. I
have no hesitation in saying that the general impression gained by the study of single
specimens of the different genera from a reef is usually an erroneous one, because it cannot
give due weight to the important facts of the possible variations of a species living in
surroundings and under conditions which are, within certain limits, identical. If we were
to take at random, for example, a specimen of Tubipora from the reef off N. Celebes, and
compare it with another specimen taken at random from the reef at Ternate, it is probable
that so many differences would be found im the shape of the colony, the size of the
tubes, the distance of the platforms, etc. in the two specimens that it would be thought
necessary to make two species, one for the Celebes specimen and one for the Ternate
specimen; but anyone who walks along the N. Celebes reefs at low tide will find in one
morning not only the reputed Ternate species but a complete series leading from that up
to the reputed Celebes species, and beyond it.

When I went out to Celebes the first thing I did was to collect specimens of
Tubipora, having already obtained a knowledge of the genus in Europe from books and
museums, but the conclusion I came to was that there is only one species which varies
on the one reef within limits almost as wide as the limits of all the hitherto described
species of the genus.

G. I. 61

474 SYDNEY J. HICKSON.

The collection made by Mr Gardiner in the Maldives affords an opportunity of judging
of the extent of variation possible in one group of islands of some other genera that are
widely distributed.

In the case of the genera Spongodes and Solenocaulon, for example, it is clear that
there is a very great range of variation in form, colour, and in the size and shape of
the spicules. They exhibit a number of “facies” which form an almost uninterrupted
series between the types that have been described as distinct species. Now these facies
are not confined to any one island of the Maldive archipelago, nor are they clearly
separated from similar facies found in other parts of the tropical Indian and _ Pacific
Oceans. If each facies represents a distinct species, then we have the very remarkable
phenomenon of a large number of closely related species distributed over a very wide
geographical area and competing in the struggle for existence, with approximately equal
success, in many localities of this wide area. If, on the other hand, all the facies together
represent but one species, then we have an example of a species that is capable of an
extraordinary range of variation in circumstances which are approximately identical.

?

As it is impossible to prove either one of these propositions without recourse to
experiments in cross breeding, the only course to pursue, as a temporary measure, is to
place in a separate species those facies which appear to be discontinuous with others in
their form and distribution, and to collect into one species those facies that are connected
by sufficiently continuous series of intermediate links.

This is the course I have adopted in this paper.

As an example of the method I may refer to the genus Solenocaulon, in which we have
no less than eight specimens from different islands of the Maldive archipelago connected
together to form a series, and one specimen which I have called S. ramosa distinguished from
them in the tube-formation of the stem, the retractility of the polyps and other characters.
This last I regard as a distinct species at present, but I think it is quite possible that
a series of intermediate forms between it and S. tortwoswm may be found which will ultimately
necessitate the abandonment of the specific name.

It is a question of very great importance as bearing upon the species problem, whether
the varying forms and colours of these Alcyonarians are due to sporadic variation irrespective
of their environment or to distinct inherited characters.

If we were to trace the history of 20 eggs of a brown divaricate Spongodes fertilised
by the spermatozoa of another brown divaricate Spongodes, and settled down within a
restricted area of a single reef, should we expect to find that all the young Spongodes
developed from them were brown divaricates (BD), or should we expect to find that some
of them were red conglomerates (RC), others yellow flabellates (YF), etc.? If they are all
BD, with of course variations in minor particulars, then we should be justified in believing
that the brown colour and the divaricate manner of branching are not only transmissible
by inheritance, but are so transmitted because they are of some importance to the species
in the struggle for existence. If on the other hand the young Spongodes are of the facies
BD, RC, YF, etc., then we should be justified in believing that the characters of the colour
and mode of branching are of little or no importance for the species in the struggle for
existence, but are due to the direct action of the environment (e.g. sunlight, food supply,
strength of currents, proximity of neighbouring zoophytes, presence of epizoic crustacea, etc.).

THE ALCYONARIA OF THE MALDIVES. 475

Conversely, we may argue that if the colour and mode of branching are of little importance
in the struggle for existence of the species the offspring of a single pair of parents will
probably show a great range of variation in these characters, BD x BD giving rise to
BD's, RC’s, YF’s, etc. in varying proportions.

It is impossible to believe that the struggle for existence has much influence upon
the colour of these Alcyonarians. There is no evidence that the colours can be of assist-
ance to the species either as apatetic colours, hiding the individuals from their enemies by
resembling the surroundings or some other dangerous zoophyte, or as aposematic colours,
signalising some dangerous sting or abhorrent taste.

Spongodes are, so far as my experience goes, extremely conspicuous in their proper
surroundings, and do not mimic any other zoophyte that is usually associated with them
on the same reef. If they were of the nature of warning colours we should expect to
find them on the Miillerian principle of one or two distinct kinds, and arranged in simple
patterns instead of the endless shades of tone and tint and kaleidoscopic arrangement that
they exhibit. It is difficult to suggest any use that these colours can be to the Alcyonarian,
and they should be regarded as being of a purely physical and chemical character, as the
green of chlorophyll and the red of haemoglobin, and having no biological significance.

The exact form and mode of branching of the colonies must in my opinion depend
very largely upon the external conditions of the individuals. Just as a tree or shrub will
vary in form according to its exposure to the wind, the character of the soil, and the proximity
of other shrubs and trees, so will a Spongodes vary in form according to the strength of the
tides flowing over it, the amount of food these tides bring, the proximity to other zoophytes,
the exposure to sunlight, and other external conditions. The external conditions that may
influence the growth of a Spongodes are so many that it is not surprising that we find
so many differences in the style of ramification. One specimen may be free to expand in
any direction and receive the full benefit of all the sun’s rays that penetrate to the depth
of water in which it lives, whereas another not a foot away may be impeded in one
direction or another by the branches of a great Madrepore and sheltered from the heat
and light of the sun by the shadows which the great coral casts. It appears to me there-
fore far more probable that the critical condition of life for a Spongodes is not that it
should branch in any one pattern that is inherited from its parents, but that it should
be able to adapt itself to any one combination of the great variety of conditions which
it may happen to meet in the course of its development. Apart from the purely physical
conditions determining the shape of a colony, there is one biological condition which is
probably of very great importance. It is well known to those who have handled museum
specimens of Spongodes and many other Alcyonarians that Crustacea of various kinds are
frequently found entangled in the branches; and Mr Gardiner has called my attention in
his manuscript notes to these Crustacea and the striking similarity of their colours to the
colours of their host. It is very probable that these Crustacea render an important service
to the Spongodes in keeping their surfaces free from the larvae of Sponges, Barnacles,
sedentary worms and other animals, and it is also probable that the character of the
ultimate ramifications is due in no small measure to the size and shape of these important
epizoites. The manner in which the epizoic crabs affect the growth of Pocillopora, Seria-
topora, and other corals, and form galls, is well known. There is some reason to believe, as
stated on p. 495, that the tubes of Solenocaulon are caused by the Alpheids which infest

61—2

476 SYDNEY J. HICKSON.

this Briareid in a similar manner. It is unreasonable to suggest that Spongodes is not
also influenced in its growth by its epizoic crabs. In fact it may be that the differences
known to systematists as divaricate, glomerate and foliate growth are due to the crabs, and
not to the inherent characters of the species of Spongodes.

The colour and mode of branching of these Alcyonarians being unsatisfactory characters
for purposes of specific classification, there remain the spicules and the more minute internal
anatomy to appeal to. While appreciating very fully the assistance that the examination
and measurement of spicules renders in some cases, I am convinced that in many genera the
spicules are liable to very great variation, and are very untrustworthy for our purpose. A
discussion of this point must however be postponed until my researches have been com-
pleted. It is upon the anatomical characters of the anthocodiae and the general canal
system connecting the polyps that we must ultimately base our division of the genera into
specific groups. This however involves the careful investigation of a large number of well-
preserved specimens belonging to each genus from several localities, and can only be
accomplished by degrees and in the course of many years. It is not possible, however, for
any zoologist who undertakes to write an account of such a collection of Alcyonarians
as the one made by Mr Gardiner in the Maldives to undertake this task for all the
genera he may find in it. Miss Pratt has shown in Part II. of this paper the manner in
which the work may be done in the case of two or three closely-related genera. On the
other hand it is to my mind very desirable that a general account of such collections
should be published as soon as possible, and to do this a provisional arrangement of the
specimens in specific groups should be made.

It appears to me that the proposal recently made by Bernard (3), so far as I can
understand it, is open to many objections. The use of a number instead of a name is not
only of no advantage, but it is very dull and uninspiring. The prominence that is given
to the locality in the case of these shallow water-marine forms is also to my mind
undesirable.

This Part contains an account of all the Alcyonacea, with the exception of the genera
Alcyonium, Sarcophytum, Lobophytum, and Sclerophytum, and of all the Gorgonacea Scleraxonia.

The most abundant species of Aleyonacea belong to the genera Sarcophytwm, Lobophytum,
and the new genus Sclerophytwm. These genera have been carefully studied by Miss E. M.
Pratt, and are described in a separate paper. The remainder belong to the genera Telesto,
Xenia, Spongodes, Chironephthya, Siphonogorgia, and Heliopora. The Gorgonacea Scleraxonia
are represented by the genera Solenocaulon and Melitodes. The Stolonifera appear to be rare
in the Maldives. Not a single specimen of the genus Clavularia was obtained. Mr Gardiner
informs me that the genus TZubipora is very scarce throughout the Maldives; “one of my
boys brought me a small piece (decaying) from the edge of the W. reef at Hulule. I saw
one colony at Maradu, Addu Atoll, but do not remember to have seen it growing elsewhere.”
The absence of Clavularia from the Maldives is interesting in view of its immense abundance
on the coral reefs of many parts of the Malay archipelago, Torres Straits, and elsewhere in
the S. Pacific. It does not appear to occur on the coast of Ceylon, of the Malay peninsula,
or of Singapore, nor on the other hand in the Red Sea, Zanzibar, or East Coast of Africa.

THE ALCYONARIA OF THE MALDIVES. 477

Tubipora, however, is abundant on the coast of Zanzibar, the Seychelles, and the Red
Sea, very rare in the Maldives, and I have Professor Herdman’s authority for saying that
it must be scarce or rare off the coast of Ceylon, as not a single specimen occurs in his
extensive collection from those waters. It is difficult to account for the scarcity of Tubipora
in these regions, as on most of the coral reefs of the world it is one of the most abun-
dant and luxuniant Alcyonarians. The absence of the characteristic Zanzibar genera Coelo-
gorgia and Lemnalia from the Maldives is also a noteworthy feature.

On the other hand the genus Heliopora, which is so widely distributed on the reefs
of the Malay archipelago, Torres Straits, and Pacifie Ocean, is also abundant in the
Maldives. Mr Gardiner writes that at Hulule Male “ Heliopora occurs practically everywhere
near the reef inside the boulder zone if other corals grow. It does not, however, live near
the velu (lagoonlet, Hulule being a faro or small atoll). It is not at all common to the
west, but great masses are found to the east, usually dead in the centre with plates
growing out all round, the tops just awash at low-tide springs. Some of the masses are
five to six feet across, but none have any great thickness. All now lie freely (not attached)
on the sand. The polyps are but rarely expanded in the daytime, and then only towards
the bases of the branches, that is, in the shade. The corallum does not appear to be
bored by any other organism.”

Speaking more generally he writes, “ Heliopora is nowhere of any special importance in
building the reefs. It may be found almost anywhere towards the outer sides of the atolls
on the reefs. Small colonies may grow on the reef flat, but they are very rare.”

The genus Xenia does not appear to be very abundant, nor to reach a large size. The
specimens sent to me were found at Addu in the extreme S. of the archipelago and at
N. Male in the northern half, and this suggests that the genus probably occurs sporadically
throughout the archipelago. The Maldives, however, do not seem to be as favourable for the
growth of this form as the Chagos archipelago in the S., Zanzibar and the Red Sea to the
west, or such localities as New Britain, Torres Straits, and the coast of Celebes and Ternate.

Spongodes appears to be extremely abundant and varied in character. When all the
specimens are placed side by side, however, there is no one feature to be observed to distinguish
them from the Spongodes of other reefs. The genus appears to be widely spread over the
whole of the Red Sea, Indian Ocean, Malayan seas, and the tropical Pacific Ocean. It is
remarkable that the genera Nephthya and Paraspongodes are so rare, being represented by only
one specimen of each.

The genus Chironephthya (W. and S.), which was for many years after its discovery by the
Challenger expedition known to occur at a depth of 200 fathoms in Japanese waters, has
recently been described by Hiles (12) from New Britain in shallow water, probably occurs in
shallow water at Funafuti (cf. Whitelegge 24, Siphonogorgia macrospina), occurs off Singapore
(e. coll. Bedford), off Ceylon (e. coll. Herdman), and is now shown to be abundant in depths
of about 20 fathoms in the Maldive archipelago. The genera Solenocaulon and Siphonogorgia
appear to have a distribution very similar to that of Spongodes.

478 SYDNEY J. HICKSON.

Family Xeniidae.

Genus Xenia. Savigny.

In writing upon this genus I must explain that I do not include in it those species
which belong to the genus Heterovenia of Kélliker. This explanation is necessary in view
of the recently published paper by Kiikenthal (16) in which the proposal is again brought
forward to merge Heteroxenia into Xenia.

The genus Heteroxenia was established by Kélliker for a species found at Port Denison
in Australia which exhibited dimorphism, and was named AHeteroxenia Elisabethae. Klun-
zinger (14) in 1879, upon the examination of some specimens of Xenia fuscescens which he
found in the Red Sea, threw some doubt on the validity of the genus Heteroxenia.

Up to this point no detailed account had been given of the structure of the small
polyps which Kolliker regarded as siphonozooids, and it was a question whether they were
really siphonozooids or young polyps not yet developed into their adult form.

In 1895 G. C. Bourne (4) examined and described the siphonozooids of a specimen of
a species which he attributed to Heterorenia Elisabethae (KOll.) from Zanzibar, and Ashworth (1)
subsequently examined the same specimen and confirmed Bourne’s determination of the species.

Both Bourne and Ashworth made careful anatomical examinations of the specimen,
studying in minute detail and with the best histological methods the structure of the
siphonozooids, and they came to the conclusion that there are true siphonozooids in this
species, and that they can be distinguished without difficulty from young autozooids. It
does not appear from Kiikenthal’s recent paper that he has made any anatomical investi-
gation of a species of Heteroenia, and the arguments used in defence of his proposal
show no advance on those of Klunzinger which were considered and rejected by Bourne
and Ashworth.

Whether the specimens collected by Klunzinger in the Red Sea, and named by him
Xenia fuscescens Ehr., are or are not identical with Heterorenia Elisabethae is a point
upon which I will not venture to express an opinion, but I cannot refrain from saying
that it is premature at present to assert that they are. I cannot therefore accept Kiikenthal’s
view that the genus Heteroxenia should be suppressed. It is unfortunate that no specimens
of a dimorphic Xeniid are to be found in Mr Gardiner’s collection, and that only eight
specimens of one species of Xenia were discovered in the Maldives. At present therefore
it appears that the genus Heterovenia is confined to the exceptionally hot waters of
N. Australia, the Red Sea, and Zanzibar, together with one species Heteroxenia capensis
with distinct characters from cooler water off S. Africa.

No specimens have yet been observed in the extensive collections of Aleyonaria that
I have examined from the Maldives, Ceylon, Singapore, and the Malay archipelago. If, by
any chance, the single specimen of Heterowenia Elisabethae in the Museum Godeftroy, which
Kélliker examined, was wrongly labelled, and did not originally come from Port Denison
in Australia, the geographical distribution of the genus might not seem so extraordinary
as it is. It is true there is no definite reason for doubting the locality of Kélliker’s
specimen, but it is noteworthy that notwithstanding the careful investigations of Saville

THE ALCYONARIA OF THE MALDIVES. 479

Kent, Haddon and others in Torres Straits and elsewhere the occurrence of this genus in
tropical Australian waters has not been confirmed.

In conclusion I may say that although I cannot accept Kiikenthal’s proposal to merge
Heteroxenia into the genus Xenia, I think it is just possible that this is the course which
will ultimately have to be adopted. If it can be shown that notwithstanding the difference
in character of the young autozooids and the siphonozooids the latter do develop later
the character of the adult autozooids, then the genus must go. This can only be proved
by the examination and comparison of a large number of specimens from the same locality.
Until this is done it is not only very inconvenient to the systematist but contrary to the
best principles of scientific investigation to suppress the genus.

Xenia garciae Bourne.

Two specimens, one from the reef at Maradu, Addu Atoll, and one from Hulule, Male A.,
were obtained by Mr Gardiner.

The former consists of about fourteen clumps, perfectly white in colour, attached to
what appears to be a piece of shell 60mm.x30mm. in size. Some of the clumps are
free, and represent single colonies, others are constricted or lobed, and may represent two
or more colonies fused together. Each clump or colony is very flat, rising in the centre
only 2 or 3 millimetres from the shell.

The second specimen (from Male A.) consists of 10 clumps, pale bluish-green in colour,
attached to a rounded knob of dead madrepore, coated with nullipore coral. Some of. the
clumps are free, others are attached at the base in pairs or triplets, and I think it is
probable that each pair or triplet represents one branched colony. An interesting point of
difference between the specimens is that in the last-named each clump rises 4 or 5mm.
from the base of attachment, showing a pronounced stalk (sterile portion).

A careful examination of the polyps shows that both specimens clearly belong to
Bourne’s species from Diego Garcia, and the differences in colour and height may be
accounted for by differences of age or position on the reef.

It is not necessary to publish the measurements of the polyps I made. They agree
very closely with those given for the species in Dr Ashworth’s table (2). I may say, how-
ever, that I have compared the polyps with one taken from Mr Bourne’s type specimen,
which was kindly lent to me for the purpose, and there can be no doubt of the identity
of the species.

From N. Male, 30 fathoms N. of the atoll, Mr Gardiner obtained six very small specimens

of a species of Xenia closely related to Xenia garciae. They are too young to place definitely
in any of the hitherto described species.

480 SYDNEY J. HICKSON.

Family Telestidae (Milne Edwards).

Genus Telesto (Lamaroux).

Two distinct species of this genus were found in the collection from the Maldives.
One of them is probably Telesto trichostemma (Dana), and the other is new. The new
form I propose to call Telesto rubra. The specimens of Telesto trichostemma were dredged
in 25 fathoms in Mulaku Atoll and in 24 fathoms Miladumadulu Atoll.

One fragment of the new species was found in the same dredging in Mulaku, and the
other in Mahlos Atoll, 23 fathoms. It is to be noticed that both specimens of the new
species are freely encrusted with a sponge, a Prosuberites, in the case of the Mahlos specimen,
and a Leptosia sp. n. in the case of the Mulaku specimen. The Telesto trichostemma,
however, is entirely free from sponge encrustations.

The occurrence of encrusting sponges on Telesto has been recorded by Dana (5) in the
case of Telesto fruticulosa, by Wright and Studer (23) in Telesto trichostemma from Torres
Straits, and by Miss Hiles (12) in 7. rupicola from Blanche Bay.

In a former paper (10) I suggested that Telesto should be removed from the family
Clavulariidae and placed among the Alcyonidae. At the time I wrote that paper I had
very little acquaintance with the genus, and I am now in agreement with May (17), who
suggests the retention of Milne Edwards’ family Telestidae in a position intermediate
between the Clavulariidae and the Aleyonidae.

Telesto rubra sp. nov.

Specimen A. Mulaku Atoll, 25 fathoms. Fime mud and rubble at the edge of the lagoon.
Specimen B. Mahlos Atoll, 23 fathoms, N.E. lagoon.

Both of these specimens are unfortunately fragmentary, and as the base is missing I
have no knowledge of the method by which the colony is fixed to the bottom. The
species is easily recognised, however, by the bright pink to red colour of the tubes.
The spicules of the body-walls of the axial and lateral polyps become firmly fused
together to form a solid tubular skeleton, perforated by numerous irregular and minute
channels, traversed by strands of mesogloea. These tubes have a remarkable similarity in
their construction to the tubes of the genus Tubipora, but are distinguished by the
presence of eight distinct longitudinal ridges (Fig. 2).

In Telesto trichostemma Studer states that “the spicules surrounding the elongated
digestive cavities of the primary and secondary axial polyps lie in such intimate contact
with one another that, in the older parts of the colony, they form a continuous calcareous
cylinder. Towards the base this is further strengthened by the horny substance secreted
between the spicules.” These tubes show as in 1. rubra eight longitudinal ridges. In
T. trichostemma, and in other species of Yelesto which form a tubular skeleton that remains
rigid on drying or slight treatment with caustic potash, the spicules appear to be bound
together by a horny substance developed in the mesogloea. In the new species this is
not the case, no appreciable amount of horny substance being noticed either in the sections
of decalcified or in dried specimens. The rigidity of the skeleton is entirely due to the

THE ALCYONARIA OF THE MALDIVES. 481

calcium carbonate, and prolonged boiling in caustic potash does not soften the tubes or
permit of the isolation of the spicules.

The anthocodiae are completely retractile, and when that part of the body-wall of a polyp
is examined which is situated at the junction of the anthocodia with the non-retractile
portion, a few isolated spindle-shaped spicules can be seen. These spicules are very different
in form to those of any species of Telesto that has hitherto been described, and it is
perfectly certain that they grow very rapidly in size until they become as firmly fixed or
fused together as are the spicules in the tubes of Tubipora (Figs. 5 and 6).

In Telesto trichostemma it will be noticed that according to Studer (23) it is only
in the “older parts of the colony” that continuous calcareous cylinders are formed. In
T. rubra, on the contrary, the spicules at the base of the verrucae of the youngest buds
are fused to form a continuous calcareous skeleton. The secondary branches and the young
polyps are given off from the axial polyps at an angle which is almost a right angle.
This is another feature in which the new species seems to be peculiar.

The tentacles of the retracted polyps are, in spirit, snow white or pale yellow. They
exhibit a row of about eight pinnae on each side, and each pinna is supported by a curved
spindle-shaped spicule 0714 mm. in length. There is also a series of rod to spindle-shaped
slightly tuberculated spicules on the aboral side of the rachis of each tentacle (Figs. 8 and 4).

In the specimen of TZ. trichostemma in the same collection the tentacles are quite free
from spicules.

The general surface of ectoderm of nearly all parts of the colony is destroyed by
the encrusting sponge.

I am greatly indebted to Professor Topsent of Rennes for examining these sponges,
and very kindly sending me a report upon them.

The sponge that occurs on the Mahlos atoll specimen is Prosuberites epiphytwm (Lamarck).
Professor Topsent tells me further that the blue-green colour is due to the Beggiatoaceae
which frequently invade Terpis fugaz, and were named by Carter Hypheothrix coerulea.

The sponge that occurs on the Mulaku specimen appears to be a new species of the
genus Leptosia.

The characters of the new species may be diagnosed as follows:—Colony consisting of
single upright axial polyps bearing a few secondary branches of the same nature. Several
young polyps are borne by the main stem and secondary branches. All the polyps are
completely retractile, the young polyps forming verrucae about 2 mm. in height. The axial
polyps, branches, and verrucae 1°75 mm. in diameter in the Maldive specimens. Body-wall
of the axial polyps and verrucae supported by a firm, perforated, tubular, calcareous skeleton,
which is pink in colour. Very few free spindle-shaped spicules at the growing edges of
the tubes. External surface of the tubes marked by eight longitudinal ridges. Tentacles
of the polyps white or pale yellow, with eight or nine pinnae on each side supported by
white curved spindles (‘14 x -015) and other spicules. Locality, Maldive Islands, 23—25 fathoms.

Telesto trichostemma Dana (5); Wright and Studer (22 and 23).

From Mulaku Atoll, 25 fathoms, fine mud and rubble, and Miladumadulu Atoll, 24
fathoms, from one of the eastern passages.
G. II. 62

482 SYDNEY J. HICKSON.

One of the specimens which, with some hesitation, I place in this species consists of
three stems, one of which is connected with a characteristic flattened, anastomosing stolon
growing on a sponge. The longest stem is 115 mm. in length, and a little over 1 mm. in
diameter. The stems consist of a single axial polyp, bearing numerous lateral polyps, situated
at an acute angle to it. As in 7’. trichostemma from Torres Straits, the spicules of the older
parts of the colony join to form a solid cylinder marked externally by eight longitudinal
ridges, but after boiling in caustic potash for some time the spicules can be separated
into groups or clusters, and I cannot find that they are supported by a definite horny
substance.

The spicules are very irregular in shape, but correspond in general form with those
figured by Studer, they are however decidedly larger than those of the Challenger specimens,
for whereas Studer gives 0°2 mm. as the greatest length, there are very few of the larger
spicules in my preparation less than 0°24 mm. in length.

Studer states that the Challenger specimens were covered with an encrusting sponge.
It is therefore remarkable that the Maldive specimen is not, considering the fact that both
specimens of Telesto rubra from the same locality are so covered.

The Challenger specimen of this species was obtained in shallow water in Torres Straits,
and it may be questioned whether it is really identical with the species figured and very
imperfectly described by Dana from the Fiji Islands.

In the Maldive specimens most of the polyps are fully expanded, but the remainder
show that the anthocodiae can be retracted as completely as in Telesto rubra. The tenta-
cles are quite devoid of spicules, and are of a reddish-brown colour. They bear about
12 pinnae on each side.

The second specimen from Miladumadulu Atoll is only a fragment. It is much
whiter than the Mulaku specimen, and the polyps are all retracted. Unfortunately the
bottle containing this specimen was broken in transit, and it was quite dry on arrival.

The distinctions between the species of the genus Telesto have not at present been
very clearly defined, and further anatomical investigation is desirable before any such
definition is attempted.

I have examined specimens which have been attributed to the species Telesto rupicola
from Bahia and from Jamaica in the collection of the British Museum, and have been
much impressed with the general similarity in habit to these specimens from the Maldives.
In the minute skeletal characters, however, the Maldive specimens resemble more closely
the description given of 17. trichostemma from Torres Straits by Studer, and they differ,
moreover, from 1’. rupicola in their opaque reddish-brown tentacles.

I have also examined a specimen of Telesto from the Persian Gulf, having close affini-
ties with von Koch’s Telesto prolifera from the Gulf of Siam, and a specimen labelled
T. aurantiacu from Kurrachee in the collection of the British Museum.

I am inclined to believe that all these species ought to be merged into one, a very
variable and widely distributed species, which should be called 7. rupicola (Miiller). Until,
however, further anatomical investigation has been made I think it is desirable to keep
the American form distinct as 7. rupicola, and to call the others 7. trichostemma (Dana).
There are not enough specimens in Mr Gardiner’s collection to determine the extent to

THE ALCYONARIA OF THE MALDIVES. 483

which the species varies in form in one group of islands. The species 7. aurantiaca ought
to disappear from the literature as Lamaroux’ description of it is far too incomplete to be
any guide to its determination.

Family Nephthyidae Verrill.

This family has recently been rearranged by Kiikenthal (15), who recognises now only
four genera; Ammothea, Paraspongodes, Nephthya, and Spongodes. Unfortunately we are still
in almost complete ignorance of the anatomy of any of the species, and there are many
reasons for regarding Kiikenthal’s new system as not entirely satisfactory. In the revised genera
Nephthya and Spongodes the anthocodiae are supported on the outer side by one or more than
one specially long spicule, the supporting bundle (Stiitzbiindel). These are not present in the
other two genera. The genera Ammothea and Nephthya have the anthocodiae arranged in
“catkins,” in the other two genera the anthocodiae are in bundles or isolated.

As regards the “Stiitzbiindel” there may be observed in many colonies considerable varia-
tions in different regions, some anthocodiae possessing a very well-marked Stiitzbiindel, and
others having no sign of one. On the other hand the very pronounced “catkin” arrangement
which is seen in many specimens shades off into bundles with closely-set anthocodiae in a
manner which often makes the determination of the genus a difficult one. For example, I had
some difficulty in determining whether the arrangement of the anthocodiae in the specimen
which I have referred to Nephthya inermis is or is not a “catkin” within the meaning of the
definition. For the present, however, it is best to work upon Kiikenthal’s system, which possesses
the great advantage of reducing very materially the number of ill-defined genera which appear
in our literature.

Of the genus Spongodes about 70 species have been described from the Red Sea, East
Coast of Africa, Malay Archipelago, China, Japan, Australia, New Guinea, New Britain (see
15, 17, 19). It is a remarkable fact that only one species, Spongodes (Morchellana) spinulosa,
has been described from the Indian Ocean (9), and only one species of NVephthya (NV. burmaensis
Ridley). The genus is very abundant in the Maldives, as the following notes show. I have a
specimen from the Persian Gulf. I have several specimens from Singapore. It occurs on the
coast of Ceylon, and probably is abundant all over the Indian Ocean.

Time has not permitted me to make a detailed examination of all the specimens, and I
feel that the task of attempting to identify them with any of the seventy defined species or
to describe any of them as new to science would yield results out of all proportion to the
labour involved. The presence of so many specimens of different forms of growth, of colour,
of spiculation living under very similar conditions in one group of islands suggests very forcibly
that all these specimens represent but different facies of one variable species. Anatomical in-
vestigation of a thoroughly searching character can alone determine this point, and until it is
made the task of species-making should in my opinion be suspended. A very large number
of the new species that have been described are founded upon one small specimen.

We have no information to guide us as to the changes that take place during the growth
of a colony from a height of 20 mm. to a height of 200 mm., nor the influence that variations
in depth, in the strength of tides, or the nature of the sea-bottom, the associated crab, mollusks,
and other creatures have upon the colour, form, and size of the colonies.

62—2

484 SYDNEY J. HICKSON.

We must remember that nearly all the specimens that are brought to Europe from the
tropical seas preserved in spirit are small specimens. They are probably also young specimens
of large species. Full-grown specimens may reach a size of 300 to 400 mm. or more in
diameter, but these are not easy to detach from their base of attachment, and it is, as a rule,
impossible to preserve them satisfactorily and send them home.

The study of the Spongodes collection from the Maldives throws, however, a little light on
one or two points of general interest.

In the first place we find in this one archipelago a very wide range of variation. The
colour of the spicules varies from pure white (as in the specimen from Kolumadulu) through
many shades of yellow, orange, brown, pink, red, purple, chocolate to black (as in the stalk of
the specimens from Fadifolu).

In some forms the spicules are enormously large, in others they are relatively small.

A few forms with very divaricate branches are found, but others have a remarkably compact
habit of growth.

These differences are not more pronounced between specimens found in the extreme north of
the archipelago and those in the extreme south. On the contrary, the specimens from the same
atoll are sometimes most diverse in colour. For example, the large coarse golden-yellow specimens
from N. Male, 20 fathoms S.W. of the atoll, are in marked contrast to the delicate chocolate-
coloured specimens from 24 fathoms in the shoal at the centre of the atoll, and the pale
yellow form, like Spongodes flabellifera found in the S. atoll in 25 fathoms. Three small speci-
mens from Hulule Male, 25 to 30 fathoms, are so much alike in colour, spiculation and size of
anthocodiae that they clearly belong to the same species, and yet they are of such different form
of growth that they might be delegated to different divisions of the genus. The specimen from
Miladumadulu in the north of the archipelago is very closely related to the single specimen
from Addu in the extreme S. They are much more closely related than are the two dusky-
brown specimens and the two red divaricate specimens which were obtained in the same dredging
in the Fadifolu atoll. There is nothing to show that geographically there are any limits to
the distribution of the species within the archipelago. There are probably no characteristic
features distinguishing the Spongodes of one atoll from that of another. But we can go further
than this and say that there are no characteristic features which appear to distinguish the
Spongodes of the Maldives from those of suitable localities in the other regions of the Indian
and Pacific Oceans. There are specimens from the Maldive archipelago that clearly belong to
the same species as specimens in my collections from Singapore, New Britain, and Fiji. There
can be no doubt that specimens are present in the Maldives, which under our present system
would be referred to Holm’s species (13) Spongodes spinifera, founded on a specimen obtained
in the Fiji Islands, to Kiikenthal’s species (15) Spongodes splendens, founded on a specimen
obtained in Ternate, to Studer’s species (22) S. flabellifera, obtained in Japan.

All these facts appear to me to point to the conclusion that the species of the genus
Spongodes are extremely variable, and have a very wide geographical range in tropical waters.
Whether there is only one true species is a question to which with our present want of exact
anatomical knowledge no answer can be given. I am convinced, however, that the characters
upon which new species have been founded in recent years in such large numbers are untrust-
worthy for the purposes.

Two lines of inquiry might yield interesting results. Firstly, there should be a careful

THE ALCYONARIA OF THE MALDIVES. 485

anatomical comparison of the structure of the stomodaeum, mesenterial filaments, and canal
systems of two or three distinct facies, and secondly, a numerical statement of the range of
variation in the length of the spicules of the Stiitzbiindel, the number of anthocodiae on the
secondary branches and other measurable characters in the different parts of one large colony.

The following is a list of the specimens sent to me of the genus Spongodes.

From Miladumadulu, 24 fathoms; in an E. passage. A small specimen about 30 mm. in
height, having the general form of S. splendens (Kiikenthal). Stalk and branches white,
anthocodiae pink.

Another small specimen from the same atoll 20 fathoms, off the reef. This is a little
larger than the first specimen. The colour of the anthocodiae is darker, and extends on to
the secondary branches, where it shades off into orange.

From Mahlos Atoll, 23 fathoms. A small specimen, about 30 mm. in height, with divaricate
branches, having a close resemblance to the description given of Spongodes rubra (May) from
the Philippines. The stalk and branches are red. The spicules of the tentacles white.

From the same atoll 22 fathoms, passage to N. of Fainu, two small specimens, one white
with purplish red supporting spicules, the other yellow with orange terminal branches.

From Fadifolu, 25 fathoms. Middle of N. part of atoll. Two beautiful but small specimens
of a form similar to that described as Spongodes flabellifera (W. and S.). They are of a deli-
cate shade of pale yellow, with white anthocodiae.

From the same atoll 14—20 fathoms there are four specimens. Two are of a dusky-brown
colour, and are of the same facies as the two specimens mentioned. The other two are more
divaricate, with larger anthocodiae, and the colouring is very remarkable. The stalk is black
to very dark brown, the branches red, and the anthocodiae white. The form of growth is very
similar to that of S. rubra (May).

From the same atoll and dredging a very fine specimen 100 mm. in height with a stalk
30 mm. is preserved, coming nearest to Kiikenthal’s species S. splendens. The stalk is yellow,
branches yellowish-red, terminal branches red, and the crown of tentacles white.

From N. Male one specimen from 20 fathoms S.W. of atoll, and two specimens labelled
Hulule, Male A., answer to the description of Spongodes spinifera Holm.

There can be little doubt that these specimens are closely related to the specimen upon
which Holm (13) founded his species, although they appear to be older and more fully developed
examples of it. The characteristic features of the species are (1) the enormous spicules of the
walls of the branches, (2) the very long projecting chief spicule of the Stiitzbiindel, and the dark
gold or yellow ochre colour.

I may here call attention to two serious misprints in the literature of the species. In
Holm’s paper (p. 39) the length of the spicules of the wall of the branches is given as 0°7 mm.
It is quite clear that the statement on p. 40 that these spicules may attain a length of 7 mm.,
however, is the correct one.

In Kiikenthal’s table (15) of the species of Spongodes the length of these spicules is given
as 4mm., an obvious misprint for 7 mm.

The specimen from 20 fathoms is 100 mm. in height, with a stalk 35 mm. in length by
30 in diameter. The spicules of the wall of the branched portion of the colony are very

486 SYDNEY J. HICKSON.

large, and irregularly scattered. Many of them reach to the extreme length of 7 mm., and
frequently exhibit one bifurcated extremity. The colour of the colony is dark golden.

The two specimens from Hulule, Male, are rather smaller. The spicules of the walls of the
branched portion are not quite so large as in the N. Male specimens, not exceeding 6 mm. in
length, and I have noticed only a very few with a bifurcated extremity. The spicules, how-
ever, in this part of the colony and in the polyps are rather stouter in build than in the
20 fathom specimen, the largest spicules of the former specimens being about 5:5 x 0'7 mm.
and of the latter 7 x 0°7 mm.

From N. Male, 24 fathoms. Three specimens of a dark-chocolate colour, having the general
form of S. splendens (Kiikenthal), were obtained near the shoal at the centre of the atoll.

From N. Male, 25 fathoms. One specimen yellow in colour, having the form of S. flabellifera,
was found in the middle of the S. atoll.

From N. Male, 27 fathoms. <A fragment, from a form similar to the last-named facies.

From Hulule, Male Atoll, 25 to 30 fathoms. Four small specimens. One uniformly pale
yellow. Three with yellow stalks and branches, red supporting spicules, white tentacles. The
form of growth is variable in the three last-named specimens. They would be regarded perhaps
as young specimens of Spongodes splendens.

From S. Nilandu, 25 fathoms, W. Passage. This is a small specimen 30 mm. in height, of
very compact form of growth, similar in some respects to S. glomerata (May). Associated with
it were a red chiton and a pelecypod.

From Kolumadulu, 35 fathoms. This specimen, about 40 mm. in height, is perhaps the
most beautiful of the series. The stalk is white, the branches and anthocodiae also white, but
the main stem contains a few large scattered pale purple spicules. It has a close resemblance
in many respects to Holm’s species Spongodes pallida, but the lowermost expanded branches
are most delicately fringed so as to have an appearance like fine lace.

From Haddumati, 40 fathoms. A small specimen of a pale brown colour related to the
last named.

Two specimens from Suvadiva Atoll, 25 fathoms. Across the mouth of a passage. The
specimens are about 80 mm. in height, with a maximum expanse across the arms of 80 mm.
The base is missing in both specimens. From a thick main stem several branches arise, which
may again divide into several branches of the second order, which bear the clusters of antho-
codiae. There are 15—20 anthocodiae in each cluster. The colour is uniformly rose-pink in
the main stem and branches, but somewhat dark in the terminal twigs and anthocodial clusters.
The colour appears to be entirely due to the spicules, which are long, delicate spindles,
irregularly scattered on the surface branches. These spicules vary considerably in size. They
are on an average 5 mm. in length, but many are 7 mm., and a few that I have measured were
9mm. in length. These are, I believe, the longest spicules known in the group of the Alceyonaria.
In the anthocodiae one or two spicules 3 mm. in length, predominate, projecting one or two
millimetres beyond the tentacles. The rest of the spicules of the anthocodiae are small and
slightly curved. They appear to be smaller (0°75—0°5 in length) than in the type-specimen
(1:0 mm.). They are also thicker.

The specimens agree better with the description of W. and S.’s species S. macrospina
than with that of any other, but in the extraordinary length of the spicules of the branches
and in the smaller of the secondary spicules of the anthocodiae they present differences which
in the opinion of some authors might be regarded as of specific value.

THE ALCYONARIA OF THE MALDIVES. 487

From the same atoll, 31 fathoms. A specimen about 45 mm. in height, having the general
form of S. splendens (Kiikenthal), with pale yellow stalk, yellow branches, and purplish-red
supporting spicules.

From the same atoll, 39 fathoms. A small very divaricate form with white stalk and
stem, but the terminal branches and supporting spicules bright orange in colour.

From Addu Atoll, 25 fathoms. Near a passage. A small red specimen of very compact
habit, like the specimen from S. Nilandu.

In addition to the numerous specimens of the genus Spongodes in the collection there
are two specimens which must be referred to the same family, but as one of them is probably
a young colony and the other only a fragment of a colony I have not very carefully studied
them.

Nephthya inermis Holm. (?)

Kolumadulu Atoll, 22 fathoms, near W. Passage. This specimen is of a dull brown
colour. It is 50 mm. in height, with a stalk 25 mm. in length. It appears to be closely
related to Holm’s specimens from the Hirudo Straits, Japan, 36 fathoms. Holm’s largest
specimen was 550 mm. in height.

Paraspongodes pustulosa (W. and 8.) (?)

Mahlos Atoll, 23 fathoms. Inner part of passage to S.E. This specimen is apparently
only a portion of a branch. It is 50 mm. in length, with numerous irregular purplish-red
spicules in the superficial parts of the coenenchym and yellow spicules in the crown and tenta-
cles. Wright and Studer’s specimen (described under the genus Scleronephthya) came from a
depth of 18 fathoms off the Philippine Islands.

Family Siphonogorginae.
Genus Chironephthya (Figs. 8, 9, 10).

The genus Chironephthya was constituted by Wright and Studer (22) for three species
obtained by the Challenger expedition on the Hyalonema ground, Japan, in 345 fathoms.
The genus has somewhat the form of a Nephthya, but undoubtedly approaches most closely
to the genus Siphonogorgia (KGll.). According to the authors, however, it differs in that it
“is essentially distinguished from Stphonogorgia by the fact that the polyps occur along the
entire course of the branches, and are less retractile.’” As regards the first of these, the
distinction breaks down in the fact that in the species Siphonogorgia pustulosa, described by
the same authors in the supplement to the Alcyonaria and in Kiikenthal’s species S. miniacea
and S. cylindrica (15) the polyps, ie. the anthocodiae, occur all along the branches as in
Chironephthya. As regards the retractility of the polyps, a point upon which less stress
can be laid, as I have previously pointed out, it must be noted that the polyps of Scphono-
gorgia pustulosa (W. and §.) do not appear to be much more retractile than those of
Chironephthya. The increase of knowledge of these two genera shows that they are very
closely related, and should perhaps be merged into one. The three species hitherto known
do not appear to me to differ very materially from one another, and it is highly probable
that when more specimens are obtained from the same ground intermediate varieties will be
found. In 1900 Miss Hiles (12) described two fragments of a Chironephthya with characters

488 SYDNEY J. HICKSON.

most closely approximating to the description of C. scoparia. This specimen was obtained
by Dr Willey in shallow water off the coast of New Britain. I have recently discovered in
a collection made by Mr Bedford at Singapore a specimen of Chironephthya from 10 fathoms
of water, which I hope to describe in a future memoir. In Mr Gardiner’s collection from
the Maldives there are several specimens obtained in water of about 20 fathoms in depth
which present several features of particular interest.

Since the publication of the Challenger reports, then, we have eames firstly that the
genus is not confined to deep water, and secondly that it has a wide geographical distribution
in the eastern seas. It must also be noted here in anticipation of the special description
which follows that notwithstanding certain features in which the species or “facies” resemble
one another, there is such a well-marked variability in other important structures that it
is extremely difticult to determine even provisionally the range of specific characters. It really
becomes a question whether it is more convenient to consider all the specimens hitherto
obtained to belong to one widely distributed but very variable species, or to describe almost
every specimen that is obtained as a new species. Our knowledge is not yet sufficiently
advanced to adopt the former course, and the latter appears to me to be contrary to the
general trend of increasing knowledge and quite unscientific.

Chironephthya variabilis. New species.

The specimens collected by Mr Gardiner may be arranged as follows:

From Mahlos Atoll 20 fathoms, N.E. Lagoon. Two complete specimens (A, B) and
several fragments (C).

From Mahlos Atoll, 22 fathoms, N. Fainu. One complete specimen (D) and fragment (£).
From Mahlos Atoll, 23 fathoms. Several fragments (/)).
From 8. Nilandu, 24 fathoms. One complete specimen (G) and some fragments (/).
There are also smaller fragments from dredgings in nearly all the atolls.
The sizes of the complete specimens were

A, 45mm. in height.

