REPORT

TO THE GOVERNMENT OF CEYLON

ON THE

PEARL OYSTER FISHERIES

OF THE

GULF OF MANAAR,

BY

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

Professor of Natural History in the University of Liverpool.

WITH SUPPLEMENTARY REPORTS

UPON THE

MARINE BIOLOGY OF CEYLON,

BY OTHER NATURALISTS.

Part III.

PUBLISHED AT THE REQUEST OF THE

COLONIAL GOVERNMENT

BY

THE ROYAL SOCIETY.

LONDON
1905.

I ol t

[ Hi 1

CONTENTS OF PART III.

PEARL OYSTER REPORT.

Page

Preface v

The Pearl Fishery of 1904 1

The Present Condition of the Pearl Banks 37

Further Report on the Parasites of the Pearl Oyster. By A. E. Shipley,

M.A., F.R.S., and James Hornell, F.L.S. (One Plate) 49

SUPPLEMENTARY REPORTS.

XVIII. — On the Sponges. By Arthur Dendy, D.Sc. (Sixteen Plates) 57

XIX. — On some AlcyonhDjE. By Edith M. Pratt, D.Sc. (Three Plates) . . . 247

XX.— On the other Alcyonaria. By J. Arthur Thomson, M.A., and

W. D. Henderson, B.Sc. (Six Plates) 269

XXL— On the Opisthobranchia. By G. P. Farran, B.A. (Six Plates) ... 329

XXII. — On the Ostracoda. By Andrew Scott, A.L.S. (Two Plates) 365

a 2

[ v ]

PREFACE.

In this Part III., I have thought it best to place proportionately more of the
Supplementary Reports and less of the Pearl-oyster work than was the case in
Parts I. and II., for the following reasons : —

1. The Supplementary Reports here printed were ready and it might have been
unfair to the authors and unfortunate for Science if there had been avoidable delay in
publication by excluding any from this Part.

2. Those sections of the Pearl-oyster work which were completed, or nearly so, are
of such a nature that the results can be communicated in writing to the Authorities
in Ceylon, and a delay of a few months in making them known to the public seemed
comparatively unimportant.

3. It seemed advisable, for obvious reasons, to have as much of the Pearl-oyster
work as possible placed together in Part IV., which will, I hope, be the final volume
of the series. This plan, moreover, presents the further advantage of allowing of
observations and conclusions to be drawn from the coming fishery in the spring
of 1905 — probably the last which will be available for the purposes of this Report.

Consequently the sections on Pearl-formation and Pearl-distribution and on the
different kinds of pearls and their natural classification have been postponed, and I
have inserted in this Part only (l) a discussion of the recent fishery (1904) based on
a report sent to me by Mr. Hornell, (2) a statement as to the present condition and
future prospects of the beds of oysters on the Banks, based on Mr. Hornell 's letters
and other documents, and (3) a further account of the internal Parasites of the Pearl
Oyster, by Mr. Shipley and Mr. Hornell, in which several new species of Cestode,
Trematode and Nematode worms are described and figured.

The Supplementary Reports now issued deal with the Sponges, the Alcyoniidse
and the other Alcyonarians, the Opisthobranchiate Mollusca and the Ostracoda.

The very considerable collection of Sponges (146 species, of which 77 were new to
science) has enabled Professor Dendy to make an extensive report, which is a notable
contribution to the taxonomy of the group. At least two of the Ceylon sponges
are of economic importance. The sponge of commerce, found in abundance at
Trincomalee, is of good form and consistence and might prove worth further cultivation;
and one of Professor Dendy's new species, Clionci margamtiferce, is the boring sponge

[ vi ]

which ruins the shell of the pearl oyster and occasionally causes great mortality.
Other sponges from the Gulf of Manaar must probably be regarded as enemies, from
the damage they do in smothering the pearl oysters they encrust and surround.

Another report of a comprehensive nature, dealing with about 77 species, is
that by Professor J. Arthur Thomson and Mr. W. D. Henderson on the Alcyonaria
— with the exception of the family Alcyoniidre, reported on by Miss E. M. Pratt.
Professor Thomson has kindly provided the beautiful coloured drawings, by
Mr. G. Davidson, from which his plates have been produced. But I am really
much indebted to all the authors for the careful and excellent drawings with which
they have illustrated their reports. Mr. Farran's report deals with a considerable
number of species of a group (the Nudibranchiata) which I had hoped to have been
able to examine myself. Consecpaently, when collecting in the Gulf of Manaar, I
made some colour notes and sketches from the living animals which Mr. Farran
has been able to make use of in his descriptions. Mr. Andrew Scott records 77
species of Ostracoda, 35 of which are new to science. To all these authors I am very
much obliged for their kindness in undertaking the work and for the skill they have
exercised in carrying it out.

It may be well to indicate here the Supplementary Reports that remain. They are
all, I think, well advanced, and ought to be ready for publication some time in the
summer : — Corals, by Mr. G. C. Bourne ; Medusae, by Mr. E. T. Browne ;
Antipatharia, by Professor J. Arthur Thomson ; Polyzoa, by Miss L. R. Thornely ;
Polychseta, by Dr. Arthur Willey ; Schizopoda and Stomatopoda, by Mr. W.
Tattersall ; Macrura, by Mr. J. Pearson ; Brachyura, by Mr. D. Laurie ; Isopoda,
by Rev. T. R R. Stebbing ; Molluscan Shells, by Mr. R. Standen and Mr. A.
Leicester ; Marine Insects, by Professor G. Carpenter, and the Tunicata, by myself
The remaining sections of the Pearl-oyster work, and our Summary of Results and
Recommendations as to the Conservation of 'the Banks, should then complete the
Report.

The records of observations sent to me by Mr. Hornell during the latter part of
1904, and especially his detailed reports upon the great spring fishery and the
November inspection — upon which I have based the two first articles in this volume
(see p. 1 and p. 37) — show that very considerable changes have taken place in
regard to the condition of the beds of oysters on the paars and the prospects of
fisheries during the next few years. Under the heading " Conclusions," in the
article on the "History of the Principal Pearl Banks" in the last volume (Part II.,
p. 36) it was stated : — " Finally, there are prospects of good fisheries both next year
and in 1906 on the Modragams and several divisions of the Cheval Paar — possibly
also on the Muttuvaratu and the Dutch Modragam in the latter year. The results
in 1907 and the immediately succeeding years will, so far as we can now see, depend
upon whether large measures of transplantation are adopted without delay." In
writing of the Cheval Paar (p. 20), also, I recommended that young oysters be

r v" i

transplanted there from the Periya Paar "at the earliest opportunity and in sufficient
quantities."

Following upon that recommendation, Mr. Hornrll was authorised by the Ceylon
Government to make preparations for transplanting young oysters in quantity from
the Periya Paar to the Cheval Paar in November, 1 904, should it still be found
necessary. This was the only wise course to take. A great part of the Cheval Paar
was denuded, and there was no certainty that a fresh fall of spat would occur.

It is essential that the Inspector of the Pearl Banks should have the power, and
the means, of ordering transplantation to be carried out whenever it may be required.
In the present instance, however, the necessity has been obviated. In the late
summer, or early autumn, an abundant deposit of spat has fallen over the greater
part of the Cheval Paar, clothing the barren sandy wastes and covering the patches
of old oysters and the rock alike, so that they are said to be indistinguishable to the
diver. These young oysters, now about 4 months old, cover an enormous area,
and must be present (judging from the samples examined) in numbers that run to
thousands of millions. No doubt the vast majority of them will die off from natural
causes during the next few years, but a fair proportion should remain to yield
fisheries in 1908 and succeeding years. In the meantime, the barren areas of the
Cheval Paar denuded by the two last fisheries are once more occupied. Transplanting
is, for the moment, unnecessary, but cultching of some parts of the ground is urgently
required. The position of affairs and the work necessary to improve the condition of
the banks are fully discussed in the articles that follow.

The other change in conditions which has to be announced is that, of the various
beds of adult oysters on the Cheval and Modragam paars, some are now dying off, or
will probably do so during the coming year, while others are apparently healthy and
undiminished. Under these circumstances certain beds (detailed in the article on the
Present Condition of the Pearl Banks, p. 44) are clearly indicated for fishing first, and
fortunately there seem to be sufficient adult oysters in sight to yield at least two and
possibly three successive good fisheries. But, although the prospects are so good, it
is the Biologist's duty to point out that now — when there are so many beds of oysters
of different ages living together on the Banks — is especially the time when vigilance
and effort must on no account be relaxed, when inspections must be frequent and
complete and the scientific examination of sample oysters must be thorough, when
cultching of sandy ground must be extensively undertaken, when thinning out of
overcrowded beds should be carried out, and when the Inspector should be authorised
to dredge up and transplant young oysters whenever it may seem desirable.

Next after the healthy development of beds of oysters comes the question of
infection with the parasites necessary for pearl production. In the preface to
Part II. the further work that lay before us in tracing the Fish-host upon which this
infection must depend was pointed out, and Mr. Hornell has continued to take
every opportunity to get information and specimens bearing on the matter. Since

[ viii ]

the November inspection he has devoted some time to work amongst the fishermen on
the shores of Dutch Bay (between Colombo and the Pearl Banks) with the view of
obtaining for dissection some of the larger Elasmobranch fish caught by the natives.
In this he has been very successful, and recent letters inform me (l) that he has
now determined the Trygon-Rays which eat the oysters as being the " Walwadi
tirikkai " of the Tamils, Rhinoptera javanica, and the still larger " Mundakanni
tirikkai," Rhinoptera adspersa ; and (2) that amongst the Entozoa he obtained from
the former species was a Tetrarhynchid agreeing in the armature of its proboscides
with the larval Tetrarhynehus unioniformis from the Pearl oyster. As the result
of this work with the fishermen, Mr. Hornell has recently made a large and varied
collection of parasites which will be the subject of a final report by Mr. Shipley
and himself in Part IV.

W. A. HERDMAN.
The University, Liverpool,
January, 1905.

REPORT ON THE PEARL OYSTER FISHERIES
OF THE GULF OF MANAAR.-PART III.

THE PEARL FISHERY OF 1904.

[Based on a Manuscript Report and Letters from Mr. James Hornell, F.L.S.,
Marine Biologist and Inspector of Pearl Banks.']

Since this present investigation was commenced in the spring of 1902, the
Government of Ceylon, after an interval of eleven barren years, has held two most
successful pearl fisheries — the last of which, that in the spring of 1904, yielded
over forty-one millions of oysters, and established a record from having given receipts
amounting to well over a million of rupees. This is, however, by no means the whole
of the value of the fishery, since the divers carry off as their share one-third of the
spoil, and the benefits conferred directly and indirectly upon thousands of attendants
and traders may be said to affect the prosperity of a large district. But, dealing
only with nett proceeds, these two last fisheries combined have brought in to the
Government of Ceylon considerably over £100,000. The prospects for another good
fishery in the spring of 1905 are now excellent, and the adult oysters at present
known to us on various parts of the Cheval Paar ought, if no unforeseen calamity
occurs, to yield fisheries in 1906 and 1907. If cultching of the barren parts of the
< 'lieval Paar and transplantation of young oysters from the Periya Paar, as recom-
mended in the earlier parts of this Pieport, be carried on with vigour there is every
reason to believe that still other fisheries will follow. Spat may fall naturally upon
suitable ground at any time, or it may be prevented from so falling, for a series of
years, by natural causes beyond the control of man. It cannot be trusted to come
when and where it is wanted and we must not depend upon it.* But if a continuity
of profitable fisheries, or even a fairly constant succession, can be ensured by the

* Since the above was written, Mr. Hornell's report on the November inspection shows that a fall of
spat has taken place on the Cheval Paar ; but I leave my words unaltered, as I believe them to be generally
applicable, and their truth is not affected by this unexpected, somewhat exceptional, but very welcome,
occurrence. — W. A. H., January 2nd, 1905.

B

•2 CEYLON PEARL OYSTER REPORT.

expenditure of a few thousand pounds annually in cultching the hanks and trans-
planting the young oysters, it is evident that the proceeding will he economically a

sound <>ne.

Mr. Hornell, since I left him on the pearl hanks in April, 1902, has kept me
supplied with frequent letters and occasional longer reports — constituting, in fact, a
journal of his doings and a discussion of his observations. He has now furnished me
with a very full report upon the record fishery of 1 904, and from this, supplemented
by the letters received from him during the progress of the fishery, I have drawn up
the following account. This Report upon our pearl oyster investigation would, 1
feel, have been incomplete without some description of the conditions under which a
great fishery is held and of the methods adopted, both administrative and scientific,
on the occasion.

Mr. Hornell's MS. report, though giving most valuable information, was not in a
form intended for publication. It was sent to me as material to mule use of. Parts
of it, such as details of the " Racial types represented among the divers," are more or
less irrelevant to our present purpose ; other parts, such as discussions of the value of
" Beacon marks," lists of " Casualties," and record of " Services of Staff," are of purely
local interest from the administrative point of view. Consequently a good deal has
been omitted and much has been re-cast. I have frequently, however, as will be
seen from the quotation marks, employed Mr. Hornell's own words. Elsewhere his
account has been used as the basis of the statements, or as the text which has called
forth remarks or upon which an argument has been founded. From the knowledge
which I gained of the locality and of the conditions of life during my visits to the
pearl banks in 1902, on the " Lady Havelock" and the " Rangasami Puravi," I find it
easy to follow and to realise the changes in the oyster beds; and I have no hesitation,
in the light of the fresh evidence, in pressing my former recommendations that the
"paars" be cleaned, thinned and improved by dredging and cultching, and that
deficiencies in spat deposit be made good on all possible occasions by the trans-
plantation of young oysters.

THE PREPARATORY INSPECTION OF THE NORTHERN PEARL

BANKS.

From the 26th January, 1904, on which date Mr. Hornell was appointed
Inspector of Pearl Banks in addition to his duties as Marine Biologist, until the 18th
of the following month, his time was largely occupied with the multifarious prepara-
tions necessary before sailing for the pearl banks. Boats had to be repaired and
painted, old stores overhauled and new ones procured, and at the same time a general
supervision given to the fitting up of the steamer ' Ready," which was to serve as an
oyster-dredger. However, by February 18th, the last stores were on board, inclusive
of over 6900 gallons of fresh water for the 70 men (divers, munducks, boatmen, &c.)

THE PEARL FISHERY OF L904. 3

lor whom provision had to he made, and the Inspector left Colombo in the native
barque " Mohideen Bux" in tow of the steam dredger.

On arrival at Marichchukaddi, near the mouth of the Modragam River, the place
selected as the centre of the pearl fishery and the site of the camp, on the morning of
February 20th, it was found that — " no Government divers had arrived. The next
day, seven Manaar men having put in an appearance, I resolved that the urgency of
immediate inspection was too great to permit waiting longer for the laggards and
left at G.40 a.m. to do the best possible with these few divers — half the usual
complement. The first step was to locate Captain Donnan's sunken tanks marking
the ' shoal buoy ' position on the tail of Karativo shoal, and I was so fortunate as to
find them within 10 minutes after anchoring on the place where I reckoned them to
be according to the bearings of landmarks ashore. After laying down a large beacon
buoy, as is usual at the beginning of inspections, I proceeded to take up my first
inspection position at the centre of the South-west Cheval Paar."

Although the banks have been inspected on the previous November, it is always
necessary to make an inspection of the beds of oysters which it is proposed to fish
immediately before a fishery opens, so that the Inspector may be in a position to
mark off the ground and regulate the number of boats and days allotted to particular
areas. The method of inspection adopted was that introduced by Captain Donnan
(who was Inspector of the banks during our visit in 1902), modified in some details.
The essential features are as follows : — Three flag-buoys are laid out by the attendant
launch or tug-boat in the direction of each cardinal point of the compass, at distances
apart of \ mile, the inmost buoys taking their distance from the inspection vessel,
which is anchored to serve as a pivot mark in the centre of the circular area to be
inspected (see fig. 1, p. 4).

Four inspection boats (modified whaleboats), each manned by a crew of six, together
with three divers and two munducks, under the charge of an experienced coxswain,
take up equidistant positions between the ship and the first flag-buoy on the north
radial line, and row slowly round the ship, retaining with wonderful accuracy their
relative positions the while. At regular intervals the crew rest on their oars to allow
the divers opportunity to make descents and bring up oysters if any are present.
The result of each dive is reported to the coxswain of the boat, who records
the condition of bottom and oysters upon a diagram form with which he is provided.
The oysters are retained in the boat for the Inspector to examine. The four
boats, having each made a complete circuit, are next ranged in line abreast in the
same manner as before, between the j-mile and the J-rnile flag-buoys, and each then
makes a second circuit. The day's work is completed by a third and last series of
circles, in this case between the buoys distant respectively \ mile and \ mile from
the ship. The four boats thus make a total of twelve concentric circuits, each boat
making three. The results shown upon the coxswain's diagrams — each of which has
three concentric circles drawn upon it representing the three circular paths covered —

B 2

CEYLON PEARL OYSTER REPORT.

are transferred by the Inspector to a final diagram or plan furnished with twelve
concentric circles. When this has been done, the distribution of old and of young
oysters is graphically shown for a circular area having a diameter of 1^ miles (fig. 1
represents two such inspection areas on the Muttuvaratu Paar in 1902).

Fig. 1. Inspection charts of (A) northern and (B) southern parts of Muttuvaratu Paar in November,
1902. There are four concentric circles made by the divers' boats between the centre and the
-£-mile buoy, four between that and the i-mile buoy, and four from that to the f-mile. Each
complete area is therefore 1£ mile in diameter. The numbers enclosed in rings indicate depths
in fathoms. The numbers on the concentric circles give the quantities of oysters brought up at
a dive on that spot. The cones indicate dives on a rocky bottom with no oysters. The dotted
line therefore surrounds the oyster-bearing area.

After calculating in square yards the area occupied by oysters, the approximate
number of oysters thereon may be estimated by taking the average number of
oysters per dive (ascertained by examining the divers' results) in conjunction with
the average amount of ground which a diver is credited with being able to clear
at one descent. Usually this area is considered, on normal ground, to be from
2^ square yards to 3 square yards. By assuming the area per dive to be 3 square
yards, the danger of an overestimate is avoided. Inspection estimates are usually
less than the total number of oysters obtained at the ensuing fishery. " Thus, at
the 1904 fishery, I estimated the number of mature oysters to be 35,000,000, whereas
the number actually fished from the area in question was roughly 37,000,000
(41,000,000 minus the amount of immature oysters fished on April 2nd)."

The North and South Modragam Paars were examined on March 5th, and in
addition a series of diving and dredging traverses was made further north, in the
region of the Vankali Paar, in fruitless search for the oysters found there during our

THE PEARL FISHERY OF 1904. 5

visit in 1902. "The absence of good landmarks in the Aripu district made the
inspection of this paar difficult, and I had to rely almost entirely on biological data,
such as the nature of the bottom and the fauna, for the verification of my position.
Without this help I could not have arrived at any certain conclusion."

How the Valuation Sample of Pearls is Procured. — Before a fishery it is also
necessary to obtain samples of the oysters to be fished in order that their pearls
may be officially valued for purposes of advertising the fishery. Consequently, during
intervals in the inspection, three large samples of oysters of fishable age were
collected from the three western blocks of the Cheval Paar, partly by means of
divers and partly by the use of dredges. The results obtained showed the sample

Pig. 2. Valuation sample of pearl oysters from the Cheval Paar, being brought on board the
" Rangasami Puravi," from the inspection boats — from a photograph by Mr. HoRNELL.

of the South-west Cheval to be well ahead of the other two lots in the weight and
the number of pearls to each 1,000 oysters; the Mid-west was extremely poor in
both respects. The washing of the sample oysters and the sifting and subsecpient
valuation of the pearls was carried out in the usual way, the two former processes
by the Government divers and munducks, the latter by three Mohammedan pearl
merchants ; all under the constant supervision of Mr. Hornell, assisted by Mr. V.
Vraspillai, the able Adigar of Musali.