B, 34mm. in height.

D, 50 mm. in height.

G, 65 mm. in height.

The most striking feature in connection with these specimens is the extraordinary
variability in colour. The colour of the spicules of the anthocodiae is usually in marked
contrast to that of the general coepenchym. The spicules of the anthocodiae are arranged
in a well-developed circlet round the base of the tentacles, which I shall for brevity call
the “crown,” and in a triangle guarding the lower part of the aboral surface of the tentacles.
These triangles I shall call for the same reason the “points” (Figs. 9 and 10).

The range of colours is as follows:

A. General coenenchym speckled red and yellow at the base, yellow in the upper part
of the stem and branches, very pale yellow crown and points.

B. G. C. dark red at the base fading into pink in the upper part of the stem and
branches; crown and points dark purple.

THE ALCYONARIA OF THE MALDIVES. 489

C. G. C. yellow, shading into pink at the termination of the branches, crown and points
purple.

D. G. C. red, with a few yellow spicules occurring irregularly among the red ones;
crown and points yellow fading into white, the spicules of the crown being more frequently
yellow than those of the points.

E. G. C. yellow; crown and points purple.

F. G. C. pink; crown and points purple. (Some fragments have G. C. yellow.)

G. G.C. of the stem red, fading into yellow in the main branches; crown and points orange.

The specimen D (Fig. 8) having a total height of 50 mm. has a stalk (barren stem) 23 ™m.
in length, it is slightly expanded at the base, where it is 8mm. in diameter and 6°75 mm.
at the level of the lowest branch. It bears ten small branches, given off irregularly from
all sides of the stem, and some of the larger of these bear rudimentary branchlets. The
anthocodiae are crowded at the termination of the branches, but occur more scattered on
all parts of the branches and main stem above the stalk. Specimen G, having a total height
of 65 mm., has a stalk 36 mm. in length. The expanded base of attachment is 11:25 mm.
wide, and the stalk narrows to a diameter of 3°55 mm. at the level of the lowest branch.
There are only six short branches, the longest being 10 mm. in extreme length.

Further description of these complete specimens is unnecessary, as it is perfectly obvious
that they are young colonies which would have altered very materially in form and_pro-
bably in colour also had they grown to the full size.

The fragments (H) from the same locality are in all probability branches of a larger
colony which may have reached its maximum size. On these we find a cluster of four or
five anthocodiae at the extremity, and others scattered at intervals of 83—5 mm.

It is impossible to say from the evidence of this collection to what size the colonies
may grow. I have seen a specimen obtained by Professor Herdman off Ceylon which is
over 1 foot in diameter.

In none of the specimens can it be said that there is a very clearly defined spiral
arrangement of the anthocodiae, although on some branches an ill-defined spiral arrangement
can be observed. I do not consider that the spiral arrangement of the anthocodiae can be
used as a character of any specific importance, as it is clearly subject to considerable variation.

In order to discover if possible some character which would be useful as a guide to
specific identification I made a careful examination of the anthocodiae of several specimens
and branches, carefully comparing the arrangement of the spicules of the crown and points of
neighbouring tentacles and polyps. The result was rather disappointing, as a considerable varia-
tion was observed throughout. The arrangements of the spicules of the point are not alike
even in tentacles of the same polyp, and the range of variation in anthocodiae of the same
branch is surprising. However, the investigation has yielded certain general results which may
be stated as follows. In the crowns of Chironephthya variabilis (Figs. 9, 10) there are usually
about ten rows of spicules arranged parallel with one another, and at right angles to the long
axis of the anthocodiae.. In the points there are four large and prominent spicules arranged
en chevron, of these the two outer ones are usually larger than the imner. The outer and
larger spicules are frequently bent slightly in the middle. The two outer spicules of the
chevron are the largest spicules of the anthocodiae. In the angle of the chevron other
spicules intrude, the number of which varies with the age of the anthocodiae.

G. IL 63

490 SYDNEY J. HICKSON.

The size and shape of the spicules vary considerably. In young anthocodiae, as for
example on D (Fig. 9), the largest spicules of the points were 0°5 mm. x 0‘04mm., and
the spicules of the crown 0°5mm.x0:015mm. In older anthocodiae, as for example on F,
they were respectively 0°6 x ‘08 and 0°5 x 0-4, and in £ 0°77 x 0:085 and 0°5 x 0:05 (Fig. 10).

The spicules of the points in H might be described as tightly packed, a large number
being wedged into the basal part of the triangle, and in the crown the greater stoutness
of each spicule causes them to le much closer together than in the crown of D. This
tight packing of the spicules of the older specimens renders the anthocodiae much less
retractile than in the younger specimens. There is another series which is less regular.
The spicules of the anthocodiae of the youngest specimens are usually white, or pale yellow,
those of the oldest purple. It is usually in the purple specimens that we find the tightest
packing of the spicules. Thus in D the spicules are white, in #’ orange, and in £ purple.
It seems a reasonable proposition to suggest that the colouring matter of these spicules
changes from white to purple as the colony approaches the light, or as they increase in age,
and that the noticeable variations in the series are due to local differences in surroundings.

Before comparing Chironephthya variabilis with the species C. dipsacea, C. scoparia, and
C. crassa of Wright and Studer (22), it may be pointed out that these authors do not
state definitely that the measurements given below refer to the largest spicules of the polyps,
but it is probable that this was their intention.

C. variabilis (mihi) C. dipsacea C. scoparia C. crassa
D E F
Crown 05x “O15 0:5 x 0°05 0°5 x 0:04 02 x003 0-7 x0-04 0°67 x 0:05
Spicules
_Point 0°5 x 0:04 0°7 x 0°85 0°6 x 0:08 012 x 0017 0°54 x 0:03 09 x013
Spicules

If Wright and Studer’s measurements refer either to the greatest size or to the average
size of the spicules in these situations it is clear that the range is much greater between
C. dipsacea and OC, crassa than it is between any two examples that I have studied in
this collection. As regards the arrangement of the spicules Wright and Studer state that
in (. dipsacea and C. scoparia the spicules of the points are arranged in a A-shaped
manner, but no detailed description or figure has been given.

I may say, however, that from a careful examination of the type specimens in the
British Museum I am convinced that the arrangement of the spicules of the points in all
three of the Challenger species is essentially the same as it is in C. variabilis mihi. There
are variations in size and in crowding as in the Maldive specimens, but the four principal
spicules of the points can be recognised in most of the anthocodiae I examined. Turning
now to the question of the colour. A glance at the type specimens of the Challenger
species shows how variable is the colour of the spicules. The description given by the
authors is accurate so far as it goes, but as a matter of opinion it does not appear to
me to give enough detail of the differences in colour between branch and branch of the
same specimen, or the gradual fading of one colour into another on the stem, as to do
full justice to the subject. I think that any impartial naturalist who carefully examined
these specimens would say that colour is a character of very little value in the attempt
to separate these three species.

It is not my purpose to go fully into the accounts given of the arrangement of the

THE ALCYONARIA OF THE MALDIVES. 491

anthocodiae, the mode of branching, the size of the spicules of the coenenchym and the
other characters of the Challenger specimens; but I think it will be found in the future
that it is not advisable to recognise in them more than one species. I propose to give a
new specific name to the Maldive specimens for various reasons, but I am prepared to admit
that the specific distinctions between them and the Challenger specimens are slight and
unimportant.

I think, however, it will be found that in the specimens proposed to be named
C. variabilis the spicules are on an average larger and more regularly arranged in the
coenenchym, that the colour, branching and arrangement of the anthocodiae are more irregular
(ve. less in conformity with a spiral arrangement), and the anthocodiae rather larger and
more retractile than in the Challenger species. It is better, however, to leave the matter in
an undecided position until more specimens are obtained from a depth of over 200 fathoms.

The next question is, How are we to distinguish the genus Stphonogorgia from Chiro-
nephthya now that the two characters given by Wright and Studer as diagnostic of the
latter genus have broken down ?

Having carefully compared the specimens of Chironephthya variabilis with Siphonogorgia
mirabilis from the Maldive archipelago, I would suggest that the name Chironephthya be
retained for the species or facies with a form and mode of branching resembling that of
the genus Nephthya, with anthocodiae rarely completely retracted in preserved specimens, and
with four principal spicules arranged en chevron in the points of the anthocodiae (Figs. 9
and 10); and that the name Siphonogorgia be retained for species or facies of more massive
Gorgonia-like form of growth, with anthocodiae capable of complete retraction within the
general coenenchym and with spicules irregularly placed or arranged in a fan-like manner in
the points of the anthocodiae (Fig. 11).

As regards the second of these characters it may be pointed out that in Siphonogorgia
pendula (W. and 8.) and the variety of Stphonogorgia pendula described by Kiikenthal as
ternatana, the anthocodiae do not appear to be completely retractile. But Kiikenthal’s variety
would be more properly referred to the genus Chironephthya, and the type must be regarded
as an intermediate form between the two genera which it is convenient for the present to
retain in the genus Siphonogorgia on account of its Gorgonia-like form.

As regards the third character, Siphonogorgia pendula (W. and 8.) and Kiikenthal’s
variety ternatana have the spicules of the “points” arranged in a manner very similar to
that of Chironephthya. There is no very careful description or figure of this feature, in other
species, owing to the retractility of the anthocodiae rendering the observation a difficult one.

In the specimens from the Maldives which I have referred to the species S. mirabilis
an examination of nearly a hundred anthocodiae showed that only in a few rare cases there
are four prominent spicules arranged en chevron as in Chironephthya:; but in the majority
of cases there is one prominent median spicule with four or five bent spicules arranged fan-
wise on each side of it, as shown in Fig. 11.

The species Stphonogorgia macrospina described by Whitelegge (24) from Funafuti from
a few fragments appears to me to be more correctly referred to the genus Chironephthya.

It is probable from the literature of the subject that in most of the species of Sipho-
nogorgia the arrangement of these spicules is similar to this, and it is also probable that
this arrangement is intimately associated with the power of retraction of the anthocodiae.

63—2

492 SYDNEY J. HICKSON,

Siphonogorgia mirabilis Klz. (Figs. 11 and 12.) Klunzinger (14) and Kiikenthal (15).

Two fragments, A, from Suvadiva Atoll, 25 fathoms across the mouth of the passage, B,
from Fadifolu Atoll, 70 fathoms outside atoll N.W. on steep slope.

These two fragments differ in some important particulars from Klunzinger’s species and
from one another, but it will be more convenient at present to name them Siphonogorgia
mirabilis Klz., and to point out the chief points of disagreement with the type specimen than
to refer them to a new species.

Fragment A from Suvadiva may be only a primary branch of a large colony. It is
110 mm. in length, with nine branches on one side and five on the other. Only one of these
branches shows any tendency to further ramification.

In specimen B it is possible we have an almost complete specimen with only the base
missing. It is 100 mm. in length, and is very much broken up, but some of the longer branches
exhibit secondary and even tertiary ramification.

Both of these fragments have a dirty-blackish crust, formed by an epizoic sponge, cover-
ing both stem and branches. In specimen B this crust is so thick on many of the branches
and a part of the main stem as to completely hide the red colour of the coenenchymal spicules,
but in A the red colour can be seen through the sponge, as also in the main stem and some
of the branches of B. The sponge belongs to the genus Halisarca, and I am informed by
Professor Topsent it probably belongs to the species H. Dujardini. In both specimens antho-
codiae can be clearly seen, and are particularly noticeable owing to their bright yellow colour.

The specimens agree with the original specimen of Siphonogorgia mirabilis in the following
particulars.

The colonies branch in one plane, sometimes dichotomously. The stem and branches are
almost cylindrical, the finest branches irregularly angular. The extremities of the finer branches
are blunt, and bear from 4—6 anthocodiae. The anthocodiae occur on all sides of the branches
and twigs, but are scarce on the larger branches, and fail altogether on the lower part of the
stem (specimen B).

The spicules of the coenenchym are spindles covered with warts and knobs, and of a deep
red colour. They are irregularly arranged on the main stem and larger branches, but almost
parallel with the axes on the smaller branches. The spicules of the anthocodiae are yellow in
colour, slender spindles, with sparse warts and tubercles.

The most important point of difference to be observed between the Maldive forms and the
type lies in the size of the spicules. According to Klunzinger the largest spicules of the
coenenchym are 0°5 mm. x 0128 broad. In the Maldive forms some of the largest spicules are
4 mm. x 02 mm., but the average is about 2mm.x015 mm. The size of these large spicules
varies so much in some of the species that it is doubtful whether much reliance can be laid
upon the figures given by some authors as being by any means exclusive. Kolliker in his
original description of Siphonogorgia godeffroyi states that the spicules vary from 1—4 mm. in
length. In the Challenger series the sizes of these spicules are given by Wright and Studer
as follows, S. pallida 2 x °3, S. pustulosa 2 x°38, S. pendula 2°8 x 224, S. Kéllikeri 1:2 x 0-2,
and the spicules of Kiikenthal’s species were S. cylindrata 3:7 x 04, S. mimacea 1:3 x 0:14.
It is to be observed, however, that in none of the descriptions of the species these spicules
are said to be so small as 0°5 mm. in length as stated to be the case by Klunzinger in S. mira-
bilis. From the fact that Klunzinger’s excellent work is unfortunately hampered by many

THE ALCYONARIA OF THE MALDIVES. 493

misprints, coupled with a study of the lowermost of his four figures marked d, which appears
to be the point of a large spicule 2 or more mm. in length, and that the author does not
comment on the small size of these spicules as compared with those of S. godeffroyi, a doubt
may arise as to whether 0°5 mm. in his description is not a misprint for 5 mm.

Even if the spicules are so small as they are stated to be in Klunzinger’s specimen it does
not seem to me that this is a sufficient reason for making the Maldive specimens into a new
species. The size of the spicules of Alcyonarians varies very much in different localities and
conditions of growth, and there might well be a considerable range in this respect in one
species that occurs over such a wide range as the Red Sea and the Maldives.

Klunzinger makes a note at the end of his description that “Beim Trocknen tritt die
Bindesubstanz durch ihre weisse Farbe vor und giebt dem Ganzen ein staub- oder schaum-
artiges Ansehen oder wie weiss und roth gesprenkelt und das Roth tritt zuriick.”

I have found that on drying a small portion of one branch of specimen B that the en-
crusting sponge turns white, and gives the specimen very much the appearance described by
Klunzinger. It is quite possible, therefore, that what this author describes as the superficial
connective tissue is in reality an encrusting sponge.

The bright yellow colour of the spicules of the anthocodiae which forms such a very con-
spicuous feature of these specimens is a minor point of resemblance with the type of S. mirabilis.
In Siphonogorgia pustulosa (W. and S.) we have a similar combination of yellow spicules in
the anthocodiae with red spicules in the coenenchym. In S. cylindrata (Kiik.), the spicules of
the anthocodiae are also pale yellow, the spicules of the rest of the colony being yellow.

The slender spicules of the crown and points of the anthocodiae vary in size, but average
about 3 mm. in length, we. only a trifle longer than those of the type specimen. The arrange-
ment of the spicules in the points, ue. on the aboral side of the tentacles, has been described
above under the genus Chironephthya.

The geographical distribution of the genus Stphonogorgia, as it is at present known, is
interesting. Specimens have been obtained from the Pelew Islands, Ellice Islands (Funafuti),
N.W. Australia, New Hebrides, Malay archipelago, and one specimen hitherto found in the
Red Sea. The discovery of Siphonogorgia in the Maldives fills up a gap between the extreme
East and West limits, and suggests that it may occur in other suitable localities of the Indian
Ocean.

Family Briareidae.
Genus Solenocaulon (Gray).

The general features of this genus were first described by Gray (8). Subsequently
Genth (6) described with considerable detail an Alcyonarian from the Philppines, which he
named Solenogorgia tubulosa. Later on Studer (21) in describing a new species (Solenocaulon
grayi), gave reasons for including Genth’s specimen in Gray’s original genus Solenocaulon.
Six years ago Germanos (7) in his paper on the Gorgonacea of Ternate gave a historical
review of the genus, and added three new species.

In Germanos’ interesting and important contribution to our knowledge of the genus he
proposed to arrange the six known species into two groups as follows:

494 SYDNEY J. HICKSON.

1. Sub-genus Sclerosolenocaulon.
Solenocaulon tubulosa, Genth.
Solenocaulon sterrokloniwm, Germ.

Solenocaulon diplokalyx, Germ.

2. Sub-genus Malacosolenocaulon.
Solenocaulon tortuosum, Gray.
Solenocaulon akalyx, Ternate.

Solenocaulon grayi, Studer.

A careful consideration of Germanos’ laborious investigation in the light thrown upon
the subject by Mr Gardiner’s fine collection and the type specimens in the British Museum
has led me to the conclusion that this arrangement of the species is not satisfactory. The
first and most important character that is used by this author for the separation of the
sub-genus Sclerosolenocaulon from Malacosolenocaulon is the presence of a stalk. It appears to
me to be still a matter of doubt whether there are any specimens of Solenocaulon without
a stalk, since the original specimen described by Gray certainly has a stalk, which is solid,
and the specimen attributed to the same genus by Studer was incomplete at the base, and
therefore afforded no evidence on the subject. The specimen upon which Germanos bases
his species S. akalyx consisted only of “zwei grosse Stiicke,” and as no description, figure,
or even statement is given of the base of attachment, it is reasonable to suppose that it
was missing. The stalk is also missing in the only specimen that has hitherto been described
of the species S. grayi, Studer. It is difficult to understand how Germanos could have over-
looked these facts in drawing up his classification, unless it is supposed he uses the word
stalk “Stiel” in a sense which is not usual. As regards the character of the axis which
plays an important part in this classification, I may say that Studer's description of the
axis “aus durch Hornmasse verbundenen, losen, spindel-férmigen Spicula bestehen” accurately
describes the axis of Mr Gardiner’s specimens. I cannot understand what is meant by the
statement that in the branches of Solenocaulon tubulosa, S. sterroklonium, and S. diplokalyx
the spicules are fused to form a solid axis. If we are to understand that the calcareous
spicules are fused together to form a solid calcareous rod as in the axis of Corallium or
the tubes of Zubipora, then it may be regarded as a character of specific or even of sub-
generic importance, but Germanos brings little evidence to prove that this is the case. In
many of the smaller branches of my specimens I find the spicules of the axis so tightly
packed together as to form a porous calcareous rod as figured by Genth, which does not
break up on boiling in 5°/, potash, but as Genth adds, “Die verschmolzenen Spicula gehen
ohne merkliche Grenze in die freien iiber.’ Whether the spicules do or do not become so
tightly packed as this in the branches depends upon the rapidity of growth of the indi-
vidual branch, and is not of the slightest systematic importance. If the spicules fused in
some forms to form a solid axis in the stem or stalk and not in others, we should have
a character of prime importance, but the axis of the stem and stalk is admittedly the same
in all species.

The species Solenocaulon tortwosum Gray and Solenogorgia (Solenocaulon) tubulosa Genth
differ from one another principally in the fact that the latter gives off from the main
tubular stems several short grooved blunt branches bearing numerous anthocodiae. In the

THE ALCYONARIA OF THE MALDIVES. 495

former these branches are rather longer and narrower. In S. grayi (Studer) the grooved
branchlets are much longer, more delicate, and pointed or knobbed at the extremity. S. tubu-
losa is therefore as regards its form intermediate in character between S. tortuosum and
S. grayt. If we elimimate these characters of form there is, it appears to me, very little
to distinguish the species. Now in considering the growth of a form like this we must
not leave out of consideration the fact that spirit specimens prove that usually and perhaps
always, the tubes of Solenogorgia contain epizoic Crustacea belonging to the genus Alpheus.
The general appearance of the groove on one side of the non-tubular terminal branchlets,
in specimens similar in form to that described by Studer as S. grayi, suggests that these
Crustacea are in the habit of running along them in search of food, and retreating to the
tubular parts for protection. The general characters of the groove are the smoothness of
the surface as compared with the outer convex side of the branch, the absence of antho-
codiae, and the relatively pale colour. These are characteristics which we should expect to
find if any part of the surface of a coral were constantly irritated by the feet of an active
little crustacean. Now it is well known that in many forms of corals such as Pocillopora,
Seriatopora, Mullepora, etc. one effect of a symbiotic crab on the growth is the formation
of a spherical or oval gall, the margins of the affected part being hypertrophied and growing
round and enclosing the animal that causes the irritation. Similarly, barnacles of various
species cause gall-like growths of the coenenchym in various Madreporaria, Millepora, Gor-
gonia, and other Alcyonaria. An examination of the surface enclosed or about to be enclosed
by the galls of Pocillopora, Seriatopora, and Millepora shows a smoothness of the surface,
and, in Millepora, an absence of polyps similar to that shown in the groove of the Soleno-
caulon branchlets.

It appears to me therefore very probable that the tubular character of Solenocaulon is
of the nature of a gall produced by an active crustacean, and that the degree to which
it affects the secondary and tertiary branches depends, not upon any inherent specific
character of the Solenocaulon, but upon the number or activity of the symbiotic crustacea.

Whether this explanation is true or not, it is not satisfactory from a systematic point
of view to distinguish as distinct species specimens which differ from one another principally
in their form of growth. The many striking similarities that there are between S. tortwosum
of Gray, S. tubulosa of Genth, and S. grayi of Studer in the shape and arrangement of the
spicules and other characters suggest very forcibly that they all belong to one widely
distributed species.

This conclusion receives considerable support from the examination of the very rich and
interesting collection of the specimens of this genus made by Mr Gardiner, and I shall
therefore place them all, except one which I believe to be the type of a new species, in one
species.

Solenocaulon tortuosum Gray. (Fig. 13.)

It will be convenient to give in the first place a list of the specimens, with a short
description of each one.
A. A small specimen of a pale pink colour, About 145 mm. in height. Stalk
75 mm. long, 5mm. in diameter. Most of the lateral flattened branchlets
short. Suvadiva, 43 fathoms. Lagoon. Hard rubble.

496 SYDNEY J. HICKSON.

B. A dried specimen rather broken, but of approximately the same size and form
of growth as A. Pale cream colour. Suvadiva, 37 fathoms. Near a passage.
Hard sand.

C. A specimen about 170 mm. in height with axis divided into two main branches
(Fig. 13). Stalk 60mm., bifid at base. There are few narrow grooved
branchlets. Colour, orange-brown. Suvadiva, 38 fathoms. In a_ passage.
Rubble bottom.

D. Two fragments, with rather longer grooved branchlets than in A. Colour, a
rich pink, darker in tone than A. Fadifolu, off W. side, 22 fathoms. Weed,

sand, and small coral.

E. A specimen in fragments. It was probably twice as large as A when complete.
In this there are several long, deeply grooved branchlets over 15 mm. in
length. The parts that form complete tubes are very restricted in extent.
Dull pale brown in colour. Mulaku, 22 fathoms. Weed and sand.

F. Another somewhat smaller specimen, with several very long almost flattened
branchlets. None of the branches among these fragments show the com-
pletely tubular condition. Colour, dull pink. Mulaku, 32 fathoms. Mud,
weed, and sponges.

G. A specimen also in fragments, with several fairly long grooved branchlets, but
freely tubular on the branches. The colour varies from dull pink to dull
pinkish brown. S. Male, 28 fathoms. Hard white sand.

H. A complete specimen. About 180 mm. in length, with a very long stalk broken
at the extremity and a divided axis as in C (Fig. 13). Very few long
grooved branchlets. Freely tubular. Colour, pinkish brown. Stalk fading to
white at base. §S. Nilandu, 25 fathoms. W. of atoll.

K. The largest specimen in the collection, but very much broken, The stalk is
relatively short. The axis and primary branches form complete tubes through-
out nearly the whole of their extent, but there are some large projecting
flattened or grooved branchlets 80 mm. and more in length. The colour is
brown. N. Male, 27 fathoms. N. of atoll. Broken shells.

L. A small fragment of the same general character as K, but rather more pink
in colour. N. Male, 28 fathoms. Near shoal, W. centre. Sand.

M. A very small fragment of same general characters as K, but bright orange in
colour. Mahlos Atoll, 23 fathoms. Passage. N.E. lagoon.

In this series of specimens from the same group of islands we have a series of links
connecting together the well-known and well-described species belonging to the genus. The
careful examination of the specimens A, B, and C from Suvadiva, and a comparison of the
spicules and tentacles, sections of the stalk and axis with the description given by Genth
convinced me that these specimens belong to the species S. tubulosum.

A comparison however of specimens G and K with the type specimens of S. tortwoswm in
the British Museum exhibited no differences either in general mode of growth or spiculation
that would justify a separation of them from the original species.

THE ALCYONARIA OF THE MALDIVES. 497

The fragment L moreover exhibits branches which have a remarkable resemblance in
their form and the manner in which the anthocodiae are distributed upon them to those
of Studer’s species S. grayt.

The specimen F from Mulaku I should place with Gray’s species Lewcoella cervicornis,
were it not for specimen E, which is intermediate in character between Leucoella cervi-
cornis and Solenocaulon grayt. The spicules of F cannot be distinguished from those of the
type specimen of Leucoella cervicornis. The other specimens cannot be so easily placed in
any known species, but appear to form intermediate stages of growth or modification between
those already referred to. It appears to me that the facts above mentioned point to the
conclusion that all these specimens of the genus Solenocaulon from the Maldives belong to
one species, and by the general custom of priority should be referred to Gray’s species
Solenocaulon tortuosum. Such a conclusion, however, would not be justified without reference
to other characters than the mode of growth.

That there are differences between the spicules in my preparations cannot be denied, but
when due allowance is made for the variation in size and shape of the spicules of the various
parts of one specimen there remains such a strong family likeness between them all, including
Leucoella cervicornis, that the differences cannot be expressed in words or figures. When
the different accounts of the spicules of the genus are compared there is a surprising similarity
rather than difference in their character to be noticed. I have collected, for example, the
length-measurements given of the long-thorned spicules of the axis, and they are as follows.

S. tubulosum (Genth) 0°36 circa. S. grayt (Studer) 0°37—0-4. 8S. tortuoswm (Studer) 0:3—
0-4. S. tortuosum (Ridley) 0°38. Leucoella cervicornis (Ridley) 0°35—0:42. 8. sterroklonium
(Germanos) 0°2—0'38. These figures are sufficient to show that in the matter of this measure-
ment there is a remarkable uniformity even in specimens so far apart as Torres Straits and
the Maldive archipelago, and as other measurements show corresponding results the spicules
do not afford sufficient grounds for splitting the genus up into species.

The next feature it was necessary to investigate was the character of the verrucae. I
carefully selected branches from several of the specimens from the Maldives, and dried them
slowly by exposure to the air in the laboratory, and then examined them with a 1 inch
objective. In all cases some of the polyps were contracted so much that the spicules of the
tentacles were below the level of the verrucae. In others which were not so completely re-
tracted there could be seen an arrangement of the spicules at the back of the tentacles and
in the neck of the anthocodiae (7.e. what I have called on p. 488 the crown and point spicules)
similar to that described and figured (Fig. 17) by Genth. Another point of considerable
importance I found in studying one of the polyps of a specimen, and that is the presence of
rough and somewhat irregular spindle-shaped spicules supporting the pinnules, in the manner
described and figured by Genth, 0'2 mm. in length. The presence of these spicules in Genth’s
specimen from the Philippines, and in this specimen from the Maldives, is especially note-
worthy.

In conclusion I may say that I propose to name the specimens of Solenocaulon in the
collection as follows:

ABC — Solenocaulon tortuosum, Facies tubulosum Genth,
DGH KM Solenocaulon tortuosum, Facies tortuosum Gray,
LE Solenocaulon tortuosum, Facies grayi Studer,
F Solenocaulon tortuosum, Facies cervicornis Gray.
G. 1. 64

498 SYDNEY J. HICKSON.

Solenocaulon ramosa. New species. (Figs 14, 15, 16, 17.)
Locality. Suvadiva Lagoon. Coarse sand and rubble. 39 fathoms. One specimen.

With the usual system I have little doubt this specimen would be made the type of a
mew genus. but after the examination of the other specimens of the genus it is only with
some hesitation that I propose to constitute a new species for it.

The characters by which it is distinguished from all the other specimens of Solencaulon
in the collection, and indeed from the other described species of the genus (including Leucoella),
might be accounted for as local variations But, bearing im mind, that no less than three
specimens A. B, and C of the ordinary kind came from the Suvadiva lagoon at the same depth,
this consideration loses a great deal of weight. Nevertheless it is just possible that the
character of the currents ai the spot where this was found influencing the symbiotic crus-
tacea Im some way or another may have produced this abnormal form of growth. It is possible
that a different and larger species of crustacean may have sheliered im the branches, driving
away the normal species, thereby causing the large belt-like sheliers, and preventing the
formation of the narrower tubes. I do not think that any such causes are so likely to have
affected the other features m which it differs from the other specimens of the collection.

The specimen is very much broken, but by putting the fragments together I have
reckoned that when taken it was about 280mm. im height. It is probable, however, that
this measurement is considerably short of the total length of the specimen when it was alive.
There is no sialk or base of attachment.

The most characieristic feature of the specimen is the presence of three complete belis,
and one incomplete belt on the axis, and one meompleie belt on a branch. These belts
are formed in the same manner. and have the same significance as the tubes of Soleno-
caulon tortuosum. They are simply flattened wing-like outgrowths from the axis or branch
which embrace a cavity and fuse where they meet at their extremities (Figs. 15, 16).

These belis are from 15—20 mm. in diameter above, and from 22—23 mm im diameter
below, and about 20mm m length. They are situaied on the axis at imiervals of about
30mm. The imeompleie belis are very similar to the complete ones im size and form, but
the sides are not fused opposite to the axis.

In ome or two of the fragmenis there are places where a branch is very much flattened
and expanded, and these are undoubiedly to be regarded as commencing belts (Fig. 15).

Unfortunately there is no direct evidence as to what these belis contained, but it is
very probable that they sheltered some epizcic crustacean or other animal. The resemblance
of the imperfect belts to the growing crab-galls of a Seriatepora or Pocillopora is very
striking.

Tn its thickest part (nearest to the missing stalk) the axis is roughly cylindrical im form,
and about 8 mm im diameter. Tt is scored with regularly arranged longitudinal grooves.
Above the lowest belt it is more flattened. and about 5 mm. im thickness, but from this pomt
upwards the extent of flatienmg is very irregular, some parts being much broader than
others.

The structure of the axis appears to be precisely similar to that of S. tortuosum, the
corky substance of the central part and the softer vascular bark having the same general

appearance in the two species.

THE ALCYONARIA OF THE MALDIVES. 499

The spicules of the axis and of the branches are so remarkably similar to those of the
other species that I cannot distinguish them. A very striking feature, however, of the new
species is the fact that none of the polyps are completely retracted into the verrucae, and
this gives the specimen a much more spiny or hirsute appearance.

The incomplete retractility of the polyps is due to the fact that the spicules of the
crown and the points are not only more numerous and therefore more closely set, but distinctly
thicker than they are in the other species.

I have very carefully compared some polyps of nearly all the specimens of Solenocaulon
from the Maldives, and in none of them could I find the same number or character of these
spicules as in the new species.

The pimnulae of the tentacles are supported by spindle-shaped spicules, but these again
are stouter than in the corresponding spicules of S. tortuosum.

The colour of the specimen is dark brown. It will be observed that this colour is very
different to the pale pink, cream, and orange colours of the other specimens of Solenocaulon
from Suvadiva. ;

Family Melitodidae.
Preliminary note on the genus Melitodes (Vermill).

In the collection from the Maldives these several specimens belonging to the genus
Melitodes and the species problem presents itself again with exaggerated difficulties.

The specimens are all very brittle and considerably broken, so that the exact form of the
whole colony is difficult to determime, but exceedingly well-marked differences between indi-
vidual specimens can be observed in the size of the nodes, the length and thickness of the
internodes, and the colour of both nodes and internodes.

From Hulule, Male Atoll, alone I have already noticed the five principal varieties given in
the following table.

| Nodes Tniernodes

Colour Diameter Colour Length Diameter
—
| 1 | Yellow | 4mm. swollen Red | 7-Bam | 2mm.
| 2 Yellow | 2mm. not swollen White 22 34mm | 2mm.
is | Bed ~ | 1mm. swollen White 5 mm 5 mm
4 | Yellow ) 3 mm. slightly swollen Pink 15 mm. 2mm.

15 | Red 1 mm. slightly swollen Gray 10mm | 75 mm.

From Addu Atoll. There are four varieties, apparently quite distinct from one another and
from the Hulule, Male, varieties.

oi—2

500 SYDNEY J. HICKSON.

From these two atolls no less than nine species might be made which could be fairly well
defined by measurements of nodes and internodes and by colour; but on the other hand the
three or four varieties from Miladumadulu, one from North Male, two from Suvadiva, and
one from Felidu, make up a series which is almost complete between the extremes. The
examination of the spicules does not help us in the matter very materially. There is so much
variation in the size and shape of the spicules in every individual specimen that the com-
parison between one specimen and another is very difficult, but of the many preparations of
spicules that I have made there is not one that can be recognised as belonging to a distinct
form of growth. The preparations prove that the spicules of all varieties have a general
family resemblance, but beyond that it is impossible to speak definitely.

The problem, however, is complicated in this case by the fact that Mr Gardiner from
observations on the reef came to the conclusion that there is more than one species. I feel,
therefore, that it is necessary, before definitely expressing my own opinion, to submit some of
the specimens to a further anatomical examination. The publication at the present moment,
however, of the fact that this interesting series of Melitodes varieties does occur, appears to
me to be advisable, although the detailed description of them must be necessarily delayed.

LITERATURE.

1. J. H. AsuwortH. “The Structure of Xenia Hicksoni.” Quart. Journ. Micr. Sci. 42.

2. “The Xeniidae.” Willey’s Results, Pt. Iv.

3. H. M. Bernarp. “On the Unit of Classification for Systematic Biology.” Proc.
Camb. Phil. Soc. X1. 4.

4. G. C. Bourne. “On the Structure and Affinities of Heliopora coerulea, etc.” Phil.
Trans. 1895.

5. J. D. Dana. “Zoophytes.” Philadelphia, 1846.
6. C. GentH. “Ueber Solenogorgia tubulosa.” Z. f. Wiss. Zool. Xvit. 1867.

7. N. K. Germanos. “Gorgonaceen von Ternate.” Kiikenthal’s Ergebnisse, 1896, and
Abhand. Senck. nat. Ges. xxi. 1.

8. J.E. Gray. “Description of Two New Genera of Zoophytes (Solenocaulon and Bello-
nella).” Proc. Zool. Soc. 1862.

9. J. E. Gray. “Description of New Species of Spongodes.” P. Z. S. 1862, p. 30.
10. S. J. Hickson. “The Alcyonaria (Stolonifera).” Trans. Zool. Soc. Vol. x1m. (1894).

?

ik, “The Aleyonaria and Hydrocorallinae of the Cape of Good Hope.’
Cape Town, 1900.

12. I. Hives. “The Stolonifera and Aleyonacea.” Willey’s Results, Part u. 1899.

13. O. Hot. “Beitrige zur Kenntniss der Alcyonidengattung Spongodes.” Zool. Jahrb.
vil. 1895.

THE ALCYONARIA OF THE MALDIVES. 501

14. C. B. Kuunzincer. “Die Korallthiere des rothen Meeres.” Part 1. 1847.

15. W. KUKENTHAL. “ Alcyonaceen von Ternate.” Kiikenthal’s Ergebnisse. Abhand. Sench.
nat. Ges. Xxiu. 1.

16. W. KtKentHat. “Versuch einer Revision der Alcyonarien.” Zool. Jahrb. System.
Xv. 6.

17. W. May. “Beitrage zur Systematik und Chorologie der Alcyonaceen.” Jen. Zeits.
J. Naturwiss. Bd. xxxi. 1899.

18. F. Mttier. “Ueber Balanus armata, etc.” Arch. f. Naturg. Jg. xxxut. 1867.

19. H. Ptrrer. “Alcyonaceen des Breslauer Museums.” Zool. Jahrbiicher. System.,
Bd. xin. Hft. 5, 1900.

20. S. O. Ripiey. “Contributions to our Knowledge of the Alcyonaria.” A. M. H.
March 1882.

21. TH. SrupEer. “Uebersicht der Anthozoa Alcyonaria der Gazelle.” Monats. k. Akad.
Wiss. Berlin, 1878.

22, E. P. Wricut and TH. StupER. “Alcyonaria of Challenger.” Vol. Xxxt.

23.

——— Supplement to “Alcyonaria of Challenger.” Vol. xxxm.

24. TH. WHITELEGGE. “The Alcyonaria of Funafuti.” Memoirs Aust. Museum, v. 1897.

PLATE XXVI.

Fie. 1. TZelesto rubra, n. sp. A portion of a colony with four secondary branches (6). At wv.
are seen the verrucae of the lateral polyps. g. a barnacle gall. All the parts that are
shaded are covered with the encrusting sponge Prosuberites, which is coloured blue-green
with Beggiatoaceae. The unshaded parts show the red colour of the skeleton through
a very thin lamina of the sponge. x 1} ca.

A portion of the calcareous skeleton of Telesto rubra after treatment with caustic potash
to show the ridges (7.7.) and the irregular perforation of the tubes.

Fig.

i)

Fic. 3. Side view of one of the tentacles of Yelesto rubra to show the eight pinnae (p.p.), each
supported by one spicule, and the smaller spicules supporting the axis of the tentacle sp.

Fig. 4. One of the spicules of the pinnae of Telesto rubra isolated. x 200.

Fic. 5. One of the red spindle-shaped spicules of Z'elesto rubra at the base of the anthocodia, which
increase rapidly in size, fuse, and form the solid tubular skeleton. x 200.

Fic. 6. Two spicules of the same region of Telesto rubra very much younger and smaller than
the one in Fig. 5. x 200.

Fic. 7. <A characteristic spicule of Zelesto trichostemma, x 150.

Fic. 8. Chironephthya variabilis, n. sp. A sketch to show the general form of the colony. x 1} ca.

502 SYDNEY J. HICKSON.

Fic. 9. An anthocodia of specimen D of Chironephthya variabilis showing the arrangement of the
spicules in the crown (C) and points (P). This arrangement is characteristic of what
appears to be young polyps.

Fie. 10. An anthocodia of specimen E of Chironephthya variabilis, showing the spicules in both
crown (() and points (7) larger and more crowded than in specimen D.

Fic. 11. Arrangement of the spicules in the points of Siphonogorgia mirabilis, showing one pro-
minent median spicule and the others smaller and arranged in the manner of a fan on
each side of it.

Fie. 12. A small piece of the terminal branch of Siphonogorgia mirabilis, the surface being covered
with the sponge Halisarca. Sp. cut edge of the sponge. Coen. the large spicules of
the coenenchym as seen when the sponge is scraped off. Anth. anthocodiae projecting
through the sponge.

Fie. 13. Solenocaulon tortuwosum Gray. Specimen C from Suvadiva. Nat. size. It shows ¢. the
tubular secondary branches, and b.b. the very short free terminal branchlets.

PLATE XXVII.

Fic. 14. Solenocaulon ramosa, n. sp. A small portion of a large colony. It shows at B.&. two
large belt-like growths on the main stems, and at b.b. the long free terminal non-tubular
branches.

Fics. 15, 16. Two stages in the formation of the belts of Solenocaulon ramosa.

Fie. 17. A terminal branch of Solenocaulon ramosa, n. sp., showing the non-retractile anthocodiae,
Anth.

Fic. 18. A terminal branch of Solenocaulon tortwoswum from the Maldives for comparison with
Fig. 17.

Fauna and Geography Maldives and Laccadives. Plate XXVI.

EMD.« S SHdel £.Wiison, Cambridge.

Hickson — Alcyonaria.

Fauna and Geography Maldives and Laccadives. Plate XXVII.

Sete pe, ey ee
EMD. del £. Wilson, Cambridge

Hickson — Alcyonaria

THE ALCYONARIA OF THE MALDIVES.

1 Ege aid be

THE GENERA SARCOPHYTUM, LOBOPHYTUM, SCLEROPHYTUM,
AND ALCYONIUM.

By Epita M. Prart, M.Sc. Vicr., Late Honorary Research Fellow of the
Owens College, Manchester.

(With Plates XX VIJI—XXXI1.)

CONTENTS.
PAGE

Me INTRODUCTION, WITH OBSERVATIONS ON THE GENERA ; F ; i ; § ; 504

II. Genus SarcopHyTuM : : é . : ; ; . . ; : 5 : 505
(1) S. ehrenbergi. (2) S. glaucum. (3) S. latum. (4) S. boettgeri. (5) S. roseum n. sp.

(6) S. tenuis n. sp.

III. Genus Losopnytum . : ; : : . 3 . : : . : : 5 514
(7) L. pauciflorum.

IV. Genus SCLEROPHYTUM n. gen. . : : : : > : . c : : : 516
(8) Sc. capitale n. sp. (9) Se. densum. (10) Sc. hirtum n. sp. (11) Se. poly-
dactylum. (12) Se. palmatum n. sp. (13) Sc. gardineri n. sp. (14) Se. durum
n. sp. (15) Sc. querciforme n. sp.

VY. Tue RELATIONS OF THE ABOVE GENERA . : 6 : 5 : c : . . 532

VI. Genus ALcyonium . . : : 5 , : . 3 9 c ; 5 : 534
(16) A. pachyclados.

535

VII. Literature

504 EDITH M. PRATT.

I. INTRODUCTION.

THE genus Sarcophytum was founded by Lesson in 1831, its distinguishing feature being
the mushroom-shape of the colony. Klunzinger in 1877 included in the genus a form which
is not mushroom-shaped, and which he termed Sarcophytum pauciflorum. Marenzeller in 1886
separated those forms which are mushroom-shaped from those which are not, and for the
reception of the latter he established the genus Lobophytwm, of which he described several
new species. This genus was based upon the form of the colony, the dimorphic character
of the zooids, and the form and size of the spicules.

When examining examples of several species in the collection which have been attributed
to Lobophytum, I observed that while all the specimens have the lobed form characteristic
of the genus, some differ from others in that they are extremely hard and brittle. This
I found is due to the presence of enormous spicules which occur throughout the colony in
very considerable numbers. In some cases siphonozooids were observed to be very minute,
and could only be distinguished with the aid of the microscope, while in a few instances
they were apparently absent.

In addition to the abundant supply of excellently preserved material in the collection
from the Maldive Islands, the zoological laboratories of the Owens College contained other
collections of similar material, which through the kindness of Professor Hickson were placed
at my disposal. These included Mr Stanley Gardiner’s collection from Funafuti, Dr Willey’s
collection from New Guinea, New Britain and Lifu, and Professor Haddon’s collection from
the Torres Straits. I have also been allowed to make a cursory examination of Professor
Herdman’s collection from Ceylon. It was, therefore, possible for me to make an investi-
gation of the anatomy of the genus Lobophytum, which hitherto had not been attempted,
and also to make a comparative study of the anatomy of many species, which have been
attributed to this genus, from many different localities. The general results of these in-
vestigations are contained in the present paper. A more detailed account of the comparative
minute anatomy will, I hope, be published later.