" When the samples are brought into the Government Kottu, a palisaded enclosure,
the sacks of oysters are emptied into dug-out canoes or ' ballams,' and covered
carefully with mats. In this state they are left under guard for a period of from

f, CKYLON PEART, OYSTER REPORT.

7 to 10 days. Bacterial putrefaction alone is not relied upon to get rid of the flesh,
the maggots of a species of blow-fly being the agency most in evidence and most
desired. Under favourable conditions, at the end of 7 to 10 days, the so-called rotting
period, the fly maggots should have eaten their way through everything, and have
left nothing but empty shells and pearls behind. When the oysters are rotted in
sacks, the time for washing is rendered evident by the pile of brown chrysalis cases-
full and empty — that litter the ground immediately around the pile of sacks as well
as encrust the sacks themselves.

" The time to wash having arrived, the covers are removed from the ballam, and
coolies fill it to the brim with water. As the water level rises to the edge, a mad
race for the points above the flood ensues among the maggots which float up in
myriads from the lower recesses. The washers range themselves in line along
either side, squatting on anything convenient — a canoe outrigger, an overturned tub,
empty kerosine tins, and the like. They are stripped to the loin-cloth, and are not
allowed to take their hands out of the water save to drop out the empty shells.
The first process is to rinse the shells thoroughly, to separate the valves, and, by
rubbing the outside of one valve against that of the other, to remove any detritus
in which a pearl might lodge. The men scrutinize the nacreous lining for attached
or shell pearls, placing any found in a special basket. The other shells, after a final
rinse, are dropped outside the ballam at the washers' feet.

" After the quantity is reduced somewhat, the floating maggots are skimmed off,
lifted by hand, and some of the water is baled out through a sieve, any material
that remains therein being carefully returned to the ballam lest a pearl might be
contained or entangled in the dirt. More water is then added, and the process of
washing the shells is continued. At last all the shells have been removed and the
men are then free to stand up and stretch their cramped limbs.

" A fresh supply of water is now poured in till the ballam overflows — a rough method
of elutriation. Time after time this is repeated till the bulk of the lighter filth is
got rid of. Then the remainder of the water is decanted, and the heavy debris in
which the pearls are mingled is exposed on the bottom. More water is repeatedly
poured in, the detritus or ' sarraku ' the while being kneaded and turned over and
over again.

' When this apparently interminable cleansing process comes eventually to an end,
every scrap of sarraku is removed with scrupulous care to a cotton cloth and bundled
up. One ballam full of oysters will usually furnish from two to three such bundles
of pearl-containing dirt. The bundles of sarraku are opened later and spread to dry in
the sun, undergoing a preliminary search at the beginning, and if wished at intervals
during the drying. These early search ings yield usually the largest and therefore
the most conspicuous of the contained pearls.

"When dry, the material is sifted out into several grades, and each is gone over
time after time by the men employed (fig. 3). The final search — after it appears

THE PEARL FISHERY OF 1004.

that even the dust-pearls, the ' masi-tul,' have been removed — is carried on by women
and children, and it is amazing to see what a large quantity of small pearls their
keen eyes and fine touch enable them to obtain. After the pearls are picked out it is

Fig. 3. Government divers picking out pearls from oyster washings. — Photo, by Mr. Hornell.

the custom to offer the apparently exhausted dirt for sale, and ready buyers can
always be found."

After the valuation was completed, and the results had been telegraphed to the
Colonial Secretary for the purpose of advertisement, Mr. Hornell utilized the few
remaining days, prior to the advertised opening of the fishery on the 14th March, in
an examination of the Karativo and Dutch Modraoam Paars.

Method of Marking out the Fishing Ground. — On the 13th March Mr. Hornell
returned to the Cheval Paar and took up a position on the south-west corner of the
south-west section, and marked out the ground for fishing in the following manner :—

" As the plan adopted was a novel one, the Government Agent kindly acceded to
my request to notify the method by beat of tom-tom in the divers' quarters,
intimating that the boundaries of the fishing ground would be marked by buoys
bearing red flags, while a series of white flag buoys would be placed wherever
fishable oysters were present. The divers were instructed to cluster their boats
round the various white flags.

" These oyster buoys were placed according to the distribution of oysters mapped
out after the inspection, but to guard against the possibility of mistake I verified the
abundance of oysters at selected spots by preliminary dives before anchoring the
buoys in position. The system proved a success, as by placing but a few white flags
down at a time 1 was able to control the movements of the fleet and ensure the bank

8 CEYLON PEARL OYSTER REPORT.

being fished in a methodical manner. At the next fishery I propose to introduce
further improvements in this flag system, among others the marking of the flags
individually with distinctive numbers and signs.

" Discovery of Brood Oysters on the Periya Paar. — The advertised date of opening the
fishery found but a handful of divers assembled at the Fishery Camp at Marichchu-
kaddi. The commencement being in consequence postponed, I took the opportunity
afforded to examine the Periya Paar, and ascertained that over an area measuring
at least 5 miles in length by lj miles in width the bank bore its accustomed
brood of young oysters.* As oysters have but once come to maturity upon any
part of this paar during the past century, and that but a small patch, I hastened
to recommend to Government a wholesale transplantation to the Eastern Cheval —
the area cleared of oysters during the 1903 fishery — requesting the use of the dredging
steamer for the purpose. This was granted, and I inspected a likely area on the
East Cheval and laid down on the selected ground a set of five mark buoys for the
guidance of the Master of the ' Ready ' when engaged upon the distribution of the
dredged oysters. Unfortunately for these plans, the dredger, by reason of rough
weather, was able to put in but 2^ days at this work prior to April 9th, on
which date I received instructions to place the ' Ready ' at Mr. Dixon's disposal
for the purpose of dredging mature oysters, and when, later on, the vessel again
reverted to me, continuous bad weather had set in, rendering it impossible to work
upon the exposed Periya Paar. As a consequence I had to cancel the cultching
arrangements I had made to prepare the ground, and transplanting operations are in
consecmence postponed probably for at least a year.

' With regard to the future need of transplantation, 1 can but endorse Professor
Herdman's reiterated statement that in ///is lies the true salvation of the pearl
banks. This being so, if Government approves of such recommendation, it will be
absolutely necessary, in order to ensure success, to give me permission to arrange for
and carry out an extensive programme of transplantation and cultching and to
provide me with adequate means for the purpose, that is to say, I should have a
dredging steamer at my disposal from the very date of finding young oysters fit for
transplanting, and this steamer should be devoted solely to this work for the whole of
I lie fine weather available and suitable for dredging. Only by transplantation on the
greatest possible scale can success be assured. We have to remember the largeness
of the object in view — nothing less than the formation of a bed of oysters of sufficient
extent to allow for extensive wastage, due to a dozen different causes, and after
such wastage to give a fishery of mature oysters calculable in millions.

" Necessity for Extensive Cultching.- —Our observations show that the sections of the
Cheval and Modragam paars giving the finest oysters and the most valuable pearl-
yield are those possessing a well-cultched sandy surface, with somewhat restricted
rocky outcrops. Unfortunately such areas are comparatively limited in extent, and
* The condition of these oysters will be discussed further on.

THE PEARL FISHERY OF 1904. 9

tlie major part of the Oheval, the north, north-east, mid-cast and part of the south-
east and south-west, with all the three central sections, consists of sandy hottom
inadequately supplied with material suitahle to serve as foothold to oysters.

We now have sufficieni knowledge of our local conditions to see that trans-
plantation must go hand in hand with extensive cultching if we are to obtain a full
return for the labour and money expended on the former operation — indeed we may
go further and say that the transplantation of young pearl oysters will be labour
wasted unless adequate cultching operations he carried out concurrently. Appreciation
of the vital importance of this fact caused me to apply for authority to obtain several
hundred tons of broken rubble for cultching purposes as soon as transplantation was
seen to be advisable last March. The proposal was at once sanctioned, and while
waiting to begin transplanting I arranged, by the kindness of the Public Works
Department, for the preparation of a trial shipment of 100 tons of broken calcareous
sandstone from Kalpitiya. The stone was to be roughly broken to the size of the
fist — about Sh X 3^ X 2^ inches. The full quantity had been prepared when the
transplanting scheme was arrested and now lies ready for shipment whenever it be
next required.

" It may be useful to add here that whenever good prices cannot be obtained for the
pearl oyster shells accumulated during a fishery, the best purpose they can be put to
is to relay them upon the depleted banks, where the ground is in need of cultch.

;' I calculate that at the very lowest estimate 1 ton of shells will furnish as much
cultch as 3 tons of Kalpitiya stone or rubble. As this quantity of the latter
costs Rs. 11.25 (three tons at Rs. 3.7ft per ton) I believe that it will be more
economical for Government to refuse to sell the shells at any less price than Rs. 10
per ton, and to use them as cultch unless this minimum price be obtained.

"Clean Banks Essential to Successful Cultivation. — In this, as in the matter of cultching,
we may with great advantage profit by the experience of European oyster culturists.
They find it necessary to check the growth of many other organisms upon the banks,
not only those that are active enemies of the oyster — star-fishes, whelks and the like-
hut also such animals as curtail the area which oysters may occupy or which consume
food that would otherwise fall to the oyster. Seaweeds, too, are for the most part
weeded out. As a consequence much can be done in cleaning ' foul ' beds by means
of the dredge. If the beds are in preparation to receive spat, all harmful matter is
taken ashore — both direct enemies, such as star-fish, whelks, mussels, and also the
various other animals that compete for food and so may be termed the indirect enemies
of the oyster.

" Fortunately the Ceylon pearl banks are comparatively ' clean.' No fact struck
me so forcibly when first I began to make diving descents as the paucity of
injurious life on the better quality of the oyster-bearing paars. Wherever oysters
were found thriving, wherever they were seen in vigorous growth and perfect health,
the bank appeared to be comparatively free from deleterious matters.

c

10 CEYLON PEARL OYSTER REPORT.

" Where ' foul ' areas are found, i.e., where Sponges, Corals, Alcyonarians, Echino-
derms and Ascidians abound, as on many parts of the Western Cheval and Muttu-
varatu paars, the oysters, while numerous enough, are stunted and poor, suffering by
competition with the host of creatures living upon the same diet of microscopical
organisms. The only means of cleaning such a bed is to dredge it thoroughly,
separating and removing the materials brought up."

On the Ceylon pearl banks the beds are too extensive to permit of dredging being
undertaken with this sole object in view, but as this cleaning can and should go on
concurrently with the dredging of spat for transplantation or of mature oysters for
sale, we have herein an additional argument in favour of taking up dredging on
a scale of considerable magnitude. We must not lose sight of the fact that dredging
has a fourfold utility, namely, (a) in fishing oysters, (b) in cleaning the ground and
removing enemies, (c) in thinning out overcrowded beds, and (d) in spat transplanta-
tion. Its value is not properly assessed if account be taken of the first item alone, or
even of the first and the last. Every live coral removed and replaced by a fragment
of clean cultch may mean the addition of three oysters at the next fishery ; every
star-fish destroyed probably means scores of oysters saved from destruction ; every
Cliona-rlddled block of coral bleached on the shore will tend to reduce the widespread
havoc this inconspicuous sponge causes amongst the oysters. The immense advantage
that accrues from keeping the banks in a state of thorough cleanliness can well be
appreciated by an agriculturist, who knows how his crops fall off if weeds be left
unchecked, and if fungoid and insect pests be ignored.

Beacon Marks. — Sir William Twynam has pointed out the need of fixed, well-
defined landmarks from which bearings may be taken at all parts of the banks ; and
he states his belief (p. 20 of his 1902 Report) that this want of adequate landmarks
resulted in the wrong bank being fished in 1836 — two beds of young oysters being
fished instead of the one proposed. Again be states that the fishery of 1860 on the
Modragam was all but lost for the same reason — a long continued search of 3 days
being necessary before the bed was re-discovered.

A notable advance was made this year by the erection of a permanent beacon on
the tail of Karativo shoal, immediately to the south of the South Cheval Paar, to
serve as the point of departure at the inspection of the whole Cheval area. Many
years ago (1878) Captain Donnan sank an iron tank close to this spot and subsequently
(1883) he added a second. It was his custom at the beginning of each inspection to
locate these tanks by bearings taken from Kudiramalai and by means of the inspection
divers, and, when found, to mark the position tenqjorarily during the inspection by a
large beacon buoy — the so-called shoal buoy.

Captain Donnan's successor, Captain Legge, appreciating both the loss of time
involved in the oft recurrent hunt for sunken tanks, and the danger of taking
incorrect bearings when these tanks could not be located, requested sanction to have
a permanent beacon erected. This was granted, but all efforts to erect it last year

THE PEARL FISHERY OF 1004. 11

were rendered futile by unfavourable weather. Before leaving for inspection this
year, Mr. Hornell pressed upon the attention of Government the importance and
urgency of another effort being made to place the beacon in position. Accordingly it
was sent up once more under the care of Mr. Lindsay, the general foreman of the
Harbour Works Department, who, after very great difficulties due to the heavy
ground swell always present on the shoal, eventually completed the erection
successfully. One great safeguard has thus been created against the loss, by oversight,
of oyster patches in the Cheval region, and valuable time will also be saved at
inspections. " This Karativo shoal beacon consists of four iron piles driven 6 feet
into the sand, braced together securely and bearing an iron telegraph pole secured by
four wire stays. The apex of the pole bears two circular metal discs set vertically at
right angles, the top being 43 feet 6 inches above sea-level, and therefore visible
at a distance of Hi miles to an observer whose eye is 12 feet above sea-level. The
depth of water where and when it was erected was 25 feet 6 inches at low tide. The
position of the sunken tanks is latitude 8° 39' 15'' N., longitude 79° 45' 0" E. ; and
the new beacon is distant therefrom 233 feet N., 3° W."

Mr. Hornell states : — " My experience during the last two years, and especially
during the last inspection, makes clear to me that a beacon in the Aripu district,
together with an improvement by heightening of the present Kudiramalai beacon, is
essential to efficient inspections. With respect to the cost of these, I would point out
that whatever it might be, the amount would be as nothing compared with the loss
of revenue entailed by the loss of a bank of fishable oysters through lack of proper
beacon marks — a mischance which has occurred in the past probably oftener than we
imagine. At a later date the scheme of land-marks might be completed and perfected
by the erection of a beacon at the Doric and by the raising of Kallar Tower another
20 feet. For the present, however, urgency centres in the raising of Kudiramalai
Tower and the erection of a beacon on Aripu Reef."

PARTICULARS OF THE FISHING OPERATIONS.

The news that a pearl fishery is advertised to take place in the Gulf of Manaar is
conveyed throughout India and other parts of the coast of the East with amazing
rapidity, and at the appointed time a heterogeneous population of from 20,000 to
perhaps 40,000 natives gathers in a few days on what is usually an uninhabited
desert — a bare sandy coast with jungle extending for miles inland. A temporary
town, covering a site of about one square mile, and having streets (Main Street, Old
and New Moor Street, Tank Street, Divers' Street, Chetty Street, Muttu Bazaar, and
the like) of houses and shops, Government buildings, such as court house, prison,
hospitals, markets, and offices of various kinds, is rapidly run up. ' Camp Town ' is
erected under the supervision of the Adigar of Musali, largely of bamboos and palm
leaves (fig. 4), and is a prosperous and very busy centre of industry and trade during

C 2

12

CEYLON PEARL OYSTER REPORT.

the few weeks of its existence. In the neighbourhood huge water-tanks are
constructed, some for the drinking supply, others for washing and for the divers to

Fig. 4. Beachmaster's (Muttutamby) house, tom-tom beater and other natives. Pile of eadjans
(palm-leaf mats) for building in foreground. — Photo, by Mr. Hornell.

bathe in when they return from the day's work. At the same time the fleet of some
200 fishing boats assembles on the shore (fig. 5), and a port with the necessary officials is
constituted. The whole organisation is under the Government Agent of the province.

Fig. 5. Part of the pearl-divers' fleet at Marichchukaddi. — Photo, by Mr. Hornell.

" Four distinct racial types were represented among the divers both at last year's
fishery and at this — Coast Tamils, chiefly from the Madras Presidency ; Moormen, or

THE PEAKL FISHERY OF 1904. 13

Lubbais, drafted largely from Kilakarai and the neighbouring villages on the Madura
Coast of the same Presidency ; Malayalam men, from the southern portion of the
Malabar Coast (Travancore principally), and lastly a compact body of some 250
so-called Arabs from Colombo and Jaffna. The first two groups comprised the bulk
of the diving community (including both divers and munducks), accounting for 1989
and 3732 respectively out of a total of 7408 who attended at last year's fishery
(vide ' Report on Fishery of 1903,' by Mr. R. W. Ievers, C.M.G.)."

Most extravagant statements are current in regard to the time a diver can remain
under water. The Arabs are far ahead of all the others in endurance, their time
under water being generally from 70 to 85 seconds. The Tamil and other divers have
a lower average, varying from 35 to 50 seconds. The maximum for the best divers,
in our experience, is not over 90 seconds. The diving goes on from daybreak
(about 6 a.m.) to noon, and the divers' boats set sail for the banks soon after midnight,
taking advantage of the land wind which then blows off shore, so as to be readv to
start work with the light.

" On the 15th March, twelve Colombo and Jaffna boats, manned entirely with
Arab divers, had taken up positions by daybreak around the barque. As is the habit
of Arabs, these men fished quietly and methodically, and with great energy. By
noon they were tired and satisfied with their catch, an average of 22,811 oysters per
boat, and, hoisting sail, made for the camp. The next day a number of Tamil and
Moorish divers appeared, and on the 5th day of the fishery the number of boats
actually at work had increased to 148, while for several days in the height of the
fishery just over 200 boats (202) were employed.

' The boats used by the divers at this fishery were of at least five distinct types —
narrow single-masted canoes, with an outrigger, from Kilakarai on the Madura Coast ;
larger 3-masted canoes, from the Tanjore Coast ; large ' Dhoneys,' from Jaftha and
the north of Ceylon ; large sailing lighters, from Tuticorin ; and ' luggers,' with
square sterns, from Paumban. Some of the large boats accommodated as many as
30 divers, and with the ' munducks ' (the men who attend to the divers' ropes) and
crew must have had about 65 men on board.

" Fishing on the South-west Cheval went on methodically and in a perfectly
satisfactory manner for 17 days, by which time the bank proper was practically
cleared of oysters. On the 18th day of the fishery the fleet, instead of settling down
to glean over the fished ground, flocked to a patch of mixed oysters on the north-east
corner of the south-west section, and as a heavy swell was rolling in from the south-
west I did not judge it advisable to attempt to move them away. This day's catch
of oysters proved phenomenal, totalling close upon 4,000,000 (3,919,712), of which
the Government share, 2,609,445, sold for Rs. 39,058.79, being at an average rate of
nearly Rs. 15 per 1,000. The oysters on this patch lay extremely thickly strewn in
bunches upon a sandy bottom ; I had not included them in my estimate of the south-
west bed, considering it preferable to leave them for next year's fishery. However,

14 CEYLON PEARL OYSTER REPORT.

the thinning out which this day's fishing entailed will have excellent results and
conduce to more rapid growth during the intervening 12 months.