An interesting and important result of my research was the discovery of a large and
extremely well-marked superficial canal system in those species of the genus Lobophytum
with minute siphonozooids, which occurs also in those with apparently no siphonozooids. In
the former this canal system was observed to be in intimate connection with the minute
siphonozooids, but it is absent or only very feebly developed im those species with large
siphonozooids. Further investigation led me to the conclusion that the present definition
of the genus, which is based upon the form of the colony, dimorphism and the character
of the spicules, is artificial and inefficient, because it is founded upon insufficient evidence.

All the species of the genus Lobophytum divide naturally into two groups; one includes
all those species whose anatomy is similar to the species pauciflorum, which has no special
superficial canal system, but is provided with comparatively large siphonozooids; the other
group includes all the remaining species of the genus, whose general anatomy is similar to
that of the species densum, a form with minute siphonozooids and a large superficial canal
system. The important differences between the two groups of species clearly indicate that
they are genetically as distinct from each other as each is distinct from the genus Sarco-
phytum or any other genus of the Alcyonaria. It is, therefore, necessary to separate the

THE ALCYONARIA OF THE MALDIVES. 505

group of species with minute siphonozooids from the genus Lobophytum, and to establish a
new genus for its reception. The characteristic hardness of many of the species suggests
the generic title Sclerophytum.

In the collection from the Maldive Islands were young colonies of Lobophytum pauci-
florum (fig. 12) and Sclerophytum densum (fig. 18), which were mushroom-shaped. Several
similar colonies of Lobophytum and Sclerophytwm occur also in Professor Herdman’s collection
from Ceylon. The mushroom-like form of young specimens of these two genera is interesting,
and may lead to confusion in the identification of the genus, for the mushroom-like form
of the colony is held to be characteristic of the genus Sarcophytum. We have now evidence,
however, that it is not confined to that genus.

It has already been shown that well-grown specimens of Lobophytuwm and the new genus
Sclerophytum may be very similar in form and appearance, but very different in anatomical
structure. Specimens of Sclerophytum occur also in the collection, which differ considerably
in outward form, but which are identical in structure (Sclerophytum durum, figs. 29 and 30).
The form of the colony alone cannot, therefore, be recognised as a generic or even a specific
character.

In many cases the spicules of Sarcophytum, Lobophytum, and the new genus Sclerophytwm
are distimct from each other, and may be easily recognised, but this cannot be regarded as
an established rule, for the zoned spicules supposed to be characteristic of Lobophytum occur
also in Sarcophytum latum and roseum (sp. nov.); the tuberculate spicule occurring throughout
Sclerophytum is to be found also in Sarcophytum in the species plicatum, boettgeri, and
apparently in nigrum.

Fhe following species which I have examined would be retained in the genus Lobo-
phytum :—pauciflorum, crassum, and hedleyi. The new genus Sclerophytum (p. 516) will include
the species denswm, confertum, marenzelleri, tuberculosum, which have hitherto been attributed

to Lobophytum.

All the specimens of the Alcyonidae in the Maldive Islands collections were examined
for reproductive organs. In the majority of cases they were apparently absent. The re-
maining specimens contained ova usually exhibiting several stages of development in a single
colony. There does not appear to be a single specimen with male cells in the whole
collection.

II. GENUS SARCOPHYTUM (PI. XXVIII. figs. 1—6, 9—11).

This genus was first founded by Lesson in 1831 (p. 92), and forty-four years appear
then to have elapsed without any further account of it. Moseley in 1875 (p. 19, Pl. VIII.
fig. 2, Pl. IX. fig. 9) carefully described the general anatomy of a species belonging to this
genus. Lesson’s definition was amended by Marenzeller, 1886 (p. 351), and again by Wright
and Studer, 1889 (p. 244).

After carefully comparing in detail the anatomy and histology of several species included
in the genus I find that Wright and Studer’s definition holds good, as far as it goes, for
all the species which I have examined. It is necessary to mention, however, that the
mushroom-like form of the colony is not confined to this genus, but occurs among young
specimens of the genera Lobophytwm and Sclerophytum. Moseley’s account of the general

G. I. 65

506 EDITH M. PRATT.

anatomy is valuable as a basis for the comparison of other species, and is of material
assistance in the establishment of certain generic features.

Six species of Sarcophytwm occur in the collection from the Maldive Islands, three of
which have been recorded from other localities, one is a new variety of an old species, and
two species are new. The species are:—glaucum, latum, boettgeri, var. nov., rosewm, sp. NOV.,
tenuis, sp. nov.

Siphonozooids. Owing to the difference in contraction of preserved specimens it 1s
difficult to obtain any specific standard of measurement of the autozooids. A series of
measurements of the siphonozooids of several specimens belonging to the same species, how-
ever, show them to be of almost uniform size. The size of a siphonozooid in the preserved
condition therefore appears to be almost constant for a species, and may be looked upon
as a specific character. The length of the stomodaeum of a siphonozooid varies to a certain
extent in different species, but is almost constant for a single species. It will be seen from
the appended table that the stomodaeum of the siphonozooids is an important factor in the
determination of the species’.

S. glaucum |S. ehrenbergi| S.latum | S. plicatum?| S. tenuis S. roseum
Surface diameter i ; “90 2 4 :
of siphonozooids { 2 mm. 37 mm. 33 mm. 3 mm 3 mm. 45 mm.
Transverse length } . 4 A ‘ “19
Bietoniedsenmt 1 mm. 18 mm. 1 mm. 1 mm 1 mm. 12 mm,
Vertical length of 4 95 : : “6 F ¢
Bae ten 16 mm. 32 mm. 3 mm. 28 mm. 2 mm. 4 mm.

By introducing lamp-black into the sea-water about living colonies of Sarcophytum,
Mr Gardiner observed currents entering the siphonozooids. After some time the blackened
water was ejected by the autozooids, and in some cases by the same siphonozooids, showing
that a reversal of the current had taken place within the colony.

The presence of numerous, well-marked, and regularly arranged siphonozooids in this
genus as well as in the amended genus Lobophytwm, enables the effectual aération of the
superficial portions of the colony, where metabolism is greatest, and is doubtless correlated
with the absence of a special superficial canal system so well marked in Sclerophytum, in
which the siphonozooids are very minute.

Autozooids. The autozooids are large, and are very similar to those of Lobophytum. The
tentacles are much shorter and broader in proportion to their length than those of Alcyoniwm.
They are of the simple pinnate character, with a single row of pimnules down each side.

In some cases the pinnules are so small as to be easily overlooked. They are very similar
to those of Lobophytum.

The stomodaeum is usually long and in the preserved state is convoluted. The siphono-
glyph is not well-marked in any of the specimens in the collection, and can only be

1 The measurements were obtained by measuring several

siphonozooids from different parts of a colony and taking
an average.

2 S. plicatum was not in the collection from the Maldive
Islands.

THE ALCYONARIA OF THE MALDIVES. 507

distinguished in stained preparations. The mesenteries are large and well developed in all
the species which I have examined. The ventral and lateral mesenteries are provided with
very stout, rounded mesenterial filaments, which are similar to those of the genus Lobophytum,
but are larger and much more strongly marked than in the genus Sclerophytum. The
dorsal mesenterial filaments are very long, and are grooved and ciliated as in many other
genera of the Alcyonaria.

Zoochlorellae are fairly numerous in the endodermal tissues, canals and zooid cavities in
all the species. In addition, rounded cells of a yellow colour occur in glawewm and latum.

Canal System. The canal system of this genus has been described and figured by
Moseley (1881). It is similar to that of Lobophytwm, but differs from Sclerophytum in the
absence of a special superficial system. It is to a certain extent homologous with’ the
internal system of Sclerophytum. In Sarcophytum the principal longitudinal vessels are direct
prolongations of the siphonozooids. In Sclerophytwm, however, the siphonozooids terminate in
the large vessels of the special superficial canal system, and as a rule are only indirectly
connected with the longitudinal vessels of the internal canal system (p. 519).

The longitudinal vessels run more or less parallel with the longitudinal axis of the
autozooids, and send off numerous branches, which may terminate in the coelenteric cavity
of a neighbouring zooid, or ramify and anastomose in the substance of the mesogloea, and
form a complete network. Some of the branches end in solid cords and strands in the
mesogloea, in a manner described by Ashworth in Xenia (1899, p. 190). The strands and
cords are not indicated in Moseley’s figure. I found it impossible to perceive cilia on the
walls of the vessels of the canal system.

In addition to this system of canals neighbouring autozooids are connected by means
of short, straight canals which are provided with fairly long cilia (fig. 3). These vessels are
described and figured by Moseley, but are not described as being ciliated. I have observed
similar ciliated vessels in Lobophytum, in living Alcyoniwm, in which they are less numerous,
and in Sclerophytwm, in which they are very few in number.

Growth. In all the species of the genus which I have examined the autozooids are
most numerous at the margin of the capitulum (fig. 1), but in this portion of the colony
they are usually smaller and appear to be younger than elsewhere. Growth, by the budding
of new autozooids, is most vigorous in the same part. Where it is very rapid the margin
is often thrown into folds as in Schenk’s figure S. plicatum (1896, Taf. 1. fig. 12). Often
a young colony can be distinguished from an older one by the rounded, unbroken contour
of the margin, but in both specimens of tenuis, which have attained a considerable size, the
capitulum is flattened, slightly concave, and the margin unbroken by folds (fig. 9). The
siphonozooids are not very numerous at the margin, but are most common in the middle
of the capitulum. They are apparently budded off from the endodermal canals usually in
the proximity of the food-supplying autozooids. Young specimens of Sarcophytwm and Lobo-
phytum are usually mushroom-shaped, and are very similar in appearance, so that at first
sight a young Lobophytum may be easily mistaken for a Sarcophytum. After carefully
examining all the young specimens in this and in the other collections I find in every case
that in Sarcophytum autozooids are most numerous at the margin of the capitulum (fig. 1),
while in Lobophytuwm autozooids are least numerous, and siphonozooids most numerous at the
margin (fig. 7). As growth of the capitulum proceeds in both genera chiefly by the budding
of new autozooids, the growth of the colony in Sarcophytum is most vigorous at the edge

65—2

508 EDITH M. PRATT.

of the capitulum, so that the colony retains its mushroom-like form throughout life. In
Lobophytum, however, new autozooids are formed from any portion of the capitulum, and
growth at the margin where autozooids are fewest is necessarily very slow, consequently as
a colony grows it loses its primitive mushroom-shape and assumes a lobulated appearance’,

1. Sarcophytum ehrenbergi (Pl. XXVIII. figs. 1 and 2). Marenzeller, 1886 (p. 356,
Taf. 1x. figs. 3 and 4).

Four small specimens were taken on the reefs at Hulule, Male Atoll, and two larger
incomplete specimens at Goidu, Gorfurfehendu Atoll.

All the specimens agree fairly well with Marenzeller’s description. The margin of the
capitulum in young specimens is incurved, rounded and unbroken, in older colonies it is
often convoluted. The colour in spirit is light greenish-brown. The autozooids are large
and well-marked, and are more numerous round the margin, but this character is not so
strongly marked as in tenwis (fig. 9) and latum.

Autozooids. The anthocodiae of fairly expanded autozooids are about 5 mm. in length.
The surface diameter of an autozooid cavity is ‘8mm. The tentacles are short and pinnate
(fig. 2), and are about 1 mm. in length. Spicules are present in the form of minute, slender,
tuberculate spindles averaging ‘1 mm. in length. These are arranged more or less longi-
tudinally on the walls of the tentacles (fig. 2 sp.), and not en chevron as in many of the
species. The stomodaewm is very long and much convoluted. The mesenteries are large, the
ventral and lateral ones being provided with very stout and rounded mesenterial filaments,
which in the preserved condition are much convoluted at the beginning of their course.
I have been unable to find reproductive organs in the mesenteries of any of the specimens.

The siphonozooids are very well marked and extend for a considerable distance into
the interior of the colony, but they are not so numerous as in S. latum and many other
species. They are regularly arranged in rings round the autozooids, and are most numerous
in the middle of the capitulum, where from four to five may be counted in a straight
line between two autozooids. The average surface diameter of a siphonozooid is 37 mm.,
and is greater than in S. glaucum. The stomodaeum may be seen in surface view as a key-
hole shaped aperture. Its average length is "32 mm., which is quite twice that of S. glaucum.
The eight mesenteries are not so well marked as S. latum. The spicules have been fully
described and figured by Marenzeller (1886). They are very abundant near the surface of
the colony, but are relatively scarce in the coenenchym,

Zoochlorellae are present in this as in other species, but I have not been able to find
any of the rounded cells with yellow contents which are present in S. glaucum, ete.

S. ehrenbergi, then, is‘ characterised by:—(1) the form and distribution of the spicules
which are numerous near the surface, but few in the coenenchym; (2) the tough, firm texture
of the colony; (3) the large size of the autozooids, which have large, stout, rounded ventral
and lateral mesenterial filaments; and (4) the clearly defined large siphonozooids, apparently
less numerous than in other species, with a long stomodaeum fully twice the length of that
of S. glaucum.

1 A young mushroom-shaped colony of Sclerophytum may _ they can never be distinguished without the aid of a micro-
be easily distinguished from Sarcophytum or Lobophytum by scope.
the minute size or absence of siphonozooids. Where present,

THE ALCYONARIA OF THE MALDIVES. 509

2. Sarcophytum glaucum (Pl. XXVIII. figs. 3 and 4), 1833. Quoy et Gaimard (p. 270,
Pl. XXII. figs. 11 and 12), 1886. Marenzeller (p. 352, Taf. 1x. figs. 1 and 2).

A single well-preserved specimen was taken on the reef of Hulule, Male Atoll.
Mr Gardiner observed this species to have large lobes much covered with mucus, with
the polyps not generally completely contracted when living, even if exposed to low tide.

Specimens from different localities vary considerably in size. This specimen is 22 mm.
high, the capitulum is 59 mm. x45 mm. in diameter, and in the middle is 9 mm. in thickness.
The margin is thrown into four large folds. In the middle of the disc the autozooids, which
when expanded are about 2mm. across the crown, are from 8—5 mm. apart, but on the
margin are younger, smaller and only about Imm. apart. The siphonozooids are very
regularly arranged between the autozooids, and are most numerous in the middle of the
capitulum, where there may be from 6—13 siphonozooids in a straight line between two
autozooids; on the margin there are only from 1—3 between two autozooids. The colour
of the specimen in spirit is of a greenish stone colour, in the living condition the green
colour is more intense. The specimen yields readily to the touch, but is fairly tough in
texture. The stalk is short, thick and cylindrical, its surface is smooth and unwrinkled.

The species has so far been described from Australia, Red Sea, Tonga Island, Viti
Island, and now Maldive Islands.

The autozooids are large, well-marked, and of a deep cream colour in spirit. Many
of them are completely retracted. The average surface diameter of an autozooid cavity is
1—15 mm. The tentacles when expanded are long and slender with short simple pinnules
in a single row down each side. Small spicules in the form of slender spindles are im-
bedded in the walls of the tentacles, and are so arranged that when an autozooid contracts
the tentacles come together to form a cone, and the delicate fleshy parts are enclosed in
and protected by a conical cap of spicules. The average diameter through the contracted
crown of tentacles is 1-4 mm.

In the preserved condition the stomodaeum is much convoluted; the siphonoglyph is not
readily seen, and, indeed, can only be distinguished in stained sections. The eight mesenteries
are well-marked and provided with large mesenterial filaments, which are much convoluted
at the beginning of their course. Each of the two dorsal mesenterial filaments has a groove
running down the middle, lined by ciliated cells. The ventral and lateral mesenteries are
rounded in cross section as in Alcyonium. I have observed no generative organs in this
specimen, but in a specimen taken by Willey in Blanche Bay typical Alcyonarian ova
measuring from ‘05—'25 mm. in diameter were seen on the mesenteries.

The siphonozooids are rounded on their external surface, with the mouth sunk in a
slight depression in the middle. Their surface diameter, which appears to be constant, is
about ‘2 mm., and they are usually arranged with the broad diameter of the stomodaeum
in a definite direction. Their length varies according to age or to the proximity of an
autozooid. When seen in longitudinal section a siphonozooid is conical in shape, the external
surface being the base of a cone. The coelenteric cavity is prolonged at its apex into a
longitudinal endodermal canal as in the species described by Moseley (1875, Pl. XIII. fig. 2).
The stomodaeum is very short and broad in this form, the average length being ‘16 mm.,
which is only half the length in certain other species (see table, p. 506). It is T-shaped
in cross section, the siphonoglyph being well marked.

510 EDITH M. PRATT.

The mesenteries are not so pronounced as in S. latum. The two dorsal have grooved
and ciliated filaments, while the ventral have a very short, free edge, but no filaments. The
lateral appear to end with the stomodaeum, and consequently present no free edge. The
coelenteric cavities of neighbouring autozooids and siphonozooids are placed in direct com-
munication with one another by means of short ciliated canals.

In addition to the zoochlorellae which are present in considerable numbers, rounded cells
of a yellow colour (fig. 4) occur in the ectodermal as well as endodermal tissues. The cells
have a nucleus, and are filled with a granular yellow matter, which is insoluble in spirit and
apparently unaffected by weak nitric or hydrochloric acids, borax-carmine, haematoxylin, oil
of cloves or of cedar wood, but is stained black by iron-brazilien. These cells will be more
fully described and figured in a later publication.

The spicules are most numerous near the surface of the capitulum, but they are not
so densely packed as in Alcyonium. In this portion of the colony they have the form of
tiny clubs with small heads, spindles, or spiny rods. The clubs are ‘084—224 mm. in length,
and the rods and spindles about ‘5 mm. by ‘02—‘04 mm. broad. Rods and spindles are
numerous on the anthocodiae of the autozooids, and are arranged in regular oblique lines,
so that when an autozooid is contracted, the soft parts are enclosed and protected by a
conical cap of spicules (p. 509), which recalls the operculum of certain fossil corals.

The spicules of the coenenchym are slender spindles similar in form to those occurring
in the autozooids. They are from 56 mm. long and about ‘018 mm. broad. Spicules ‘07 mm.
broad are also present; they are about the same length, and covered with simple spines. The
spicules of the stalk are larger and stouter than those of the capitulum, being *56—7 mm.
long and -12 mm. broad. They are tuberculate, warted spindles, similar in form and sculpturing
to those of the new genus Sclerophytum (fig. 20), but shghtly smaller.

S. glaucum, then, is characterised by (1) its abundant mucous secretion, which gives
to the colony in its living condition a slimy appearance; (2) its green colour, which how-
ever is somewhat soluble in spirit; (3) its soft, flexible texture; (4) the large size of the
autozooids; and (5) the clearly defined siphonozooids, which differ from all other species
which I have examined in the shortness of their stomodaea, and feebly marked mesenteries.

3. Sarcophytum latum, Dana, 1849 (Pl. LVIII. figs. 6 and 7), Dana, 1859, p. 125.
Whitelegge, 1897, p. 215.

A single fragment of a colony was taken at Hulule, Male, Maldive Islands. It consists
of a portion of the capitulum measuring 68 mm. x 42 mm. in diameter and about 10 mm.
in thickness, and a portion of the stalk. As in Mr Gardiner’s specimen from Funafuti
the autozooids are usually about 1 mm. apart, and are much more numerous at the margin
of the capitulum, where the autozooids are younger, and are consequently smaller than
elsewhere.

The siphonozooids are fairly large, the average surface diameter of a siphonozooid being
‘33mm. There are usually from 3 to 4 in a straight line between two autozooids. The
siphonozooids are often arranged with the ventral edges of the stomodaea pointing in the
same direction, but here and there may be seen in alternating rows with the ventral edges
pointing in opposite directions. The stomodaeum is longer than in S. glawcum, and is ‘3 mm.
in length. The eight mesenteries are all well-marked, the ventral and lateral mesenteries

THE ALCYONARIA OF THE MALDIVES. tal

have free edges, but no filaments; the former are a trifle longer than the latter. The
dorsal mesenteries have grooved ciliated filaments. Ciliated transverse vessels between the
zooids are very numerous. The Internal Canal System is well marked. Yellow cells are
present as in S. glaucum.

The autozooids are fairly large with well-marked characters. The diameter through the
crown of a completely retracted zooid is ‘8—'9 mm., and of a fairly expanded zooid 1:2 mm.
The diameter gradually diminishes posteriorly so that some distance below the surface it is
sometimes difficult to distinguish between the cavity of an autozooid and that of a siphonozooid,
except in sections. The anthocodia of expanded autozooids average 3mm. The tentacles are
short, pinnate, and blunt at the tip, averaging only 56 mm. in length. The stomodaeum
is long, with folded walls, and the siphonoglyph is not conspicuous. The eight mesenteries
are extremely well marked, the muscle bands being clearly seen in sections. Ova are present
on the mesenteries, measuring from ‘03—5 mm. in diameter.

This species is interesting because of its lobed capitulum (Dana, fig. 7), and its warted
and zoned spicules (Whitelegge, Plate X. fig. 1). In both these respects it approaches the
genus Lobophytum.

4, Sarcophytum boettgert. Schenk, 1896 (p. 72, Taf. u. fig. 9, Taf. Iv. figs. 29 and 30), var.

A single complete specimen was taken at the Suvadiva Atoll from a depth of 31 fathoms.
It agrees in many respects with Schenk’s description of the species from the Moluccas, but
differs from his specimen in size. The stalk is longer and more slender, the capitulum has
a smaller diameter, and is much thinner than in the type species. The colour is of a warmer
brownish shade, and the autozooids are slightly larger.

The specimen is 50 mm. high. The stalk is 39 mm. long, 25 x 14mm. broad at the
base, and gradually diminishing in diameter to 9 x 7 mm.

The autozooids are about 2°5 mm. apart in the middle of the capitulum, but near the
margin they are much more closely set. Many are expanded and measure about 1 mm.
through the crown of tentacles. The siphonozooids are numerous, but so minute that they
are invisible to the naked eye.

Owing to the presence of numerous large spicules near the surface of the capitulum
and throughout the stalk the colony is fairly hard to the touch, and is only very slightly
flexible. Near the base of the stalk there is a rhizome-like outgrowth which is attached
at its distal end to a fragment of a shell. This probably gives additional support to the
colony. A similar outgrowth is to be seen in a specimen of Sclerophytum durum (sp. nov.
p- 528) from 24 fathoms, S. Nilandu. The colour in spirit is of a pale greyish-brown.

The spicules near the surface of the capitulum are very closely packed. They are of
the tuberculate warted type found throughout the genus Sclerophytum (figs. 20 and 31). They
vary in size, but the largest are about 1 mm. long by ‘1 mm. in diameter. They are similar
in form to those figured by Schenk, 1896 (fig. 28), but slightly larger. The spicules of
the coenenchym and of the autozooids are long, slender rods and spindles, some of the
latter with a few short spies. The largest of these spicules are about ‘5 mm. long by
‘04mm. broad. I have observed this form of spicule in every species of Sarcophytum which
I have examined (fig. 8). Most of the spicules of the stalk are like those near the surface
of the capitulum, but among them may be seen a few slender spiny spindles like those of
the coenenchym of the capitulum. They resemble those figured by Schenk (figs. 29 and 30).

ILD EDITH M. PRATT.

5. Sarcophytum roseum (Pl. XXIX. figs. 10 and 11), sp. nov.

A single large fragment of a female colony consisting of a portion of the capitulum
was taken at a depth of 25 fathoms to the west of South Nilandu Atoll. In spirit the
specimen is of a beautiful brownish-pink colour, which was more vivid in the living con-
dition. When taken it was associated with a red Chiton and a Lamellibranch. In the
complete state the capitulum would be apparently cup-shaped.

The specimen yields slightly to the touch, and the surface is somewhat granular. The
capitulum is apparently thickest in the middle, where it measures 11 mm. Towards the
margin it thins out, so that a radial section through the capitulum is wedge-shaped. In
all other specimens of the genus in the collection the margin of the capitulum is curved
downwards and inwards to a more or less degree. In S. latwm the fold or seam is very
shght, but in this species it is entirely absent (fig. 11). The capitulum is irregularly lobed
at the margin, and its under surface is wrinkled and granular. Through the ectodermal
tissues marginal zooids may be seen running parallel with the surface.

Autozooids. On that portion of the capitulum which is farthest from the margin the
autozooids are 7 to 10 mm. apart; near the margin they are much more numerous and are
often less than ‘5 mm. apart. Most of the autozooids are completely retracted, but here and
there one may be seen with tentacles slightly expanded. Its colour is then seen to be of
a deep rich pink showing up clearly against the pale pink background of the colony. The
tentacles are short, simple and pinnate. The stomodaeum is convoluted in the preserved
condition, and the mesenteries are well marked. Ova at various stages of development are
present on the mesenteries, some of the largest being fully *5 mm. in diameter.

The siphonozooids are clearly marked and can be easily seen without a lens; as in
other species they are most numerous where the autozooids are fewest and vice versd. Their
average surface diameter is about ‘45 mm. Their stomodaea are longer than in any other
species which I have examined, and have an average length of ‘4 mm.

The spicules are more numerous near the surface than elsewhere, in this position being
in the form of small clubs and spindles averaging 13 mm. in length and ‘06 mm. in breadth
(fig. 10, 1, 1, m1). In the coenenchym they are slender, pointed spindles with proportionately
large tuberculate warts, which have a tendency to be arranged in zones as in S. latum and
in Lobophytum. The spindles measure °2—4 mm. long by ‘06—07 mm. broad, and a very few
crosses 3 x *2 mm. are also present in the coenenchym. In shape the pointed spindles resemble
those of S. ehrenbergi, but in that species the warts are smaller and have not a zonate
arrangement. In this respect this species approaches S. Jatuwm, and the allied genus Lobo-
phytum, but the spicules of S. latwm are much broader in proportion to their length than
those of this form.

6. Sarcophytum tenuis (Pl. XXVIII. fig. 6 and Pl. XXIX. fig. 9), sp. nov.

Two specimens, one with the lower portion of the stalk missing, and another more
fragmentary, were taken well within the South Male lagoon at a depth of thirty fathoms.
The colony is in the form of a stalked dise 63x58 mm. in diameter, and 4 mm. thick,
excluding the anthocodiae of the autozooids. The portion of the stalk is 4mm. long, and
15x 5mm. in diameter. The autozooids and siphonozooids are borne only on the upper
surface and margin of the disc. The specimen is soft to the touch, but is of a tough texture,
and of a dull brown colour.

THE ALCYONARIA OF THE MALDIVES. 513

Autozooids. The autozooids are most numerous round the margin, and most are almost
completely expanded. In the middle of the disc they are from 5—10 mm. apart, but on
the margin they are crowded together in radiating lines. The anthocodia, or free distal
portion of an expanded autozooid, is 4—4°5 mm. in length with an average diameter of 13mm.
through the crown of tentacles. The tentacles are of the simple type (with a single row
of pinnules on each side), and are about 1 mm. in length, so that they appear small com-
pared with the large size of the anthocodia. The stomadaeum is long, and its walls are
much folded. As in other species of Sarcophytum the siphonoglyph is not well marked.
The mesenteries are very large, and are much convoluted in the preserved condition.

The siphonozooids are narrow and long, where their growth in length is not impeded
by the proximity of autozooids. The surface diameter averages “3mm., so that they may
be fairly easily seen with a lens. They apparently fill up all the spaces between the
autozooids, and are consequently most numerous near the middle of the disc, where there
may be as many as from 10—16 siphonozooids in a straight line between two autozooids.
Nearer the margin there may be from 4—6, but the autozooids are so crowded on the
margin that the intervening siphonozooids cannot be counted except in sections. The stomo-
daeum is short, but is slightly longer than that of S. glawcwm; its average length is 2 mm.
The mesenteries are apparently more conspicuous in this than in any other species which
I have examined.

Spicules. This species is interesting because of its small, slender spicules, which are
very few in number. It is very improbable that spicules have been dissolved out by an
acid preservative because they are most numerous in the peripheral tissues, and in decalcified
preparations the holes left by the spicules are small and few in number. As the specimens
were taken in fairly deep water the sparsity of spicules may to a certain extent be due
to their habitat.

The spicules are somewhat more numerous in the stalk than in the capitulum, but they
appear to occur even there only near the surface, where they have the form of minute
clubs 17mm. long and ‘03mm. broad. These spicules are similar to those near the surface of
the capitulum. The spicules of the coenenchym throughout the specimens are very slender,
pointed spindles, straight or slightly curved and dotted with a few minute spines (fig. 6);
they are about ‘5 mm. long by ‘03 mm. broad. Long, slender clubs, about *5 mm. in length,
with small heads also occur. A few of the slender spindles are regularly arranged en chevron
in the walls of the anthocodiae of the autozooids, while shorter spindles occur on the
tentacles.

The species differs from all others which I have examined, in its soft, tough texture,
in the sparseness and small size of its spicules, in its flattened disc-like form, and in the
large size of its autozooids, which are extremely numerous on the margin of the disc. The
siphonozooids appear to differ from those of other species in the length of their stomodaea
(2 mm.) and in their strongly marked mesenteries and filaments.

514 EDITH M. PRATT.

Ill. GENUS LOBOPHYTUM.

Lobophytum (amended genus)! (Pl. XXVIII. fig. 7, Pl. XXIX. figs. 12—14).

As in Sarcophytum and Sclerophytum a colony consists of two parts:—the capitulum
and the stalk. The capitulum is the superior portion of the colony, and bears the antho-
codiae; on its+surface are the apertures of the zooids. The capitulum is supported by the
stalk, by means of which the colony is attached to the substratum. The stalk has no zooid
apertures of any description on its surface, but is penetrated to the base of the colony by
the cavities of the primary autozooids.

As in Sarcophytum, all the species are dimorphic, autozooids and siphonozooids being
both well marked. Young colonies are often mushroom-shaped (fig. 12), the capitulum repre-
senting the pileus of the mushroom. As the colony grows older irregular excrescences appear
on apparently any portion of the surface of the capitulum. These often attain a con-
siderable size, so that the original resemblance to a mushroom is entirely lost, and the
capitulum has an irregularly-lobed appearance (fig. 7). Calcareous spicules, imbedded in the
mesogloeal and often ectodermal tissues, are to be found throughout the colony. They are
usually very small, short and stout, or long and slender, and generally studded with large
warts, arranged in zones*. Throughout the genus they are apparently similar to those
figured by Marenzeller (1886) in his original description. When clubs are present they are
few in number, and are not crowded near the surface as in the new genus Sclerophytum,
in Acrophytum and in Aleyonium.

The autozooids are very well-marked, and are similar in form and general structure
to those of Sarcophytum, but much larger than Sclerophytum (figs. 2, 18, 22). The eight
tentacles are slightly longer and more slender than in any species of Sarcophytwm which
I have examined. They are of the simple pinnate type with a single row of pinnules
down each side, the pinnules apparently less numerous in the younger than in the older
autozooids.

The stomodaewm is very well-marked, but the siphonoglyph is not easy to distinguish
in the preserved state except in stained sections. All the mesenteries have mesenterial
filaments. The two dorsal filaments are similar in form to those of Alcyoniwm, and to those
of the siphonozooids, in that they are ciliated and grooved. The ventral and lateral mesen-
terial filaments are usually rounded in cross section, but here and there may be seen a
filament which is slightly grooved down the middle. The ova are similar in size and shape
to those of other Alcyonaria. I have been unable to find male cells in any of the specimens.

The siphonozooids are very similar in form and function to those of Sarcophytum.
They have no tentacles, and only the two dorsal mesenteries bear mesenterial filaments.
These zooids have no special digestive cells, and do not bear reproductive organs. They
appear to be respiratory and excretory in function.

The canal system is well developed and is similar in general structure to that of
Sarcophytum. It differs from Sclerophytum in the absence of a special superficial canal
system, which is so well marked in that genus. As in Sarcophytwm transverse ciliated

1 A detailed account of the general and minute anatomy this genus, for similar spicules are present in Sarcophytum

of this genus will be published later. latum.
2 The zoned character of the spicules is not confined to

THE ALCYONARIA OF THE MALDIVES. 515-

channels connect the coelenteric cavities of the zooids (fig. 3), which, by enabling the circu-
lation of sea-water through every zooid, effectively aerate the internal parts of the colony.

7. Lobophytum pauciflorum. Marenzeller (Pl. XXVIII. fig. 7 and Pl. XXIX. figs. 12—14).

Two fragments of a large colony were taken in shallow water on the Hulule, Male, reef.
Two young complete colonies, which had retained to some extent their mushroom-shape
(fig. 12), were also dredged, one from 20 fathoms, Mahlos Atoll, and the other from 24
fathoms in the S.-E. passage of Kolumadulu Atoll.

This appears to be the most common as well as the most widely distributed species
of the genus. It was first taken in the Red Sea and subsequently described by Ehrenberg
(1834). It has since been recorded by many other writers: 1848, Dana, Alcyoniwm pauci-
florum; 1877, Klunzinger, Sarcophytum pauciflora; 1886, Marenzeller, Lobophytum pauciflorum.
Also var. validum; 1897, Whitelegge, var. validwm; 1900, Hickson and Hiles, Lobophytum
pauciflorum. In distribution it is now known from the Red Sea, New Zealand, Andamans,
Tonga, Moluccas, Funafuti, New Britain, and now Maldive Islands.

The spicules have been described by Klunzinger and by Marenzeller. As in all other
species of the genus they are extremely small, elliptical, or cask-shaped, usually three times
as long as broad, but some longer and some shorter. Tuberculated warts are usually arranged
in zones on their surfaces, but here and there a spicule may be seen without warts. Spicules
of the coenenchym are ‘16—24 mm. in length, and are ‘064—08 mm. broad. Those near the
surface are similar in form, but smaller, with only two rows of warts; they are ‘(08—128 mm.
long and ‘032 mm. broad.

In all the specimens most of the autozooids are contracted, and in many cases are
withdrawn to a considerable distance below the surface of the colony. Owing to the different
degrees of contraction of individual autozooids in the preserved state their dimensions vary
considerably. The average diameter of a fairly expanded autozooid through the crown is
18 mm., and through the base of the tentacles 14mm. New autozooids are budded off
from the endodermal canals near the surface at any portion of the capitulum, this giving
to the colony its lobed form. As in Sarcophytum siphonozooids are formed also by budding
from the endodermal canals, chiefly in the vicinity of the autozooids, except on the margin,
where the siphonozooids are very numerous, and the autozooids very few.

The tentacles are fairly long and conical (fig. 13, tent.), their average length when fairly
expanded being “8mm. They are of a simple pinnate character, each having down either
side a single row of pinnules, averaging in a fairly expanded tentacle 13 mm. in length.
The pinnules are less numerous in the young than in the older zooids. In a mature
autozooid the pinnules of a single tentacle are almost uniform in size, but in young auto-
zooids they are largest at the tip, and gradually become smaller towards the base of a
tentacle. The number of pinnules on the tentacle does not appear to be a constant feature,
sometimes a tentacle has more pinnules on one side than on the other, but in such cases
the autozooids may not be mature.

The siphonozooids (figs. 13 and 14, s?.) are universally scattered in more or less straight
lines between the autozooids over the whole of the upper surface of the capitulum. They
are about ‘8mm. in surface diameter, and can therefore be seen with the unassisted eye.
On the tips of the lobes there may be from two to five siphonozooids in a straight line
between two autozooids, but on the sides sometimes as many as 10 to 15. On the extreme

66—2

516 EDITH M. PRATT.

margin of the capitulum the siphonozooids are extremely numerous, a contrast to Sarco-
phytum, in which the autozooids are most numerous in this position. The stomodaeum
(fig. 14, s.) is short and straight, averaging -26 mm. in length. The siphonoglyph is well
marked.

As a rule, the length of a siphonozooid varies according to age, the young siphonozooids
are extremely short, but the older ones may extend for quite a considerable distance into
the interior of the colony. The diameter of a siphonozooid is greatest at the surface, and
gradually decreases inwardly. Some distance below the surface the cavity is sometimes so
narrow that, but for the presence of the dorsal mesentery filaments, it might be easily
mistaken for an endodermal canal. Sooner or later a siphonozooid terminates posteriorly in
the endodermal canal, from which as a bud it originated. The keyhole shaped mouth
aperture of the stomodaeum, with well-marked ciliated siphonoglyph, and the eight mesenteries
of a siphonozooid, may be easily distinguished in surface view (fig. 13).

IV. GENUS SCLEROPHYTUM.

Sclerophytum, gen. nov. (Pl. XXVIII. fig. 8, Pl. XXIX. figs. 15—18, Pls. XXX. and XXXI.).

In the Introduction to this paper (pp. 504—5) I have given the most important reasons
for the establishment of this genus, and for the inclusion in it of the following species,
which have hitherto been attributed to the genus Lobophytum: L. densum = Sclerophytum
densum; L. densum, var. = Sclerophytwm, sp. nov.1; L. confertwm = Sclerophytum confertum’;
L. marenzelleri= Sclerophytum marenzelleri’; L. tuberculosum=Sclerophytum tuberculosum'. In
addition to these six new species were collected by Mr Gardiner on the Maldive reefs.

The genus Sinularia (W. May, 1899, p. 100), which is founded entirely upon the form
of the colony and the character of the spicules, cannot be maintained. Further investigation
of its single species S. brassica will doubtless result in its inclusion in the present genus.

Brief Diagnosis of Genus. Colony lobed and sometimes similar in form to Lobophytwm.
Corallum often hard and granular (exceptions given in table, p. 531). Colonies usually
dimorphic, but the siphonozooids are very minute and exhibit unmistakable signs of degenera-
tion. Autozooids are usually smaller than in Sarcophytum, Lobophytum and Alcyoniwm, and
in some cases slightly degenerate. There are two well-marked canal systems, (i) superficial,
and (ii) internal. The superficial canals im most species large and numerous, but this system
absent or only feebly marked in Lobophytum, Sarcophytum and Alcyonium. The internal
canal system is similar in many respects to that of Sarcophytum and Lobophytwm, and
penetrates down to base of colony. In Alcyoniwm the internal canal system is not well
marked in the deeper portions of the colony.

Spicules—usually distinct from those of Sarcophytum or Lobophytuwm; they are generally
large and of varied form. Tuberculate spicules occur throughout the genus. Minute clubs
and spindles are crowded near the surface, and form a very much thinner crust than in
Alcyonium or Acrophytum.

Great diversity of form is to be found among the representatives of the genus in the
collection. The young colonies are more or less mushroom-shaped (fig. 18). Some specimens
have a cup-shaped capitulum with marginal lappets—these are usually fairly young colonies

1 These species are not present in the collection.

THE ALCYONARIA OF THE MALDIVES. SyiL7/

(fig. 30). Owing to the growth of new lobes, in older specimens the cuplike form is often
obliterated (fig. 29). Many exhibit a more or less well-marked lateral compression. Some
of the specimens have a short, broad stalk and a much lobed convex capitulum as in
Sc. capitale (fig. 8), in which the branches are closely set. In some cases the lobes are
long, slender, branched, and digitate as in Se. polydactylum. Some specimens are long-stalked
and erect with slender or stout, more or less arborescent branches as in Sc. querciforme
and Sc. palmatum (figs. 33 and 26).

The mesogloea in which the spicules are imbedded varies in density in different species.
It is most abundant in the soft fleshy forms. In species with very large spicules it is
always very dense and stains readily.

The spicules are large compared with those of allied genera. Spicules 2—4 mm. in
length, are common in many species, but in Se. querciforme they attain a length of 6 mm.
and in Se. durum are even larger, a fair number attaining 7 mm. in length by 1‘7 mm. in

breadth.

Most of the spicules have tuberculate warts on their surface (fig. 20) and vary consider-
ably in shape. The most common type is the spindle with blunt or sharply-pointed ends.
This form is found throughout, but is not confined to the genus, it occurs also in Sarco-
phytum (S. plicatum and S. boettgeri), and also in Sinularia, which, however, cannot be regarded
as a true genus (p. 516). I have not observed this form of spicule in any species of the
amended Lobophytum.

Some of the tuberculate spicules are curved, some are crescentic, kidney-shaped, or ridged
so that they are triangular in cross section, as in Sc. durum (fig. 31). In many species
they are branched and are Y or K shaped, or have several forks as in Sc. capitale (fig. 16).
They are sometimes irregularly dumb-bell shaped (fig. 32 /), but this form of spicule is rare.
The tubercles are often arranged in more or less straight lines. Usually they are very
numerous but in a few cases are sparsely scattered over the surface of a spicule. In Se.
durum and other species the tubercles are so closely set that a mosaic is formed (fig. 31 6).

In every species of the genus, which I have examined, minute clubs and spindles are
more or less crowded near the surface, but they form a much thinner crust than in Aleyonium
or Acrophytum. These clubs often vary in shape to a considerable extent in a single specimen
(fig. 32 a, b, c,d, e). In some cases, the head is simply a knotted prolongation of the handle,
but many of the clubs are bi- or tri-ramous. In all cases tuberculate warts are scattered
in varying numbers over their surface. The spindles near the surface are usually very small,
and dotted with a few simple spines (fig. 32g). Small spindles or rods generally occur in
the outer walls and tentacles of the autozooids, and form a protective, calcareous, more or
less conical cap which covers the aperture when the autozooid is partially retracted as in
Sarcophytum.

Zooids. Of the eight species represented in the collection, four are dimorphic, but in
all these the siphonozooids show unmistakable signs of degeneration. In two species the latter
has proceeded further, and only vestiges of siphonozooids are present. In the remaining two
species siphonozooids are absent.

In some of the species slight signs of degeneration are discernible in the autozooids.

The autozooids on the whole are smaller in this genus than they are in either Sarco-
phytum or Lobophytum. They are usually most numerous on the tips of the lobes.

518 EDITH M. PRATT.

Throughout the genus the tentacles are short, blunt, and stumpy, compared with other
genera. They vary slightly in size in different species (Table, p. 531) and in some cases are
extremely small. They are only slightly pinnate, and in Sc. gardinert the pinnules are ex-
tremely rudimentary. Owing to the contracted condition of many of the specimens it was
found impossible to determine the actual length and pinnate character of the tentacles of
some of the species. In Sc. capitale and Sc. palmatum, which have comparatively large
tentacles, some of the autozooids are expanded. In both of these species a double row
of very small pinnules was seen on each side of the tentacles (figs. 17 and 27); in this
respect this species approaches the genus Xena.

The stomodaeum is comparatively long and has a well-marked siphonoglyph in every
species of the genus. The mesenteries differ in size, in Sc. palmatum, Sc. hirtum, and Se.
capitale being well marked, in Sc. densum, Sc. polydactylum, Sc. durum and Se. querciforme
long and slender, but in Se. gardineri only the dorsal ones well marked. The dorsal
mesenterial filaments are long, broad, grooved and ciliated in every species, but the ventral
and lateral vary in size in different species. These filaments are often very short and feebly
developed and are never long or well marked (Table, p. 531); im Se. gardineri they are
apparently absent. The small size or entire absence of ventral or lateral mesenterial filaments
is probably correlated with the presence of numerous zoochlorellae, for they are most numerous
in Se. gardineri, which has no ventral or lateral filaments.

Reproductive cells are not present in all the specimens, and where they do occur they
consist entirely of female cells. I have not been successful in obtaining a single specimen
with male cells. The ova appear to be of the typical Alcyonarian type, but the occurrence
of ova on the dorsal mesenteries of Sc. gardineri is interesting (pp. 527—8).

Siphonozooids are present and are numerous in many but not in all species of the
genus. They are very small compared with those of Sarcophytwm and Lobophytum and show
unmistakable signs of degeneration. They are largest in Sc. capitale where they have an
average length of 3 mm. (figs. 17 and 19, sz), but they are not difficult to recognise in
stained preparations or sections of Sc. denswm, Sc. hirtum, and Sc. palmatum, in which they
vary from ‘1 mm.—17 mm. in length. In all these species they have small stomodaea which
usually open to the exterior (figs. 19, 21, 22, 25, m.ap.), but this is often difficult to see
except in stained sections. Traces of mesenteries occur in many of the siphonozooids. As
a rule a siphonozooid terminates posteriorly in a large transverse vessel of the superficial
canal system.

In Se. durum and Se. polydactylum the degeneration of the siphonozooids has proceeded
much further than in Sc. capitale, Sc. densum, Sc. hirtum, and Sc. palmatum. In these two
species the siphonozooids are represented by very short vertical caeca from the superficial,
transverse canals. Traces of a stomodaeum may often be seen in an aggregation of ectoderm
cells between the terminal portion of the caeca and the surface of the colony, but I have
not observed any apertures to the exterior, nor any traces of mesenteries. Sc. gardinert and
Sc. querciforme are similar anatomically in many respects to other species of this genus,
but are more degenerate in character, and have completely lost their siphonozooids. In Se.
gardinert the autozooids also exhibit slight signs of degeneration.

Canal-systems and formation of zooids by budding. ‘The transverse superficial
and the internal canal systems can be easily distinguished in decalcified moderately thin
free-hand sections, mounted simply in glycerine jelly or stained with borax carmine or

THE ALCYONARIA OF THE MALDIVES. 519

brazilin, but the superficial canal system is best observed in stained microtome sections about
20 w in thickness. These sections however are too thick for histological purposes. The trans-
verse superficial canal system is extremely well-marked in most species. The principal vessels
are comparatively broad channels, which run also immediately below but parallel with the
surface (fig. 22 t.s.c.); these I have named the “transverse canals.” The vessels of this
system ramify and anastomose throughout the peripheral tissues, they communicate with the
terminal portions of vessels of the internal system, and have numerous apertures into the
coelenteric cavities of neighbouring autozooids (fig. 22 ap.c.). From these vessels buds arise
towards the surface, which in Sc. gardineri and Sc. querciforme develop into autozooids. In
all the other species in the collection buds are extremely numerous and are very regularly
arranged near the capitular surface, and differ from autozooid buds (1) in that they are all
arrested at a very early stage, which appears to be constant for the species, and (2) differ
also to a certain extent in development. There can be no doubt that these buds are
homologous to the siphonozooids of Sarcophytum and Lobophytwm, from which they differ in
that their growth and development are arrested at a much earlier stage. These rudimentary
siphonozooids retain throughout life their intimate connection posteriorly with the transverse
vessels from which they originated (fig. 22). In Sc. capitale, Sc. densum, Sc. palmatum, and
Sc. hirtum development is not arrested until the rudiments of a stomodaeum with mouth
aperture and rudiments of mesenteries have been formed. In Sc. durum and Se. polydactylum
development is arrested at an earlier stage, a few ectoderm cells representing the stomodaeum
wander into the mesogloea about the terminal portion of a bud, but there is no mouth
aperture, and the bud is simply a caecum from a transverse canal, with only traces of
mesenteries. In dimorphic species the number of siphonozooids arising from a transverse canal
varies, but in Sc. capitale seven or eight siphonozooids terminating in a single vessel may
often be counted between two autozooids.

In the presence of a special superficial canal system this genus is completely separated
from Lobophytum, Sarcophytum, and Alcyonium, but in this respect it approaches Xenia.
Zoochlorellae are numerous in the endoderm of the superficial canals, and are present in
varying numbers in the lumen of the canals.

The internal canal system is strongly marked and is very similar to that of Xenia. It
differs from that of Sarcophytum and Lobophytum in that the siphonozooids do not terminate
posteriorly in the longitudinal vessels of the system, but in the transverse vessels (fig. 19).
The principal vessels run more or less parallel with the longitudinal direction of the auto-
zooids and have numerous apertures into their coelenteric cavities (fig. 22 ap.c.). They
ramify and anastomose throughout the internal and especially in the basal portions of the
colony. Many vessels terminate near the periphery in the vessels of the superficial canal
system. Zoochlorellae are only to be found in this system in those canals which lie near
the surface.

Zoochlorellae are present in all the specimens of the genus in the collection, but they
vary in number in different species. Their numbers, however, do not appear to be affected
by bathymetric variations within certain limits, for they are fairly numerous in specimens
which have been taken at a depth of from 24—34 fathoms (Table I. p. 531). They are
most plentiful in Sc. gardineri and least in Se. capitale, both of which species were taken
only in shallow water.

I have suggested (p. 528) that their presence in enormous numbers in Se. gardineri is

520 EDITH M. PRATT.

correlated with the reduced size of the tentacles and the absence of ventral and _ lateral
mesenterial filaments.

8. Sclerophytum capitale sp. nov. (Pl. XXVIII. fig. 8, Pl. XXIX. figs. 15—17).

Two specimens were obtained from the reef at Hulule, Male Atoll, from shallow water.
One specimen is complete (fig. 8), the other has a portion of the stalk and capitulum
missing. The complete specimen is attached basally to a fragment of hard coral.

The colony is fairly large, broad and laterally compressed, with a beautifully lobed
convex capitulum. The lobes are numerous, many of them are laterally compressed; some
are branched. The colony is 65 mm. high, and 70 x 35 mm. broad. The stalk is irregular
in outline at the base, but the sides are rounded unbroken, and granular to the touch.
The height of the stalk is 27 mm. and the breadth is 43x29mm. The colour of both
specimens in spirit is yellowish-grey.

Spicules. Clubs and slender spindles are crowded near the surface as in other species
of Sclerophytum. Most of the clubs have small knotted heads (fig. 15) and are about 18 mm.
in length. A few clubs occur, which are bi- or tri-ramous; they are slightly longer and
more slender than those of Sc. densum. The tuberculate spicules (fig. 16) are smaller and
less numerous than in many species of the genus. In the capitulum they are 1—2‘5 mm.
in length, but they are a little larger in the stalk, where some of them attain a length
of 3mm. These spicules vary considerably in shape as well as size (fig. 16); regularly and
irregularly curved spindles are numerous. Crescentic, Y and K-shaped spicules also occur,
as well as a few spicules irregularly forked at both ends.

The autozooids are larger in this than in many species (Table I. p. 531), an ex-
panded zooid measuring about 1 mm. across the crown. They are very regularly distributed
over the surface of the capitulum, but they are slightly more numerous on the tips of the
lobes, where growth is apparently most vigorous. In vertical sections they appear to be
very long and slender, broadest at the surface and tapering gradually posteriorly.

The tentacles are apparently longer than in any species except Sc. palmatum, when fully
expanded measuring about “45 mm. in length by *2 mm. in breadth. They are broadest at
the tip and are not bluntly conical in the preserved condition as in other species. The
tentacles like those of Sc. palmatum are interesting because they have a double row of
pinnules down each side (fig. 17 tent.), in this respect approaching the genus Xenia.

The stomodaeum is long and convoluted, and the mesenteries are all well-marked. All
have mesenterial filaments. The dorsal ones are similar to those of other species. The
ventral and lateral mesenterial filaments are short but are clearly marked. A_ partially
digested mass of food material containing a few zoochlorellae in the process of disintegration
was observed in the coelenteric cavity of one of the autozooids.

Siphonozooids are present in this species (figs. 17 and 19, sz.), but they are rudimentary.
They are, however, larger in this than in any other species of the genus which I have
examined. Many of them have an extremely small stomodaeum, which in sections is seen
to open to the exterior. They terminate posteriorly in the transverse vessels of the superficial
canal system. Zoochlorellae occur in the endoderm and the superficial canal system, but
they are not numerous in the coelentera nor in the lumen of the canals.

THE ALCYONARIA OF THE MALDIVES. 521

Canal Systems. The superficial and internal canal systems are both extremely well-
marked. Near the surface the vessels are very large (fig. 17) compared with those of other
species. They are much branched and form an intricate network. The large transverse
vessels run immediately below the siphonozooids and their lumen is continuous with their
coelenteric cavities (fig. 19). They also communicate with the vessels of the internal canal
system and directly or otherwise with the coelenteric cavities of neighbouring autozooids.
The breadth of the transverse vessels is ‘07 —3 mm.

The species is characterised by (1) the tough fleshy texture of the colony; (2) the
small size of the tuberculate spicules; (3) the siphonozooids less degenerate than in other
species; (4) the superficial and internal canal systems extremely well-marked; (5) zoochlorellae
less numerous than in any other species in the collection.

9. Sclerophytum densum (= Lobophytum densum Whitelegge, 1897, p. 219, figs. 19—22),
Pl. XXIX. fig. 18, Pl. XXX. figs. 20—22.

One complete specimen over the edge of a shoal from Addu Atoll, at a depth of
11 fathoms. Several fragmentary specimens from shallow water from the western reef of
Hulule, Male Atoll. Two fragments from 25 fathoms, Miladumadulu Atoll. One complete
young specimen (fig. 18) from 34 fathoms, Felidu Atoll (this specimen has not yet lost its
mushroom-like form although lobes have grown out from the margin).

The specimens vary considerably in shape and size, are all hard and brittle, and their
surfaces rough to the touch. The capitula of mature specimens have numerous lobes, which
vary in size in specimens from different localities. In many cases the margin of the
capitulum has no seam, so that but for the presence of zooid apertures on its surface, the
capitulum passes almost imperceptibly into the stalk.

I have compared these specimens with others from Funafuti, China Seas, and British
New Guinea and find them to agree fairly well in all essential particulars. The species
is the commonest in the collection and appears to be the most widely distributed of the
genus. Its bathymetric range from shallow to fairly deep water is very interesting. The
colour in spirit varies in individual specimens, some are dark brown, some are dark yellow,
while others are greyish-brown in colour. The spirit in which the specimens had been
preserved had a distinct greenish tinge, which may be due to the solution of green colouring
matter in the zoochlorellae, as they are very numerous in the canals and endodermal tissues
near the surface of the colony.

The spicules of this species are described and figured by Whitelegge. As they do not
differ in any important respect from those of other species, they may be regarded as being
fairly typical of the genus (fig. 20) if we except Sc. durum which apparently has no branched
tubereulate spicules. Minute clubs and spindles are crowded near the surface. The most
common tuberculate spicules are 2—3 mm. in length.

The mesogloea is of a dense tough nature, staims very readily, and is of a yellowish-
brown colour in spirit specimens.

The autozooids are very numerous, regularly arranged on the lobes, and apparently
just as abundant in the furrows between the lobes. The diameter of an autozooid cavity
(fig. 22) is greatest just below the surface and averages ‘75 mm., in the stalk being slightly
reduced.

G. IL 67

522 EDITH M. PRATT.

The tentacles (fig. 22 tent.) are very small but are slightly larger than those of Sc. durum,
being about “4mm. long and bluntly conical. They are pinnate, but I have been unable
to see more than a single row of pinnules down each side.

The stoumodaeum (fig. 22s.) is long, and somewhat convoluted and is 6—85 mm. in
length. The siphonoglyph is clearly shown in transverse sections.

The mesenteries are long and slender (fig. 22); the ventral and lateral mesenterial
filaments are short and broad and measure about ‘18 mm. in length (fig. 22 v. m. f, l. m. f).
The dorsal mesenteries are much longer and are broader than the others, have broad grooved
ciliated filaments (fig. 22) and are similar to those of other Alcyonaria.

Ova in all stages of development are present in the mesenteries of one of the frag-
mentary specimens from shallow water. They are of the typical Alcyonarian character and
are of a yellowish-brown colour. The largest are easily seen with the naked eye.

Siphonozooids. Whitelegge (1897, p. 220) says of this species, “The Siphonozooids are
minute and the orifices difficult to distinguish even with a lens.”

If a thin strip of tissue be taken from the surface of the colony and examined under
the microscope the siphonozooids are seen to occupy minute rounded areas surrounded by
a spicular tissue containing minute clubs and spindles, as in Sc. capitale (fig. 17). It is
only after decalcification and staining (borax-carmine answers very well) that the stomodaea
and their apertures can be distinguished (fig. 21). The aperture is often rounded, but it
is occasionally keyhole-shaped?. As in other dimorphic species the siphonozooids terminate
posteriorly in the transverse vessels of the superficial canal system (fig. 22). As these vessels
lie very near the surface of the colony the siphonozooids are necessarily very short, averaging
‘17 mm. long.

The stomodaeum is usually a tiny tube one cell in thickness and one or two cells in
length. Sometimes it is a solid cord of cells without lumen so that the siphonozooids
have no aperture to the exterior. Traces of mesenteries are not difficult to perceive in some
of the siphonozooids.

Canal Systems. Both superficial and internal canal systems are clearly marked (fig. 22),
and may be regarded as being typical of the genus (pp. 518-19). The vessels ramify and
anastomose in all directions, so that a complete circulation of fluids is ensured throughout
the colony. The vessels near the surface are crowded with zoochlorellae. The transverse
vessels lie very near the surface, they are not so broad as the transverse vessels of Sc. capitale,
their average diameter being ‘08 mm.

10. Sclerophytum hirtum sp. nov. (Pl. XXX. figs. 283—25).

A single well-preserved specimen from Naifaro reef, Fadifolu Atoll, having a portion of
the stalk with basal attachment missing (fig. 23), 46 mm. high and 66 x 35 mm. broad.

The stalk is short, broad, laterally compressed, and is much wrinkled vertically, its surface
is granular and it yields slightly to pressure. Its height (incomplete) is 22 mm. and breadth
45 x 25mm. The capitulum is irregularly divided into numerous and slightly flexible lobes.
The primary lobes or branches vary to a considerable extent in height, breadth, and number
of secondary lobes. The mode of branching is irregular; di- or tri-chotomy is frequent,
but here and there lateral branches are given off from a common stem, some of these branches

! These apertures should not be confused with the lobes left by spicules after decalcification.

THE ALCYONARIA OF THE MALDIVES. 523

ramify dichotomously. The primary branches are 10—22 mm. high and 10—17 x 7—8 mm.
broad. In spirit the specimen is of a dull brownish-grey colour. The specimen is firmer
in consistency and less fleshy than Sc. capitale, but not so hard and brittle as Sc. densum,
Sc. durum, Sc. querciforme, ete.

The spicules are very numerous, clubs are closely packed near the surface, and most
of the tuberculate spicules are very densely set with warts (fig. 24). The clubs have slender
handles, but vary considerably in shape and size. They may have from one to four tuber-
culate heads. They are ‘1—33 mm. in length and ‘02—12 mm. in diameter through the
head. Very slender clubs and spindles are present in the tentacles. In a partially retracted
condition of the autozooid, they form an outer protecting conical wall to the soft fleshy
tissues of the autozooid. The tuberculate spicules show great diversity in form and size
(fig. 24 a, b, c,d), they are 1:1—4 mm. in length and -18—1 mm. in breadth. Forked, curved,
and branched spicules are present similar to those of Sc. capitale (fig. 16). Irregular spindles
constricted in the middle (fig. 24d) also occur, but they are not common. Some of the
spindles have few warts with very small tubercles, so that they seem to be beset with
small spines.

The autozooids are very regularly arranged over the whole of the capitular surface,
and are 1—1‘5 mm. apart. Many of them are expanded, measuring about ‘8 mm. across the
crown. The tentacles are short, broad, and bluntly conical, there are indications of a double
row of small pinnules down each side. Spicules in the form of slender clubs and spindles
are regularly arranged on the outer walls. The stomodaeum is long and convoluted, and the
siphonoglyph is strongly marked. The mesenteries are larger and stouter structures than in
many species of the genus. The ventral and lateral mesenterial filaments are longer and
broader than in Sc. denswm and Sc. durum. The dorsal mesenterial filaments are similar
to those of other species. Reproductive cells were not observed in the specimen.

Siphonozooids (fig. 25). The siphonozooids are very rudimentary, and are so minute
that they cannot be distinguished without the aid of a microscope. Their average length
is ‘1mm. and breadth (045mm. If a thin strip of tissue be carefully removed from the
surface of the colony and stained with borax-carmine, the siphonozooids are seen to occupy
more or less hexagonal or rounded areas, and are separated from each other by a spicular
tissue containing minute clubs and slender spindles as in Sc. capitale (fig. 17). The siphono-
zooids appear to be devoid of spicules, but with high powers there may be seen a few
extremely minute ones imbedded in the ectoderm cells. The rudimentary stomodaeum with
its aperture to the exterior (fig. 25) is seen in surface view to be keyhole-shaped, in a few
cases rounded, but sometimes the aperture cannot be distinguished even in sections. Traces
of mesenteries are to be found in some of the siphonozooids.

Canal Systems. As in other species, the transverse superficial canal system is well-
marked, the average diameter of the large transverse vessels being ‘07 mm. If allowance be
made for the great amount of space taken up by the numerous tuberculate spicules the
internal canal system may also be said to be well-marked. The average diameter of the
longitudinal canals near the surface is (04mm. near the base of the colony, the canals are
broader and have an average diameter of ‘07 mm.

Zoochlorellae are numerous in that portion of the endoderm of the canals which lies
nearest to the surface of the colony. They are not numerous in the lumen of the canals.

67—2

524 EDITH M. PRATT.

The species approaches Sc. capitale and Sc. denswm in that the siphonozooids are less
degenerate than in the other species of the genus which I have examined. They are smaller
than those of Sc. capitale, but agree with that species in having indications of a double
row of pinnules on the tentacles and in that the mesenteries and mesenterial filaments are
fairly well-marked, and are longer and broader than those of Sc. denswm.

11. Sclerophytum polydactylum (= Alcyonium polydactylum Dana, = Lobularia polydacty-
lum Ehrenberg).

A single complete and excellently preserved specimen, attached basally to a piece of
hard coral, was taken at 25 fathoms, Miladumadulu Atoll. The colony has a marked lateral
compression and is 48mm. high. The capitulum is 80 x 34mm. broad. The base of the
stalk is irregular in outline, the greatest height of the stalk is 26 mm., its breadth is
58 x 31mm. The specimen is of a tough fleshy consistency yielding slightly to pressure.
The lobes are long and slender compared with other species, and are slightly flexible. The
colour in spirit is a pale brownish-grey.

The appearance of the colony is very similar to the specimen figured by Klunzinger,
1879 (p. 26, Taf. 1, fig. 6) with which it also agrees in the form and character of the
spicules. For many years the species has been included in the genus Alcyonium. A study
of its anatomy however reveals its generic identity with the newly established genus
Sclerophytwm.

The spicules are described and figured by Klunzinger, many of them are of the tuber-
culated warted type. The spindles average 2 mm. in length and ‘45 mm. in breadth. Crosses,
K and Y-shaped spicules, and a few spiny spindles occur. Clubs are closely packed near
the surface as in other species.

There are no true siphonozooids in the species.

The autozooids are very numerous and are regularly arranged on the many lobes. In
the furrows between the lobes they are less numerous. The autozooids are very small, their
cavities averaging ‘5mm. in diameter near the surface. When partially retracted, they are
conical in surface view, the tentacles bending over towards the middle. Regularly arranged
minute spicules in the form of slender spindles are imbedded in their outer walls, forming
conical shields which effectively protect the soft delicate tissues within. Young autozooids
arising from endodermal canals are numerous on the tips of the lobes. The tentacles are
pinnate but they are all more or less retracted so that it is impossible to count the rows
of pinnules.

The stomodaeum and siphonoglyph are similar to those of other species. The mesenteries
are all long and slender, but the dorsal ones are the most prominent. The ventral and
lateral mesenteries have short, feebly-marked mesenterial filaments, but the dorsal mesenterial
filaments are broad, grooved, and of considerable length. I have been unable to find male
or female cells in the specimen.

Canal Systems. The transverse superficial canal system is extremely well-marked.
Short, broad, vertical caeca are given off from the large vessels of this system towards the
surface, but do not communicate directly with the exterior. There can be no doubt that
these caeca are identical in origin and to a certain extent in structure with the degenerate
siphonozooids of Sc. capitale, densum, hirtum, and palmatum, but in this species degeneration
has proceeded so far that the siphonozooids have no stomodaeal apertures. Traces of stomodaea

THE ALCYONARIA OF THE MALDIVES. 525

may, however, in many cases be seen as aggregations of ectoderm cells about the upper
terminal portions of the caeca. The average diameter of the transverse vessels is ‘16 mm.

The internal canal system does not differ in any important respect from that of other
species. The vessels of this system gradually become larger as they approach the surface.
Many of the longitudinal vessels terminate in the superficial canals.

Zoochlorellae are present in considerable numbers in the endoderm and lumen of the
superficial canals, but they are not so numerous as in Se. gardineri.

This species is interesting because in many respects it forms an important link between
Se. capitale, denswm, hirtum, and palmatwm, which have clearly marked but degenerate
siphonozooids, and Se. gardinerit which has no siphonozooids. In Sc. polydactylum the auto-
zooids are small and appear to be slightly degenerate in the small size of the tentacles,
and in the feebly-marked ventral and lateral mesenterial filaments. There are no siphonozooids
which are functional as such, but traces of these individuals remain as the vertical caeca
from the transverse vessels of the superficial canal system. Zoochlorellae are abundant in
the endoderm and lumen of the superficial canals, and to a certain extent in the endo-
dermal tissues of the zooids, but they are not nearly so numerous as in the more degenerate
Se. gardineri.

12. Sclerophytum palmatum sp. nov. (Pl. XXX. fig. 26 and Pl. XXXI. fig. 27).

A single well-preserved specimen was taken in shallow water from Hulule, Male Atoll
(fig. 26).

The colony is erect, branched, and shows a marked lateral compression. It is 95 mm.
high and the capitulum is 35x17mm. in diameter. The stalk is 53 mm. in length and
is 19 x 85 mm. in diameter. The capitulum is divided near its base into two large primary
branches, one of which is 36mm. in length, and 10x11 mm. in diameter, the other is
33 mm. in length and 14x 11 mm. in diameter. These branches give off numerous secondary
branches which vary in length and thickness. The stalk, the capitulum, and all the fairly
large lobes exhibit the same lateral compression. The specimen is tough and fleshy in
consistency. The lobes are soft and flexible.

The colour in spirit is bluish-green, with irregular brown patches on the stalk and
about the base of the lobes. The bluish-green colour is confined to the surface of the colony
and is only to be observed between the autozooids. The colour is not affected by dilute
borax-carmine, but takes up Delafield’s haematoxylin and alum, iron brazilin, and strong
borax-carmine stains fairly readily.

The spicules are similar in form and size to those of other species of the genus (fig. 20).
Immediately below the surface there is a thin layer of clubs averaging ‘13 mm. in length
by 003 mm. in breadth, and slender spindles averaging ‘07 x ‘003mm. The spicules of the
coenenchym are not numerous and are all of the tuberculate warted type characteristic of
the genus. The spindles average 3 mm. in length and ‘4mm. in breadth; a few are irregu-
larly branched. The spicules of the stalk are similar to those of the capitulum.

The autozooids are of a pale brown colour, and are almost uniformly scattered over
the whole of the capitular surface (fig. 26). There are usually from 7—8 to 10mm. Many
of the autozooids are almost fully expanded, and are situated on small raised rounded areas,
which are not indicated in the drawing.

526 EDITH M. PRATT.

The autozooids are slightly larger than those of many species. The anthocodial portion
in the preserved state is very short, being only about “6 mm. in length, so that the crown
with its eight radiating tentacles appears to be apposed to the surface of the capitulum.,
The average diameter of a fully expanded crown is 15 mm. Below the surface the diameter
of an autozooid gradually diminishes posteriorly.

The tentacles are long compared with those of other species, and are about ‘7 mm. in
length when extended. They are blunt at the tip and are almost of uniform breadth.
When carefully examined microscopically a double row of rudimentary pinnules may be seen
down each side (fig. 27). In this respect this form resembles Sc. capitale and approaches Xenia.

The stomodaeum is long and convoluted, and the stphonoglyph is provided with long
cilia. The mesenteries are larger and stouter in this form than in any other species in the
collection. The dorsal mesenteries have large and broad, grooved mesenterial filaments, and
the ventral and lateral mesenteries have longer and broader filaments than those of other
species. The coelenteric cavities of neighbouring autozooids are often connected by means
of short transverse vessels as in Sarcophytum (fig. 3). I have been unable to find male
or female cells in the specimen of this species.

Siphonozooids are present but are clearly degenerate in this species, being represented
by minute caeca from the large transverse vessels of the superficial canal system. They are
regularly distributed in the peripheral portions of the capitulum between the autozooids. They
apparently have no aperture to the exterior and therefore cannot be functional as true
siphonozooids. In decalcified staimed slips of tissue from the surface of the colony the caeca
may be clearly seen in surface view through the ectodermal tissues, and average ‘07 mm. in
diameter. There is little doubt that the caeca are really very degenerate siphonozooids, for
they are identical in origin with the less degenerate siphonozooids of Sc. capitale, Sc. densum,
and Se. hirtum, in which species the siphonozooids possess a small stomodaeum and traces of
mesenteries.

Canal Systems. The transverse vessels of the superficial canal system are very pro-
nounced but they are not quite so broad as in Sc. capitale, averaging ‘1mm. and lying
about 15 mm. below the surface. The vessels of the internal system are fairly numerous,
and are very broad near the surface, where as in other species they usually terminate in
the transverse vessels of the superficial canal system.

Zoochlorellae are very numerous in that portion of the endoderm of the transverse
vessels which les near the surface. It is so distended by their presence as to almost
obliterate the lumen of the canals. They are not numerous in the vessels of the internal
canal system, except where they approach the surface. They are also present in considerable
numbers in the endoderm of the autozooids, but they are not so numerous in any part
of the colony as they are in Se. gardineri.

The mesogloea is diffuse and does not stain very readily with borax-carmine or haema-
toxylin owing to the sparcity of spicules in the lobes, it is more abundant there than in
many other species.

The interesting features of this species are (1) the colony is tough and fleshy in texture ;
(2) the spicules are not very numerous, and are small compared with those of many other
species; (8) the autozooids are fairly large. The tentacles are longer than those of any
other species in the collection, and have a double row of minute pinnules down each side.

THE ALCYONARIA OF THE MALDIVES. 527

The mesenteries and mesenterial filaments are more pronounced in this than in any other
species; (4) the siphonozooids are extremely degenerate and have no apertures to the ex-
terior. They are represented by vertical caeca from the transverse vessels of the superficial
canal system; (5) the transverse and internal canal systems are well-marked; (6) zoochlorellae
are fairly numerous.

‘This species resembles Sc. capitale in certain respects: the tough fleshy texture of the
colony; the fairly large autozooids; and the presence of a double row of rudimentary pinnules
on the tentacles. It differs from that species, however, in that the mesenteries and mesenterial
filaments are more strongly marked, the siphonozooids are more degenerate, and the canal
system is not so clearly defined.

13. Sclerophytum gardineri sp. nov. (Pl. XXXI. fig. 28).

A single complete and well-preserved specimen was taken in shallow water from Hulule,
Male Atoll.

The colony is 40 mm. high and is 48 x 26 mm. broad, irregularly cup-shaped and laterally
compressed (fig. 28). The stalk is 20 mm. high and is 39 x 20 mm. broad, its surface is
somewhat wrinkled and is hard, granular and unyielding to slight pressure. The capitulum
is irregularly cup-shaped with irregular long and slender lobes, arising from the middle and
edges of the cup. The cup is 40 x31 mm. in diameter and in the middle is 20 mm. deep;
there are two clefts in the sides of the cup. The lobes are brittle and are often branched.
The largest lobe (/.) arises slightly obliquely from the middle of the cup and is 28 mm. high
and 9 x 65 mm. in diameter near the base. It gives off numerous digitate branches which
vary in length from 2 to 7 mm., but having an almost uniform breadth of 3mm. Owing
to the presence of numerous spicules the surface of the capitulum like that of the stalk
is hard and granular. In spirit the specimen is of a dark cream colour.

Spicules. The hard and brittle nature of the coral is due to the presence of in-
numerable spicules. Clubs and slender spindles, similar to those of other species and measuring
-1—2 mm. in length and 05—06 mm. in breadth, are crowded near the surface. The
tuberculate spicules are similar to those of other species, the majority are spindles about
3 mm. long and ‘8—1l mm. broad. A few are slightly forked as in Se. densum and Se. hirtum.
The outer walls of the anthocodiae are studded with minute spindles as in many other species.

Autozooids only are present and the species may be regarded as having lost its
siphonozooids. The autozooids are very small, and as many of them are almost completely
retracted only a few can be distinguished on the surface of the capitulum (fig. 28). A fairly
expanded autozooid measures “4—‘5 mm. across the crown. The tentacles are very minute,
and when extended average only “3mm. in length. The pinnules are represented by very
slight indentations of the lateral walls and are not easy to distinguish. The stomodaeum
is fairly long, with much convoluted walls; its siphonoglyph is large and is provided with
long cilia.

Mesenteries. The dorsal mesenteries are clearly marked and are very long; they are
provided with long, broad grooved and ciliated mesenterial filaments, which are much con-
voluted at the beginning of their course. Ventral and lateral mesenteries are present but
below the stomodaeum diminish rapidly in size, and consequently have a very short free
edge with no filaments. Ova in all stages of development are present, and are frequently
to be found on the dorsal mesenteries. In this respect this species is interesting, for ova

528 EDITH M. PRATT.

are usually only found on the ventral and lateral mesenteries, and I believe have not been
observed hitherto on the dorsal mesenteries in any species of the Alcyonaria.

Zoochlorellae are very numerous in the endoderm of the tentacles, and are present, but
less abundant, in the endoderm of the mesenteries and in the coelenteric cavity of the zooids.

Canal Systems. The vessels of the superficial canal system are broad and _ fairly
numerous. The transverse vessels le in the mesogloea about ‘07 mm. below the surface and
are about ‘1 mm. in diameter. Both endoderm and lumen of these canals are densely crowded
with zoochlorellae, which are also present but are not so numerous in the vessels near the
surface of the internal system, which is similar to that of other species.

A study of the anatomy of this species reveals many interesting features, which may
be briefly summarised. The autozooids are very small and are less numerous than in many
other species. In certain respects they show unmistakable signs of degeneration:—(1) the
tentacles are very short, blunt, inflexible, and are only very slightly pinnate; (2) the ventral
and lateral mesenteries are very feebly marked, and (3) have no mesenterial filaments. On
the other hand the siphonoglyph of the stomodaeum, the dorsal mesenteries and their filaments
are all strongly marked. The occurrence of ova on the dorsal mesenteries is also interesting,
but is probably an aberrant feature, as they have not to my knowledge been previously
observed on the dorsal mesenteries in any species of the Aleyonaria.

The degenerate character of the food-capturing and digestive portions of the zooids is
doubtless correlated with the presence of vast numbers of zoochlorellae in the endoderm and
lumen of the superficial canals and in the endoderm of the zooids, for in no other species
are the zoochlorellae so numerous nor the degeneration of the zooids so marked. It is
possible that the absence of siphonozooids may also be correlated with the strongly marked
character of the ciliated portions of the autozooids. In the absence of ventral and _ lateral
mesenterial filaments, and of siphonozooids, and the presence of a well-marked superficial
canal system this species resembles the genus Xenia.

14. Sclerophytum durum sp. nov. (Pl. XXXI. figs, 29—82).

Four specimens were taken. Three of them have a cup-shaped capitulum with marginal
lappets; one of these has a rhizome-like outgrowth from the stalk which serves for additional
fixation (fig. 8307h). The fourth specimen (fig. 29) is the largest and probably the oldest,
for ova are present at an advanced stage of development. This specimen is not cup-shaped,
but is proportionately shorter and stouter than the others and has a much lobed capitulum,
so that the specific identity of the four specimens was not easy to recognise. Careful and
minute examination, however, shows the difference to be one of facies only, for the anatomy
of all the specimens proved to be identical. All the specimens are extremely hard and
brittle. Enormous spicules (fig. 31 sp) can be seen lying parallel to the surface through the
ectodermal tissues.

One cup-shaped specimen creamy-white in colour (in spirit) was taken near Nilandu at
a depth of 24 fathoms. Two cup-shaped specimens, one of which has a double cup, dirty
yellow in colour (in spirit), were taken on the Mahlos Atoll in shallow water. The lobed
specimen of a pale brownish-grey colour in spirit was taken on Addu Atoll, also in shallow
water. The specimens vary from 30 to 50 mm. in height, their capitula from 27 to 45 mm,
in diameter. Two of the specimens have slender, and two have broad stalks, the stalk of
the lobed specimen being proportionately shorter and broader than that of the rest.

THE ALCYONARIA OF THE MALDIVES. 529

Spicules. The species is characterised by the enormous spicules which occur through-
out the colony and can be seen with ease through the ectodermal tissues (figs. 29 and 30 sp).
The largest are all spindles of the tuberculate type (fig. 31). These spicules have an im-
portant bearing upon the form of the colony in that they materially affect its outline. Small
clubs are numerous near the surface as in other species. The tuberculate spindles vary in
size from 15 x ‘5 mm. to 7 x 17 mm. (fig. 31). They are often somewhat irregular in form;
they may be sharply pointed or blunt at each end, many are ridged and are three-sided,
some are crescentic, many are curved. All the large spicules are unbranched, in this respect
the species differing from any other in the collection.

When a spicule is seen with the naked eye the general outline appears smooth and
unbroken, when examined with a lens the surface appears granular (fig. 31), but when
subjected to higher magnifications the granules are seen to be tiny tuberculate warts regu-
larly arranged (fig. 31 6, and c), averaging ‘09 mm. across. A few spindles occur here and there
with a few tubercles.

Packed closely near the surface are the small clubs and a few minute spindles with
few tubercles, and a very few irregular tuberculate dumb-bells about -1 mm. in length and
‘08 mm. in breadth. The clubs are knotted and of varied shape, -12—24 mm. long and
about ‘02 mm. broad (fig. 32a, b,c,d,e). They are apparently the only branched spicules
in the species.

The autozooids, most of which are completely retracted in all the specimens, are more
numerous on the lobed (fig. 29) than the cup-shaped specimens, but are proportionately fewer in
this than in any other species of the genus which I have examined. This may, to a great
extent, be due to the presence of numerous large spicules which run immediately below but
parallel with the surface. The autozooids are very minute, and in the retracted state their
apertures are difficult to distinguish even with a lens. In the cup-shaped specimens the
polyps appear to be most numerous on the margin of the capitulum. The diameter of a
polyp cavity when the polyp is completely retracted is -°25—5 mm. The tentacles are ex-
tremely short and stumpy, and only very slightly pinnate. They possess spicules in the
form of extremely small spindles. The stomodaewm is long, measuring about °85 mm. It is
slightly convoluted in the preserved state; the siphonoglyph is fairly well-marked. The 8
mesenteries are very slender, but are clearly marked. All of them have mesenterial filaments.

The two dorsal mesenterial filaments are of considerable length and are grooved and
ciliated as in other Aleyonaria. The ventral and lateral mesenterial filaments are extremely
short and are confined to that portion of the autozooid immediately below the stomodaeum.
They are similar in form and anatomy to those of other Aleyonaria. Typical Alcyonarian
ova, ‘3—5 mm. in diameter, are present in the mesenteries.

The siphonozooids are extremely degenerate in this species, and are represented by
small caeca from the superficial transverse canals.

Canal Systems. The superficial canal system is strongly marked. The large transverse
vessels run in a sinuous course immediately below the ectodermal tissues. Many of the
canals are closely apposed on their internal side to the enormous spicules which in this
species le so near the surface. These canals have very short vertical caeca towards the
surface but they are shorter than Sc. polydactylum and are homologous to the degenerate
siphonozooids in other species. The superficial canal system has numerous communications
with the internal canal system, which is not so strongly marked as in other species.

feo aa 68

530 EDITH M. PRATT.

Zoochlorellae are present in that portion of the endoderm of the transverse canals which
hes nearest to the surface, but they are not numerous, few are to be found in the lumen
of the canals.

Owing to the great amount of space occupied by the spicules the supply of mesogloea
is less abundant than in other species, but it is of a very dense nature and stains very readily.

15. Sclerophytum querciforme sp. nov. (Pl. XXXI. fig. 33).

Two complete colonies were taken in shallow water on Hulule, Male Atoll. Both
specimens are arborescent and are very similar in their manner of growth. Each is com-
paratively tall and slender, with a much branched capitulum, many of the branches having
long and slender lobes. From near the base of the larger specimen a second stalk grows
out bearimg a small branched capitulum (fig. 33).

The larger colony is 46mm. high. The stalk is cylindrical above the junction of the
secondary stalk and elliptical in cross section below. At the top the stalk is branched and
passes almost imperceptibly into the capitulum, which is 20 mm. high and 25 x 18 mm. in
diameter, being slightly compressed laterally. The branches vary very considerably in size,
but the two principal branches are 10 and 125mm. long and about 4mm. broad. The
lobes are very slightly flexible. The surface of the capitulum and stalk is granular and
fairly hard to the touch, but yields to slight pressure. The specimens in spirit were of a
pale greenish-grey which gradually changed to a deep cream colour.

The spicules are very similar in form and structure to those of other species. Clubs
are very numerous near the surface, °2—26 mm. long by about ‘008 mm. in greatest breadth.
The tuberculate spicules vary in size, the largest being fully 6 mm. long by 1 mm, broad.
A few forked spicules occur but are not numerous.

There are no stphonozooids in this species.

The autozooids are fairly numerous and are all of one kind. Most of them are com-
pletely retracted. An autozooid cavity is "4—5 mm. in diameter just below the surface.
The tentacles are very short and pinnate. The pinnules are proportionately longer in this
than in any other species in the collection.

The stomodaeum is long; the siphonoglyph is well-marked and is provided with very
long cilia. The mesenteries are all strongly marked and are all provided with mesenterial
filaments; the dorsal mesenterial filaments are long, grooved and ciliated as in other species.
The ventral and lateral mesenterial filaments are not so long nor so broad as the dorsal
filaments; they are more or less rounded in cross section and are similar in form and
structure to those of Sc. palmatum, capitale, hirtum, and of many other Aleyonaria, but both
mesenteries and filaments are longer and broader than in many species of the genus.

Canal Systems. The vessels of the superficial system are not so large nor so numerous
as in other species in the collection, The transverse vessels lie about ‘25 mm. below the
surface and have no vertical caeca, but follow a course almost parallel with the surface.
As in other species, these vessels communicate with the coelenteric cavities of the autozooids
and with many vessels of the internal system, which is very similar to that of other species,
but less strongly marked.

Zoochlorellae are fairly numerous in the endoderm and lumen of the canals and in the
endodermal tissues of the autozooids,

531

THE ALCYONARIA OF THE MALDIVES.

AOT[BYS

SULOUYVI FG OF
IOWA MOT[BYS ULOTF

S[BUBO JO TATANT
pus muepopue
Ur snore ApIVey

ueuIN] Ut Moy
‘snoraumu 419A you

seloeds 1aq40 UI SB
PayAUL [JA os you

paxpreur [aa Apres

setoads 1ayj0 UI SB
pexAVUT [faa Os you

pRoaiq ‘uLUT GQ.
ODBIIAB STASSOA
‘paxytBu [AAs

quasqe

S[BRUBO ASTOASUBIY
WIOA} BOARD 4.LOYS
Araa Xq poyuasoada.t
sproozouoy dis

pearasqo you

yuasaid Bao

SULOYYRT FE OF Sp10d0a
IdJBM MOTTBYS TOYVM AMOTLVIS Sule I0JBA AMOT[BYS BLAS TOJBM MOT[B
4 Tet i Te UVF SG + Tey 19}VM MOT[VYS UOT 9 Teds oujaurdyyeg,
91B.1941[qO
es acaos areca s[BuRo JO MeUINT s[Buvo Jo uatINy aoeyans
Rida te UI puv ULapopua snorauina ApATey Ur Me} 49Nq UEP | eon snoraunt SnOIeUINT Jou aRl[aqo[yoo7
f I Ul snosetmnu ATdA -opua Ul snotemnu |” 3
-lauanu Ajauia.1}xo 2 ;
aoRjINS wou SvOra mig aoRJINS TIM asIvy
payareur [Jaa poyreur [TAA payjrwur [paar IOPBOIG S[ASSAA s[9ss0a paxvaut trom AIOA PUB SHOIEUINT [Buteyay
*paxUBUd [TOA | per tl S[OSSOA payIVUL [[AA.
Bor iq * -—L0-
pvorq “wut T. pvoaq ‘uu T. pBorq “wt QT. pvoaq ‘wu 10. Rene EOrd (Oleh uetees ep /y
3 i WO 80. a5 BIOAB snoreunu ATdA |
ODBIVAB STOSSAA ADBIDAB S[ASSAA ASBIBAB STOSSAA OSGIOAB SSNOTSUENU |) ee Tee s[ossaa ‘payxeu Tetoyaedng
‘pexyaBur ‘payaBul [TAAL ‘poyreur [jaa Area | Spassaa ‘payavur [aA Re i i
pex Tres Pr II pox, Il T ser Ul S[essea ‘payreuryfaa| [[aa\ ATauteayxa
aa = sasvo eee
aged aUlOS UT SaDBAy BI} S0BIY Salle} Masayy
pedeys pepuno.r
. sainjiede om toe ejoyAoy soanqaode | AT[ensn sarngaode Axequoutpnr UinaBpouloyg
‘ATey UO TAI put ‘Arequeurpnd AIA
S[BUBO OSIOA qgsuey
qyysue] ut -SUBIY UIOIZ BOBTO yysu0e] (suey ur Ul “UIUL ¢. OFBIDAB
quasqs ; ie 5 3 0 ; 4 aZIg
UUM LT. aBIGAB Aq paqyuesaidar Ul ‘WUT. 9dBIDAB CUI JT. A5BIOAB satoads raq40 :
sproozouoydis AUB Ul UBYY 0FIv]
sana] Uuasaul SUBSIO
° ; asqo AAIAsqoO jou S95BqS [[B UL BA ara
1vsiop Wo .4n990 VAD paartasqo you paadesqo you paadasqo 4 48 [[B UL BAO pearasqo you eayjonpoidoy

prorq peyreur prorq pedojasap soioeds Auvut : ; squemery erteg | | |
pur suo, Aparey Ayqaey ‘yaoys Aros eee pure suoy ApArey ATqoaoy “4104s UL URY} 1o5U0| ES ous -TOSOUT [B10 A,
ayte OL sotoads ayo , BOL
poyreut [TAA Jepues[s Ayxe ae es op A{u0 ur ueyy ne Japueys ‘suo, eee ae 3u0] Japuals puw Su0y payer [faa sala} Uaso py
ue Arveyuomripusr £19A AOI ayquop acd AOI a[SUIS AOI a[qnop solu
oi or “wut ¢. “uu }. “WU F- “MUL GF. yysuary
pazoeajar payovayat oi Br payor.yar [eordoo Aunyq yBormoo ATyUNTq dy 48 ysaprorq $a)90]UAT,
aoBJANS MOT
a on 7 fs boo snf proozojne
“MUI G.—F. “UU G.— cz. wu G. wut G). 4 parents
jo rojomerg
mae vas % “8 F ee A - uAOIO popurd
UIUI G.—F. UU G. TT, UI g. Ulu T -xa Jo 10,0081
satoads ray ayoyidpa
T[vus yrems AoA T1ems Ata Ul Wey 1aSIey yyems A10a ueryy deypemts asiey Ayres aZzig
quasqe ayBiauasap quasqe ayeiauasap
sploozouoydrs AI9A sproozouoydis sploozouoydts Suge eer ga AIOA Sproozouoydis MBO Pusey recent EE
4duey Ut “UU )— sue] que uhh
Yysuey ur wot g—T Weg a RN I Raat eauaeAaulll it re es Buoy “wu Z ynoqe Buoy “Wut F—T DUO, “WIM F—F Ul "WU G.Z—T. sonords
‘esle] snorewnU A Serey £ inl g £ pe ae ElOPEICRS ‘Tyeus snosouinu ‘sno1oumnu ‘snodemmnu A138A ‘TyBuIs pus CMC MORGILN
snoreumnu A104 snosamantm AIOA sno.ount you snozatanu 410 30u
eovyins
sno1ewnd snosomnt pepAoso ATasuap SnoIeTINU Ssay sno19emn snoJeumnu snoretint sno.euinU Iau soppurds
pas sqnip
Aysey aT4IIq pue F , ; Aysoy pue Auojoo
Aysis seqoy psy prey Ajeutesyx9 eB eaEaaH Aube SHON} Aqseg ‘y8n0} OE EE) SUERSAE MET) OI qos AToatyBaeduroo | jo arnyxay,
auwofravand uninp qaurpavb ungougod unhyovophjod UNL wnsuap appyidva sotoadg

‘NOILOATION AHL NI WALJAHdAOUATOY AO SHIOMdG AHL AO SUSONDVIC, AHL YNIAIN ‘T WIAVY,

sumajsX{g [eae

sproozouoydig

2

68

sproozojny

sornordg

532 EDITH M. PRATT.

The mesogloea is fairly dense owing to the presence of numerous large spicules, it is
less abundant than in the soft, fleshy species.