" Judging it time to abandon the south-west region, I moved the ship to the bed
upon the North-west Cheval on the 3rd April, laying out a series of red flag buoys
along the boundary and white flag buoys as fishing centres, as upon the South-west,
and accordingly the fleet began work there upon the following day. A break in the
hitherto excellent fishing weather unfortunately coincided, and from the 5th April
more or less unfavourable weather prevailed steadily and practically without inter-
mission, rendering fishing increasingly difficult. As a consequence the north-west bed
was not exploited in the thorough manner characteristic of the fishery on the south-
west section, and whereas 26,000,000 5-year-old oysters were taken from the latter
bed as against an estimate of 19,500,000, upon the north only 4,500,000 were fished
as against the estimate of 9,000,000. This less satisfactory result was, however, not
due entirely to rough weather ; many large patches were so thinly furnished with
oysters that 5, 7, and 10 oysters per dive were the best possible — a condition of
matters ill-suiting men who had been accustomed for over a fortnight to averages of
25 and 35 per dive.

" It became expedient therefore to open the last section — the mid- west — to fishing
on April 12th, and on that date the boats lifted 1,244,070 oysters from this bed.
From this date the wind increased greatly in force and, being accompanied by a
heavier swell, it became impossible to set definite limits to the ground to be fished
upon. The number of the boats at work began to diminish rapidly, and by April 1 6th
the fishery terminated, so far as the fleet as a whole was concerned.

" From the 16th to the 18th April the weather was too boisterous to permit of any
fishing. On the latter date, some of the Arab divers being willing, it was arranged
that they should be sent out aboard the steamers ' Ready ' and ' Serendib ' to carry
on diving from these vessels. In all 44 responded to this emergency call, and
I superintended their work on April 19th, beginning fishing at 10.10 A.M. and
ceasing at 3 p.m. During the whole time the weather conditions were the worst
possible, a heavy sea running, accompanied by a chill, cutting wind. The men worked
most pluckily and with the utmost good humour. The majority did not touch food
all the time they were aboard, from 5.30 a.m. to nigh on 6 p.m. This day's work by
these 44 Arabs stirred up emulation in the breasts of their co-religionists, the
Kilakarai Moormen, and enabled the fishery to be carried on in spite of the boisterous
character of the weather for 4 days longer and with a gain to Government of
Rs. 15,580.

" By this time the great majority of the diyers' boats had left Marichchukaddi,
and on the 23rd April it was seen to be impossible to attempt to further prolong the
fishery, which thus expired, after a nominal length of 33 fishing days ; on 7 days,
however less than 200,000 oysters fell to Government share, owing to the unfavour-
able nature of the weather on these days ; so under these circumstances the grand

THE PEARL FISHERY OF 1904.

15

total of just over 41,000,000 oysters was an extremely satisfactory result. Last
year's fishery produced an almost identical number, 41,169,637, spread over 38 fishing
days of continuously fine weather. Had similarly favourable conditions prevailed this
year the divers would certainly have fished another 8,000,000 oysters, and would
probably have raised the grand total to 50,000,000."

It is said that even the large amount of pearls obtained from these oysters is not
the total number fished up, as during the 3 or 4 hours' sail home from the banks
opportunity is given to the divers and their boats' crews to abstract some of the finest
from an occasional gaping shell. These illicit pearls, concealed it may be in the
eyes, or in the cavities of the nose, are often discovered by the official searchers
before the men leave the Government kottus. So there is some ground for supposing
that the divers really carry off a good deal more than the one-third of a fishery to
which they are entitled. The kottus where the oysters are received fx-om the boats
and separated into piles are large enclosures close to the beach, surrounded by high
fences of wattling (fig. 6) and having sheds thatched with palm leaves. Here the

Fig. 6. Government divers and munducks picking out pearls in the kottu. — Photo. l>y Mr. Hohnell.

Government officials and clerks assign a separate compartment to each boat's crew as
it arrives with a load of oysters. The first proceeding is for the divers to divide
their oysters into three heaps, one of which is destined to be their own, and there
can be no doubt that they make these piles as nearly equal as they possibly can,
since they have no means of knowing which one the official will presently point out

16

CEYLON PEARL OYSTER REPORT.

as their share. The divers now carry off their own oysters and emerge from the
kottu on the landward side after running the gauntlet of the searchers. They are at
once surrounded by a crowd of natives anxious to buy oysters in small quantities, by
the dozen or even singly. The diver usually disposes of his whole stock in this way
in a very short time, and then spends the remaining bonis of the day in bathing
and resting.

Inside the kottu the clerks are now counting the millions of oysters left as the
Government's share ; and about sunset the Government Agent puts them up for sale
by auction at the Court House. The prices bid are per thousand, and may vary
from 20 to 80 rupees, and the buyer can take one or more thousands at the price he
bids. The kottu with its valuable contents is closed and guarded during the night,
and in the morning the buyers remove the oysters to their own enclosures, and then
the lengthy and unpleasant process of washing the rotting oysters begins. Some of
our figures (fig. 3 and fig. 6) show the examination of the " sarraku" or residue for
pearls, while fig. 7 shows Tamils, outside the kottu, searching for attached or blister

Fig. 7. Tamils searching the washed shells. — From a photograph by Mr. HoKNELL.

pearls amongst the piles of shells thrown out after washing. Fig. 8 shows two pearl
merchants, who sit all day ready to buy or sell, weighing, discussing and playing
with their gems. They have basket-like sieves for grading the pearls, and curious
little scales with scarlet seeds for certain weights. These capitalists (many of them
Moormen and Bombay merchants) are an important section of the population of
" Camp Town," There are also the pearl-cutters and " fakers," who carry on their

THE PEARL FISHER V OF 1904.

17

business in the street at little three-legged wooden work-tables and with simple bow-
drills, by means of which they pierce the pearls for stringing. All these workers
who have some definite connection with the fishery or the pearl require others to

Fig. 8. Pearl merchants. — From a photograph by Mr. Hornell.

supplv them with the necessaries and the luxuries of life, and so is brought together
the large heterogeneous population. But as soon as the fishery closes the whole
town appears to dissolve in a day, the people separate in all directions as rapidly as
they gathered, and Marichchukaddi is again a solitary sandy waste.

The Oyster Dredging Experiment.

At the 1904 fishery, amongst other new departures, some of which have already
been noted, oyster dredging from a steamer was tried for the first time with a view
to supplement the work of native divers.

The experiment was, however, carried on with imperfect apparatus and under
difficulties which will certainly disappear as further experience is gained. The
steamer " Ready," engaged for this work during the greater part of the fishery, was
found to be unsuitable in several particulars, and Mr. Hornell has made a number
of recommendations in regard to the deck fittings and gear, which will lead to greater
efficiency and economy of time by working a larger number of dredges simultaneously.

Mr. Hornell has kept an exact record of the dredging results clay by day, which,
however, considering the imperfect nature of the experiment on this first occasion,

18 CEYLON PEARL OYSTER REPORT.

need not be published. After a discussion of the results he concludes as follows : —
" At 70,000 to 80,000 oysters per day I should consider the steamer to be giving very
satisfactory results, but I wish to emphasize the fact that the best results can be
hoped for only from a few of the banks, i.e., the North Modragam, certain large
regions of the Cheval, and some of the smaller and less important paars. The South
Modragam and the remainder of the Cheval occupy a somewhat doubtful position,
and further working experiments are necessary before we can pronounce a definite
verdict regarding the prospects of commercial success in fishing by means of the
dredge. I believe, however, the advantages of dredging to be greatest in regard to
spat transplantation — as has been explained elsewhere in this Report. Dredging
is also of great assistance in the preliminary or prospecting inspections — when it is
necessary to ascertain the presence or the absence of oyster patches on grounds too
extensive for detailed examination by means of divers."

Details of the Oysters Found.

The oysters found during the recent inspection were of two distinct generations,
the older being from 4^ to 4f years old, the younger from 2^ to 2^ years. The
former were of the same brood as those fished on the Eastern Cheval in 1903, and
were therefore of a very satisfactory age for fishing this year (1904). Those of the
younger brood will hereafter be termed " immature," not in the sense of sexual
immaturity, but in regard to pearl-yield.

The fishable oysters lay as a discontinuous north and south band along the seaward
or western side of the Western Cheval. Save for a scattered remnant left over from
last year's fishery (1903) on the Eastern Cheval, and for occasional individuals
forming the nuclei of bunches of younger oysters on the parts covered by the
immature generation, no other oysters of fishable age were found. Those present
last year on the Periya Paar Kerrai and Vankali Paar were gone, and the huge bed
upon the Muttuvaratu Paar, estimated by Captain Donnan, in 1902, at a total of
277,000,000 oysters, and which would have been fishable this year, had also
disappeared, being partially replaced by a later brood of 2^-year-olds. On the
Karativo Paar, also, none of the 25,000,000 oysters, 1^ to 2 years old, found there
in April, 1902, remained, the bank being bare save for a few handfuls of 2-year-olds.
A small bed of this fresh generation of oysters was also discovered on the Dutch
Modragam Paar.

The details respecting the number and distribution of the oysters present may be
summarized thus :—

Number, Distribution, and Sizes of the Fishable Oysters.

The beds of 2^- to 3-year-old oysters on the Western Cheval (comprising the
south-west, mid-west, and the north-west sections), which we saw for the first time
during our visit to the banks in February, 1902, were found to have the same general

THE PEARL FISHERY OF 1904. 19

disposition, but to have largely decreased in area and to be represented by several
detached patches thickly covered with 4^- to 5-year-olds on the South-west Cheval,
by a large area of more thinly spread oysters of the same age on the north-west
section, and by a third of rather smaller extent on the mid-west, the southern end of
the last named joining the north-west patch on the south-west section.

"The south-west patches covered in the aggregate an area of 2,754,021 square
yards, with an average of 22 oysters per dive on 1,778,676 square yards and of 21 on
the remainder. As each dive is estimated to account for an area of 3 square yards, I
estimated the total number of oysters on this section to be 19,500,000, sufficient to
give daily loads of 10,000 oysters per boat to a fishing fleet numbering 100 boats for
a period of 19| days, each boat being presumed to have a complement of 10 divers.

' The oysters on the north-west section amounted to 9,000,000, covering an area of
1,685,460 square yards, a number based upon an average of 20 oysters per dive on
716,100 square yards and 13 per dive on 969,360 square yards. Although apparently
sufficient to give fishing for 100 boats for 9 days at an average load per boat of
10,000 oysters, the result of the fishery showed that the number of oysters per dive
on certain portions of this bed was too low to give profitable results, with the
consequence that the fleet had to be moved off after only five days' fishing and the
removal of a little over 4,000,000 oysters — rather less than half of those present.

" The bed on the mid-west had an area of 992,250 square yards, whereon I
estimated there were 6,500,000 oysters, being at the rate of 20 per dive. These
oysters were smaller for their age than those of the south-west and most of those on
the north-west sections.

" The total number of fishable oysters on the three sections of the Western Cheval
I estimated at 35,000,000, equal to 35 days' fishing for 100 boats. By far the most
important of the beds found in respect of numbers as well as condition were the
patches on the south-west section, for here not only were the oysters splendidly
grown for their age, but they were absolutely free from mixture with young. Every-
where else there was more or less admixture of young of 2 to 2^ years of age,
especially upon the eastern margin of the ground where they adjoined the beds of the
immature generation.

" Over those parts of the South, South-east, and South-central Cheval, and North
Modragam occupied by oysters, the older oysters, while nearly universally present,
were in a very small minority, averaging not more than 1 to 6 of those 2^ or 2 years
old. Here and there on the northern edge of the South-east Cheval were small
patches of little mixed 4£- to 4f-year-old oysters, the remnants of the bed fished
last year.

" As was to be expected, the Periya Paar Kerrai, ravaged by ray-fish (Trygonidas)
in March, 1903, gave no results, and but a few oysters of no fishery value were
found on the Kondatchi Paar, ground that is utterly bad for oyster growth under
present circumstances {infra).

D 2

20

CEYLON PEARL OYSTER REPORT.

" The sizes of average samples from the three sections of the West Cheval are as
follows : —

Name of Bank. Height of oyster.

Length.

Breadth.

millims.

South-west Cheval . . . 76-20
Mid-west „ . . . 72-30
North-west „ ... 72-87

millims.

69-20
65 • 25
66-50

millims.
32-32
31-70
31-57

" Some exceptionally fine oysters from a patch at the extreme south end of the
Mid-west, a patch lying upon sandy bottom, rose to the high average dimensions of
76*56 millims. by 69-16 millims. by 34*44 millims. If we exclude the last-mentioned
oysters as being exceptional, we get as the general average of the 4^ to 4f-year-old
oysters fished this year 73*79 millims. by 66*98 millims. by 31 "86 millims.

"During the course of last year's fishery (1903) I found the average size
characteristic of oysters on the Eastern Cheval to be 75*51 millims. by 69*18 millims.
by 31*18 millims. Both lots originated from the same brood, and it is significant that,
although this year's oysters are a full year older, their size, except in respect of
breadth, is decidedly inferior to the average of those fished last year from the
Eastern Cheval. This marked character, due to stunted growth, has long been
associated with oysters from this particular region, and has frequently been referred
to by Sir William Twynam and by Captain Donnan. Occasionally this stunted
growth has been so extreme that the divers and merchants have referred to such
oysters as being of the ' Kottapakku' variety, a form which is undoubtedly correlated
with overcrowding on extensive stretches of rocky bottom."

Number, Distribution, and Sizes of the Immature Oysters.

" Immense quantities of immature oysters varying from 2|- to 2^ years old covered
practically the whole space on the eastern half of the Mid and North-west sections,
extending eastwards into the North-central, Central, and South-central areas. The
South and the South-east Cheval also bear very large quantities, as do also the North
and South Modragams.

" A fair-sized bed, plentifully covered with oysters, is at present on the Dutch
Modragam Paar, but they are too young to fish for at least a year, as the age of the
majority varies from 2 to 2^- years. These oysters represent the young spat which
was noted as abundant in November, 1902 ; their growth has been at the expense of
the older oysters that were present that year, and which have been killed off by the
more vigorous and numerous young.

' The size of those on the North and South Modragam paars and those on the
South, South-east, and South-central Cheval, is greater than that of those upon the

THE PEARL FISHERY OF 1904.

21

remaining sections, including also the Dutch Modragam, indicating more vigorous
growth and a better food supply. Indeed, many of the oysters have a superficial
appearance of being from 3 to 3j years old. The general superiority of these oysters
is well brought out in the following tabulation of the average size of representative
oysters from each of the localities named : — ■

Locality.

Height of oyster.

Length.

Breadth.

South-central Cheval

North west „

North-west „ (another part) . . .
Mid-west „

millims.

6<A
56

61'
56v
68^
58|4!

r,s':'

millims.
61,',

51|

57J
52|

6 4. A

55A

millims.

27 o

■s:,\ :

23f

24|

27^

25£|

24i

South Modragam

Dutch „

Karativo Paar

" There are few oysters of an age less than 2 years mixed with these — an
important favourable factor, as the fall of abundant spat upon a bed of oysters
approaching maturity is one of the most serious dangers that beset the fruition of
an oyster bed, and one that is most difficult to combat."

SlMMARY OF THE PAST HlSTORY OF THE OYSTERS FOUND.

The oysters fished this year were first seen in March, 1900, by Captain Donnan,
who reported that over the whole of the Western Cheval, an area nearly 5% miles
long by 1000 to 4000 yards wide, young oysters 3 to 9 months old were thickly
spread. The area actually covered was estimated at 5800 acres. During the
1 2 months following a great decrease took place in the number of oysters present and
in the acreage covered, the Inspector reporting that the solid phalanx of 1900 had
been broken up into three detached beds totalling only some 1685 acres bearing-
80,000,000 of oysters.

A year later (February and March, 1902), during our visit to the banks in the
" Lady Havelock " and with Captain Donnan in the " Rangasami Puravi," we saw
these oysters and noted the marked inferiority of their size to that of oysters of
similar age upon the Eastern Cheval. Captain Donnan's estimate of their numbers
was 123,357,000 oysters, spread over an area of 2170 acres. At the same time it was
noticed as a disquieting feature that the fresh fall of spat so generally distributed
over the banks had invaded the Western Cheval — a large proportion of the older
generation bearing each several young attached to the valves. These young oysters,

22

CEYLON PEARL OYSTER REPORT.

upon which we uow rely for the next immediate fisheries, appeared when first seen
(February, 1902) to be from 2 to 6 months old. Besides having invaded the Western
Cheval, immense quantities were found on the South and North-central, South and
South-east Cheval regions, and also upon the two Modragams, as well as upon many
of the southern banks extending from Negombo to Karativo.

In November, 1902, the numbers on the Western Cheval bank were given by
Captain Legge as 57,605,000, occupying 936 acres, but, owing to bad weather, the
whole of the bank was not surveyed. No other numerical estimate was made till
Mr. Hornell's inspection of this year, when a very careful estimate placed the
number remaining at 35,000,000 upon 1122 acres. Tabulation of the above facts
shows this history clearly.*

Date.

Number estimated.

Age

in years.

Acreage.

Number
per dive.

March, 1900

„ 1901

„ 1902

November, 1902 (inspection
incomplete)

February, 1904

Too great for calculation
80,000,000 (?)
123,357,000

Over 57,605,000
35,000,000

n » if

n » n

3 ,, 3j

5800
1685
2170

Over 936
1122

21 to 30
27 „ 40

13 to 22

The discrepancy between the estimates for 1901 and March, 1902, appears largely
explicable by the estimate of oysters per dive having been either too low for the
former year or too high for the latter, estimates being liable to variation with every
condition that affects a diver's powers of working. Divers have been known to bring-
up from one particular spot only 10 oysters per dive when the weather conditions were
unfavourable, and yet the following day, the sea being calm and warm, the same men
would report double the number. To arrive at accurate estimates, the weather
conditions, if adverse, should be allowed for, and the total number increased
accordingly.

In November, 1902, the encroachment of the younger generation had markedly
increased, and with the growth of the smaller ones many portions of the Western
Cheval bank — the parts where they were most abundant — had to be transferred from
the category of regions bearing maturing oysters to those monopolised by the younger
generation.

" In February, 1903, no regular inspection was made of these latter regions, but by
means of numerous descents in a diving dress I was able to obtain a fair knowledge
of the distribution of the younger oysters. On the Cheval and the North and South

* See also Part II. of this Report (1904), p. 15, and figs. 9 to 14.

THE PEARL FISHERY OF 1904.

23

Modragam the young already noted were found well grown and nourishing, and, while
their actual age was only I } to 1^ years, many, especially upon the Modragams and
South 'Cheval, had the appearance of heing 2 years old. On the Muttuvaratu and
Dutch Modragam the older oysters had made no increase in size, and were in an
unhealthy and miserable condition. Mixed with them were considerable numbers of
a younger generation, apparently over 1 year old. This year, as already stated, the
older generation proved to have died off upon the Karativo, while the younger
generation found last year had made fair progress and appeared moderately healthy.
Similar conditions, I believe, characterise the Muttuvaratu this year, judging from
the rather meagre data at my command."

Rate of Increase in Size. — So little is known with exactitude concerning the rate of
growth of oysters, that the following tabulation by Mr. Hornell of the observed
increase in size of these younger oysters has considerable value.

Table showing Rate of Growth of Oysters aged 2\ to 2| Years, March, 1904.

Date.

Age in years.

Size in millimetres.

I. South-east and South-central Cheval :-

AtoJL

March, 1902 .
November, 1902
March, 1903 .
1904 .

II. MODRARAM PAARS: —

March, 1902

1903 (North) . . .

1904 (South) . . .

III. North-west Cheval : —

March, 1902

November, 1902

March, 1904

„ 1904 (another locality)

li 1_2_

IS » 'l2
l1 l-1

■21 .. 2 J

12 » 12
l1- li

2i 2±

I 2 " T2"

ii 1 -

12 » '12

01 91

9.1 91

35-00 x 32 00 x —
50-57 x 47 -52 x —
54-41 x 49-75x20-38
67-12x61-16x27-24

37l5x 33-20 x —

6117x54-50x24-28
68-68x64-16x27-28

33-50 x 30-75 x —
47-64 x 44-36 x —
56 -60x52 -60x24 -56
61-33x57-40x23-70

When these records are extended over the remaining life-period of these oysters
we shall have, for the first time, reliable data as to the relative rate of growth from
year to year upon three of the most important of our oyster banks.