The interesting features of the species are: (1) the erect arborescent form of the colony
(fig. 33); (2) the presence of numerous very large tuberculate spicules; (8) the absence of
siphonozooids; (4) the well-marked character of the zooids, the mesenteries and their filaments
large compared with many other species; (5) the comparatively feeble development of the
superficial and internal canal systems; (6) zoochlorellae present but not in considerable
numbers.

V. THE RELATIONS OF THE ABOVE GENERA.

Sclerophytum differs from Alcyoniwm in (1) the form and size of the spicules; (2) the
small size of the zooids; (3) the occurrence of dimorphism in many species; (4) the presence
of a special superficial canal system; (5) the small size or complete absence of ventral and
lateral mesenterial filaments. These important points of difference indicate a very distant
relationship between Sclerophytwm and Aleyonium.

A study of Table II. shows that while specimens of Sclerophytum and Lobophytum may
be similar in outward form, yet there are many striking points of difference in their internal
anatomy. The general anatomy of Lobophytwm, however, is very similar in many respects
to that of Sarcophytum, with which it appears to be closely allied.

Some of the important anatomical points of difference between Sclerophytum and the
two genera Sarcophytum and Lobophytum are: (1) the presence in Sclerophytum of more
than one row of pinnules on the tentacles; (2) the ventral and lateral mesenterial filaments
are always very feebly marked (but are absent in Sc. gardineri); (3) the presence of a special
superficial canal system; (4) the absence of dimorphism in some of the species. In all these
respects Sclerophytum approaches the genus Xenia. The following diagram suggests the
probable relationship between the four genera and Alcyoniwm.

Common Ancestor

Sarcophytum Lobophytum Sclerophytum Xenia Aleyonium

Mr Gardiner has given me the following note of the distribution of the genera Sarco-
phytwm, Lobophytum, and Selerophytwm in the Maldives:—“They are found on nearly every
reef in the Maldives but are most common on the reefs which fringe the passages into
atolls. Under no circumstances are they ever found covering great areas, as Saville Kent
says is the case with Sarcophytum on the Great Barrier Reef of Australia. They do not
generally occur on shoals in the lagoonlets of secondary atolls (velu of faro), nor are they
found on any reef more than 100—150 yds. from its outer edge. In the more enclosed
atolls such as Goifurfehendu, Addu, Kolumadulu, and Haddumati, these genera are scarce,
only small colonies being found on the reefs arismg within their lagoons.

“The genera are common on the reefs of Hulule faro, Male Atoll. They live very
abundantly on the outside part of the west reef and form large spreading masses. On the

THE ALCYONARIA OF THE MALDIVES.

533

TABLE II. CoMPaARISON OF THE GENERAL ANATOMY OF THE THREE GENERA.

Genus

Sarcophytum

Lobophytum

Sclerophytum

Form of Colony

usually mushroom-shaped
throughout life (fig. 1)

young colonies may be mush-
room-shaped (fig. 12), older
forms have a much lobed
capitulum (fig. 7)

as in Lobophytum
(figs. 18 and 23)

Spicules clubs, rods, and spiny spindles | small zoned spindles, few clubs | usually large tuberculate spicules
usually not zoned of varied form; may not be
zoned, clubs usually numerous
near surface
Zooids dimorphic dimorphic often dimorphic

Growth of colony
by formation of
new autozooids

most vigorous on the margin,
hence mushroom-shape

may take place from any part
of capitular surface, gives
lobed form

as in Lobophytum

DIstRIBUTION :
(1) of autozooids

(2) of siphono-

most numerous on margin,
fewest in middle of
capitulum (fig. 9)

least numerous on margin,

more or less regularly on
surface of capitulum, least
numerous on margin

usually most numerous on

as in Lobophytum

imperceptible to naked eye

zooids elsewhere regularly margin, elsewhere regularly
arranged (fig. 9) arranged (fig. 7)
Avrozoorps : fairly large, pinnate, only similar to Sarcophytum very small, in some cases
one row of pinnules slightly degenerate
Tentacles A e ay short, stumpy, in some cases a
double row of very small pinnules
Stomodaeum long, convoluted 0 PA comparatively long
Siphonoglyph not well marked i i very well marked
Mesenteries all well marked all well marked in some cases ventral and

Dorsal mesenterial
filaments

Ventral and lateral

well marked

fairly large

as in Sarcophytum, differ

lateral feebly marked

well marked

very short and feebly marked

mesenterial slightly in size or absent
filaments
SrpHONOzooIDs : comparatively large as in Sarcophytum, differ very small and rudimentary
slightly in size or absent
Canal Systems one one two

internal well marked

ciliated channels between auto-
zooids numerous (fig. 3)

as in Sarcophytum

ciliated channels as in Sarco-
phytum but not so numerous

internal, well marked, and a
Special Superficial Canal System
intimately connected with
rudimentary siphonozooids

few ciliated channels between
autozooids

Zoochlorellae

fairly numerous in endoderm of
siphonozooids and autozooids

not numerous

extremely numerous in super-
ficial canals and endoderm
where there are no ventral and
lateral mesenterial filaments,
less numerous where they
are present

534 EDITH M. PRATT.

boulder zone W. and S.W. small clumps occur, but it is obvious that the best position of
growth is a little further out towards the atoll lagoon, where even at the lowest tide there
is 3—6 feet of water. Everywhere colonies are scarce within the boulder zone and are
never found on the sand. A brilliant green species (Sarcophytuwm glaucum) alone inhabits
the east and south sides of the faro; it ranges from the more external parts of the sand-
flat even to the extreme edge of the reef-flat where the waves break.

“Exposed to the air and sun colonies will continue to live so long as portions of them
are bathed by the water, even for 3—4 hours. Torn off their support they may be kept
alive for an hour or so in a bag or any receptacle which will prevent evaporation. In a
mixed bucket of animals from any reef (enamelled bucket) the colonies will rarely live beyond
an hour or two, but separate colonies can be kept alive for several days. Anaesthetisation
with cocaine is useless, the chemical encouraging disintegration of the tissues; alcohol some-
times yields good results, as also does chloral hydrate.”

VI. GENUS ALCYONIUM.

Wright and Studer distinguish between Alcyoniwm and Lobularia on the ground that
Alcyonium is restricted to temperate seas, while Lobularia occurs only in tropical waters.
This distinction cannot be maintained, for in the collection is a single but excellently pre-
served specimen which appears to be identical with Klunzinger’s Alcyoniwm pachyclados
(1879, p. 24, Taf. 1, fig. 5) from the Red Sea. Hickson (1900, p. 72) records from the
Cape of Good Hope the occurrence of a form similar in so many respects that he refers
it to this species. I have been able to compare these specimens with the one from the
Maldives and with another tropical form taken by Willey in New Britain. All the specimens
agree in the character and form of the spicules and apparently in their general and minute
anatomy, but differ in form and colour, so that it is very probable that they are different
facies of the same species. The generic identity of these forms from temperate and tropical
waters is indisputable, so that <Alcyonium can no longer be regarded as an_ extra-tropical
genus. The vague definition of Lobularia has led to much confusion among systematists,
some writers including it with Alcyoniwm and not regarding it as a true genus.

Alcyonium is well known to be a monomorphic genus. Whitelegge (1897, p. 220) de-
seribes a lobed form with well-marked dimorphism from Funafuti which he doubtfully refers
to Lobularia as L. ? viridis (Quoy et Gaim.) His description of this form, and his drawings
of the spicules indicate it to be a species of the genus Lobophytwm which I have attempted
to revise in the present paper (p. 514). It is probable that other species of Lobularia which
have not been attributed to Aleyoniwm will eventually be included in the new genus Sclero-
phytum (p. 516). The vague, confusing, and useless diagnosis of Lobularia makes it imperative
that it should be deleted as a genus from the family of the Alcyonidae.

16. Alcyoniwm pachyclados Kz.
A single well-preserved specimen agreeing with Klunzinger’s description (1879, p. 24,
Taf. 1, fig. 5) was taken from the reef of Naifaro, Fadifolu Atoll.

The colony from which the stalk is missing grows irregularly, and the bluntly digitate
lobes with their short, blunt, closely set branches vary considerably in length. This specimen

THE ALCYONARIA OF THE MALDIVES. 535

differs from the Cape specimens in form and colour. The Cape specimens are all more
arborescent with longer and more slender branches.

The Maldive Island specimen in spirit is of a pale brownish-grey colour, paler on the
tops of the lobes than in the furrows between. The specimens from the Cape, three in
number, are all different in colour, one white, one red, and the third yellow. This difference
is due to the difference in colour of the spicules. The form and character of the spicules
in all cases is identical. The elliptical and opaque spicules described by Klunzinger were
present in the Maldive Island specimen, but I have not been able to find them in the Cape
specimens.

The general and minute anatomy of all the specimens is very similar to that of
Alcyonium. Ova are very numerous on the mesenteries, the largest being ‘45 mm. in diameter.
The nutritive canal system differs from that of Sclerophytwm in the absence of a special
transverse superficial canal system.

The research in connection with this paper has been carried out in the Zoological
Laboratories of the Owens College, under the supervision of Professor Hickson, to whom I
am greatly indebted for much valuable assistance and advice.

My thanks are also due to Professor Herdman in allowing me to examine his valuable
collection of Alcyonidae from Ceylon.

VII. LITERATURE.

1833. Quoy ET GatmaRD. Voyage de l Astrolabe, Tome Iv. (Zoologie), Paris.
1834. EHRENBERG, C. G. Die Corallthiere des Rothen Meeres.
1848. Dana, J. D. “Zoophytes,’ Unit. St. Hapl. Exped.

1877. Kuunzincer, C. B. Die Korallthiere des Rothen Meeres; erster Theil, “Die
Aleyonarien und Malacodermen.” Berlin, 1877.

1881. Mosetey, H. N. Challenger Reports. Zoology. Vol. u. Corals (“ Heliopora and
Sarcophyton”).

1883. Hickson, 8. J. “On the Ciliated Groove (Siphonoglyph) in the Stomodaeum of
the Alcyonarians.” Phil. Trans. Part M1.

1884. Witson, E. B. “The Mesenterial Filaments of the Alcyonaria.” Mitt. Zool. Stat.
Neap. V.

1886. MARENZELLER, C. von. “Ueber die Sarcophytum bennanten Alcyoniiden.” Zool.
Jahrb. 1., Jena, 1886.

1889. Wricut, C. P., anp Stuper, T. H. Challenger Reports. Zoology. Vol. XXXt.
Alcyonaria.

1889. Famintzin, A. “Beitrag zur Symbiose von Algen und Thieren.” Mém. de l’Acad.
Imp. des Sci. St Pétersbourg, Ser. vu.

536

EDITH M. PRATT.

1895. Hickson, S. J. “The Anatomy of Alcyoniwm digitatum.” Quart. Journ. Micro.
Scr. Vol. xxxvir. Pt. 4, 1895.
1896. ScHENK, A. “Clavulariiden, Xeniiden, und Alcyoniiden von Ternate.” Frankfurt-a-M.
1896.
1897. WHITELEGGE, T. “ Aleyonaria of Funafuti.” Mem. Austral. Mus. Pt. m1. 1897.
1898. AsHworTH, J. H. “The Stomodaeum, Mesenterial Filaments, and Endoderm of
Xenia.” Proc. Roy. Soc. Vol. uxut. 1898.
1899. May WatLrHer. “Beitrige zur Systematik und Chorologie der Aleyonaceen.”
Jena Zeitsch. Naturw. Bd. xxxuu. N. F. Xxvi.
1900. Bournr, Gitpert C. “On the genus Lemnalia,’ also “Branching Systems of
Aleyonacea.” Trans. Linn. Soc. London (Zool.), Vol. vit. Pt. 10.
1900. Hickson, S. J.. AND Hues, Isa L. “Stolonifera and Alcyonacea from New
Britain.” Willey’s Zool. Results, Pt. tv. Camb. Univ. Press.
1900. Hickson, S. J. “Aleyonaria and Hydrocorallinae of Cape of Good Hope.” Marine
Investigations in South Africa, Cape Town, 1900.
1901. Hickson, 8. J. <Alcyoniwm. Liverpool Marine Biolog. Committee, London, 1901.
EXPLANATION OF PLATES XXVIII—XXXI.
ap.c. =aperture of canal. l. = lobe.
au. = contracted autozooid. m. = mesentery.
e.U. = expanded autozooid. m.ap. = mouth aperture.
C.0. =coelenteric cavity. mar.lob, = marginal lobe of capitulum.
cap. = capitulum. mg. = mesogloea.
cap.mar.=capitular margin. mg.c. =mesogloeal cells.
el. = clubs. mg.d. =dense layer of mesogloea surrounding
C.0. = ciliated vessel. siphonozooids.
d.m,f. =dorsal mesenterial filament. rh, = rhizome.
ect. = ectoderm. 8. = stomodaeum.
end. = endoderm. st. = siphonozooid.
end.b.w. =endoderm of body wall. Sp. = spicule.
end.can. = endodermal canals in mesogloea. sp.h. =hole left by spicule after decalcification.
GHes = golden cells. st. = stalk.
m.c. = vessel of internal canal system. tent. = tentacle.
in.lob. =internal lobe of capitulum. ts.s.c. = transverse superficial canal.
Le. = longitudinal canals. v.mjf. =ventral mesenterial filament.
lm.jf. = lateral mesenterial filament. 20. = zoochlorellae.
Fie. 1. Sarcophytwm ehrenbergi. x 2. The mushroom-like form is characteristic of the genus. The

autozooids are most numerous on the margin of the capitulum where the siphonozooids
are few in number. The downward and inward curve of the margin is shown. Many
of the autozooids are almost completely retracted, but a few partially expanded ones
are to be seen on the margin and upper part of the capitulum.

Fic.

Fic.

Fic.

Fic.

Fic.

Fic.

Fic.

Fic.

Fic.

Fic.

Fic.
Fic.

THE ALCYONARIA OF THE MALDIVES. 537

bo

Surface view of a portion of the capitulum of the same species showing two autozooids,
one expanded, the other contracted. Small spicules are seen imbedded in their walls.
Spicules are also numerous in superficial tissues between the zooids and are seen by
transparency of the ectoderm. The mesenteries of the siphonozooids may be seen through
the ectodermal tissues. x 20.

3. Sarcophytum glaucum. <A ciliated canal connecting the coelenteric cavity of an autozooid
with that of a siphonozooid. x 300. Running at right angles to the ciliated canal is
a longitudinal canal of the internal canal system, which in the preserved condition
appears to be unciliated; this canal also opens into the coelenteric cavity of the autozooid.

4. Group of rounded golden cells in same species with granular yellow contents in internal
canal system. x 620,

5. Sarcophytum plicatum. Spicules. x 60. (This species was not in the collection, but the
drawing is inserted because the spicules are tuberculate and approach Sclerophytum
in form.)

6. Sarcophytum tenuis sp. nov. Spicules. x 120. a@=a slender spindle to be found in the
coenenchym and on the anthocodia of the autozooids. 6&=slender clubs from the
coenenchym. c¢=small club to be found in the periphery of the capitulum and in
the stalk.

7. Lobophytum pauciflorum. A colony characteristic of the genus. x2. As all the specimens

in the collection of fairly large colonies are fragmentary, a drawing was therefore made
of a complete specimen from New Britain, which agrees with the specimens from the
Maldives in all essential details. The lobed appearance of the capitulum is due chiefly
to the formation by budding of new autozooids at any portion of the capitular surface.
In this genus the siphonozooids are most numerous on the marginal seam of the
capitulum. The marginal lobes differ from the inner lobes in the proportionately greater
number of siphonozooids compared with the number of autozooids on their surface.

8. Sclerophytum capitale sp. nov. A complete colony. x2. Many of the autozooids are
fairly expanded.

9. Sarcophytum tenuis sp. nov. x14. Only the upper surface of the slightly concave capitulum
is shown. Autozooids are very numerous round the margin forming a thick fringe. In
the middle of the capitulum they are much fewer in number. Siphonozooids are indicated
as dots between the autozooids, they are much more numerous in the actual specimen.

10. Sarcophytum roseum sp. nov. Spicules. x 120, The spicules are slightly zonate, and in
this respect approach Lobophytum. 1, U1, U1 are taken from near the surface of the
capitulum, Iv, Vv, VI and vit from the coenenchym of the capitulum.

ll. A portion of the capitulum of the same species. x14. All the autozooids are contracted.
The capitulum is thickest in the middle and thins out towards the margin, and therefore
has no capitular fold as in “ehrenbergi” (fig. 1). Siphonozooids are indicated as small
dots between the autozooids.

12. Lobophytum pauciflorum. Young mushroom-shaped colony. x 2.

13. Surface view of tip of lobe of capitulum of same species, showing three autozooids and
surrounding siphonozooids. x 30. One autozooid is fairly expanded.

G. II. 69

538
Fig.

Fie.

Fig.

Fie.

Fic.

Fic.

Fig.

Fig.

Fia.

Fia.

Fic.

EDITH M. PRATT.

14. Longitudinal section through four siphonozooids of the same species (Si,, Si,, Siz, St),
showing ciliated vessels connecting the coelenteric cavities. The section passes through
the siphonoglyph of the first siphonozooid. Only the left half of the fourth siphonozooid
is indicated.

15. Sclerophytum capitale. Clubbed spicule. These are very abundant near the surface of
the colony. x 219.

16. Different forms of the tuberculate spicules to be found in a single colony of the same
species. x 20. (Regular spindles, similar to those of other species, are also numerous,
but they are not figured.)

17. Surface view of a portion of the capitulum of the same species, showing arrangement of
siphonozooids between two autozooids. x45. One of the autozooids is partially contracted,
the other is fairly well expanded. The tentacles have a double row of small pinnules
down each side.

18. Sclerophytum densum. Young colony, showing formation of primary lobes from the margin
of the capitulum. The colony has not yet lost its mushroom-like form. The surface of
the capitulum is slightly concave. The siphonozooids are very minute and are not in-
dicated in the drawing. x lt.

19. Sclerophytwm capitale (sp. noy.). Drawing of a fairly thick free-hand section through a
lobe of the capitulum of the same species. x 65 (Cam. Lucida). The drawing shows
the rudimentary siphonozooids which terminate posteriorly in the transverse vessels of
the superficial canal system. The vessels of this system communicate with those of the
internal canal system, and directly or otherwise with the coelenteric cavity of the
neighbouring autozooid.

20. Sclerophytum densum. a, b, c, d, e. Different forms of tuberculated spicules characteristic
of the genus. x 10. f. A forked spicule with irregular tubercles. g, h. Two tubercles
seen laterally. x 300 (Cam. Luc.).

21. Longitudinal section through a single degenerate siphonozooid of same species. x 730,
The section passes through the mouth aperture of the siphonozooid. The stomodaeum
is extremely short in this specimen. A deeply staining layer of dense mesogloea sur-
rounds the siphonozooid internally. The siphonozooid passes posteriorly (on the left) into
a superficial transverse vessel. The endoderm lining the coelenteric cavity is distorted
by the presence of numerous zoochlorellae. An aggregation of endodermal cells near
the stomodaeum is probably a degenerate mesentery.

22. <A partially retracted autozooid of same species. The drawing was made from a fairly
thick freehand longitudinal section stained with borax carmine. The figure shows the
intimate connection between the coelenteric cavity of the autozooid and the lumen of
the canals, and also with the coelenteric cavity of the degenerate siphonozooids. The
long stomodaeum is shown (dotted shading). The ventral and lateral mesenterial fila-
ments are very short. The dorsal mesenterial filament is long, ciliated, and grooved.

23. Sclerophytum hirtum. Drawing of a colony to show lobation of capitulum. x 14. Most
of the autozooids are partially expanded.

24. Spicules of same species. a, 6, c, d, tuberculate spicules; e, f, clubs.

Fauna and Geography Maldives and Laccadives.

RAT L400.

end. OW:

E.M.P, & E.D.de/.

E.Wilson, Cambridge

Pratt - Aleyonaria.

Fauna and Geography Maldives and Laccadives. Plate XXIX

EMPEEDoce/ E.Wils on, C

Pratt - Alcyonaria.

Fauna and Geography Maldives and Laccadives

Plate XXX.

ori

3

To

EWilson, Cambridge.

imipeulele - Aleyonaria.

Faun

aand Geography Maldives and Laccadives Plate XXXI.

—

srerwenterh

saa |

2 & EF, DP. des E n
E.Wi/son, Cambridge.

Pratt -Aleyonaria.

Fic.

Fig.

Fic.

Fic.

Fic.

Fic.

Fic.

Fic.

Fic.

25.

29.

30.

31.

33.

THE ALCYONARIA OF THE MALDIVES. 539

Group of rudimentary siphonozooids of same species. Surface view of stained preparation.
x 90. The drawing shows the degenerate stomodaeum with mouth aperture (Cam. Luc.).

Sclerophytum palmatum sp. nov. Drawing of a colony to show branching of capitulum.
Many of the autozooids are expanded. Natural size.

A tentacle of autozooid of same species, showing double row of rudimentary pinnules
down each side. (x 60 Cam. Luc.)

Sclerophytum gardinert sp. nov. Drawing of a colony. x2. The colony is laterally
compressed and is irregularly cup-shaped, with lobes arising from the middle and edges
of the cup. The autozooids are small and are not numerous. Siphonozooids are absent.

Sclerophytum durum sp. nov. <A lobed colony. x14. All the autozooids are retracted.
The enormous spicules characteristic of this species can be seen through the ectodermal
tissues of the capitulum. They are also shown protruding from the base of the stalk.

A younger cup-shaped colony of same species with rhizome-like outgrowth from the stalk.

x 2. A few very small lobes are seen arising from the basin of the capitulum. The
autozooids in this specimen are apparently most numerous round the margin of the
capitulum.

Tuberculate spicules of same species. a. A single tuberculate spindle. x10. 6. A
portion of the tuberculate surface of the same spicule, more highly magnified. x 80.
c. Tip of a more pointed spindle showing tuberculate warts. x 8. d. x 6 (Cam. Luc.).

Spicules of same species; numerous near surface of colony. a, 6, c, d, e. Different forms
of clubbed spicules. All these forms are common. a, 6, c, x 290, d, x 270, e, x 250.
f. An irregular tuberculate dumb-bell (rare). x 270. g. An irregular spindle with small
spines. x 250.

Sclerophytum querciforme sp. nov. A colony with a secondary stalk and capitulum. x 2.
The large tuberculate spicules are seen protruding from the base of the stalk. The stalk
is branched at the top and passes almost imperceptibly into the capitulum.

69—2

NUDIBRANCHIATA,

WITH SOME REMARKS ON THE FAMILIES AND GENERA AND
DESCRIPTION OF A NEW GENUS, DORIDOMORPHA’.

By Srre Cuarzes Exior, M.A., K.C.M.G.,

Commissioner of the British East Africa Protectorate.
(With Plate XXXII.)

THe Nudibranchs described in the present paper represent the collections made by
Mr J. Stanley Gardiner in the Maldive and Laccadive Islands (1899—1900). I have added
descriptions of two other remarkable specimens with which he has furnished me, one
(Doridomorpha Gardineri) obtained by himself at Rotuma in 1896, and the other (Plewro-
leura striata) by the Skeat Expedition from Pulan Bidang in the Malay Peninsula (1899).
The specimens from Minikoi (Laccadive Islands) were all reef forms: those from the other
localities were largely dredged. In only a few cases are there any notes as to the colour
or appearance of the living animals: as preserved they have, with a few conspicuous
exceptions, lost entirely their original colour and assumed the dirty yellowish tinge common
in alcoholic specimens. They have also in many cases been strongly contracted and dis-
torted by the action of the preserving medium. I have therefore felt very diffident in
creating new species, and have perhaps erred on the side of identifying specimens with the
most similar forms already described. But I think it useless to create new species of
Nudibranchs unless it is possible to formulate the specific characters. The fact that an
alcoholic specimen does not seem quite the same as those already described is not sufficient
to give it specific rank.

Besides forms common in the Indo-Pacific region, the present collections contain a
number which confirm the descriptions by previous authors of species established on the
strength of one or a few specimens, and also offer some novelties. Of these the most
interesting is Doridomorpha, a small Dorid without gills, related to Doridora Bgh. There
are also a large and well-preserved Notodoris, which appears specifically distinct from
N. citrina, for which Bergh established the genus: a new species of Dictyodoris: a form

' [This paper was received in June, 1902, though only of some of the species noticed here, which may bear an earlier
published now. The author has since published descriptions date than the present volume but are really later. Ep.]

NUDIBRANCHIATA. 541

to which I hesitate to give generic rank and somewhat doubtfully refer to Stawrodoris,
liberally interpreted: and a remarkable specimen of the Hexabranchidae, which has been
greatly distorted, but seems at least specifically distinct. The rediscovery of Plewroleura
striata, v. Hass., is also interesting.

The list of genera and species is as follows:

HOLOHEPATICA.
1. Marionia arborescens B. 15. Discodoris meta B.
2. *Doridomorpha gardineri nov. gen., nov. sp. 16. Platydoris eurychlamys B.
3. Hexabranchus faustus B. lee 3 scabra (Q. and G.,).
4 : digitatus nov. sp. 18. ~ sp.
5. Notodoris gardineri nov. sp. 19. Dictyodoris maculata nov. sp.
6. Trevelyana crocea B. 20. Stawrodoris pecten nov. sp.
7. Nembrotha nigerrima B. ? var. 21. Trippa ornata B.
8. Chromodoris elizabethina B. 22. Doridopsis nigra Stimpson, var. nigerrima B.
9 5 rosans B. 23. em apicalis B.?
10. a pustulans B, 24. Phyllidia varicosa Lam.
ll. Ceratosoma trilobatum Gray. 25. © elegans B,
12. 56 gracillimum Semper. 26. Fryeria riippelii B.
13. Dzscodoris concinna (A. and H.). 27. Phyllidiopsis striata B.?
14, ps pardalis (A, and H.)?
CLADOHEPATICA.
28, Pleurophyllidia gracilis B. 32. Madrella ferruginosa (A. and H.).
29. *Pleuroleura striata (van Hass.). 33. Hermaea minor B.?
30. Doto indica B? * 34. Oyerce nigra B.

31. Phidiana unilineata (A. and H.)?

Of these nos. 2 and 29 (Doridomorpha and Pleurolewra), both starred, are not from the
Maldive or Laccadive Islands.

I have left undescribed eleven specimens which were too much injured to admit of
identification. Six were Dorids of soft consistency, without jaws, and with simply hamate
teeth: the remaining five were a small species of Chromodoris, possibly scurra, though, if
so, all trace of the characteristic coloration had disappeared.

Though one of the best works ever published on the Nudibranchiata (Alder and Hancock’s
Monograph of British Nudibranchiate Mollusca) is in English, little has been written in our
language on the group, and the exotic forms have received scant attention from our naturalists,
the greater part of what is known of them being due to the fruitful labours of Prof. Bergh’.

1 The reference letters S.R. stand for Prof. Bergh’s ‘ Malacologische Untersuchungen,” published in Semper’s
Reisen im Archipel der Philippinen.

542 SIR CHARLES ELIOT.

I have therefore given somewhat full descriptions of the family and generic characteristics
which I trust will not be considered out of place. In classification I have adopted Bergh’s
division into Holohepatica and Cladohepatica in preference to Pelseneer’s four tribes Tritonioidea,
Doridioidea, Aeolidioidea, Elysioidea. There is probably no group of comparatively well-known
animals in which we may look forward to as many additions to our knowledge as in the
Nudibranchiata, and to the discovery of forms which are not merely new but of morpho-
logical and systematic importance. Such are Doridoxa, Atthila and Hedyle discovered in the
last few years, and as few seas have been thoroughly investigated and the fauna of many
large marine areas (such as the tropical and southern Atlantic) remain as far as Nudi-
branchs are concerned practically unknown, many more finds may be anticipated. It is
therefore better to have as elastic a system as possible, and to arrange the group in a series
into which new members can be easily inserted, than to cut it up into rigid subdivisions
whose definitions may require alteration. Pelseneer’s Tritonioidea contain a somewhat mis-
cellaneous assemblage, and it is not clear what is the distinction between them and the
less typical Aeolidioidea. Why, for instance, does Dendronotus belong to the former and Doto
to the latter class? I have, however, departed from Bergh’s arrangement in two respects.
Firstly, I classify Tritonia among the Holohepatica, and not as he does among the Clado-
hepatica. It is true that its general anatomy allies it rather to the Cladohepatica, but it
is clearly a connecting link between the two tribes, and as the distinction between them
is the entire or divided condition of the liver, and as the liver of Tritonia is entire, it
seems more logical to regard it as a holohepatic form with affinities to the Cladohepatica
than as a cladohepatic form (which it is not) with affinities to the Holohepatica. Secondly,
I have included among the Cladohepatica the Elysioidea or Ascoglossa, which cannot in my
opinion be separated from the rest of the Nudibranchiata.

HOLOHEPATICA.

The Nudibranchiata Holohepatica are characterized by the liver being entire and not
ramified. They may be divided into two sections, the first containing the Tritoniadae and
the second all the other forms. As mentioned, the members of the first section are really
intermediate between the Holo- and Cladohepatica: in virtue of their unbranched liver they
belong to the former but in their other characters they bear more resemblance to the latter.
The other section, often conveniently called Dorids, are characterized by possessing two
spermatothecas and a blood gland. Further, in all except a few abnormal genera the branchiae
are dorsal and arranged round the vent.

Fam. Tritoniadae’.

The characters of the family may be defined as follows. Animals of moderate or even
large size with an elongate and somewhat rectangular body. A frontal veil, usually provided
with appendages and two somewhat spoon-shaped tentacles are present. | Rhinophores

1 For the study of this family the following may becon- xv. (2) Nudibr. Mollusca of North Pacific (in Dale’s
sulted. Alder and Hancock, British Nudibr. Mollusca, 1855, Results of Exploration of Alaska). (3) Beitr. zur Kenntniss
pt. vit. Vayssiére, ‘‘ Description du Marionia Berghii,’ Jour. der Japanischen Nudibr. 1. (4) Report on the Nudibranchiata
de Conchyl. 3. 8. x1x. 1879. Pelseneer, Recherches sur divers of the Challenger, 1884. (5) Opisthobranchier der Sammlung,
Opisthobranches, Bruxelles, 1893, and especially the following Plate. (6) Nudibranchiate Gasteropoda of the Ingolf Expe-
papers of Bergh. (1) Semper’s Reisen, Hefte xv., xvi. and dition, 1900.

NUDIBRANCHIATA. 543

retractile into sheaths. Branchiae consisting of arborescent tufts set in a single row along
the pallial margin. Buccal mass very large; mandibles powerful. Radula wide, with a broad
central tooth, the first lateral larger than the rest. Liver not at all ramified. Vent lateral.
No blood-gland. One spermatotheca. Reproductive system unarmed.

Setting aside the disputed genus Hancockia Gosse, the family may be considered to
comprise three genera, J'ritonia Cuv., Marionia Vayssiere, and Atthila Bergh. Although the
creator of the last named considers it sufficiently distinct to form a new family, it has
clearly great affinities to Tritonia. The characters of the latter genus are as follows. The
frontal veil bears papillae and there are two spoon-shaped tentacles. The rhinophores are
not laminate but have a single club, surrounded by pinnate filaments. There are no dorsal
appendages except the branchiae and no stomach plates. Most authors recognize as a sub-
genus Candiella Gray (Figures of Moll. Animals, tv. 1850) distinguished by having only a
few digitate processes on the frontal veil and a somewhat narrower radula. The genus
Marionia Vayssiere resembles Yritonia externally, but the stomach is armed with a series
of horny plates. Generally but not always the appendages on the frontal veil are branched.
The genus Atthila, instituted by Bergh for a single specimen obtained by the Ingolf ex-
pedition in the northern Atlantic, has the frontal veil entire, and rhinophores of the
ordinary lamellate type. On the back are three rows of appendages, probably all simple,
but the specimen was somewhat damaged. In their general characters the branchiae and
buccal parts resemble Tritonia.

The Tritoniadae are not numerous in collections, being inhabitants of fairly deep water
and only obtainable by dredging, but seem to be cosmopolitan in their distribution. They
are recorded from the Mediterranean, the Northern and Arctic Atlantic, Brazil, the Pacific
coast of Patagonia, California, the Arctic Pacific, Hawaii, the Cape, the Red Sea, and the
Malay Archipelago. Marionia is recorded from a less wide range than Yritonia, but it is
probable that several species described under the name T'ritonia should really be referred
to the latter genus.

The Tritoniadae are sluggish but voracious animals, living largely on Alcyonium, large
chunks of which are often found in their stomachs. The spawn is a simple string-like coil
deposited on seaweeds or corallines.

The present collection contains only a single representative of the family belonging to
the genus Marionia, created by Vayssiere, and characterized by the presence of a girdle
of horny plates set round the stomach.

Marionia Vayssiere?.

1. Marionia arborescens B. (S.R. Heft xvi. pp. 890—894, Pl. LX XXVIII).

One specimen from Fadifolu Atoll, Maldives, 12—25 f. (fathoms). The species was
established by Bergh for four specimens captured by Brock at Amboina. The specimen
contained in the present collection agrees well with his description but is somewhat smaller,
being only 3:5 cm. long and 1°3 broad.

1 Vayssiére, ‘‘Sur un nouveau genre de la famille des Journ. de Conchyl. 3. 8. x1x. 1879, pp. 106—118, Bergh,
Tritoniadés,” Comptes rendus de Vac. des sc. uxxy. 1877, ‘‘Beit. zu einer Monogr. der Gattung Marionia, V.” Mittheil.
pp- 299—301. Id. ‘‘ Description de la Marionia Berghii,” aus der Zool. Stat. zu Neapel tv. 2. 1883, pp. 303—326.

544 SIR CHARLES ELIOT.

The form is as usual in the genus. The colour uniform dirty yellow. From Bergh’s
specimens it appears that the living animal is yellow with green branchiae. Frontal veil
with 10 simple papillae arranged in three groups, four in the middle and three on each
side: outside the latter are the tentacles with a deep groove on the under side. The
dorsal surface is covered with minute knobs. The dorsal margin is fairly prominent and
bears 11 branchiae on each side, of which the first, tenth and eleventh are rudimentary
and the fourth, fifth and sixth the largest. The branchiae are long, arborescent and mostly
bipinnate. The rhinophore sheaths are distinctly raised and the rim is irregularly scalloped:
the rhinophores as usual, with five pectinated filaments and a simple club. The buccal
mass, as usual in this family, is enormous for the size of the animal: the jaws are large
and strong and the cutting edge bears a single row of large denticles. The radula also
as usual, consisting of about fifty rows of yellow teeth: the formula for each row about
45+1+1+1+45. The shape of the median tooth and first side tooth as usual: the
laterals diminishing towards the outside. The first stomach lamellated: the second with a
girdle of yellow horny plates. There are about 30 large plates and numerous small ones
of varying sizes between them. The other internal organs were not well preserved, but did
not indicate any deviation from the generic characters. The species is characterized by the
appendages on the frontal veil being simple and not branched and by the single row of
large denticles on the jaw.

Fam. Doridoxidae.

This family was created by Bergh in 1900 (The Danish Ingolf Expedition—Nudibranchiate
Gasteropoda—Copenhagen, 1900) to contain a remarkable mollusc dredged at a depth of
55 f. in the North Atlantic. Doridova is externally similar to the Dorids in_ its
general shape and rhinophores, but has no external gills whatever and the vent is lateral,
not dorsal. The jaws are large and strong: the radula is somewhat as in Tritonia and
Bathydoris and consists of a large central tooth with several laterals, of which the innermost
differ somewhat in form from the rest. A blood-gland is present and there are two
spermatothecas. It is most interesting to find that Mr Gardiner has captured a somewhat
similar form at Rotuma in the southern Pacific. The specimen is unfortunately small and
the internal organs are hardened into an indistinguishable mass. It is therefore impossible
to state whether the animal possesses a blood-gland and a second spermatotheca. If those
organs are really absent, it would probably be necessary to make it the type of a new
family, but meanwhile its other characters, and in particular the buccal parts, show a close
resemblance to Doridoxa. It is however in any case generically distinct from that animal,
from which it differs in its round form, smooth back, absence of tentacles and its toothed
mandibles. A third gill-less Dorid is the Heterodoris of Verrill, which has the general shape
of Triopa, a lateral vent but no mandibles or central tooth. In its external characters it
appears to form a good connecting link between T'ritonia and the Polyceradae, but in its
buccal organs it varies from the former family much more than do the Doridoxidae.

2. Doridomorpha Gardineri nov. gen., nov. sp.

One specimen from Rotuma. In the tank received by me this animal was concealed
within the epipodia of an Aplysia but the position was no doubt accidental and does not
indicate any natural association. The form (fig. 1) is oval and much rounder than Doridoza,

NUDIBRANCHIATA. 545

the length being 0°65 cm. and the breadth 0°55. Though the back is arched, the height
is small compared to the length and breadth, and the general appearance is flat. The colour
is a uniform yellowish brown. The back is perfectly smooth and the mantle edge ample
all round. The foot (fig. 2) is however not small: in the preserved specimen its edges
are turned inwards, but when extended it must be of a considerable breadth. Its anterior
margin is not grooved. The visceral mass is much narrower than either the mantle or the
foot. The rhinophores (fig. 3) are yellowish and perfoliate: they appear to be retractile
into simple holes without any rim, but in the preserved specimen both were exserted.
Above the foot and clearly divided from it is a sort of snout without any trace of tentacles.
Considering the size of the animal the buccal mass is large and strong. The edge of the
powerful mandibles (figs. 4 and 5) is denticulate, and at the base of the denticulations
is a row of blunt knobs. The radula (fig. 6) was somewhat injured in extraction but
appeared to be short and to contain about 13 teeth in each row. The character of the
teeth was, however, clear. The central tooth consists of a basal plate with a strong cusp:
the laterals of a basal portion and a hook: towards the ends of the row the teeth grow
more slender and elongate and the hook smaller. Owing to the bad preservation of the
interior and the small size of the animal it was not possible to satisfactorily determine the
character of the digestive tract and other organs, but it was quite plain that the anal
papilla was under the mantle and latero-posterior (fig. 2). A careful examination failed to
discover a blood-gland, but its absence cannot be regarded as certain. In the central nervous
system all that could be clearly made out was that the optic nerve was long and the eyes,
which were large and black, situated unusually deep. At the right anterior corner of the
foot was situated the large genital papilla with two distinct orifices (fig. 2). It was un-
fortunately impossible to ascertain the character of the internal genital organs, but the
hermaphrodite gland appeared to be large and to cover the liver. There were no signs
whatever of internal or external gills.

Fam. Hexabranchidae,

The family as at present known is composed of the single genus Heaxubranchus, with
the following characters. Large animals of a flattish shape and with ample and undulated
pallial margin. Back soft and smooth. Foot broad. Tentacles foliaceous. Rhinophores
perfoliate, retractile. Labial armature very strong. Radula composed of numerous simply
hamate laterals, without a central tooth. Branchial apparatus consisting of 6—8 arborescent
tufts, separate from one another, not retractile into a common cavity, but each capable of
contracting imto a small hollow at the base, which does not however amount to a permanent
pocket and does not hide the branchia. Anal papilla within the circle of branchial tufts,
central or subcentral. Reproductive organs unarmed. Verge exceedingly long. Nervous
system much concentrated, so that it is difficult to distinguish the ganglia. Blood-gland
single but large. Near the renal pore there is another small orifice of unknown function.

The Hexabranchidae differ remarkably from most Dorids in their habits. They are large,
active animals, capable of swimming as well as of crawling. I have generally found them
in pools left by the high tide, but think that on these occasions they are merely stranded
and naturally prefer deep water. I have also found them swimming on the surface about
a quarter of a mile from the coast. Their coloration is generally red and exceedingly
conspicuous, in no way resembling their surroundings, but, when at rest, they fold the sides of

Gant 70

546 SIR CHARLES ELIOT.

the mantle over the back (almost like Placobranchus) and conceal their gorgeous colours, thus
entirely altering their appearance. When swimming, they flap the sides of the mantle and
bring the two edges of the foot together so as to form a sharp keel. When in captivity
I have often observed them adopt an almost vertical position, standing on their heads and
walking, or at least feeling their way with their large tentacles. They are very voracious
and browse freely on sponges, sertularias, polyzoa, ete.

Bergh has justly observed that it is very difficult to determine the species of Heaa-
branchus without, or even with, an anatomical examination. Perhaps most of the so-called
species are merely varieties. Most of them are very imperfectly described, and even in those
cases where the description is adequate it is not always plain what are the specific
characteristics beyond colour. There are unfortunately no notes as to the colour of the
animals in the present collection when living. As preserved they are of an almost uniform
dirty yellow. The family is recorded only from the Indo-Pacific (EK. Africa, India, Philippines,
Polynesia, Red Sea).

The peculiar arrangement of branchiae displayed in this genus is also found in the
Bathydorididae and Kalinga. The former are deep-sea forms, with strong jaws and a radula
provided with a median tooth, and are thus allied to Doridoxa and Tritonia. The buccal
parts of Hewxabranchus, though unusually powerful, show a near approach to the ordinary
type of the cryptobranchiate Dorids. The affinities of Kalinga are very uncertain. Bergh
places it among the Polyceradae.

Gen. Hexabrunchus' Ehr.
3. Hexabranchus faustus B.

I somewhat doubtfully refer to this species a specimen from the S.W. reef of Hulule,
Maldives. It is 14 cm. long, and considerably broader across the gills (7°3) than across the
head (54). The mantle is crinkled at the sides, but quite straight and simple in front.
It is compressed against the sides of the body, but the free edge when pulled out and
extended measures 25cm. The rhinophore sheaths are slightly raised and quite smooth.
The groups of branchiae are set rather close together, and are not always distinct. They
appear to be usually 8, but might perhaps be divided differently. Each group consists of
several, generally 4, tripinnate tufts’. The anal papilla is central, wide, but not very promi-
nent. Near it are the renal pore and the other orifice alluded to. The labial armature
and radula are both strong and dark brown, the former consisting of two furrowed plates
composed of a cumpact mass of small and slightly bent rods. The latter contains about
45 rows of simply hamate teeth, there being about 70 teeth on each side of the rhachis in
every row. The tentacles are large, foliaceous, furrowed down the middle, and with crinkled
edges. Reproductive organs as usual.

4. Hewxabranchus digitatus n. sp.

One individual from Hulule S. reef, 7°8 cm. long and 5:2 broad, the breadth being
much the same in front and behind. The specimen is very much retracted and corrugated

1 Bergh in S. R. Hefte xi., xvr. and Supp. 1. and in Exploring Expedition, vol. xm. (Gould).
Ergebnisse einer Reise nach d. Pacific, Schauinsland, Opistho- 2 Prof. Bergh gives the number of branchiae as 6, but the
branchien, 1896-7. Several species are described in the plate shows 7.
voyages of the Bonite, Astrolabe, and the United States

NUDIBRANCHIATA. 547

owing to the action of alcohol, but the following characters seem certain and_ sufficient
to establish a new species. The mantle is extremely ample and much crinkled, recalling
in this respect Tridachia; when extended from the body, the free edge measures 3:4 cms.
Over the head are three digitate processes, which are not due to undulations but are
distinct and permanent projections from the mantle edge. The rhinophores are not much
bent back. The tentacles are almost arborescent; they arise from short stalks, and are
deeply indented and subdivided. The buccal parts, nervous system, and reproductive organs
are as usual in the genus. The radula and labial armature are strong and reddish-brown
in colour. Both the head and the branchiae are withdrawn to an unusual degree into the
body. The latter are partly covered by an anterior flap of skin and are crowded together,
so that the usual division into groups is not clear. The sections are however divided from
one another by eight membranous septa. If this arrangement could be considered with
certainty as structural it would afford an interesting link between the ordinary branchial
arrangement of Heaabranchus and that of the cryptobranchiate Dorids, but in view of the
generally wrinkled and contracted state of the specimen I do not think it safe to regard
this configuration as more than a distortion, and should define the species by the following
characters—mantle edge extremely ample with digitate processes over the head: tentacles
much subdivided and almost arborescent. But it must be remembered that specimens of
Hexabranchus which seem to be mere varieties exhibit great differences in the extent to
which the mantle and tentacles are developed.

Fam. Dorididae phanerobranchiatae.

A number of Dorids are conveniently grouped together as Phanerobranchiata, that is to
say, they have no distinct and permanent dorsal cavity into which the branchiae can be
retracted, though when touched the branchiae may contract so strongly as to almost dis-
appear. Bergh divides them into the Goniodorididae, which possess a buccal gizzard (ingluvies
buccalis), and the Polyceradae, which do not, an arrangement which brings together forms
of such different appearance as Lamellidoris and Ancula. It seems to me that it would
be more natural to subdivide the group into Polyceradae with limaciform bodies, sometimes
bearing appendages, and Pseudodorididae, flat Dorid-hke forms. It would appear that in
other groups closely related genera may either possess or not possess this buccal crop.
Thus it is present in Placobranchus and Tridachia, but absent in Elysia. The Pseudodori-
didae seem to be confined mainly if not entirely to northern seas.

The radula of the Phanerobranchiata is usually narrow and with differentiated teeth.
They show a greater variety of form than the Cryptobranchiata, but are far less numerous
in species.

Gen. Notodoris B2

Body limaciform, with no clear distinction between the back and sides: hard, rough,
and rugose; surface filled with spiculae. Frontal veil large. Rhinophores smooth, protected
by valves. Gills protected by a large valve, which is sometimes elaborately subdivided. No
labial armature. Radula without central tooth but with numerous hamate uniform laterals,
having a rudimentary secondary denticle below the main hook.

1 Bergh, ‘‘Neue Nacktschnecken d. Siidsee, m1.” Journal d. Mus. Godeffroy, vm. 1875.
=
70—2

548 SIR CHARLES ELIOT.

This genus was instituted by Bergh for a single specimen (N. citrina) from Rarotonga.
I have obtained another (N. minor) at Zanzibar, which seems specifically distinct, and the
present collection contains a third. All three forms are identical in their anatomy, and
are also all of a yellow colour, but they differ externally in the contour of the body and
the subdivisions of the branchial valve. It may perhaps be proved that they represent
merely varieties or different ages, but provisionally they must be classed as separate species.
N. citrina appears to have a single median ridge from the rhinophores to the branchiae,
a branchial valve with about 8 lobes, and a fairly distinct tail. In NV. minor the tail is
not distinctly marked off from the rest of the body, and the whole animal seen dorsally
somewhat resembles Phyllidia. The inconspicuous branchial valve is three-lobed, and though
the surface of the back is somewhat irregular there are no distinct longitudinal ridges.
The colour is lemon-yellow, with black lines, not spots. In N. gardineri the long tail is
plainly distinguished from the body: there are four dorsal ridges: the branchial valve consists
of three lobes, each of which is elaborately subdivided: the colour is yellowish, with a few
round black spots.

Notodoris is evidently allied to Aegirus (found on the British and E. European coasts)
and to the little known T'riopella, both of which are of hard consistence, and have the
branchiae and rhinophores protected by tubercles. This little group of three genera seems
to stand near the cryptobranchiate Dorids, inasmuch as the teeth are not differentiated
but uniform, and the branchiae capable of being concealed. They are not retractile into
a pocket, but the same result is produced by the protecting valve, which shuts down over
them. Nothing is known of the habits of these animals. In captivity N. minor was quite
inert, and did not make any movement. With Mr Gardiner’s specimen is a piece of hard,
yellow sponge. There is no note, but as the colour and consistency of the sponge closely
resemble those of the nudibranch it is highly probable that the latter frequent it.

5. WN. gardineri n. sp. (fig. 7).

One specimen from Hulule, Maldive Islands. Length 3°35 cm.: breadth at widest part ‘9:
height at highest part 1:2. Colour, uniform dirty yellow, with a few vivid black spots on
the upper but not on the under surface. Frontal veil large and rounded. The flat sole
occupies the whole under surface of the body. There is no distinct mantle edge, but a ridge
runs from each end of the frontal veil to the branchial valves. Between these are two more
ridges, which are again connected by smaller ridges, shaped like the letter Y. The depressions
between the ridges are smooth, but the tail and other parts of the body covered with rough
and rather large tubercles. The radula (as in the other species) consists of numerous uniform
hamate teeth, with a rudimentary additional denticle. Their form is exactly that given in
Bergh’s plate (J. c. Pl. IX. 39). The visceral mass and reproductive organs were too much
hardened to admit of any examination, but in NV. minor the verge appeared to be trifid at
the extremity and covered with rows of small tubercles. The rhinophore openings are closed
by their valves. The large branchial valve (fig. 8) is composed of three lobes, each elaborately
divided and furnished with numerous spatulate processes. In order to preserve the specimen
the internal parts were taken out through the foot and the frontal veil and branchial valve
not removed. Hence the gills were not examined.

NUDIBRANCHIATA. 549

Gen. Trevelyana Kelaart’.

The shape of Trevelyana is limaciform, there being no mantle margin or distinction
between the back and sides and no appendages. The rhinophores are retractile and perfoliate.
The branchiae, which are generally placed about the middle of the back, vary in number
(7—32), but are usually numerous, simply pinnate or compound. The tentacles are small.
There is no labial armature or rhachidian tooth. The radula is short and narrow: the first
and sometimes also the second tooth differs im shape from the others. There are two or three
distinct hermaphrodite glands not attached to the liver. The verge is armed.

Trevelyana is only recorded from the Indo-Pacific area, but within it is one of the
commonest of polycerids, and is frequent under stones in the littoral zone, eg. in East Africa.
It is generally found in pairs.

6. Tr. crocea B.

Two specimens from Hulule. Length 2-4 cm.: breadth 1:2. Body rather square and
stout: well preserved, but the colour lost, the whole animal being of a uniform greyish-white.
Rhinophores small, with about 12 perfoliations, and retractile into simple pits without raised
margin. Branchiae 20 in both individuals, forming a circle open behind, which is completed
by the anal papilla. The gills are bent inwards, and on their inner side are simple lamellae.
The tentacles very small: the radula is composed of about 15 rows: the two teeth next to
the rhachis are bicuspid, consisting of a basal portion, from which two hooks arise: the others
have a single hook, which is much reduced in the outermost. The nervous system is rather
large, and shows the various ganglia very distinctly divided. There are two distinct herma-
phrodite glands close together, and not spread over the liver. The verge is armed with small
spines. Prostate large.

Though there is no record of the colour of the living animal the specimens seem

referable to Trevelyana crocea on account of the number of the branchiae and the shape of the
teeth.
Gen. Nembrotha Bergh’.

Externally this genus closely resembles Trevelyana, and differs only in the gills, which
instead of being numerous and small are few in number, but very stout and strong. Internally
it differs in the structure of the radula and generative organs. There is a squarish central
tooth, with a large hamate lateral at the side. The remaining teeth, which vary in number
from 3 to 12, are also generally squarish plates without hooks. The hermaphrodite gland
is spread over the liver, and not separated into distinct masses. There is no prostate. As
a rule there is no labial armature, but a weak one exists in N. nigerrima.

7. Nembrotha nigerrima ? variety.

One specimen from Minikoi, much bent and contracted. Length 34 cm., breadth 16,
height 2:4. Colour, a uniform dirty yellow: the gills and rhinophores are much redder, and
there are a few raised reddish spots, particularly near the gills. The rhinophores small, with
about 30 perfoliations: the skin round the rhinophore openings is much contracted, but no
raised edges were discernible. The gills large, 5: two on each side and one in front. As
in other species of Nembrotha which I have seen the rhachis is very hard and strong, so

1 A. and H., Notes on a Coll. of Nudib. made in India, der Polyceraden, 1883.

1863. Bergh in S. R. Hefte x1. and xvi. Id. Beitr. zur Kenntn. > Bergh in S. R. Hefte x1, xvir., and in Harvard Bull. of
der Japanischen Nudibr. 1880. Id. Beitr. zu einer Monographie Comp. Zoology, 1894.

550 SIR CHARLES ELIOT.

that when the gills are contracted it almost covers the branchial lamellae, hke the valve
in Notodoris. No tentacles were discernible: the inner lp is brown, and bears a weak
armature of small rods: the formula of the radula appears to be 11+1+41+4+1+11. The
central tooth is squarish: its front edge is turned upwards, and bears four or five denticles.
The first lateral is large and sickle-shaped: the rest squarish without hooks, and decreasing
in size outwards. Blood-gland large. The hermaphrodite gland spread over the liver. Verge
armed with small hooks.

The typical Nembrotha nigerrima appears to be black, with bluish-green spots, red gills,
and a red mantle border. Several species of Nembrotha which I have preserved in spirits
have kept their colour well, and I do not think it likely that the ground colour of the
present specimen was ever black. I am very doubtful if this individual should be referred
to N. nigerrima, as both its shape and coloration seem different, but I cannot formulate any
separate specific characters.

Fam. Dorididae cryptobranchiatae.

The cryptobranchiate Dorids are characterized by the possession of a distinct and perma-
nent dorsal pocket, into which the branchiae can be entirely withdrawn. The rhinophores
are invariably perfoliate and retractile. There is never a buccal gizzard. A labial armature
is present or absent. The radula is broad, and the teeth either uniform or only slightly
differentiated. A central tooth is rare.

In number of species this is the richest family of all the Nudibranchiata. Though
cosmopolitan in distribution it is particularly abundant in the Tropics.

Gen. Chromodoris A. and H.

The form of this genus is generally high, narrow, and lmaciform, the mantle being
somewhat small at the sides, but expanding into a fairly ample frontal and caudal veil.
The coloration is brilliant. The foot usually projects behind the mantle. The branchiae
are simply pinnate, but are sometimes bifid and vary considerably in their arrangement.
The tentacles are retractile. The labial disk has a strong armature of minute hooks. The
rhachis of the radula is either naked or bears thickenings, very rarely true teeth. The laterals
are numerous and hamate: the first is denticulate on both sides, those which succeed it
on the external sides only, the outermost at the extremity only.

Chromodoris is the largest genus of the Cryptobranchiatae, comprising over a hundred
species. Many of these resemble one another closely in structure, but some show remarkable
variations such as Chr. scabriuscula, which has a hard tuberculate back and rhachidian teeth,
peculiarities which almost entitle it to generic rank. Also the shape is somewhat variable.
Typically, it is as described above, but some species are flatter with an ample mantle margin,
and present a quite different appearance. Many tropical Chromodorids have considerable powers
of altering their form, and alternate between a high narrow condition and a flat condition

1 Bergh in S. R. xr. Supp. Heft 1., Hefte xv., xvr., xv. Challenger Report on Nudibranchiata, 1884. von Jhering,
Bergh, ‘‘ Neue Nacktschnecken der Siidsee,” Journal des Mus. ‘‘ Beit. zur Kennt. Nudibr. Mittelmeeres,” Mal. Bliatter, N. F.
Godeffroy, Heft viz. 1875 and Heft x1v. 1878. ‘‘ Untersuchung Bd. nm. 1880, Collingwood, On some New Nudibr. Mollusca
der Chr. elegans u. Villafranca,” Mal. Blitter, xxxy. 1878. from the Eastern Seas, 1878.

NUDIBRANCHIATA. 551

in which the mantle is extended. Thorunna, Bgh., closely resembles Chromodoris, but lacks
the labial armature.

The Chromodorids are confined to the warmer seas, and are most abundant in the
Tropics, but there are some ten species in the Mediterranean, and the genus is represented
on the coasts of Japan and California. They are common on coral reefs and in pools filled
with seaweed. Their brilliant coloration, which is often dazzlingly gorgeous in the living
animal, does not appear to in any way harmonize with their environment. On the other
hand some of the most striking species are commonly found under stones where their hues
can be neither “ protective ” nor “warning.”

The species of Chromodoris are distinguished largely by their colour, though other
characters are not wanting, such as the details of the teeth and labial armature, the presence
of rhachidian thickenings, glands on the mantle edge, and the arrangement of the gills.
The number of gills however is not constant in the same species, and even the arrange-
ment seems to vary. The colour of Nudibranchs as a rule is very variable, but in this
genus the general pattern and ground colour, though they may differ in elaboration or
intensity, are mostly constant. Unfortunately there are no notes as to the living colours of
the specimens in this collection, and in only two are they recognisable, so that it is
exceedingly difficult to make any certain identifications.

8. Chr. elizabethina Bgh.

Two specimens from §S. Nilandu Atoll, Maldives, 30 7. One was dissected. The animal
was much bent, but if straightened out would have been about 2°5 cm. long by ‘6 broad.
Colour in alcohol brownish-yellow, margin of mantle whiter. From a black patch between the
rhinophores rise three broad black lines, two of which run down the sides and meet behind
the gills, while the third runs from the rhinophores to the anterior edge of the branchial
pocket. Round the sides of the body are other black lines, in one specimen three, in the
other two. In this latter specimen the lines are only on one side, having disappeared on
the other. Branchiae twelve, two being free in the middle, and the rest united at their
bases into two groups of five, one on each side. Three gills on one side and two on the
other were bifid. The radula with thickenings on the rhachis: the innermost teeth with
4 small denticles on each side. The outermost with from 2 to 4 points.

9. Chr. rosans B. (? Chr. Bullockii Coll.).

One small specimen from Minikoi. 13 cm. long and ‘5 broad. Colour a uniform reddish-
purple, with traces of a white border. Rhinophores and branchiae grey. It would appear that
the living animal is rose-coloured, with a brilliant white border and yellow rhinophores
and branchiae. Gills nine, apparently set in a horseshoe open behind, decreasing in size
as they go backwards. No thickenings or rhachis of radula: teeth as usual, outermost with
several largish denticles.

This seems to me to be a young individual of Chr. rosans B., which is probably identical
with the Chr. Bullockii described and figured by Collingwood (1. c.).

10. Chr. pustulans B.

Two specimens from Minikoi seem referable to this species. Length of largest 2°1 cm.,
breadth 1:2. The broad back is covered with low, round tubercles. The ground colour is dull
yellowish-brown, but the gills and tubercles show traces of bright yellow pigment. The

552 SIR CHARLES ELIOT,

gills are 12 (eleven in Bergh’s specimen), arranged in an incomplete circle, the vacant space
in which is filled by the anal papilla. The two median gills are separate, the others con-
nected into two groups, five on each side. A strongish labial armature of bifid hooks.
Radula without thickenings or rhachis. Innermost teeth somewhat irregularly denticulate on
both sides. Outermost very irregular in shape.

Gen. Ceratosoma Ad. and Reeve.

This genus is easily recognized by its extraordinary shape. The body rises upwards
from the head so that the branchiae are much higher than the rhinophores. Near the
branchiae are two side lappets and behind them a long dorsal process. There is a round
frontal veil, a narrow mantle edge and a narrow foot prolonged into a tail of nearly the
same length as the body. In their essential characters these animals resemble Chromodoris*.
There is a labial armature of hooks: the radula has no central tooth but numerous hamate
laterals which are sometimes denticulate. The tentacles are retractile: the branchiae simply
pinnate: the reproductive system unarmed, and as in Chromodoris there is no true stomach,

I have seen numerous specimens of Ceratosoma polyomma alive in East African waters.
They are sluggish and generally found among seaweed, where they make no attempt to
hide themselves. They have a strong, disagreeable odour. The colour varies from green
to chestnut-red and there is also some variety in the shape; one specimen had neither the
lateral nor the dorsal lobes.

11. ©. trilobatum Gray.

Two specimens from 8. Nilandu Atoll, Maldives, 30 f The animals as preserved are
quite white, except for a slight bluish tinge at the end of the dorsal process. They are
hard and of a peculiar waxy appearance. This is not the coloration of the specimens
previously described, but the species seems to be really characterized by the back being
relatively less steeply inclined than in others, by the thick mantle edge and frontal veil
and by the thick, rounded and somewhat indistinct lateral lobes. They resemble fairly well
Gray’s figure (Figures of Molluscous Animals, Pl. 67, fig. 14) of Doris trilobata. The largest
specimen is 5‘7 cm. in length and 15 in breadth across the lateral lobes. The strong
rounded tail is 3:2 cm. long and the dorsal process 1 cm. The foot is a long narrow groove.
The mantle edge, lateral lobes and frontal veil are all thick and indistinctly marked off
from the body. The tentacles are retracted: the mouth is a narrow vertical slit between
them. The labial armature consists of two light yellow plates, rather soft and composed
of little hooks. ‘The radula contains about 70 rows: in each row were numerous hamate
teeth bearing a single denticle on each side of the naked rhachis. The back is quite
smooth; immediately below the dorsal process lies the small branchial opening, which con-
tracts even further within. The rim is smooth and hardly raised. The branchial apparatus,
which is retracted into the lowest depths of the cavity, is rather large and it is a wonder
how it can even emerge from so small an opening. The branchiae appear to be 14 in
number. In front are two, quite separate. At the sides are six united together and some-
times bifid so that it is hard to say how they should be counted. The posterior ends of
these two lateral groups turn inwards,

1 Bergh, S. R. Heft x., xvu., and Supp. Heft 1., and in * Some species of Chromodoris (e.g. Ch. Semperi) have a
Nudibranchiata of Challenger. dorsal prominence behind the gills which recalls Ceratosoma.

NUDIBRANCHIATA. 553

12. ©. gracillimum Semper.

One specimen from Kolumadulu Atoll, Maldives, 33 f, of a uniform dirty yellowish-brown.
It is easily distinguished by its very slender dimensions. The length from the head to the
tip of the tail is 2°7 cm., the body being somewhat bent. The tail is very long (1°6 cm.)
and compressed, being only 2mm. thick. The breadth across the back is 7 mm., and the
height of the dorsal process also 7mm. The mantle and frontal veil are far more distinct
than in the last species and the free mantle edge measures 2mm. The apertures for the
rhinophores and branchiae are very small. The specimen was not dissected.

Two small specimens about a centimetre long from N. Male Atoll, Maldives, 27 f, are
perhaps referable to the same species though proportionally less elongate and slender than
the larger individual.

Gen. Discodoris! B.

The Discodorids are soft, flat animals, of an oval or oval-elongate form. The back is
granulated or finely tuberculate, and in the known species generally yellow or brown with
darker mottlings. The mantle edge is ample but the foot fairly wide, grooved and more
or less notched on the upper lip in front. The tentacles are digitiform, the branchial
aperture round, crenulate or stellate. There is a labial armature formed of two plates (or
a continuous ring) composed of minute rods. The radula consists of simple hamate laterals,
without a central tooth. The reproductive organs are unarmed but there are large prostates.

Discodoris is distinguished from Platydoris externally by its softer texture and internally
by the presence of a labial and the absence of a genital armature. Separate genera have
been created by Bergh for several allied forms with a labial armature but showing various
peculiarities in the reproductive system.

Discodorids are recorded from the Pacific, Indian Ocean, E. Africa, Philippines, Mediter-
ranean, and West Indies. They are therefore in all probability cosmopolitan. Like Platydoris
they are frequently found under stones between tides on coral reefs. They are lethargic
in their movements and several species have an extraordinary power of self-mutilation. When
handled, or apparently when feeling uncomfortable, they slowly detach part of the mantle,
or even the whole mantle edge in a complete ring. The remaining part of the animal
seems uninjured and lives at least several days. This self-mutilation is generally explained
as an attempt to escape from a foe who has seized hold of the mantle, but considering
how slowly it is effected and that it takes place when the animal is in captivity and is
not attacked, though perhaps suffering from the effects of bad water, it may equally well
be regarded as a sign of weakness, due to unfavourable physical conditions’.

13. D. concinna (A. and H.).
One specimen from Minikoi.
back is minutely tuberculate, yellowish-brown

Length 42 cm., breadth 2°2 cm., foot 13cm. wide. The
in colour with blackish spots. The tips of

This was the case when they were put

1 Bergh, S. R. Hefte xu., xvi., xv., Supp. Heft 1, and
Challenger Report on Nudibranchs. Alder and Hancock,
Notes on a Coll. of Nud. Moll. made in India, 1863.

2 This self-mutilation is particularly frequent in a species
of Gena common at Zanzibar. I have collected about 30
specimens and found that they invariably threw off their
tail, at least half the bulk of the entire body, within three

(Gh UU

hours after capture.
in a vessel quite alone without any other animal or object
which could alarm them. The division always took place at
exactly the same point, just in front of the last pair of dorsal
processes. The rejected half bore an extraordinary resem-
blance to a nudibranch. The foot was able to adhere and
the dorsal processes looked like rhinophores.

71

554 SIR CHARLES ELIOT.

the tubercles, both in the yellowish and in the blackish parts, are of a lighter colour than
the bases, which gives a curious effect well rendered in Alder and Hancock’s plate lc. The
foot and under edge of the mantle are spotted with purplish-brown. The rhinophore sheaths
are raised and tuberculate. The branchial opening is large and crenulate and contains six
tripinnate gills. The labial armature consists of two narrow yellowish plates, which form
an almost continuous ring. The radula is as usual in the genus and family broad, the teeth
being smaller but not denticulate at the outside.

I think this specimen may be safely referred to A. and H.'s species concinna.

14. D. pardalis, var. (A. and H.).

One specimen from N. Male Atoll, Maldives, 25 f/ Length 44 cm., breadth 2°1 cm.
Dorsal surface finely tuberculate, yellowish, with traces of orange-coloured spots. Under
surface yellowish, without spots. Foot grooved at the anterior end but the notch in the
upper lamina is hardly perceptible. Branchial opening raised, so as to be almost a papilla,
round and not notched. Branchiae six, long and slender, tripinnate. Labial armature circular
and very strong, consisting of a compact mass of yellowish rods. Radula unusually narrow:
teeth simply hamate, large and coarse and not materially diminishing at the inside or
outside of the row. Formula 13 (or 14). O. (14 or) 13. Reproductive system unarmed.
Prostates large.

I doubtfully register this specimen as D. pardalis (A. and H. J. c.) because it has obvious
resemblances to that species as described and the points of certain difference do not seem
sufficient to constitute a new species. The coloration, the slender branchiae, and the absence
or inconspicuousness of the notch on the upper lamina of the foot are all points of re-
semblance. But the branchial opening though raised is distinctly round and not stellate as
in A. and H.’s specimen, and though they give no details as to the radula it is noticeable
that they do not allude to its narrowness or the great size of the separate teeth.

15. D. meta B.

A single specimen from N. Male Atoll, Maldives, 30 7, seems referable to this species
with tolerable certainty. The animal is oval in shape, 23cm. long by 17 cm. wide: the
texture is soft and the back covered with fairly large papillae. The ground colour is yellowish
with the remains of brown and purple blotches in the middle. From between the rhino-
phores to the gill opening runs a somewhat indistinct dorsal ridge. The mantle margin is
ample and undulated. The rhinophore sheaths are projecting, with jagged edges, which in
the preserved specimen have closed over the opening. The branchial opening consists of a
transverse slit, the upper end of which overhangs the lower. Gills, 6, tripinnate. Foot
deeply grooved and notched in front. Labial tentacles reddish-yellow. Labial armature bright
blue, formed of minute rods. The radula consists of simply hamate teeth. Both the imner-
most and outermost, especially the former, are smaller than the median teeth, but none
were rudimentary. The cerebral ganglia were rather far apart.

The colour of the labial armature is unusual. In the two specimens examined by Bergh
(lc. Heft x1.) it was olive-grey.

NUDIBRANCHIATA. Dao

Gen. Platydoris B.

Flat, oval animals of a fair size and peculiar hard, leathery texture. The back is finely
granulated and rough to the touch: the mantle margin ample: the foot grooved in front
and the upper lamina deeply notched. The branchial opening is generally stellate: the
branchiae are tripinnate and usually five, six or seven in number. There is no _ labial
armature or rhachidian tooth. The radula consists of numerous hamate laterals, the outermost
of which are sometimes finely serrulate. The reproductive organs have an armature of hard
scales, bearing hooks.

The Platydorids are recorded from the Indian Ocean, E. Africa, the Pacific (including
the W. coast of America), the Mediterranean, Great Britain (Pl. planata), the West Indies,
and the Canaries. They may therefore be considered as cosmopolitan. They are frequent
on coral reefs under stones at low tide. They are exceedingly lethargic and are often found
wedged into the crevices of stones, the mantle having grown into the shape of the hollow
in a way which suggests that the animal has not moved for a considerable period. Several
of the species show the same phenomenon of self-mutilation as Discodoris.

16. Pl. eurychlamys B.

One specimen from the Kolumadulu Atoll, Maldives, 33 f The animal as preserved is
7 cm. long and 4cm. broad. The back, which is as usual flat and leathery, is blackish-
brown, with white spots, which are particularly numerous near the margin. The underside
dirty white, with black spots on the foot. The rim of the branchial pocket white; the
branchiae six, whitish with black rhachis. The numerous teeth are of the ordinary hamate
type; the outermost are very irregular in shape but not serrulate.

From Bergh’s descriptions of various specimens the colour appears to be very variable.

17. Pl. scabra Q. and G.

One small specimen from Minikoi 3°5 cm. long by 1:7 cm. broad. Not dissected. This
species, which I have frequently seen alive in the South Seas, is hard and leathery and
grows to be three or four inches long. The mantle is wavy and indented. The labial
tentacles small, white and tapering. The rhinophore sheaths somewhat projecting. The colour
of the back is white with irregular blotches of brown formed by aggregations of minute
spots. The under surface yellowish-white. The branchial aperture has six lobes and presents
a star-like appearance. The branchiae are six, yellowish, very voluminous and delicate’.

18. Platydoris sp.

Two specimens from the Maldives, Hedufuri, Mahlos Atoll. The largest is 2°2 cm. long
by ‘9 cm. broad, the form being somewhat elongate. The back is finely granulated, bluish-
white, with traces of yellowish-brown mottlings in the centre. Round the mantle margin

1 Bergh in S. R. Hefte x1., xvr., xvit., and Supp. Heft r.,
and Sammlung, Plate, Opisthobranchiata.

2 As there seems to be no description of this species except
Quoy and Gaimard’s brief notes, I add the following par-
ticulars taken from specimens examined in East Africa. The
living animal is as much as 10 em. long, very stiff, yellowish-
white, with irregular violet mottlings of varying intensity,
formed mostly by minute rings of the same colour with
yellowish centres or by spots. The under side is a clear
yellowish-white, with no markings except at the sides of the

foot which are mottled like the back. The branchiae are
light yellow with grey axes: the rhinophores, buccal mass
and viscera all bright, light yellow. The foot small and
narrow, grooved in front and notched, but not deeply: the
upper lamina thick. No trace of labial armature. Radula
about 48x60.0.60. Teeth simply hamate: innermost
smaller, two or three outermost irregular. Male branch of
reproductive organs with the characteristic hook-bearing
scales apparently set in four rows: female branch with
strong folds.

71—2

556 SIR CHARLES ELIOT.

is a row of purplish spots and the sole and sides of the foot are streaked with the same
colour. The margin of the mantle is undulated. The rhinophore sheaths not much raised
and with smooth margins. The branchial opening roundish with indistinct lobes and con-
taining five small tripinnate branchiae. No labial armature. The radula consists of simply
hamate teeth, smaller towards the outside but not denticulate. The verge is armed with
hamiferous discs.

Not having access to the descriptions of some recorded species of Platydoris, I cannot
give an opinion as to whether this form is new or not. I lay no stress on the branchial
opening being roundish. This is not unfrequently the case in alcoholic specimens, although
the opening may be typically stellate.

Gen. Dictyodoris Bergh’.

The texture of the body is leathery, the dorsal skin smooth. Foot not grooved in front
or only slightly. Branchial aperture round: branchiae few, bipinnate. Tentacles digitiform.
No labial armature. Radula of simply hamate teeth with naked rhachis. Reproductive
system unarmed. The two known species have reticulate figures on the back and only four
branchiae. In D. tesselata the outermost teeth are pectinate, in D. maculata not.

19. D. maculata n. sp.

One specimen from the Faimu-Inguradu Channel, N. Mahlos, Maldive Islands, 26 f.
Nearly round in form: length 8 mm., breadth 6 mm. Colour bluish-white, with blackish spots
which seem to be under the skin. There are several black spots of a more ordinary ap-
pearance on the under edge of the mantle and a black rim running round at the junction
of the foot and body. On the dorsal surface is a reticulate pattern formed by ridges,
but their conspicuousness is due rather to a difference of colour than to their real
prominence, which is small. I could not distinguish any groove in the anterior part of the
foot. Branchial aperture small and round: 4 bipinnate whitish gills, the posterior pair being
smaller than the others: the main and secondary axes are marked by a broad black line.
Tentacles digitate: no labial armature: radula fairly broad: no central tooth: laterals simply
hamate, the innermost small, the outermost irregular, sometimes bifid or trifid but not
consistently pectinate. Central nervous system somewhat concentrated, though not as in
Asteronotus, two pairs of ganglia being distinguishable.

This specimen bears a striking resemblance to Alder’s drawing of Doris Incii (reproduced
in Gray’s Figures of Molluscous Animals, 1850, Tab. 226, figs. 1 a, b, c), which, as Bergh
points out, may represent either a Dictyodoris or Sphaerodoris. As however we have no
account of the anatomy of this animal, it can never be possible to certainly attribute it
to either genus. In spite of the strong external similarity of the present Dorid to Alder’s
drawing I do not propose to call it D. Zncwi because the figure has (1) indistinct but
apparently simply pinnate branchiae, (2) no tentacles, (3) starlike apertures for the rhino-
phores, (4) no spots or dark rim on the under surface. The genus Dictyodoris appears to
be closely related to Halgerda, Bergh.

1 Bergh, S. R. Sup. Heft 1. 1880.

NUDIBRANCHIATA, 557

Gen. Staurodoris Bergh}.

This genus was established by Bergh for the Doris verrucosa which is common in the
Mediterranean and is characterized by him as follows. The body is flat and tuberculate:
the retractile rhinophores and branchiae are protected by valve-like tubercles. The branchiae
are simply pinnate. There is no labial armature or median tooth: the radula consists of
numerous simply hamate teeth and the reproductive system is unarmed. If this definition
of the genus is rigidly adhered to, the form described below must be made a new genus,
as the rhinophores and branchiae are not protected by valves and the branchial opening is
crescent-shaped and very peculiar. Von Jhering however describes under the name of
Staurodoris pseudoverrucosa an animal obtained at Naples in which there were no valves
round the gill cavity?, It seems therefore simpler to somewhat enlarge the definition of
the genus than to create a new one. We may therefore characterize Stawrodoris as belonging
to the large class of Dorids which have no labial armature or median tooth but a simply
hamate radula and as distinguished from the other genera by its tuberculate back and
simply pinnate branchiae. Perhaps to these characters may be added the shape of the
tentacles, which appear as two ridges above the mouth®*.

20. Staurodoris pecten nov. sp.

Two specimens from N. Male Atoll, Maldives, 27 7 The animals are nearly round, the
larger specimen being 13cm. by 1 cm., and flat. The dorsal integument is full of rod-like
spicules and studded with small conical warts or papillae, which are not connected or confluent.
The colour is quite uniform, in the one specimen yellowish-brown, in the other faded yellow.
The foot is broad (1 em. by 0°7 em.) and slightly grooved in front, the upper lamina being
notched. The rhinophores emerge in the middle of a wart, but can hardly be said to be
protected by distinct valves. The gill opening is in the shape of a crescent, the horns
pointing anteriorly. The crescent itself acts as one large valve, but its lips are not pro-
tected by any special tubercles, and when it is closed and the branchiae completely withdrawn
it is hardly distinguishable from the rest of the dorsal surface. The branchiae are disposed
in three groups. The anterior one consists of a single simply pinnate leaflet: those to the
right and left consist each of six similar leaflets, united at the base so that they might
be regarded as bipinnate. The central nervous system is much concentrated and the com-
missures very short. The tentacles seem to be ridges above the mouth, very slightly grooved
externally. There is no labial armature. The radula consists of simply hamate teeth, not
denticulate, and uniform except that the innermost and outermost are somewhat smaller than
the rest. There is no central tooth. The reproductive system is unarmed.

It is very doubtful whether this species should be regarded as a Stauwrodoris or the
type of a new genus. Probably new forms belonging to this group will be discovered and
have to be distributed between the genera Stawrodoris and Echinodoris. The latter though
strongly characterized externally by its greatly elongate papillae is really closely allied to

1 Bergh in S. R. Heft xim. Supp. Heft 1. and Supp. Heft 3 Since writing the above more than a year ago, I have
1. and in Bull. of Mus. of Comp. Zool. Harvard, 1894, pp. examined in East Africa several forms referable or allied to
160-2. Also von Jhering, Jahrbiicher der Deutschen Malak, Staurodoris, which may necessitate a further modification of
Gesellschaft, 1886, Heft m1. pp. 230—233. this definition. I postpone the discussion of their affinities
2 It had the further peculiarity of having bipinnate gills. to another paper.

358 SIR CHARLES ELIOT.

Staurodoris, from which it differs chiefly in its denticulate teeth, penial armature and _ bi-
pinnate gills. But as seen from the present specimens it is not always easy to say whether
gills are simple or bipinnate.

Collingwood (On some New Species of Nudibranchiate Mollusca from the Eastern Seas)
gives notes on a Doris pecten from Formosa. His description is not sufficiently full to
decide to what genus the animal should be referred but, as it has a crescent-shaped row
of simply pinnate branchiae and the other characters are not inconsistent with those of the
present specimens, I have indicated its possible identity in the proposed specific name. No
species of Stawrodoris is recorded with certainty from the Indo-Pacific, but Abraham’s Doris
pustulata seems closely related, though it has bipinnate branchiae.

Gen. Trippa Bergh’.

This genus, which seems identical with Phlegmodoris Bergh*, comprises a few tropical
cryptobranchiate Dorids with a soft tuberculous body and no jaws. The characters are as
follows. Body flattish, soft, sometimes gelatinous: the dorsal surface covered with tubercles,
which are generally compound and support smaller projections. Tentacles small. Foot
broadish with a shallow groove in front, the upper part united to the head. The branchial
rosette far back at the end of the body composed of five or six tripinnate plumes. No
labial armature: radula composed of uniform simply hamate teeth: a number of glandular
bodies (? glandulae ptyalinae) are inserted in a circle round the posterior extremity of the
oral tube. The reproductive system is unarmed. In the Opisthobranchier d. Sammlung Plate
(p. 527) Bergh apparently drops the character of the foot beg connected with the head.

21. TF. ornata B.

Recorded hitherto from the Philippines and the Malay Archipelago. The present col-
lection contains two specimens from Minikoi marked “mottled yellowish-brown and white,”
which agrees fairly well with Brock’s coloured sketch of the animal. The preserved specimens
are of a purplish grey; the dorsal tubercles somewhat darker, though some on the other
hand are quite white. The foot somewhat lighter. They are both about 3:5 cm. long and
2em. broad. The back is flattish and covered with numerous small tubercles, which when
closely examined are seen to be covered with little lumps. The mantle fairly wide at the
sides and over the head, with an irregularly marked dark border. The rhinophores com-
pletely retracted, but as far as can be judged bent backwards in the living animal. The
branchial orifice set far back, shallow, the branchiae being exposed in both specimens. They
are five in number and tripinnate, the secondary pinnae being long and the tertiary short.
The anal papilla is central, large and long. The radula is small and white and consists
of twenty rows (only one specimen examined), each of which contained from 32 to 37 teeth
on each side of the rhachis. The other characters are as for the genus. Except for some
small differences in the radula the specimens correspond very well with Bergh’s description.

1S. R. Heft x11. 1877, Heft xvm. 1890, Opisthobranchier 2S. R. loc. cit. Heft xm. Bull. of Mus. of Comp. Zool.
der Samml., Plate, p. 527. Harvard, 1894.

NUDIBRANCHIATA, 559

Fam. Doridopsidae’.

The Doridopsidae externally much resemble the cryptobranchiate Dorididae but differ
in their internal structure, particularly in the buccal parts. The body is soft, often gelatinous:
the back either smooth or covered with soft warts and tubercles. The rhinophores perfoliate
and retractile. The branchiae retractile into a common pocket, generally rather large, few
in number and tri- or quadripinnate. The foot broad, grooved in front with a notch in
the upper lamina, not extending behind the mantle. Tentacles small, not projecting freely
and resembling two folds, sometimes hardly distinguishable. The mouth is a fine pore,
usually situated in the anterior cleft of the foot: there is no trace of jaws, radula, or labial
armature and the buccal organs consist of a slender suctorial tube which can be protruded
as a proboscis through the mouth pore. Beneath the anterior part of this tube are two
large folliculate glands called ptyaline. The true salivary glands appear to be represented
by two nodules at the commencement of the oesophagus. The stomach is surrounded by
the liver, which is bifid behind. The nervous system is extremely concentrated. On the
interior dorsal wall of the pericardium are a number of folds, sometimes called the pericardial
gill, though their respiratory functions are not clear. The hermaphrodite duct is spread
over the liver and there are two spermatothecas. There is an armature of minute hooks
on the spermatic duct and the glans.

The Doridopsidae are clearly allied to the cryptobranchiate Dorids by their retractile
gills and general structure, but can usually be distinguished externally by their soft con-
sistency and pore-like mouth in the anterior pedal groove. The suctorial mouth allies them
to the Phyllidiadae, especially Phyllidiopsis, and the details of their anatomy (the concentrated
nervous system, the pericardial lamellae and the penial armature) present remarkable re-
semblances to that family. Whether they are phylogenetically more akin to the Dorids or
the Phyllidiadae can only be proved, if ever, by the discovery of connecting links, but the
present evidence seems to bring them nearer to the former than to the latter group. They
differ from the Dorids mainly in the buccal parts, which may be easily modified by diet,
but from the Phyllidiadae in their general shape, texture and respiratory organs. The
concentrated nervous system and reproductive armature are found in several Dorids and it
is not clear that the pericardial lamellae are a character of importance.

The family contains more than 60 species arranged by Bergh in two genera, Doriopsis
and Doriopsilla. The latter, which contain only one certain species, is characterized by the
body being hard and by the buccal ganglia being situated at the anterior not the posterior
part of the buccal tube. The genus Doriopsis (or Doridopsis), though apparently not found
in the colder seas is probably cosmopolitan in warm waters, being recorded from the
Mediterranean, the West Indies, and most parts of the Pacific and Indian Oceans. The
animals are generally abundant under stones on coral reefs. Their food is unknown. In
captivity I have never seen them take nourishment, though they crawl over stones and
sponges as if seeking for something to eat.

1 Pease, Proc. Zool. Soc. London, 1860, Amer. Journal Deutschen Malac. Gesell. v1. 1879. Id. Die Doriopsen des
Conchol. v1. 1871. Hancock, On the Anatomy of Doridopsis, Mittelmeeres, ib. vu. 1880. Id. Challenger Report on Nudi-
1865. Alder and Hancock, Notes on a Coll. of Nud. Moll. branchs, 1884. Id. in S. R. Heft x. Supp. Heft 1., Heft xv.,
Jrom Madras, 1863. Bergh, ‘‘Neue Nacktschnecken der xvi.,xvu. Id. Die Nudibranchiata holohepatica Porostomata,

Siidsee,” Journal Mus. Godeffroy, Heft vii. 1875, Heft xtv. Wien, 1892.
1878. Id. ‘‘Die Doriopsen des Atl. Meeres,” Jahr. der

560 SIR CHARLES ELIOT.

22. D. nigra Stimpson, var. nigerrima B.

Three specimens from the Maldives, Mulaku Atoll 40 f, and Mahlos Atoll 20 f. The
largest is 1'7 cms. long and “8 cm. broad. The colour is a deep bluish-black; the foot,
especially the margin, is somewhat lighter and on the back are some inconspicuous spots
of bluish-grey, looking as if the black pigment had been rubbed off. But there are no
signs of white spots, a coloured border or of white on the rhinophores or gills. The latter
are six and tripinnate. The internal organs are as usual, but even here the black tint
prevails.

This species shows great variety of coloration and in my experience the pure black
form is the rarest. At Zanzibar I have found in considerable abundance (1) a black form
with large white spots and no border, (2) a purplish form with numerous small spots and
a border of the same colour but lighter, (8) a black form with or without white spots and
with a bright red border, (4) a greyish-white form without markings; found in the company
of (8) and apparently differmg only in coloration.