24 CEYLON PEARL OYSTER REPORT.

The respective increases during the first two annual intervals are as follows : — ■

1 2 months from age 3 to "1

6 months J

12 months from age 1| to~l

Total for two years .

South-east and South
Cheval.

North-west Cheval.

Modragams.

minimis.

19-41 xl7-75x —
12-71 x 11-41 x6-86

millims.

millims.

24-02 x 21 -30 x —

7-51 x 9-66x3-00

32-12x29-16

25 • 46 x 24 ■ 25 (average) ! 31 • 53 x 30 ■ 96

It will be noted that while the growth of the Modragam and of the South-east and
South Cheval oysters has been greater and more rapid than that of those upon the
Western Cheval, the Modragam oysters, growing more quickly than those of the
South-east and South Cheval up to the age of 14; to 1|- years, and thereby attaining
larger size, suffered retardation in the rate of increase thereafter, making but
7'51 millims. X 9'66 millims. X 3'00 millims. increase, whereas the South-east and
South Cheval oysters made 1271 millims. X 11 "41 millims. X 6'86 millims. in the
same period, and so reduced the disparity in size caused by the more rapid growth of
the Modrap-am individuals at the earlier age — so much so that the difference between
them at the age of 2^ to 2\ years is but 1*56 millims. X 3"00 millims. X 0'04 millim.

This lends support to our previous statement (this Report, Part I., 1903, p. 136)
that shell growth is most vigorous during the first two years of life, or, as we may
now put it, up to the size of about 60 millims. X 54 millims. X 24 millims., and that
thereafter it slackens to a very slow rate, actually coming to a stand, or even retro-
grading as regards height and length in the case of the stunted oysters of the
Muttuvaratu and the Mid-west Cheval.

As examples of this slowness of growth in older oysters we may take those fished
this year from the South-west Cheval. These were the most vigorous and freely
grown as regards the general average, and yet the increase made from November,
1902, to March, 1904, was but 5 31 millims. in depth by 377 millims. in length, i.e.,
between the ages of say 3^- to 3 J and 4^- to 4f years (size in November, 1902, was
71 '25 millims. X 65 "39 millims.). Contrast with this the increase within the same
period (16 months) when the oyster is younger, say between the ages of \\ to l-j-2-
and 2\ to 2^ years. In the case of oysters of this age from the South Cheval, the
increase has been, as shown above, as much as 16 '5 5 millims. X 13-64 millims., and in
the case of the North-west Cheval 13'69 millims. X 13*04 millims. — a remarkable
difference indeed.

Again we see that the disparity of growth is most marked during early life — up to
the age of 2 years — so that oysters situated in exceptionally favourable conditions
attain such dimensions that they are liable to be credited with greater age than

THE PEARL FISHERY OF 1904. 25

knowledge of their origin will warrant. This circumstance may account for some of
the differences of opinion as to the ages of young oysters in the past (see, for example,
the oysters attached to the Karativo buoy in 186G, descrihed by Mr. Holdsworth
and discussed by Sir W. C. Twyn am— Report, &c, Colombo, 1900).

Causes of Decrease among the Ftshable Oysters since March, 1902.

The factors which tend to reduce the number of oysters, both old and young, have
been treated of at some length in former parts of this Report, and it is only necessary
here to discuss the causes which have led to great decrease since 1902 in the old
oysters that were fished this year on the Western Cheval and the disappearance of
the same generation from the Muttuvaratu, the Dutch Modragam and the Karativo
paars. The decrease appears to have arisen from two chief causes, overcrowding and
sand disturbance, the ill-effects of which have been intensified by several other
influences of subsidiary importance. The former affected the beds in the North-west
and Mid-west of the Cheval sections, with the Muttuvaratu and Karativo paars; the
latter those of the South-west Cheval.

Sand Disturbance did the greatest harm upon the South-west Cheval, as this area
has but few stretches of exposed rock surface, while "cultch" is not sufficiently
abundant. When first seen by Captain Doxnan in 1900, the whole area was thickly
spread with oysters on the sand as well as on the rock and on the cultch-strewn areas.
With every recurrent period of stormy weather since that time these oysters have
suffered thinning by encroachment and overwash of sand, till eventually they were
cleared from off the sandy uncultched areas and only those were left that lay on rocky
ground and on sufficiently cultch-covered bottom.

A subsidiary cause of destruction was the ravages of the boring sponge, Cliona
marga/ritif&rce, which appeared specially active on this section, riddling the valves
and apparently sapping the vitality of the oyster by the drain it caused upon nacre-
secretion. Out of 400 individuals examined 310 were affected by Cliona — less than
23 per cent, were free from the infection.

Apart from this the oysters of the South-west Cheval were the healthiest, the
bodies being plump and well developed, whereas a considerable proportion of those
from the Mid- west and North-west sections were thin and of poor appearance.

The Effects of Overcrowding. — Sand disturbance produced comparatively little
numerical reduction upon the North-west and Mid-west sections, these regions
possessing more extensive stretches of rocky bottom than any others in this neighbour-
hood. The reduction here was due primarily to the effects of overcrowding, not,
perhaps, so much because of their own numbers, great as these were originally, but
rather because of deposits of spat which fell in July and December of 1901 upon
the older oysters along the eastern part of these sections. For the first year the
effects were little marked, but as the myriads of new-comers flourished and increased

e

26 CEYLON PEARL OYSTER REPORT.

in size the well-being of the older generation was affected, a result that showed
itself first by arrest of the growth of the shell and then, later, by starvation, lowered
powers of vitality and reproduction, and wide-spread mortality.

Regarding the arrest of growth, we noted that the older oysters on the Mid-west
Cheval measured 69"42 millims. X 64"54 millims. in March, 1902, when they were
2j to 2f years old, while survivors on less thickly populated parts of the bank were
only 72-30 millims. X 65"25 millims. in March, 1904, an increase in size of but
2 '8 8 millims. in height by 071 millim. in length in 2 years, a growth so slight as
to justify the conclusion that oysters from the Western Cheval practically attain
their maximum height and length upon reaching the age of 3 years, increasing
thereafter in no appreciable extent save in thickness, which is due to deposits of
nacre upon the inner surface of the valves.

In the case of the Muttuvaratu, the Karativo and the Dutch Modragam paars, the
disappearance of the oysters has been total, caused not so much by the competition
introduced by the presence of myriads of vigorous young — though this also existed
and no doubt had some effect — as by the overcrowding and diseased condition set up
by the immense number of adults. This has had such baneful effect that when
Mr. Hornell saw them in March of last year they showed marked deterioration upon
their condition 4 months previously, while their average size showed actually a
decrease in height and length (58-84 millims. X 54-32 millims. in November, 1902, as
against 57 '54 millims. X 54'UO millims, in March, 1903). These oysters are in a
shrunken, poorly nourished condition, and out of 227 examined, 25, or over 11 per
cent., were affected with the "yellow" disease — a malady marking the imminence of
wide-spread mortality, and characterised by the invasion of most of the tissues by
immense numbers of leucocytes crowded with yellow granules.

That the Muttuvaratu oysters should have been entirely cleared away, while a
relatively large number of the same generation is left upon the Western Cheval, is
undoubtedly due to the overcrowding having been less acute and less wide-spread
in the latter case.

Subsidiary Causes. — The general enemies of the pearl oyster were present as usual,
such as star-fish, boring molluscs and the smaller oyster-eating fishes (Batistes, &c).
The two former, the star-fishes especially, do much harm, and their reduction in
number is one of the benefits to be derived from dredging on a large scale. The
smaller oyster-eating fishes have not been excessively numerous upon the Cheval and
are not at present a destructive agency, seeing that the oysters are now too large and
too strong to be crushed by the small jaws of these fishes. Their ravages are only
to be feared when a bank is covered with spat ; then these fishes gather from all
cpaarters and, if the deposit is limited and the nature of the bottom sufficiently
smooth, the attack may involve partial or even entire annihilation of the bed. The
boring-annelid, Polydora sp., was markedly abundant on the South-west Cheval,
but comparatively few bad cases could be traced to its tunnelling. Far otherwise

THE PEARL FISHERY OF 1904. 27

was it with the inroads of the boring sponge, Cliona ma/rgwitiferce, already alluded
to. Hundreds of valves were examined where not only was the substance of the
shell riddled by its honeycomb-like tunnels (see fig. 9), but, a more serious matter, the

Fig. 9. Pearl-oyster shell honeycombed by Cliona inargaritifine, Dendy.

insertion scar of the adductor muscle was tuberculated and diseased, a condition due
to the inferior powers of nacre-production of the epidermal layer in this region. " So
great indeed was the nacreous tuberculation in some cases, that I have seen the natives
save these shells with a view to cutting the tubercles off to use in the manufacture of
native pearl jewellery." Whenever the inroads of Cliona were extensive, the
sub-epidermal tissue in particular, and the other tissues in general, were thin and
diseased-looking. A fairly large proportion of the dead shells found during the
fishery on the South-west Cheval was badly affected ; Cliona must undoubtedly be
included among the more destructive agencies with which we have to contend.

" There was no evidence of any depredations on the part of Ray fishes this
year, like the destruction of the Periya Paar Kerrai oysters last year. Rays were,
however, several times seen ; on the 27th February a large shoal, heading south,
passed the ship at anchor on the North-west Cheval. The sight was a remarkable
one, the sea for over 20 yards square being one patchwork of yellow disk-like bodies.
Fair-sized exan^les of Trygon walga were caught from time to time. None of them
furnished direct incriminating evidence, the stomachs being filled with crustacean
debris and fragments of lainellibranch shells other than those of the pearl oyster.
Like the trigger fishes, these smaller rays would appear to find the present size and
strength of the oysters too great for the power of their teeth and jaws."