23. D. apicalis? B.

One specimen from the Suvadiva Atoll, Maldives, 42 f Length 5-4 cm., breadth across
the rhinophores 1:1 cm., across the gills 233 em. The form is thus unusually long and
tapering. The back is smooth and the mantle edge narrow. The colour uniform dirty grey
with some whitish mottlings on the back but no dark spots or markings. Foot, dirty yellow,
smaller before than behind. Rhinophores small, straight with about 20 perfoliations. Gills
six, tripinnate. The blood-gland appears to be double and the intestine gives off a small
diverticulum near the pylorus. This last character and the elongate tapering shape of the
animal incline me to identify it with D. apicalis B., recorded from Amboina. But the colour
is not quite the same, the shape is not peculiar to this species but found elsewhere (e.g. in
D. vidua) and it remains to be seen whether the diverticulum is a consistent specific character.

Fam. Phyllidiadae’.

The genus Phyllidia was created by Cuvier and Fryeria by Gray. Our knowledge of
the group is however mainly due to the writings of Bergh’. The members of the family
are perhaps the most easily recognizable of Nudibranchs and can at once be identified by
their curious leathery consistency and the brightly coloured lumps or ridges which usually
adorn their dorsal surface. They may be more accurately characterized as follows:

The texture of the animals is coriaceous and the form flattish, elongate oval. The
margin of the mantle covers the foot all round, though the latter is fairly broad. There
is no distinct head but merely a minute oral pore beneath two small digitate tentacles
which are united at the base so as to form a rudimentary veil, and are slightly furrowed
on the outer side. Seen from above, the shape of the anterior and posterior extremities is
much the same. The dorsal surface bears tubercles or ridges, which are generally brightly

1 Bergh: (1) ‘Bidr. til en Monogr. af Phyllidierne,” (4) Neue Beitriige zur Kenntniss der Phyllidiaden, Wien,

Naturhist. Tidsskr. 3 R. 1869. 1876 (aus den Verhandl. der k. k. zoologisch-botanischen
(2) ““Neue Nacktsch. der Siidsee,” Journal des Mus. Gesellschaft in Wien besonders abgedruckt).

Godeffroy, Heft 1. 1873. (5) Die Nudibranchiata holohepatica Porostomata, Wien,

(3) Malacol, Untersuch. in Semper’s Reisen, Heft x. 1892 (do. do.).
1876, Heft xvr. 1889 and Heft xvir. 1890.

NUDIBRANCHIATA. 561

coloured. The rhinophores are perfoliate and completely retractile. They usually emerge from
two of the anterior tubercles. Under the mantle edge are a large number of branchial
lamellae which form a more or less complete circuit interrupted only by the mouth and
genital orifice and in Fryeria by the vent. The mouth parts consist of a minute poriform
aperture leading into a narrow tube, which passes into a large mass of glands that impart
a thick conical appearance to the anterior portion of the digestive tract (except in
Phyllidiopsis). Behind this glandular mass the pharynx reemerges as a thin tube, which
subsequently dilates and enters the liver. There is no exsertile proboscis as in Doridopsis.
The vent is usually dorsal and perforates one of the posterior tubercles. Only in Fryeria
is it under the mantle rim. There are no jaws, radula or buccal armature of any kind.
The liver is not divided posteriorly. The verge is armed with rows of hooks. As in the
Doridopsidae the dorsal wall of the pericardium is furnished with a series of lamellae, and
the nervous system is much concentrated. A single blood-gland is present.

The Phyllidiadae are Dorids in virtue of their unbranched liver, their two spermato-
thecas, their retractile rhinophores, and their usually dorsal vent. The suctorial mouth, the
concentrated nervous system and the so-called pericardial gill ally them to the Doridopsidae,
but still the group Porostomata proposed for the two families by Bergh seems, as he himself
admits, somewhat artificial, for the absence of the ordinary branchial rosette and the position
of the branchial lamellae under the pallial margin separate the Phyllidiadae decisively from
all Dorids except the Corambidae. From that family however they are equally decisively
separated by the structure of the mouth parts and the general habitus.

The family is divided by Bergh into four genera, Phyllidia, Phyllidiella, Fryeria and
Phyllidiopsis, Fryeria is sharply distinguished from the others, the vent not being dorsal
but terminal and under the mantle rim. Phyllidiopsis is also well characterized, as the
mouth parts resemble those of Doridopsis and the anterior part of the oral tube is not
surrounded by a large glandular mass as in the other genera. The distinction between
Phyllidia and Phyllidiella is less satisfactory. The former is said to have the dorsal
tubercles arranged in longitudinal rows and the latter in groups or quincunxes. But there
is much variety in respect of this arrangement. I have seen specimens of Phyllidiella nobilis
in which the tubercles appear to be arranged in rows, and the present collection contains
a specimen which has all the characters of Phyllidia varicosa, but in which there are only
seven tubercles scattered irregularly on the back and certainly not arranged in lines. In
dissecting a considerable number of specimens I have not been able to observe the difference
noted by Bergh in the mouth parts; in Phyllidia forma tubi oralis glandulé ptyalina obtecti
symmetrica and in Phyllidiella asymmetrica. The remarkable Phyllidiopsis papilligerat, which
has elongated papillae on the back, seems worthy to be the type of a new genus.

The Phyllidiadae are abundant in the Red Sea, Indian Ocean, the Philippines and
Malay Archipelago, Polynesia and East Africa. They are not recorded from the more northern
or southern parts of the Pacific, from the West Coast of America, North or South, from
the Mediterranean or Atlantic, except Phyllidiopsis papilligera from the Gulf of Mexico.
They seem therefore to be tropical forms and so little is known of the Nudibranchs of the
equatorial Atlantic that it is probable they will be found there. They appear to be mainly
littoral and inhabitants of coral reefs. Hardly anything is known of their manner of life,

1 Bergh, Bull. of the Mus. of Comp. Zoology of Harvard posed the genus Ceratophyllidia to include this form and
College, x1x. 3.1890. Since writing the above, I have pro- another which I have discovered in East Africa.

Galle 2

562 SIR CHARLES ELIOT.

but from the extreme narrowness of the mouth pore and the oral tube it may be concluded
that they live by suction, probably on animal juices. Most authors repeat the statement of
Quoy and Gaimard that they are extremely sluggish and almost motionless. I cannot support
this from my own observations, as Phyllidia varicosa when kept in a basin crept continually
round and round as actively as any Dorid. Collingwood (Nudibranchiate Mollusca from the
Eastern Seas, 1878) describes the spawn as a long irregular narrow ribbon of a whitish
colour. When touched the skin of most Phyllidiadae secretes a liquid with a disagreeable
smell.

Gen. Phyllidia Cuv.

Vent dorsal: tentacles free. Tubercles in centre of back arranged in longitudinal lines:
Phyllidia sensu stricto. Tubercles in centre of back not arranged in lines but in groups:
Phyllidiella.

24. Phyllidia varicosa Lamarck.

Back with at least three median lines of tubercles, more or less confluent and forming
united ridges. Lateral tubercles varying, but generally two or three form an irregular line
or ridge running in from the mantle edge to the central dorsal area. Form somewhat
elongate; foot broad, darkish in colour, with a distinct black line down the middle. Four
specimens from N. Male and Hulule, Maldives.

The colour of this species, as well as the disposition of the tubercles, seem very variable.
The living animals, which I have seen at Zanzibar, were of a slaty-blue with brilliant orange
tubercles. The specimens preserved in the present collection’ show three distinct varieties.
First, one magnificent specimen from the Maldive Islands, 6 cm. long; the smooth parts of
the back are of a brilliant black, the projections lilac, but considerably varying in intensity.
The median tubercles are fused into three distinct ridges, not uniting anteriorly or posteriorly,
the tubercles appearing merely as peaks on the ridge. Lateral tubercles more or less fused
into ridges, but irregular. Secondly, three specimens from the Maldive Islands from 3 to 4 em.
in length. Tubercles and ridges much as in the previous variety: colour black and dirty
white, probably representing blue and yellow in life. Thirdly, six specimens from Rotuma
with much the same coloration as the last variety. The tubercles are not fused into ridges,
but are sometimes quite separate, sometimes connected by a low black elevation. Five of
the specimens show a tendency to have the tubercles arranged in seven longitudinal lines,
one median and three lateral on each side. In one specimen this arrangement is quite
symmetrical and complete. In the others the outer lateral rows are confused and indistinct.
The sixth specimen, as already mentioned, is quite abnormal and has merely seven large
and indistinctly compound tubercles scattered over the back, though its other characters appear
to be those of the species in every respect.

25. Phyllidia elegans B.

One specimen from Kolumadulu Atoll, Maldives, 40 7 Form elongate, 2 cm. long.
Tentacles short and thick. Foot grey, with a thick black line down the centre. The ground
colour of the back appears to be black but there are eleven patches of a lighter colour
(in the preserved specimen dirty white), two median, one anterior, one posterior, four on
the right side and three on the left. They are studded with low confluent tubercles, of

1 Including six from Rotuma in the South Pacific.

NUDIBRANCHIATA. 563

very irregular shape. Those on the median patches are arranged in three not very distinct
lines.
Gen. Fryeria Gray.

The shape of the animal and arrangement of the tubercles much as in Phyllidia but
the vent is terminal, beneath the mantle edge, interrupting the branchial circuit.

26. Fryeria Riippellii Bergh.

One specimen from the W. side of Fadifolu Atoll, Maldives, 70 f, 15cm. long. The
form and mouth parts are much as in Ph. varicosa. There are three longitudinal lines of
tubercles on the back, two of which are fused into ridges, while those of the left-hand line
are mostly separate from one another. There are also twelve lateral groups of ridges, running
inwards from the mantle edge, one anterior, one posterior, and five on each side. The ground
colour is blackish, and the tubercles and ridges dirty white, perhaps representing blue and
yellow in life. The foot and the rhinophores are yellowish. The animal is undoubtedly a
Fryeria, and allowing for the variability of markings so common in this family corresponds
fairly well with Bergh’s description of Fr. Riippellii.

Gen. Phyllidiopsis Bergh.

External appearance much as in Phyllidia but the tentacles are not free but attached
along their entire length. The buccal apparatus resembles that of Doridopsis, that is to
say it is not involved in a glandular mass. The oral pore gives entrance to a fairly wide
chamber with laminated walls, into which opens the anterior portion of the alimentary tube
which is cylindrical and receives two small salivary glands. Beneath it is a single gland
probably corresponding to the ptyaline gland of other Nudibranchs.

27. Phyllidiopsis striata ?

One specimen from Haddumati Atoll, Maldives, 40 f Mouth parts as described above.
The upper surface of the preserved specimen is blackish, with three white lines down the
centre and two white patches one at each extremity. Near the edge are some irregular
white markings. There are no tubercles or ridges. The foot and branchiae are somewhat
lighter than the back. The form is long and narrow, 14cm. by 5 mm. broad and very flat.

The markings of this specimen do not entirely correspond with Bergh’s figure (Semper’s
Reisen, Xvi. Plate 84, fig. 23) of an individual captured by Moebius at the Ile aux Fouquets,
Mauritius, but the differences do not seem sufficient to warrant the creation of a new species.

CLADOHEPATICA.

The Nudibranchiata Cladohepatica are characterized by the liver being ramified or at
least divided. The dorsal surface generally bears appendages, into which the ramifications
of the liver pass. As a rule, well-developed mandibles and a radula are present. In the
more specialized forms the latter is frequently reduced to a single row of teeth. There is
no blood-gland and only one spermatotheca.

The numerically largest family of the Cladohepatica, which also shows all the character-
istics of the division in a typical form, is that of the Aeolididae, but there are a number

72—2

564 SIR CHARLES ELIOT.

of other families obviously related to Tritonia. These mostly show an advance in the
direction of the Aeolid type, but do not seem capable of arrangement in a_ progressive
series culminating in Aeolis, as most of them have marked peculiarities of their own which
are not found in Aeolis or other families. Such are the Lomanotidae, Dendronotidae,
Scyllaeidae, Bornellidae, Tethymelibidae, Phylliroidae, Hedylidae and Pleurophyllidiadae. The
first six of these are not represented in the present collections, which is somewhat remarkable,
as Melibe, Scyllaea and Bornella are all common in the Indo-Pacific.

Fam. Pleurophyllidiadae.

These are animals of a remarkable external appearance, which it is easier to recognize
than describe. The form is elongate—oval or tongue-shaped. The back is somewhat coriaceous
and bounded by a distinct mantle rim which extends beyond the foot. The dorsal surface
is marked with longitudinal stripes or tubercles. The head is distinct and divided from
the body by a space which may be conveniently called the neck. In front of this neck
is a tentacular shield, the corners of which are usually drawn out into points. Behind the
neck and immediately under the anterior margin of the mantle are two small perfoliate
rhinophores. In front of them are often a wart (caruncula nuchalis) or several papillae.
Except in Plewroleura there is a branchial apparatus. Behind the head and underneath the
mantle a cleft runs inwards in which are set a number of branchial lamellae, and the sides
of the body underneath the mantle edge also bear another series of lamellae. The vent,
genital and renal openings are all on the right side. Jaws strong and denticulate. Radula
with a strong median tooth, cusped and denticulate. Laterals varying in form, denticulate
or not. Liver much ramified, sending branches to the back and sides, which penetrate into
the lateral lamellae. On the margin of the mantle are cnidopores, often visible to the
naked eye, from which cnidae can be emitted. The reproductive system is unarmed.

The family appears to be cosmopolitan, but is recorded in especial abundance of species
from the Indian Ocean. The systematic position of the Pleurophyllidiadae has given rise
to some discussion. The older naturalists created a division, Inferobranchiata, to contain this
family, the Phyllidiadae and Hypobranchiata (= Corambe), but it is now generally admitted
that this association of forms is artificial and untenable. The presence of cnidae and the
ramification of the liver are points of connection with the Aeolids, and Pelseneer (Recherches
sur divers Opisthobranches, 1893) actually places the Pleurophyllidiadae among the Aeolidioidea.
But their dentition allies them to the Tritonids, and they have distinctive peculiarities which
separate them from both Aeolids and Tritonids. No connecting links have yet been discovered
which throw light on their relations to other groups.

Gen. Pleurophyllidia Meckel?.

Pleurophyllidia possesses lamellae under the mantle both at the sides and in a cleft.
It is distinguished from the other genera which possess them by the following characters.

1 Bergh, Bidrag til en Monogr. af Plewrophyllidierne, Kenntniss der Nudib. der brasilianischen Kiiste,” Jahrb. d.
1866. Id. S. R. vr. 1874. Id. Beitriéige zur Kenntniss der Malaco. Gesel. xm. 1886. Eydoux et Souleyet (under name
Japanischen Nudibranchien, 1880. Id. Weitere Beit. zur Diphyllidia, Voyage de la Bonite, 1841).

Kenntniss der Pleurophyllidien, 1890. von Jhering, ‘ Zur

NUDIBRANCHIATA. 565

The body is elongate. The mantle edge is continuous in front: the rhinophores are placed
close together: on the neck in front of the rhinophores is a tubercle (caruncula nuchalis)
or a series of papillae. Cnidocysts are present. The genera Linguella Bly. (= Sancara Bgh.)
and Camarga Bergh are closely related and are perhaps only subgenera.

About 20 species of Pleurophyllidia are recorded from the N. Atlantic, Brazil, Mediter-
ranean, Indian Ocean, E. Africa, Malay Archipelago, Polynesia, and California. They generally
inhabit mud or sand and their peculiar shape seems correlated with a burrowing habit. I
have observed Pl. californica in captivity on the West Coast of N. America. It buried
itself in the sand with the head and tail exposed and bent upwards, so that the body
assumed a crescent-like shape. A continuous current of water was kept passing from the
front backwards under the mantle, the edge of which was apparently applied to the sides
so as to form a tube over the lamellae. At night however the animals became very lively
and moved rapidly round and round the glass dish in which they were confined.

The species of Pleuwrophyllidia are more distinct than is common among Nudibranchs.
A useful synopsis will be found in von Jhering’s paper cited above. They may be divided
into the species which have tubercles on the back (Pl. pustulosa and marmorata), and the
larger number which have longitudinal stripes. This latter class may again be divided into
those which have papillae on the neck and those which have merely one more or less well-
developed tubercle before the rhinophores. The several species have usually a characteristic
dentition. Another point of difference is that some species have numerous separate lamellae
at the sides of the body, while others have these lamellae united into three (or in Pl.
pallida into only one) continuous bands.

28. Pl. gracilis Bgh.

One mutilated specimen from Minikoi seems referable to this species, which was in-
stituted by Bergh for two individuals found at Bombay (S. R. vi. p. 250). There was a
large rent which had carried away much of the back, most of the left side and nearly all
the internal organs. The length is 23cm. but the body was somewhat bent: the breadth
‘Sem. What remained of the dorsal surface was almost colourless, with the remains of
numerous yellowish longitudinal bands and occasional traces of brownish pigment. The neck
was thickly studded with conical colourless papillae. The gills were fairly numerous and
of different sizes. The lamellae on the mght side of the body formed three continuous
crinkled and sometimes indented lines. The uppermost extended nearly the whole length
of the body: the second was shorter, and the third shorter still. The jaws were strong,
the cutting edge armed with numerous rows of minute serrulations. The rhachidian tooth
of the radula was as usual squarish: there were 7—8 denticles on each side of the central
cusp and sometimes two more near its point. The first lateral was also squarish, denticulate
on the outer and bearing traces of minute serrulations on the inner edge. The subsequent
laterals are denticulate on the outer edge only, and these denticulations decrease as the teeth
are further from the centre. The last 10 or 15 teeth are simply hamate and the outermost
small and rudimentary.

It is risky to identify a mutilated specimen, but the essential characters of Pl. gracilis
seem to be (1) numerous nuchal papillae, (2) small outermost teeth, and the present animal
possesses both.

566 SIR CHARLES ELIOT.

Gen. Pleuroleura Bgh? (= Dermatobranchus van Hasselt).

The genus Pleuroleura, which is practically Plewrophyllidia without gills or side lamellae,
is made by Prof. Bergh the type of a separate family. But as it differs from Plewro-
phyllidia only in the point mentioned and agrees with it in numerous distinctive peculiarities,
it seems better to include both in one family, like Tethys and Melibe, of which Tethys has
gills while Melibe has none’.

The history of the genus is somewhat remarkable. Van Hasselt described three species
of Dermatobranchus in 1824. In 1874 Bergh created the genus Pleurolewra for a fourth
species (ornata) and held that there was not sufficient evidence to identify it with Dermato-
branchus as van Hasselt’s specimens were lost. He subsequently discovered them in the
Leyden Museum and in 1888 (J.c.) expressed the opinion that the identity was practically
certain, but maintained the name Pleuroleuwra all the same. By the rule of priority there
can be little doubt that the genus should be called Dermatobranchus, but that rule is not
an end in itself but a means to an end, namely to find one name for one genus or species.
When two names have approximately equal claims the earliest ought certainly to be taken.
But when hardly anything is known of the animal under the earlier name and _ practically
all our knowledge comes from descriptions published under the later name, it appears to
me that the principle at stake is not sufficiently important to outweigh the inconvenience
of resuscitating the original designation, a reference to which would give an enquirer very
little information.

Pleuroleura strongly resembles Pleurophyllidia in appearance, but the known species
seem somewhat smaller and have no warts or papillae on the neck. The tentacular shield
is perhaps not always so distinct as in the other genus. The rhinophores are set close
together, but emerge from separate short sheaths and bear longitudinal folds. The sides of
the body bear no appendages, and there are no branchial clefts. The jaws, radula, digestive
and reproductive apparatus are all similar to those of Plewrophyllidia. The mantle margin
is perforated by enidopores.

Bergh gives as a family character of the Pleuroleuridae, Notaeum antice in nucham
transiens. If I am right in supposing this to mean that the mantle margin is not con-
tinuous in front but interrupted and that the dorsal surface and neck are consequently
continuous’, the character is not possessed by the specimen here examined and apparently
not by Pleuroleura fortunata (Bergh, l.c. 1888, Pl. X. figs. 1 and 3).

Pleuroleura is recorded hitherto only from the Philippines, the Malay Archipelago, East
Africa, and Spitzbergen*. Nothing whatever is known of its habits.
29. Pleuroleura striata (van Hasselt).

Mr Gardiner furnished me with a single specimen found by the Skeat Expedition in
Pulau Bidang. It is externally well preserved but very hard and brittle. The form is rather

1 yan Hasselt, Allg. Konst en letterbode, 1824, 2. 3. 4. has no branchial clefts but side lamellae only.

Bergh, S. R. v1. p. 276 ff. 1874. Id. ‘* Die Pleuroleuriden,” % This appears to be the case in Linguella, vy. Bergh,
Zool. Jahrbiicher: Abth. fiir Systematik, 3 Band, 3 Heft, Anatomisk Undersggelse af Sancara quadrilateralis, Plate x11.
Jena, 1888. figs. 1 and 5.

2 T have since discovered in East Africa a form which I 4 Pleuroleura Walteri, Krause, ‘‘ Moll. von Ost-Spitz-

call Pleurophyllidiella and which seems to be a connecting bergen,” in Zool. Jahr. Jena, 1892, p. 366.
link between Pleurophyllidia and Pleuroleura inasmuch as it

NUDIBRANCHIATA, 567

short and broad, 9°5 mm. long, 6 mm. broad across the back: foot 3mm. broad. The colour
of the dorsal surface and of the tentacular shield is blackish. Down the middle of the
back runs one fairly straight line. On each side of it are eleven or twelve other lines,
not parallel to it, but arising with it at a point behind the rhinophores and going off
towards the mantle edge at various angles in a posterior direction. All these lines as well
as the borders of the mantle of the tentacular shield are raised, very distinct, and of a
whitish-yellow, which contrasts strongly with the ground colour. Numerous ecnidopores were
distinctly visible on the mantle edge. The anterior corners of the foot are produced. The
anterior mantle edge is continuous, but deeply notched above the rhinophores.

The jaws are as usual in the genus, with several (at least six) rows of denticles on
the cutting edge. The radula of about 20 fully developed rows. The rhachidian tooth of
the usual shape with a long central cusp and about 6 denticles on each side. The first
lateral squarish, with about 5 denticles on the outer side. The other laterals were 9—11
in number in the various rows and, as far as I could see, not denticulate. The inner are
broad and somewhat clumsily shaped hooks: the outer narrower and of the ordinary hamate
type. The intestines were much hardened and little could be made out of them, but the
ramifications of the liver seemed discernible.

Van Hasselt’s description of his D. striatus is as follows’. “Corpore nigro, flavo fuscoque
variegatissimo: dorso mucoso sulcato, sulcis simplicibus longitudinalibus, rectis parallelis:
capitis et pedis margine flavo: tentaculis apice nigris: pallio antice emargimato. Long. pallii
0-011: lat. 0-003. Corporis totius 0-1. Hab. Anjer-baie (Sunda Strait) vulgaris.” This
description seems applicable to the present animal except that the dorsal stripes are not
strictly speaking straight and parallel. A comparison of several specimens might show some
variation in this respect, and as van Hasselt says the animal is common they may be
found in the future. Meanwhile, I see no reason why this animal should not bear the
name striata.

Fam. Dotonidae’.

The Dotonidae are closely allied to the Aeolididae in both external and internal structure.
The rhinophores however are simple and retractile into wide cup-shaped sheaths. The cerata
are arranged in two rows and as a rule are large, ovate, and at least some of them
compound. They contain ample and many lobed hepatic ramifications. As a rule there are
no anterior tentacles, but a rounded frontal veil. The vent is latero-dorsal. Cnidocysts are
absent. The jaws are weak and not denticulate. The radula uniseriate (as a rule): the
teeth with a central cusp and more or less distinct lateral denticles. The reproductive
system is unarmed.

The Dotonidae as a family are distinguished from the Aeolididae by one character
only—the rhinophore sheaths. The genus Doto is also marked by its ovate compound cerata
and its frontal veil, but a formation of the head analogous to the latter is also found in
Embletonia, and the nodules on the cerata of Capellinia and Proctonotus seem a_ step in
the direction of the composite condition. The other genera of the family are less well known
and differ from Doto chiefly in having some or all of the cerata simple and not compound.

1 I quote from Bergh’s quotation S. R. vi. p. 277, not Porto di Genova, 1881. Vayssiére, Recherches sur les Mol-

having access here (East Africa) to the original work. lusques Opisthobranches de Marseilles, 1888. Bergh, Beitr.
2 A. and H., Brit. Nud. Moll. Trinchese, Aeolididae del z. Kenntniss der Aeolidiaden, v1. Id. S. R. xvi.

568 SIR CHARLES ELIOT.

The Dotonidae are mostly minute animals and the forms described come chiefly from the
northern Atlantic and Mediterranean. At least one species however comes from East Africa,
and the family is probably cosmopolitan.

30. Doto indica Bgh. ?

One small specimen from Minikoi in the Laccadives. Length 7 mm., breadth 3 mm.,
colour uniform dirty yellow. There were apparently five pairs of cerata on the back, but
all have fallen off. One, however, which doubtless belonged to this specimen, was found
in the same bottle. It was fusiform and bore three rings of tubercles. The frontal veil
was ample and expanded at the side into lappets. The rhinophore sheaths with an uneven
indented margin: on the outside of each is an undulated membranous process. The foot
rounded in front, the tail short. The jaws weak. The radula uniseriate. The teeth of the
form usual in the genus. The central cusp not very prominent, with 3—5 irregular lateral
denticulations.

There is, strictly speaking, no proof that this specimen belongs to the genus Doto,
since the cerata are not preserved, but I think it probable that it is Bergh’s Doto indica
(S. R. xvi. p. 795) recorded from Mauritius. The identification is supported by the shape
of the rhinophore sheaths and the membranous process.

Fam. Aeolididae.

The Aeolids are limaciform animals with tentacles', rhinophores either simple or perfoliate
but always non-retractile and without sheaths, and simple dorsal cerata containing ramifica-
tions of the liver. Mandibles are always present, and the radula has never more than one
or three teeth in a row.

The Aeolids are the largest family of the Cladohepatica and comprise about forty genera.
It is therefore somewhat remarkable that they should be represented in the present col-
lection by only one small specimen. It has often been noticed that they are far more
abundant in Northern than in tropical waters—or at least in collections from the North
than in those from the tropics—but still the proportion here is unusually small. My own
experience after collecting continuously for a year in East Africa and Zanzibar is that the
genera of Aeolids are numerous but the individuals few and hard to get. Dorids can be
found under almost every stone on a coral reef, but Aeolids are obtained either in fairly
deep water or among seaweed where they are not easily seen.

Gen. Phidiana Gray’.

The characters of the genus are as follows. Body elongate and slender. Rhinophores
perfoliate. Foot rounded or truncated in front. The masticatory edge bears a single row
of denticles. Radula uniseriate. Verge armed with a hook.

Phidiana is one of the many genera of Aeolididae, which have an elongate body and
perfoliate rhinophores. It is separated from most of them by the fact that the anterior

1 The strange Pseudovermis Kovalewski (Mém. de U'Ac. de 2 A. and H., Notes on a Coll. of Nudibr. Moll. made in

S. Pét. Ser. vii. vol. x11. no. 4, 1901) is described as having India. Bergh, Beitr. z. Kenntniss der Aeolidiaden, 1. and v1.
neither tentacles, rhinophores, nor cerata. Id. Opisthobranchier der Sammlung, Plate.

NUDIBRANCHIATA. 569

margin of the foot is round or square and not produced into tentacular prolongations. In
its external characters it closely approaches Speirilla but differs in the jaws, teeth, and
reproductive organs. The genus is recorded from the Indian Ocean and several parts of the
Pacific, including the West Coast of South America, and from the Atlantic.

31. Phidiana unilineata (A. and H.)?

One small specimen from Fadifolu Atoll, Maldives, 25 f, imperfectly preserved. Length
9mm., breadth 25 mm. Colour uniform yellowish-white. Cerata partly lost, but apparently
arranged in seven indistinct oblique rows, leaving only a narrow space bare in the middle
of the back. Foot rounded in front. Tentacles long: rhinophores much shorter. Jaws long
and narrow, with a single row of rather distinct denticles. The radula consisting of about
25 teeth, bearig from 5 to 8 lateral denticles, two or three of which are not on the
basal portion but on the sides of the cusp itself. The penial hook was small but appeared
to be shaped as in Bergh’s figures of Ph. selencae.

This specimen perhaps belongs to A. and H.’s Ph. unilineata (l.c.) in view of the habitat
and of the fact that there is nothing in its external or internal characters which conflicts
with the identification. Still, the anatomical details given by A. and H. do not afford many
specific characters, and all trace of the colour has disappeared in this specimen. It is
therefore impossible to be sure of the identification.

Fam. Janidae Bgh

The Janidae closely resemble the true Aeolids, but are distinguished externally by the
cerata extending not only all round the sides and end of the back as usual but also in
front of the rhinophores along the anterior mantle margin. This character is accompanied
by a corresponding arrangement of the hepatic ramifications which supply the cerata. Further,
in most members of the family the radula is multiseriate; there is a crest between the
rhinophores (except in Proctonotus) and the vent is median and dorsal. The genus repre-
sented in the present collection, Madrella, does not possess these peculiarities and serves
to connect such forms as Janus and Proctonotus with the ordinary Aeolids. The radula
has only three teeth in a row, and the vent is lateral, not dorsal. In all the known genera
the reproductive system is unarmed and there are no cnidocysts.

Gen. Madrella A. and H.
32. Madrella ferruginosa A. and H.

One small and immature specimen obtained off a large <Antedon laevissima from
S. Nilandu Atoll, Maldives, 30 f Length 6 mm., breadth 4mm. The animal is oval and
flat, of a deep copper colour. Round the dorsal margin is set a not very thick band of
cerata and on the central area of the notaeum are a few scattered papillae. Both the
mantle edge and the foot are wide, but between them there is a deep groove. The vent
is at the posterior and the generative opening at the anterior end of the right side. The
form of the rhimophores is unusual: they are not perfoliate, but there is a circle of papillae
1 A. and H., Monograph of Brit. Nud. Moll. Id. Notice Kennt. d. Aeolidiaden, 1.1873. Id. Report on Challenger Nud.
of a Coll. of Nud. Moll. made in India. Bergh, Beitriige z. Trinchese, Aeolididae del porto di Genova, 1881.
Ce 106 73

570 SIR CHARLES ELIOT.

round the club much as in Zritonia. The mantle overhangs the head: there are no anterior
tentacles, but the head is broad and crescent-shaped with produced ends. The mouth is
ventral: the front of the foot square, without tentacular prolongations. The jaws are large
and not denticulate. The middle tooth of the radula has a central cusp with 7—8 denticles
on either side, and the laterals are denticulate on the inner margin. The ganglia of the
central nervous system are distinct.

This genus is recorded from the Indian Ocean (A. and H.) and Zanzibar, where I
have captured many specimens. The external characters of the animal vary somewhat with
the size, that is presumably with age. The smaller specimens are as described above. The
larger have many more cerata, and the papillae in the centre of the back (which apparently
do not contain hepatic ramifications) pass up between the rhinophores, forming a sort of
rudimentary crest. They have also a white patch between the rhinophores, which is absent in
the smaller individuals. When kept in captivity Madrella pours forth a peculiar ferruginous
secretion, which discolours the water round. It appears to proceed from the whole surface
of the body, not from any one gland, and is of much the same colour as the skin, so that
the creature looks as if it were dissolving.

There are four more families of the Cladohepatica which are sometimes classed together
as the tribe Elysioidea (Pelseneer) and sometimes united with Oxynoe and Lobiger as the
Ascoglossa (Bergh) or Saccoglossa (v. Jhering). They are (1) Hermaeidae, much like
Aeolids externally but without oral tentacles. (2) Phyllobranchidae, with peculiar flat
leaf-like cerata. (3) Elysiadae (including Placobranchus), with no cerata, but lateral ex-
pansions or wings. (4) Limapontiadae, small forms with no cerata, wings or appendages,
and often with neither tentacles nor rhinophores. All these families have a ramified liver
and also the following anatomical characters. (1) There are no jaws. The radula consists
of a single series of teeth, each fitting into the next. The front teeth instead of becoming
worn away and destroyed fall into a peculiar pouch in which they are preserved. This
form of radula may be conveniently termed “ascoglossan.” (2) In the nervous system the
cerebral and pleural ganglia are fused, the pedal ganglia are approximated and the visceral
commissure bears three ganglia (but in Limapontia only two). (8) The reproductive organs,
though imperfectly understood, are extremely complicated, provided with accessory glands
and ramified throughout the body. The male and female orifices are far apart, and there
are generally two of the latter.

Fam. Hermaeidae'.

The Hermaeidae closely resemble Aeolids and can hardly be distinguished by any ex-
ternal character except that the oral tentacles are not developed. Stiliger Mariae was
described by its discoverers, Meyer and Mobius, as an Embletonia. Internally, the family
is distinguished from the Aeolids by the following characters: (1) Jaws are absent and the
radula is ascoglossan, (2) the nervous system is formed of seven ganglia, (3) the hepatic
system is somewhat as in Fiona, there being no posterior hepatic trunk: the lateral branches
are often much ramified in the cerata and dorsal integuments. The vent as a rule is dorsal
and anterior. (4) The genital apparatus is much complicated by the presence of accessory

1 Alder and Hancock, Brit. Nudibr. Mollusca. Bergh, Golfe de Marseille, 1888. Pelseneer, Recherches sur divers

Beitr. zur Kennt. der Aeolidiaden, 1v., v. and via. Id. S. R. Opisthobranches, 1893:
Heft xvi. Vayssiére, Recherches sur les Moll. Opisth. du

NUDIBRANCHIATA. 571

glands, which are often diffused and ramified throughout the body. But the details of

structure seem to vary in the different genera and even species.

Gen. Hermaea Lov.

Small animals resembling Aeolids in shape but without oral tentacles. Foot long and
narrow with round expanded anterior corners. Rhinophores canaliculate. Cerata elongate or
slightly fusiform, arranged somewhat irregularly at the sides of the back. Hepatic trunks
elaborately ramified in the cerata and dorsal integuments. Anal papilla large, medio-dorsal.
Genital orifices distinct, at base of rhinophores. Albumen-gland diffused and in some species
at least ramified in cerata. Verge generally armed with a spine’.

33. Hermaea? minor B.

One specimen from the Laccadives, Minikoi, Length 3 mm.: breadth 4 mm. On the
label is written “Green nudibranch”: the preserved specimen is of the usual dirty yellow.
The centre of the back bears no cerata, but on the sides are about 20 rows containing
2 to 4 papillae each. The cerata are elongate, fusiform, and besides the hepatic ramifica-
tions contain other glands, the openings of which form tubercles on the exterior surface.
The teeth appeared to be 11 in all, exclusive of those in the bag, transparent, and very
narrow. I could not discover more than two genital openings, and there appears to be no
hook or spine on the verge.

I think this specimen probably belongs to a species which I have found abundantly
at Zanzibar, and which is perhaps identical with Bergh’s H. minor from Mauritius. The
colour does not quite correspond with the descriptions of Bergh’s two specimens, which are
said to have had the cerata tipped with red. My specimens varied from brown to white,
and in captivity tended to assume the latter colour. The prevailing tint of the newly-
captured animals was however green, the body walls being pale translucent green, and the
hepatic ramifications dark bright green. The eyes were distinct on a yellow spot. The
cerata were not tipped with red, but some specimens had splashes of carmine on the back.
I append a drawing made from life by Mr Crossland (fig. 9).

It is possible that these specimens may after all not be identical with H. minor, for
they had never more than 4 cerata in a row, whereas the latter is said to have as many
as 10.

Fam. Phyllobranchidae’.

The animals belonging to this family can easily be recognised by their flat shape and
leaf-like cerata.

There is a small frontal veil, above which rise two bifid canaliculate rhinophores.
Small canaliculate oval tentacles are generally present. The sides of the back are covered
with thick rows of large, flat, leaf-like cerata, which generally, if not always, contain hepatic
diverticula. Anal papilla large, dorsal, or lateral. No cnidocysts. Genital apertures two: ~ on

1 The genus has been subdivided as follows: (1) Hermaea
sens. stric. Teeth smooth, liver ramified within cerata: vent
median dorsal. (2) Hermaeina Trin. Teeth denticulate on
lower edge. (3) Placida Trin. Hepatic lobes not ramified
within cerata. (4) Hermaeopsis A. da Costa. Vent lateral.

Pelseneer has also pointed out remarkable differences between
Hermaea dendritica and Hermaea bifida.

2 A. and H., Notes on a Coll. of Nud. Moll. made in India.
Bergh, S. R. 11, xvr. Beitr. zur Kennt. der Aeolid. tv.
Pelseneer, Recherches sur divers Opisthobranches, p. 51.

73—2

Die, SIR CHARLES ELIOT.

right side of head and ¢ behind it. No jaws: a gizzard attached to the buccal mass:
radula ascoglossan: teeth generally denticulate: only a very few functional at a time.
Genital apparatus exceedingly complicated.

There are three genera (1) Phyllobranchus A. and H. Foot undivided: teeth denticulate,
forming a spiral in the radula bag. Gizzard long. Vent lateral. Verge unarmed. Recorded
from the tropical parts of the Pacific and Indian Ocean and from the Caribbean Sea.
(2) Cyerce B. Foot transversely divided: teeth denticulate, forming an irregular heap in
bag. Gizzard short. Vent dorsal. Verge armed. Recorded from the tropical Pacific and
Indian Ocean. (3) Caliphylla Da Costa. Foot undivided: no oral denticles: teeth not
denticulate, forming a spiral in the bag. Vent latero-dorsal. Verge unarmed. Two species
from the Mediterranean. It is possible that there may be really a fourth genus, since
A. and H. describe their Ph. orientalis as having cerata round the head, apparently like
Janus. Bergh thinks this is merely a mistake.

The present collection contains two specimens referable to Cyerce. There is some doubt
as to the relations of the cerata, and the liver in this genus. Pelseneer (/. c. speaking of
Cyerce Jheringi, of which I have not seen a detailed description) states that the liver is
entirely contained in the visceral mass, and does not extend into the cerata. Bergh (1871)
seems satisfied that the papillae do contain hepatic ramifications, though he admits that
the structure of the liver is obscure, and that the “hyaline folliculi” in the cerata can only
doubtfully be homologised with the contents of the cerata of Phyllobranchus. The present
specimens throw no light on the matter.

34. Cyerce nigra B.

Two specimens from Minikoi, which had entirely lost their cerata, only one small
detached leaflet being found with them. Apart from this the animals were well preserved
externally, but the internal organs were hardened into a uniform mass, in which nothing
could be distinguished. What remained of the colour was black with reddish-yellow spots,
and lines consisting of a border round the foot, another round the mantle edge, and two
lines behind the moderately large pericardial prominence. The body was very plump. The
length of the largest specimen 271 cm., the breadth 1 cm., the height 9 mm. The oval
tentacles small, canaliculate and contimuous with the ends of the moderately-large frontal
veil, from which a sort of ridge runs down to the anterior margin of the foot. The foot
is broad, and very deeply grooved in front. The transverse division of the under surface
is quite distinct, but not very deep. The part of the foot before the division is 5 mm.
long. The anal papilla in front of the pericardium large and high. The rhinophores strong,
bifid and canaliculate. Of the internal organs nothing could be ascertained except that
the teeth are long and denticulate. The single small detached papilla, which was preserved,
was black, with a reddish-yellow margin and spots of the same colour on both sides. It
may therefore be concluded that at least one and probably both the specimens belonged
to the variety ocellata B.

This collection of Nudibranchs does not in any way indicate that the fauna of the
Maldive and Laceadive Islands contains any element foreign to the rest of the Indo-Pacific.
All the genera except the very doubtful Stawrodoris are typical of this area, and all the
species, which are not new, are more or less certainly recorded from it. Two points however
are worthy of notice. Firstly, a large number of common and characteristic Indo-Pacific

Fauna and Geography Maldives and Laccadives Plate XXXII

E. Wilson, Cambridge

Eliot Nudibrancmata

NUDIBRANCHIATA. SYS

genera are absent, such as Asteronotus, Kentrodoris, Casella, Melibe, Bornella, and Placo-
branchus!, and secondly, the specimens are on an average considerably smaller than those
recorded from the Philippines, India, or East Africa. This latter feature is also noticeable
in the Tectibranchs (particularly the Pleurobranchoids) collected by Mr Gardiner, and I have
observed it in the specimens collected by myself at Samoa. It would seem as if the
marine fauna of small islands which have never been connected with a conti-
nent is less both in number and size than that of the mainland and its adjacent
islands. I am not however sure that a thorough examination of the Polynesian fauna
would bear out this conclusion, particularly as regards the number of genera and _ species.

EXPLANATION OF PLATE XXXII.

Doridomorpha gardineri nov. gen., nov. sp.

1. Dorsal view of a preserved specimen. 2. Ventral view of the same. 3. Rhinophore pulled
out of its pocket after removal of mantle. 4. Mandible. 5, Edge of mandible more highly
magnified. 6. Teeth (a) central, (+) first lateral, (c) from centre of row, and (d) outermost.

Notodoris gardineri nov. sp.

oa

7. Animal preserved in spirit about twice the natural size. 8. One lobe of the branchial valve,
Hermaea minor Bergh.
9. Living animal. x 10. 10, One of the cerata. x 30.

1 These genera are recorded from the Philippines, Indian Archipelago, and South Seas, and I have found them
all abundantly on the East Coast of Africa.

MARINE CRUSTACEANS.

IX. THE SPONGE-CRABS (DROMIACEA).

By L. A. Borraparis, M.A., Lecturer in Natural Sciences at Selwyn
College, Cambridge.

(With Plate XXXIIL)

THE members of the Dromiacea are separated from the other groups of Brachyura by
a number of characters, most of which are primitive and show that they are nearer than
the rest of the Crabs to the macrurous ancestors from which the tribe arose. The most
important of these features are: the presence of a pair of vestigial limbs on the first
abdominal segment of the female, and often on the sixth in both sexes’, the epipodites,
which are often found on one or more of the pairs of legs, the number of gills, which is
usually large, the more or less square shape of the mouth, the large, usually free, basal joint
of the antennal stalk, and the often incomplete orbits. To these must be added a highly
characteristic, but not primitive, feature—the small size, dorsal position, and usually prehensile
shape of one or both of the hinder pairs of legs (Pl. XX XIII).

This conformation of the hinder legs is connected with the most characteristic peculiarity
in the habits of the Dromiacea. Most of them are in the habit of holding over their backs
masses of some sessile animal—usually a sponge, but at times an ascidian, an alcyonarian,
or even half a bivalve shell, on which a sea-anemone is often seated—by means of which
they are completely hidden*.

The most important part of the prehensile apparatus by which this is accomplished is
the last joint of the legs of the hinder two pairs. This joint is sharp and hooked in most
genera, and is buried deeply in the substance of the sponge. There is often also a thorn
on the propodite, which makes with the last joint a more or less perfect chela (Pl. XXXIII,
figs. 3b, 46). In many cases, at least, the crab does not confine its choice to one species
of sponge. Whether the latter is at first merely a fragment broken off from a larger mass,
or is always a whole colony, perhaps specially adapted to this life and not found elsewhere,
is not known, but in any case the initiative in forming the partnership must come from

1 Bouvier (Bull. Soc. Philomath. Paris, 1896) throws doubt of Macrura.

on the correctness of the view which homologises the struc- 2 This curious habit is also found in the Dorippidae
tures on this segment in Dromiidae with the last pair of limbs among the Oxystomata (see above, p. 439, vol. 1.).

MARINE CRUSTACEANS. 575

the crab, for the sponge is quite free, and is merely held in position by its host. On the
other hand, it must live thus for some time, and grow meanwhile, for it will always be
found to have a deep hollow on the underside, closely fitting the back of the crab. No
doubt this helps to keep it from slipping, as does also, perhaps, the hairy or woolly coat
with which the decapod is usually provided. The hairs of this coat are often very diverse
and complicated in shape—pointed or clubbed, feathered or simple (Pl. XXXIII, figs. 1b—d,
3c—f1)—though it is hard to see what end these can serve.

Of the two groups into which the Dromiacea are subdivided, the Homolidea are princi-
pally deep-water forms, and were not met with by the expedition. The Dromiidea, on the
other hand, are fairly common on the reef, where they are to be found in all possible
positions, holding on tightly with the sharp tips of their very mobile walking-legs (Pl.
XXXIIT). In their habits they are slow and heavy, trusting rather to concealment than
to active avoidance and defence.

In the systematy of the Dromiudea much use is made of the sternal grooves of the
female. These are a pair of furrows, one on each side, which run longitudinally over the
thoracic sternites of the female, beginning on the hindermost and ending on that which
corresponds to the first, second, or third pair of legs. Their forward ends may be raised
on knobs, and are either wide apart, side by side but not meeting, or curving round to
meet. They correspond closely with the position in which the long abdominal limbs of the
male are carried, and are therefore probably used for some sexual purpose. Other features
of systematic importance in this group are: the presence or absence of epipodites, the pro-
portions of the body and the extent to which furrows are present on the back, dividing
it into regions, the shape and relative length of the legs, ete. For specific distinctions, the
shape, number and arrangement of the teeth on the front, anterolateral edge, underside of
the body, and legs are chiefly used.

The following keys, introduced here in performance of the promise given in the article
on the classification of the crabs (p. 426 above), will enable members of the legions and
families of the Dromiacea to be recognised :

Key to the legions of the Dromiacea.

I. Sternum of female with longitudinal grooves. Vestiges of 6th pair of abdominal limbs
usually present. Gills 14—20 on each side. Eyes usually completely sheltered by orbits when re-
tracted. No lineae anomuricae. Dromiidea.

II. Sternum of female without longitudinal grooves. No vestiges of 6th pair of abdominal
limbs. Gills 8—14 on each side. Eyes incompletely or not at all sheltered by orbits when with-
drawn against the body. JLineae anomuricae usually present. Homolidea.

Key to the families of the Dromiidea.

I. No vestige of 6th abdominal limb. Carapace longer than broad, with ill-marked side edge.
[First three legs with epipodites, fourth and fifth small, subdorsal, and prehensile.] Homolodromiidae.

1 For the investigation of this interesting point I am indebted to Mr Edwin Wilson’s enthusiasm for the subject.

576 _L. A. BORRADAILE.

II. A vestige of the 6th abdominal limb present [except in Hypoconcha, where also no epi-
podites]. Carapace usually not longer than broad, with well-marked side edge.

A. An epipodite on the first leg (cheliped) only or on none. Fourth and fifth legs small,
subdorsal, and usually prehensile. Dromidae.

B. Epipodites on all the first three pairs of legs. Fifth leg only small and _ subdorsal.
Dynomenidae.

Key to the families of the Homolidea.

I. Gills 13 or 14 on each side. Epipodites on first one or three pairs of legs. First joint
of eyestalk not much longer than second. Homolidae.

II. Gills 8 on each side. Epipodites not found on any legs. First joint of eyestalk much
longer than second. Latreillidae.

Of the seven species in the present collection three are new, and all the remainder
are already recorded from the Indian region. Many of the specimens bear sponges, but, as
the species of these have not been determined, no further allusion will be made to them.

Family Dromiidae.

Generic names in this family are used here in the sense in which they have lately
been defined by the writer (Ann. Mag. Nat. Hist. (7) x1. p. 298).

Genus Dromia Fabr., 1798.

1. Dromia rumphi Fabr., 1798 (Pl. XXXIII, fig. 1). Alcock, v. p. 187.

In young specimens, such as the one figured on Pl. XXXIII, there is a distinct tooth
at the inner orbital angle, and the teeth on the wrist are also much sharper. The species
was dredged in Haddumati Atoll in 40 fathoms.

Genus Dromidiopsis Borradaile, 1900.
2. Dromidiopsis australiensis (Hasw.), 1882.
Dromia (Dromidia) australiensis, Alcock, Cat. Ind. Decap. Crust. Ind. Mus. 1.1. p. 76 (1901).

Var. bidens n. Specimens of the variety described by de Man [Zool. Jahrb. Syst. rx.
p-. 372], with only two of the three teeth behind the orbital angle in the type, were taken
in the Maldives. I propose to name this variety bidens. A specimen from Fiji in the
Cambridge Museum belongs to the same variety, but there is another Rotuman specimen,
and one from Fiji, which have only one of the three side teeth remaiming. This I propose
to call var. wnidens.

Taken on the reef at Hulule, Male Atoll, and dredged in 25 fathoms in South
Nilandu Atoll.

3. Dromidiopsis tridentatus n. sp. (Pl. X XXIII, fig. 2).

Diagnosis: “A Dromdiopsis whose carapace is slightly longer than broad; the body and
limbs smooth, covered with a fine down; the front triangular with the tip just visible from
above, deeply grooved and with low swellings of the sides to form a pair of indistinct
teeth; no trace of a tooth at the upper, inner orbital angle, the outer orbital angle strong,

MARINE CRUSTACEANS. SWAT

but not forming a tooth, a deep cleft on the orbital edge just below the outer orbital angle ;
three small blunt knobs on the anterolateral edge behind the outer orbital angle, two broad
notches behind these teeth corresponding to the branchial groove (which is present) and
probably to the cervical groove (which is absent); the 2nd and 3rd legs with strong, hooked
end-joints, armed with a few thorns underneath, the fourth leg short, subchelate, and the
fifth leg as long as the third, also subchelate.”

Length: 11 mm. Breadth: 10mm. Colour in spirit: pale yellow.
One female from Hulule, Male Atoll.

This species much resembles Dromidiopsis australiensis, but may be distinguished by
the more distant and more outstanding teeth of the anterolateral edge, which in D. austral-
ensis are broader at their base, by the sternal grooves of the female, which meet and end
on the segment of the first walking-leg, and by the presence of a spine on the hinder
edge of the propodite of the fifth leg. The cheliped has an epipodite.

Genus Dromides Borradaile, 1903.
For a diagnosis of this genus see the paper cited on p. 576 above.

4. Dromides hilgendorfi (de Man), 1887. Alcock, v. p. 145°.
Dredged in Suvadiva and South Nilandu Atolls, in 30 and 19 fathoms respectively.

Genus Cryptodromia Stimps., 1858.

5. Cryptodromia bullifera Alc., 1898. Alcock, v. p. 143.
Dredged in South Nilandu Atoll in 30 fathoms.

6. Cryptodromia hirsuta n. sp. (Pl. XX XIII, fig. 3).

Diagnosis: “A Cryptodromia with the body broader than long, smooth and hairy, the
regions indistinct; the front prominent and bearing three fairly sharp teeth, of which the
middle one is somewhat longer than those at the sides; the upper rim of the orbit finely
beaded and bearing a tooth near its inner end, a tooth at the outer and another at the
lower inner orbital angle; one tooth on the anterolateral edge and a low swelling behind
this tooth, a small blunt tooth behind the branchial groove, and a strong conical subhepatic
tooth; the carpopodite and propodite of the cheliped and second two pairs of legs knobbed,
but without sharp thorns, and the last two pairs of legs short, subequal, and feebly chelate.”

Length: 7mm. Breadth: 8mm. Colour in spirit: dull yellow, legs often orange.

The sternal grooves of the female end widely apart on the line between the sternites
of the first and second walking-legs. There is no epipodite on the cheliped.

Taken on the reef in Minikoi and Male Atolls.

1 For the principle on which references are given in this Indian Dromiacea here referred to appeared in Journ. As.
series of papers see above, Pt. 1. p. 192. Alcock’s paper on Soc. Bengal, uxyrtt. ii. (1899).

Ga iL 74

578 L. A. BORRADAILE.

Genus Cryptodromiopsis Borradaile, 1903.
For a diagnosis of this genus see the paper cited on p. 576 above.

7. Cryptodromiopsis tridens n. sp. (Pl. XX XIII, fig. 4).

Diagnosis: “A Cryptodromiopsis with the body broader than long, smooth and _ hairy,
the grooves between the regions almost wholly lost, a smooth, naked, oval boss on each
side of the cardiac region; the three teeth of the front of almost equal length and all
sharp; teeth at the upper and lower inner, and at the outer orbital angles; two anterolateral
teeth but no tooth at the branchial groove; a subhepatic tooth and a prominent angle of
the mouth; the fingers of the chelipeds gaping and two blunt teeth on the wrist; a thorn
at the end of the carpopodite in the first two pairs of walking-legs, and another at the
end of the propodite in the last two pairs, making with the end-joint a strong chela.”

Length: 6mm. Breadth: 8 mm. Colour in spirit: dull yeilow.

There is no epipodite on the chelped, and the sternal grooves of the female end side
by side on the cheliped segment without actually meeting.

Taken on the reef in Male and Minikoi Atolls.

EXPLANATION OF PLATE XXXIII.

Fie. 1. Dromia rumphi, young individual; a. whole animal, 6. one of the long hairs on the back ;
c, d. short hairs on the back.

Fic.

to

Dromidiopsis tridentatus; a. whole animal, from above. The carapace is rather longer
than is shown in the figure.

Fic. 3. Cryptodromia hirsuta; a. whole animal, 6. end of the fifth leg, c. one of the hairs on
the ridge overhanging the furrow just behind the front, d. small hairs on the back,
e. long hair on the back, f hairs on the legs.

Fic. 4. Cryptodromiopsis tridens; a. whole animal, 6. end of the fifth leg.

Fauna and Geography, Maldives and Laccadives Plate XXXIII

EB, W, del.

BORRADAILE——DROMIACEA.

To face ~. 578.

Mi
-

ON A LAND PLANARIAN FROM HULULE, MALE ATOLL,

WITH A NOTE ON LEPTOPLANA PARDALIS LAIDLAW.
By Frank Fortescue Larpiaw, B.A. (Cantab.).

A SINGLE specimen was found of a small species of Rhynchodemus under decaying wood.
Its total length is about 25 mm. and breadth a little under 2 mm.

After some hesitation I have referred it to von Graft’s Rhynchodemus ceylonicus’, a species
of which a single specimen was collected by the brothers Sarasin in Ceylon. This the type
specimen is considerably larger than that found by Mr Gardiner (48 mm. long by 3 mm. in
breadth), but the colouring agrees fairly well, although owing to the fact that the specimen
under discussion has been treated with corrosive sublimate it has rather faded.

There is a somewhat obscure median dorsal dark band and two feebly marked broader
lateral bands. Near the middle of the body the small bell-shaped pharynx is protruded.

No land planarian has hitherto been recorded from any coral island in the Indian Ocean.
Other islands remote from continental land in this Ocean have, however, a small number of
inhabitants belonging to this group. Such are the Seychelles with two species of Amblyplana,
and Rodriguez with one species Geoplana whartoni of Gulliver, who also observed a specimen
of land planarian on Mauritius. The planarian fauna of these islands is considered by von
Graff to be derived from that of the Aethiopian region, whilst the Maldivian species is
clearly derived from Ceylon where the genus Rhynchodemus is fairly abundant.

Certain Oriental species of the genus have a wide distribution, such for example is
Rhynchodemus nematoides Loman, which is recorded from Ceylon and Java by von Graff in
his monograph of the land planarians, whilst I have examined specimens from a height of
2500 ft. from a hill in the north-east of the Malay Peninsula.

Species of land planarians are known to occur on at least two coral islands in the
Pacific. One of these is Rhynchodemus putzei v. Gr. from the Marquesas Is., the other is
Geoplana pelewensis v. Gr. from the Pelew Is.

1 von Graff, Monographie die Turbellaria ut. Tricladida terricola.

580 F. F. LAIDLAW.

Note ON LEPTOPLANA PARDALIS LAIDLAW.

Since my account? of Mr Gardiner’s marine planarians was published, I find that I had
inadvertently overlooked the description of Leptoplana subviridis v. Plehn, given by von Plehn
in Semon’s Zool. Forschungreisen, 5, 11. p. 330, Pl. XXIII. figs. 11, 12. Since three of the
four species described and figured in this paper are also described with the same figures
in von Plehn’s previous paper in the Jena Zeitschrift, I had carelessly put the former on
one side, supposing that all the species described in it were also described in the latter.

I find now that the species which I overlooked is identical with my Leptoplana pardalis,
which therefore becomes merely a synonym of ZL. subviridis. I find, however, that there are
certain discrepancies between our accounts especially as concerns the male apparatus. I have
re-examined my sections in consequence and give the following comments.

As regards the vesicula seminalis, von Plehn states that the vasa deferentia have muscular
walls and open together into a vesicula seminalis. In my account I stated that the lower
ends of the vasa deferentia were dilated to form a pair of vesiculae which opened into
a ductus ejaculatorius. In favour of this latter view I find that in my specimens at any
rate, the ends of the vasa deferentia are distinctly dilated, and that the vessel called the
vesicula by von Plehn is narrow and of the same diameter throughout, until it decreases
on approaching the penis. In support of von Plehn’s view is the fact that her interpretation
is more in accord with the condition usually found in Leptoplanas. I am inclined to believe
that both views may be right in this case, ie. that either the lower ends of the vasa
deferentia, or the part which I have called the ductus ejaculatorius may function as
a vesicula seminalis.

Von Plehn’s specimen does not appear to have shown the very distinct narrowing of the
two parts of the accessory vesicle and their subsequent dilatation to form a receptaculum
seminis on either side, so that each receptaculum is placed as it were on a distinct stalk.

The synonymy of this species then stands as follows:
Leptoplana subviridis v. Plehn.
L. subviridis von Plehn. Semon’s Zool. Forschungreisen, 5, 11. p. 330. Tf. XXIII. figs. 14, 15.
L. pardalis Laidlaw. Faun. and Geogr. of the Maldives and Laccadives, 1. 3, p. 287,
Pl. XIV. fig. 9, Pl. XV. fig. 14. Lbid., Proc. Lit. and Phil. Soc. Manchester, Xuvit. (1903), no. 5.
I take the opportunity of thanking Mr Stanley Gardiner for permission to call attention
to my oversight, which I greatly regret to have committed.

1 “The Marine Turbellaria, ete.” Fauna and Geography, Maldives and Laccadives, vol. 1. pp. 281—310 (1902).

LAGOON DEPOSITS.

I. GENERAL ACCOUNT.

By J. Stantey GARDINER, M.A.

II. REPORT ON CERTAIN DEPOSITS.
By Sir Jonn Murray, K.C.B., F.R.S.
(With Plate XXXIV.)

I. GENERAL ACCOUNT (by J. Srantey GarpDINER)}.

In Appendix A, Section II. of my article on “The Maldive and Laccadive Groups”
(vol. I. pp. 317—323), I have already referred to some of the deposits found in the atolls
and on the banks which we visited. I am greatly indebted to Sir John Murray for kindly
examining and giving me the accompanying Report on certain deposits from the lagoons of
the atolls which I submitted to him. Perhaps the Section above referred to should be read
a second time in connection with this Report, though the latter does not, so far as I can
see, invalidate any conclusions therein expressed as to the formation of lagoons, but on the
contrary lends them additional important support.

On the most open banks, such as Miladumadulu and North Mahlos, we found, generally
speaking, either a hard bottom without any deposit or else a deposit formed by very coarse
broken coral and shell fragments, leaves of Halimeda and allied algae, together with a few
bottom-living Foraminifera of large size. As might have been expected, there did not appear
to have been anywhere a deposit of any thickness except to leeward of certain reefs, where
it might have the characters of nos. 11 and 12 of Sir John Murray’s Report.

Within shallow atolls, or banks which had assumed the atoil shape, the bottom, except
along the lines of the currents, was commonly covered by a deposit of white sand. This
was always quite loose and never coherent in any way. It consisted of fine fragments of
the same materials as on the more open banks, all more or less rounded or smoothed. Except
that the remains of pelagic organisms were rarer, the deposit was the same as that which
formed the sand of the lagoon-ward parts of the islands (nos. 14 and 15). The grains varied
in size, being in enclosed situations extremely small, such as the centre of Minikoi (no. 16),
except for leaves of Halimeda and broken shells. In this position no doubt a great part
of the deposit was formed by the washings off the land, but that from the velu (lagoon)
of Bodu Faro, North Mahlos, a quite open faro (atollon) 13 fathoms in depth, approaches
rather to a mud (no. 13). Nos. 17—19 are also examples of coral sands from the neighbour-
hood of passages into the lagoon of Suvadiva, where currents of greater or lesser force are felt.

1 Sir John Murray suggested that I should give some possess. Consequently, I alone am responsible for any

general account of these deposits, as I necessarily had a statements or opinions contained in this Section.
local knowledge of the region, which he did not himself

582 J. STANLEY GARDINER.

It is on this sand deposit that the greatest amount of free animal life is found, while
sedentary animals tend to settle upon the more current-swept areas. Among the latter forms
are certain Polyzoa and Polytrema, which completely grow round small fragments of stone,
turning them into more or less round balls, various Alecyonaria, a few sponges and tunicates
and corals of such genera as Dendrophyllia, Alveopora, Goniopora, ete. Of the free forms
in the coral sand small lamellibranchs and gastropods are numerous, as well as young echinids.
Crustaceans, particularly Portunidae and Oxystomata as well as Macrura, abound. Certain
holothurians, sipunculids and _polychaets, as well as Ptychodera and Cephalochorda in certain
parts of the shallower waters, are quite numerous. Corals are found only where the sand
is moderately coarse, and there is some current, but then such forms as Diaseris, Goniopora
stokesi, Heteropsammia and Stephanoseris were remarkably abundant locally. Indeed, dead
coralla of the first-named were found to be forming the greater part of a series of con-
solidating masses of rock, dredged from an eddy in one of the northern passages of Suvadiva
atoll, of which no. 23 was a specimen.

Lastly, within the most enclosed parts of certain banks and within the lagoons of the
deeper atolls a greenish grey ooze, almost a mud, was found (nos. 1—10, 20, 21, 24).
A similar deposit may possibly have existed in quite shallow water and have been over-
looked by us, but, so far as our soundings and dredgings went, there would almost seem
to have been a critical depth at about 30 fathoms below which it was for some reason
unable to form. It was generally a fairly clean mud, but below 40 fathoms, when fresh,
unctuous, and sometimes smelling slightly of sulphuretted hydrogen. In Suvadiva atoll it
was largely composed of the shells of Pteropoda, in which connection it is interesting to
remark that one morning at the beginning of November, 1899, the water in the lagoon
of Goifurfehendu was positively thick with the shells of Clio. Pelagic Foraminifera were
common, but bottom-living species were by no means numerous. Sedentary organic life,
except sponges, scarcely existed upon it, while free life was only represented by a few species
of Crustaceans and a still smaller variety of flat fishes. A holothurian was dredged from
this class of bottom (p. 339); it was of peculiar interest on account of the relatively large
fragments of coral contained in its gut (no. 25).

The most interesting point about this muddy deposit from Suvadiva is the presence in
it of large quantities of casts of various organisms, more especially Foraminifera. Its minera-
logical constituents are small, and are for the most part those found in the open ocean,
being of yoleanic origin. In connection with this, one may point out that on the shores
of many of the islands there are lines of pumice, which the natives state were washed up
about 1885, and would hence have probably owed their origin to the eruption of Krakatoa
in 1883. In addition half-decomposed pumice is found, in places at some considerable
distance inland, which evidently belonged to an earlier period. The analogy between these
casts and those of Greensand appears to me to be very close, and I would suggest that
here we have an incipient formation of the latter deposit, which is generally supposed to
be largely terrigenous in origin. Greensand essentially owes its origin to the filling in of
the shells, particularly of the Foraminifera, by a deposit. It was shown in the Challenger
Reports to occur between 100 and 900 fathoms along bold and exposed coasts, where pelagic
conditions might be expected to closely approach the shores, though all the elements of its
essential constituent, glauconite, may also be present in pumice or the ordinary oceanic,
volcanic deposits. That there is any real homology between this deposit and that of Greensand
is by no means certain, but, if there be, either the formation of Greensand must depend
on physical conditions rather than the proximity of continental land, or Suvadiva atoll at
least must have been closely connected with or perhaps founded on the remains of a continental
land mass.

LAGOON DEPOSITS. 583

So far as I am aware this is the first recorded occurrence of a true pelagic . deposit,
we. Pteropod Ooze, within the lagoon of an atoll. It is accompanied by much fine matter, |
the inorganic part of which consists almost entirely of carbonate of lime. The Challenger
Report recorded the presence of a similar deposit off many coral reefs and oceanic islands,
that around the Fiji Group being particularly noticeable, though in my own experience it
is not found in the lagoons of the atoll or barrier reefs of those islands. The presence of
this deposit in Suvadiva lagoon at depths over 40 fathoms is interesting as it is not found
elsewhere in the shallower lagoons we investigated, and it would naturally be expected that
the calcareous remains of such small pelagic organisms would have been rapidly dissolved
in the sea-water and removed. A considerable proportion of the deposit being formed of casts
shows that its origin cannot have been due to merely temporary causes. It might reasonably
be deduced that the lagoon of Suvadiva below 35 or 40 fathoms is a relatively still one
with little circulation in its sea-water, but the reason of this state is not obvious, as that
atoll has 38 well-defined passages into its lagoon, of which 7 are about or over 35 fathoms
in depth, and 17 about or over 20 fathoms. It is useless, however, theorising on the subject,
when it is obvious that practically nothing is known of the physical conditions of this large
lagoon. In connection with the question it will be interesting to see what deposits Prof.
Agassiz found around the Maldive banks at moderate depths. The presence or absence of
manganese nodules or deposits, indicated in no. 24 below, will no doubt be considered also by
the same author.

I have appended a chart to Sir John Murray’s Report (Pl. XXXIV), underlining the
positions from which the samples from Suvadiva atoll were obtained. I add from the same
Report a list of the animals of such groups as will otherwise not be treated of in this
publication. Names in brackets have been added by me, the bracketed genera of the
Foraminifera having been kindly identified by Mr J. J. Lister.

Pelagic Foraminifera :

Globigerina bulloides.

Globigerina aequilateralis.

Pulvinulina tumida.

. sacculifera. ei dubia. a menardit.
conglobata. Orbulina universa. a canarinensis.
3 digitata. Sphaeroidina dehiscens. Cymbalopora.
5 rubra. Pullenia obliquiloculata.
Bottom Foraminifera :
Biloculina. Rheophax. Polymorphina. (Gypsina).
(Miliolina). (Textularia). Uvigerina. Amphistegina.
Spiroloculina. Cassidulina. Pullenia sphaeroides. (Polystomella).
Peneroplis. Lagena. Rotalinae. (Operculina).
(Orbitolites). Nodosaria. Polytrema. Heterostegina depressa.
(Alveolina).
Mollusca Pteropoda:
Clio subula. Limacina.
» australis. Cavolinia.
Mollusca Heteropoda:
Atlanta. Carinaria. (Oxygurus). (Pterotrachea).
Other Mollusca :
Pleurotoma. Dentalium.
Diatomaceae :
Pleurosigma. Surirella. Coscinodiscus.

Il. REPORT ON CERTAIN DEPOSITS (by Str Jonn Murray).

Number

Depth in
fathoms

Designation and Physical
Characters

Carbonate of Calcium

Per cent.

Organisms

42

43

44

44

46

48

PTEROPOD 00ZE, greyish green
in colour, slightly coherent,
fine grained, contains many
Pteropod fragments.

Residue, green, flocculent.

PTEROPOD 0028, green in colour,
fine grained, coherent.

Residue, heavy, black or dark
green.

CoRAL SAND, cream in colour,
incoherent.
Residue, brown, flocculent.

CoRAL MUD, greenish grey, fine
grained, incoherent.
Residue, green.

PTEROPOD 00ZE, green in colour,
incoherent.

Residue, dark green, floc-
culent.

PTEROPOD 00ZE, brownish grey,
incoherent.
Residue, brown, flocculent.

PTEROPOD OOZE, greenish grey,
incoherent.

Residue,
flocculent.

greenish brown,

80

70

90

96

90

80

90

Pteropods (Clio subula, C. australis, Cavolinia) and frag-

ments, pelagic Foraminifera (Globigerina bulloides, G.
sacculifera, G. conglobata, G. digitata, G. rubra, G. aequi-
lateralis, Pulvinulina tumida, P. menardii, P. canari-
nensis, Pullenia obliquiloculata, Orbulina universa, Sphae-
roidina dehiscens), bottom-living Foraminifera (Vodosaria,
Rotalidae, Biloculina, Rheophax, Uvigerina), Otoliths,
Ostracodes, Echini spines, larval Molluscs, Heteropods
(Atlanta), calcareous coprolites, coccoliths and rhabdo-
liths.

Pteropods (Clio subula, C. australis, Cavolinia), Heteropods

(Atlanta, Carinaria), pelagic Foraminifera (Globigerina
bulloides, G'. sacculifera, G. digitata, G. aequilateralis,
G. conglobata, G. rubra, G. dubia, Orbulina universa,
Sphaeroidina dehiscens, Pullenia obliquiloculata, Pul-
vinulina tumida, P. menardii, P. canarinensis), bottom-
living Foraminifera (Uvigerina, Rotalidae, Biloculina,
Nodosaria), larval Gastropods and Lamellibranchs, Ostra-
codes, Otoliths, Echini spines, Serpula, calcareous copro-
lites, fish teeth, coccoliths, rhabdoliths and coccospheres.

Coral fragments, one or two small bivalve shells, larval

Lamellibranchs and Gastropods, Pteropods (Clio austra-
lis, C. subula), Echini spines, bottom-living Foraminifera
(Amphistegina, Rotalidae, Spiroloculina, Biloculina, No-
dosaria), Otoliths of fish, Ostracodes, pelagic Foraminifera
(Globigerina bulloides, G. rubra, Sphaeroidina dehiscens,
Pullenia obliquiloculata, Orbulina universa), Tunicate
spicules.

Coral fragments, Pteropods (Clio subula, C. australis, Cavo-

linia), Heteropods (Atlanta), larval Lamellibranchs and
Gastropods (Pleurotoma, Dentalivm), Echini spines and
shells, bottom-living Foraminifera (Vodosaria, Rotalidae),
fragments of Crustacean tests and claws, pelagic Fora-
minifera (Globigerina bulloides), Ostracodes, Otoliths of
fish, Tunicate spicules, Serpula, coccoliths.

Pteropods (Clio subula, C. australis, Cavolinia), Heteropods

(Atlanta), pelagic Foraminifera (Globigerina bulloides,
G. conglobata, G. sacculifera, G. digitata, G. aequilateralis,
G. dubia, G. rubra, Orbulina universa, Sphaeroidina
dehiscens, Pulvinulina tumida, P. menardii, Pullenia ob-
liquiloculata, Cymbalopora), bottom-living Foraminifera
(Spiroloculina, Cassidulina, Pullenia sphaeroides, Amphi-
steyina, Rotalidae, Uvigerina), larval Lamellibranchs and
Gastropods, Otoliths, Ostracodes, rhabdoliths, coccoliths
and coccospheres.

Pteropods (Clio subula, C. australis, Limacina, Cavolinia),

Heteropods (Atlanta, Carinaria), pelagic Foraminifera
(Globigerina bulloides, G. conglobata, G. aequilateralis,
G, sacculifera, Orbulina universa, Sphaeroidina dehiscens,
Pullenia obliquiloculata, Pulvinulina tumida, P. menar-
dvi), Echini spines, larval Lamellibranchs and Gastro-
pods, Ostracodes, Otoliths of fish, bottom-living Fora-
minifera (Amphistegina, Rotalidae), Tunicate spicules,
rhabdoliths.

Pteropods (Clio subula, C. australis), Heteropods (Atlanta),

pelagic Foraminifera (Globigerina bulloides, G. rubra,
G. sacculifera, G. conglobata, G. aequilateralis, Pulvinu-
lina menardii, P. tumida, Orbulina universa, Sphaeroi-
dina dehiscens), bottom-living Foraminifera (Uvigerina,
‘assidulina, Polymorphina, Biloculina), Echini spines,
Tunicate spicules, larval Gastropods and Lamellibranchs,
Ostracodes, coccoliths, calcareous coprolites.

585
Residue
Position
Per cent. Minerals Siliceous Organisms Fine Washings
20 (2°/,), angular, mean dia- | (3°/,), arenaceous Fora-

30

10

10

20

10

G.

meter 0°15 mm., felspar,
hornblende and magnetite.

(1°/,), angular, mean dia-
meter 0-1 mm., felspar and
magnetite. No manganese
observed.

(1°/,), angular, mean dia-
meter 0°06 mm., felspar
only observed.

(1°/,), angular, mean dia-
meter 0°08 mm., fragments
of glassand felspar, one par-
ticle of felspar measured
0O-14mm.; flakes of iron
rust, probably off the ma-
chine used in collecting the
sample, were also observed.
No manganese observed.

(2°/.), angular, mean dia-
meter 015 mm., felspar
and glass, many casts of
Foraminifera, etc. but not
so perfect as in the dry
sample (no. 4). No man-
ganese.

(2°/,), angular, mean dia-
meter 0-07 mm., felspar
only observed.

(1°/.), angular, mean dia-
meter 0°07 mm., felspar,
hornblende and magnetite.

minifera, sponge spicules,

many perfect green clayey |

casts of Foraminifera, etc.,
Radiolaria.

(5 °/.), sponge spicules,
Radiolaria, many clayey
casts of Foraminifera,

clayey coprolites, diatoms.

3°/,), diatoms, sponge spi-
5 , Sponge sp
cules.

(2 °/.), sponge spicules, Ra-

diolaria, Diatoms (Pleuro- |

sigma, Surirella, Coscino-
discus),and many beautiful
green clayey casts of Gas-
tropods, Echini spines,
Foraminifera, ete., copro-
lites, no casts in Pyrite.

(5°/.), sponge spicules, Ra-
diolaria, small sponge, are-
naceous Foraminifera, Di-
atoms.

(4°/,), sponge spicules, Ra-
diolaria, Diatoms, one or
two white clayey casts.

(3°/,), sponge spicules, Ra-
diolaria, Diatoms.

| (6°/.)

(15°/.). amorphous clayey
matter with minute frag-
ments of siliceous organ-
isms and minerals less

(24°/,), amorphous clayey
matter with minute mine-
ral particles less than
0:05 mm. in diameter, etc.

(6°/,.), amorphous clayey
matter, with one or two
minute mineral particles.

(1°/,), amorphous clayey
| matter, with minute frag-
ments of siliceous organ-
isms, and small mineral
particles lessthan0°05mm.
in diameter.

(3°/.), amorphous clayey
matter, small casts, ete.

(14°/.), amorphous clayey
matter with minute mine-
ral particles and fragments
of siliceous organisms.

amorphous clayey
matter with minute mine-
ral particles less than
0:05 mm. in diameter and
fragments of siliceous or-
ganisms.

than 0°05mm. in diameter. |

Suvadiva, N.N.E.
2} miles S. b. W.
Mametu.

Suvadiva, N.E.
43 miles W. Digura.

Suvadiva, S.E.
2 miles N. Noor-
bhai I.

| Suvadiva, S.E.
1? miles N.W. Fulu
Ie

Suvadiva, S. 2
miles W.S.W. Mafe
1.

Suvadiva, 8. 3
miles further W.
than last.

Suvadiva, 8. 2
miles further W.
than last.

Il.

75

586

Carbonate of Calcium
| Number | Depth in Designation and Physical ‘
| fathoms Characters
| Per cent. Organisms
8s. 44 PTrEROPOD 0OZE, greenish grey, 90 Pteropods (Clio subula, C. australis, Limacina), Heteropods
fine grained, incoherent. (Atlanta), Coral fragments, bottom Foraminifera (/ota-
Residue, one or two black lidae, Amphistegina), pelagic Foraminifera (Globigerina
particles, very little clayey bulloides, G. conglobata, Pulvinulina menardii, Cymbalo-
matter, green in colour. pora), Echini spines, larval Lamellibranchs and Gastro-
pods.
9. 43 PTrEROPOD 00ZE, greenish grey 90 Pteropods (Clio australis, C. subula, Limacina, Cavolinia),
in colour, fine grained, inco- Heteropods (Atlanta, Carinaria), pelagic Foraminifera
herent. (Globigerina bulloides, G. conglobata, G'. sacculifera, G.
Residue, green casts, very aequilateralis, G', rubra, Orbulina universa, Sphaeroidina
little floceulent matter. dehiscens, Pulvinulina menardii, P. tumida, Cymbalopora,
Pullenia obliquiloculata), bottom-living Foraminifera
(Rotalidae, Pullenia sphaeroides, Uvigerina, Amphiste-
gina), Echini spines, Otoliths, Ostracodes, Aleyonarian
spicules, Tunicate spicules, larval Lamellibranchs and
Gastropods.
10. 30 CoRAL SAND, greenish grey, in- 90 Fragments of Coral and Mollusc shells, larval Lamelli-
coherent. branchs and Gastropods, Otoliths of fish, Ostracodes,
Residue, green. bottom-living Foraminifera (Rotalidae, Amphistegina,
Spiroloculina, Lagena, Biloculina), pelagic Foraminifera
(Globigerina bulloides, G. conglobata), Pteropods (Clio
subula), Echini spines, Tunicate spicules, Aleyonarian
spicules, Coccoliths.
11. Z8 CoRAL SAND, cream coloured, 98 Coral fragments, Echini spines, calcareous Algae, Aleyo-
fine grained, incoherent. narian spicules, bottom-living Foraminifera (Lagena,
Residue, grey, no clayey Amphistegina), Tunicate spicules, larval Lamellibranchs
matter. and Gastropods, Crustacean claws, Ostracodes, and Oto-
liths of fish. y
12. 20 CoraL sAND, cream coloured, | 97 Coral fragments, Aleyonarian spicules, bottom-living Fora-
incoherent. minifera (Orbitolites, Amphistegina, Rotalidae), calcareous
Residue, brown. Algae.
13. 8 Corat Mvp, white, finely granu- 97 Fragments of Coral and calcareous Algae, Molluscs and
lar, unctuous. fragments, bottom-living Foraminifera (Miliolidae, Ro-
Residue, a few grey particles, talidae, Peneroplis, etc.), Tunicate spicules, Ostracodes,
no clayey matter. Otoliths of fish, also vertebrae and teeth of fish, Aleyo-
narian spicules, Echini spines, Polyzoa, pelagic Fora-
minifera (only Cymbalopora observed).
14. 0 CoRAL SAND, loose, incoherent, 90 Coral fragments, larval Lamellibranchs and Gastropods,
granular, grey in colour. bottom-living Foraminifera (Orbitolites, Spiroloculina,
Residue, black. Lagena), Echini spines, Aleyonarian spicules, Polyzoa,
| calcareous Algae, Ostracodes, Otoliths of fish, pelagic
| | Foraminifera (Globigerina conglobata, G. aequilateralis).
15. (0) CoRAL SAND, cream colour, in- 97 Coral fragments, Mollusc shell fragments, larval Lamelli-
coherent, gritty, largest piece branchs and Gastropods, Echini spines, Alcyonarian
of coral 13 mm. in diameter. | spicules, bottom-living Foraminifera (Orbitolites, Amphi-
Residue, brown vegetable stegina, Rotalia).
matter. |
|
16. 8 CoRAL SAND, white, granular, 98 Largely made up of fragments of calcareous Algae with
incoherent. coral fragments, Molluscs and fragments, bottom-living
Residue, grey particles. | Foraminifera (Miliolidae), Tunicate spicules.

587

Residue

Per cent.

Minerals

Siliceous Organisms

Fine Washings

Position

10

10

10

10

(1°/,), angular, mean dia-
meter 0°6 mm., felspar and
magnetite; no manganese
or sulphide of iron ob-
served.

(1°/,), angular, mean dia-
meter 0°6 mm., felspar and
magnetite; no manganese
grains observed.

(3°/,), angular, mean dia-
meter 03 mm., volcanic
glass and felspar, one par-
ticle of felspar observed
measured 0°55 mm. in dia-
meter, and a fragment of
voleanie glass measured
2mm. in length. A small
but perfect crystal, clear
yellow in colour, and of a
wedge shape, probably
sphene (?), was also ob-
served.

(1°/,), angular, mean dia-
meter 0718 mm., particles
of felspar only observed.

(2°/.), angular, mean dia-
meter 0'2 mm., felspar,
hornblende and volcanic
glass.

(1°/.), one or two particles
of felspar and glass.

meter 0°18 mm., particles
of felspar observed only.

None.

None.

(1°/.), angular, mean dia-

(3°/,), sponge spicules, a
few green clayey casts of
Foraminifera ; Coprolites ;
Diatoms.

(5°/.), green clayey casts of |
Foraminifera, etc.; grey
and black clayey Copro- |
lites ; no manganese grains
or sulphide of iron.

(2°/,), fragments of Radio-
laria; Diatoms.

(1°/.,), sponge spicules.

(1°/,), sponge spicules,

(2°/.), sponge spicules, Ra-
diolaria, Diatoms (Suri-
rella, etc.).

(1°/,), sponge spicules.

|

| (1°/.), sponge spicules.
|

|
(1°/,), one or two sponge |
spicules.

(6°/,), amorphous clayey
matter with minute frag-
ments of siliceous organ-
isms, casts and mineral
particles lessthan0-05 mm.
in diameter.

(4 °/,), amorphous clayey
ratter, with minute mine-

ral particles less than
0:05 mm. in diameter.

(5 °/,), amorphous clayey
matter with fragments of
minerals and small Dia-
toms.

Practically none.

None.

None.

(8°/,), made up principally

of vegetable fibre and a |

great many small brown

cocoons of larvae (insect ?). |

(2°/.,), vegetable fibre.

(1°/.), a few fragments of
vegetable matter.

Suvadiva, W. 1
mile W. b. S. of Ve-
lana I.

Suvadiva, W. 3
miles W. of Havaru
Tinadu.

Felidu, middle of
atoll.

North Mahlos,

middle east side.

North Mahlos,
S.W. Close off Fus-
mundu.

North Mahlos,
W. Within lagoon
of Bodu Furo.

Minikoi. Soil of
low planting ground,
middle of island.

| Minikoi. Soilfrom
| 3-4 feet deep near
Lighthouse.

Minikoi.

of lagoon.

Centre

588 SIR JOHN MURRAY.

17. Coratsanpd. 317. Suvadiva, W. 1 mile #. of S. end of Hondedu.

Very little material which appears to have been washed, consisting of small fragments of Coral and
Mollusc shells, one or two bottom Foraminifera (Amphistegina, Rotalidae, Biloculina, etc.), Serpula.

18. Cora. sanp. 307. Suvadiva, W. Opposite passage between Hondedu and Kandudu.

Very little material, appears to have had the finer material washed away; is grey in colour, granular,
and made up mostly of Coral fragments, bottom Foraminifera (Amphistegina, Rotalidae, etc.), broken Mollusc
shells, Diatoms, etc.

19. Coran sanp. 337. Suvadiva, W. 1 mile N. of last (not shown on Chart).

White, gritty, no fine material, made up of Coral fragments, bottom Foraminifera (Amphistegina, Hetero-
stegina depressa), Serpula, Polyzoa.

20. Cora sanp. 447. Suvadiva, W. 2 miles E. of Kandudu.

Cream colour, gritty, made up largely of fragments of Coral, bottom Foraminifera (Amphistegina, Rotalidae),
Echini spines, larval Lamellibranchs and Gastropods, Pteropods (Clio subula), Alcyonarian spicules.

21. Prerorop oozr. 427. Suvadiva, W. 1} miles N.E. of last.

Cream colour, incoherent, appears to be slightly washed, fine grained, contains Pteropods (Clio subula,
C. australis), Heteropods (Atlanta, Carinaria), pelagic Foraminifera (Globigerina bulloides, G. conglobata,
G. aequilateralis, G. rubra, Sphaeroidina dehiscens, Orbulina uwniversa), bottom Foraminifera (Uvigerina, Cassidu-
lina, Biloculina, Spiroloculina, Lagena, Rotalidae, Amphistegina), Aleyonarian spicules, Tunicate spicules, Coral
fragments, larval Lamellibranchs and Gastropods, Polyzoa, Otoliths of fish, Ostracodes.

22. Cora sanp. 34f. Suvadiva, W.N.W. 1 mile W. of Thackaru I.

Loose, gritty, grey in colour, made up of angular and rounded fragments of Coral and Mollusc shells,
Polyzoa, Serpula, calcareous Algae, larval Lamellibranchs and Gastropods, Alcyonarian spicules, bottom-living
Foraminifera (Amphistegina, Polytrema, Rotalidae, Uvigerina, Nodosaria), pelagic Foraminifera (Globigerina
bulloides, G. conglobata), Tunicate spicules, Otoliths, Ostracodes, Sponge spicules.

23. Corat conctomerate. 37—8 7. Suvadiva, N.N.W. In Huradu passage.

Piece of coral rock 44 by 3 inches in diameter consisting of an intergrowth of cup corals, etc., covered
with calcareous Algae, Serpula, Polyzoa, attached Mollusc shells, bottom Foraminifera (Amphistegina, Rotalidae,
Polytrema, Uvigerina, ete.), larval Lamellibranchs and Gastropods.

24. 367 Felidu. Centre of S.E. horn.

The material is grey in colour with black particles, shells, etc. coated with Manganese, but no grains
of Manganese observed; is gritty and incoherent, appears to be washed, contains small greyish black rock
particles, the largest observed being about 3 mm. in maximum diameter. Larval Lamellibranchs and Gastro-
pods (Pleurotoma and Dentaliwm), Echini spines, Polyzoa, Coral fragments, Aleyonarian spicules, Otoliths of
fish, Ostracodes, pelagic Foraminifera (Globigerina conglobata), bottom Foraminifera (Amphistegina, Rotalidae),
Pteropods (Clio subula), Tunicate spicules.

25. SAMPLE TAKEN FROM THE GUT OF A LARGE HonorHuRIAN. Suvadiva, between nos. 21 and 8
above.

The material is coarse and gritty, cream in colour, and consists of angular and rounded fragments of
coral, Mollusc shells, Polyzoa, calcareous Algae, Echini spines, Alcyonarian spicules, bottom Foraminifera
(Orbitolites, Rotalidae, Polytrema, Amphistegina, Uvigerina, Nodosaria, Biloculina, Spiroloculina, Cassidulina),
pelagic Foraminifera (Globigerina bulloides, G. conglobata, Sphaeroidina dehiscens), Pteropods (Clio subula, C.
australis), Heteropods (Atlanta), Ostracodes, Otoliths of fish, Serpula, larval Lamellibranchs and Gastropods,
Sponge spicules, Radiolaria, Diatoms and minute particles of minerals, mean diameter 0°06 mm, consisting
felspar and magnetite; one or two particles of vegetable matter and fibre were observed.

Fauna and Geography, Maldives and Laccadives Plate XXXIV

oro™

Great Central Flat, covered

with mud 40-50 Fathoms

°

SUVADIVA ATOLL.
(Soundings in the lagoon by the Expedition. Samples described from those underlined.)

To face p. 588.

aie es