No sharks were seen on the fishing grounds. Mr. HoRNELL says : " During all the
months I have spent upon the pearl banks during the last 2^ years, I have

E 2

28 CEYLON PEARL OYSTER REPORT.

never had a glimpse of a shark dangerous to man. Several times the boatmen have
caught ' basking-sharks'* of considerable size, but all were of a species that lives
almost entirely upon small fishes and mollusca."

Condition of Spat on the Periya Paar in 1904.

" March 14th was devoted to an inspection of the Periya Paar, with the result that
an immense bed of oyster spatt was there discovered. The young oysters were
incredibly abundant, spread densely over an area fully 5 miles in length by from
1 mile to 1^ miles in width. The age was under 3 months, so that they must have
been spawned not earlier than December of 1903, more probably towards the end of
the month. Alike on sandy and on rocky ground was this profusion seen, but on
the rock the numbers were notably greater. On the sandy stretches the oysters
formed bunches often of large size, the individuals attached in most cases to one
another, but generally having a nuclear fragment of small cultch which most frequently
was a cockle shell, a nullipore ball (IAthothamnion) or a small fragment of 'paar.'
The spinning mussel (Modiola barbata), ' Suran ' as it is called by the Tamil divers,
was in places abundant, especially upon the sandy ground, but not sufficiently so
to constitute a grave danger to the bed. This discovery of spat paralleled in extent
and nature that made by Professor Herdman in February, 1902. when he estimated
approximately the numbers of the young oysters at 100,000,000,000, and which, it
will be remembered, had all disappeared when I next examined the bank in November
of the same year.

" A very remarkable fact was that some of the largest of these young oysters when
last seen, 3rd April, had already developed gonads tilled with reproductive products
fully formed and ready to be shed. The average size of these sexually mature
individuals was 25^ millims. X 26^ millims. X 9 millims. The smallest was
24f millims. X 25 millims. X 8| millims. The extreme age of these cannot be more
than 3^ months, and we are fully warranted in considering that sexual maturity is
reached by the pearl oyster soon after the age of 3 months. Previous to this record,
the age of 0 months was the earliest at which reproduction had been observed."

Sexual maturity at such a small size and early age, in itself remarkable, has a
practical bearing upon the replenishing of the banks from time to time — a subject to
be treated of again.

Principal Spawning Seasons. — The gonads of this year's fishable oysters were nearly
all in a spent condition during February and March, and this taken in conjunction
with the presence of enormous multitudes of very young oysters — so young indeed
that they could not have been set free earlier than the middle of the December

* The sketch of one of these which Mr. Hoknell has sent home has been identified by Mr. G. A.
Boulenger as Stegodoma tigrinum (Gmel.).

+ The term "spat" is used m these pajes in a wide sense to compiiseall sizes of attached oysters
under the age of 4 months.

THE PEARL FISHERY OF 1904. 29

preceding — points to December and January as being the period ot one of the
maxima of reproductive activity in the pearl oyster of Ceylon.

In April the gonads began to show signs of increase, and a small quantity of spat
was seen. It now seems that the two maxima do not coincide, as we at first supposed,
with the intervals of line weather and variable winds between the monsoons,
i.e., November in the one case, March and April in the other. This year's observations
point distinctly to the chief spatting seasons having Deeemher and January, and
June, July or August, as the central periods respectively — times when the monsoon
winds are blowing steadily and strongly. It may he noted that a wide dispersal of
the embryos can lie more effectually attained during these boisterous periods of strong
currents than if they were expelled during the inter-monsoon lulls, characterised by
light winds and erratic surface drift.

Prospects of Future Fisheries.*

" The prospects of a large fishery, both next year and in 1906, upon the Cheval and
Modragam paars are excellent ; the immense quantities of young, 2 to 2|- years old,
occupying the Mbdragams, the South-east, South, Mid-west, North-west, and the
greater part of the three central blocks of the Cheval Paar being ample for 2 years'
fisheries. Those on the Modragams, and on the South, South-east, and South-central
Cheval will, I hope, be fit to fish next March, being more advanced in growth than
those on the Mid-west and North-west Cheval. Apart from this reason it will be
advisable to fish the former at as early a date as the valuation of samples may show
to be profitable, because, as many of the oysters on the beds named lie loose in
bunches on a sandy bottom, they are thereby exposed to the danger of destruction by
currents of exceptional force — a contingency infrequent but not unknown {e.g., fishery
of 1888). By March of 1905 these oysters will be about 3g years old, and, judging
from the rapidity of their growth and their present large size, by that date they
should be sufficiently matured to give a profitable fishery. They will then be
practically of the same age as those of the 1903 fishery, which realised just under
Rs. 30 per 1000, average price.

" If the fishery be postponed till 1906, the pearl yield would be greater, but if this
were done, even supposing the oysters remained in safety till March, 1906, upon the
banks, there would then be too great a quantity to fish in one year, and a portion
would be left for the succeeding year, 19o7. But in 1907 the remainder would be
o£ years old, and under the conditions prevailing on the Western and Median Cheval,
where these oysters are, it is most improbable that they would be still alive. A
definite decision cannot be arrived at till the next valuation has been madet in
November of the present year.

* I give this as Mr. HORNELL's opinion in the Summer of 1904. It must be mid in conjunction with
the next article (p. 37) based upon the inspection of lust November.
t Sec below . p. t3.

30 CEYLON PEARL OYSTER REPORT.

" The oysters at present on the Muttuvaratu and the Dutch Modragatn paars

should also he ready to fish in 1906, but I do not build confidently upon this ; even
now these oysters are extremely stunted and sickly looking, and are certain to decrease
very much before they are ripe for fishing.

"Beyond 1906, or at latest 1907, there is likely to be a break in the fishery
sequence, a more or less prolonged blank* whereof the duration will be to some extent
dependent upon whether or not large measures of transplantation are resorted to,
seeing that there has been no fall of spat upon the Cheval, Modragam, or Muttuvaratu
paars within the last two years, that is to say, since June, 1902. The only other
hope lies in the possibility that a younger generation may be growing up on some of
the paars that have not been inspected of recent years. I therefore trust that I may
be given the opportunity to make exhaustive inspections during the next 2 years.
The more I learn at first hand about the banks aud the more I read of the old records,
the more convinced I am that limited and narrow-compassed inspections have been
the cause of the loss of many valuable fisheries. To give a concrete instance of the
essential need for thorough periodical examinations of the whole pearl-bank area, the
following extract from Captain Donnan's Report of the 1878 inspection furnishes
significant reading : —

" ' While inspecting the Cheval I discovered oysters on a new bank, 2| miles to the
westward of the Cheval, where it has generally been considered that there was
nothing but an extensive sandy flat. I saw it recorded in the diary of inspection
by Captain LAcrGHTON, in March. 1802, that he found oysters of 4 or 5 years in
abundance in 9 fathoms, rocky bottom, Kudiramalai bearing S.E. \ E. ; and being
anxious to test if rocky bottom really existed in that neighbourhood, I took all the
inspection boats in tow of the launch out to the veige of the bank of soundings,
and directed them to dive all the way back to the inspection vessel, so as to cross
the position noted by Captain Laughton, and they came upon rocky ground and
oysters, where they placed a buoy. On subsequent examination this bed of oysters
proved to be about 1^ miles in its longest direction N.N.W. and S.S.E., aud about
f mile broad at its broadest part, having a depth of water over it of from 8^ to
94; fathoms, and a sandy flat of 9 fathoms between it and the Cheval. I believe it
to be an extension of the Periya Paar in a S.E. direction. I have duly recorded
its position in my diary. The superficial area of the bed is 3,845,935 yards, and
the oysters on it average 22 to a dive. I estimate the total number of oysters on
it at 3,500,000, sufficient to give 10 days' fishing for 50 boats with loads of
7000 oysters each. I found, however, 13 per cent, of dead shells, which is an
unfavourable sign, and a large quantity of seaweed growing on the oysters. I have
been much puzzled about the age of these oysters, as they appear to be old, although
of small size. I believe them to be of the description known to the natives as

* The position of affairs is now altered by a great fall of spat on the Cheval Paar (see below, p. 37).

THE PEARL FISHERY OF 1904. 31

" kottapakku," which are said to he a small class of oyster but rich in pearls. They
' resemble the oysters of the Chilaw banks, which I know to be 4 years old now, and
' therefore they ought to be fit for fishing in March next. I can find no record of
' this bed having ever been inspected since 1802, and it is a curious coincidence that
• what was then recorded of its condition might be recorded now on its most probably
' first examination since that time.' "

The oysters referred to in the above extract were fished the following year, 1879,
and the question is at once suggested, " How man}' times between 1802 and 1878, the
interval during which this paar received no attention, did oysters mature unknown
upon this ground ; and, further, how many other oyster-bearing patches have been
left unnoticed for want of more extended inspections?"

Parasites of the Pearl Oyster.

" Entozoa were plentiful in the oysters fished this year, both nematode worms and
encysted cestode larvae being found : the former especially were more numerous than
1 had before seen, and many oysters contained two individuals. A solitary example
of the trematode Aspidogaster margaritlferw was found, located as usual in the
pericardium. Cestode larvae were particularly abundant in the oysters from the
North-west Cheval, and this fact will be referred to when we come to treat of the
relative pearl yield of the three sections.

"More specimens are still needed to enable us to complete our knowledge of the
life-history of the pearl-inducing parasite, and till 1 have an opportunity to dissect a
number of individuals of the large oyster-eating sting-rays, the Walwadi tirikkai
of the Tamils, there can be no hope of filling up this important gap in our
investigation. No opportunity offered to obtain one of these rays at the late fishery,
my position 12 miles from land precluding access to the fishermen of Pukkulam
and Marichchukaddi, the only possible places where I could hope to procure
specimens. Several small individuals of Trygon walga were caught by the inspection
boatmen. This species does not, however, possess teeth sufficiently powerful to
destroy oysters more than a few months old and no parasites belonging to the species
found in the pearl oyster were discovered when these fish were dissected. Besides
pearl oysters a large number of fishes were dissected in search for the adult stage of
the pearl-parasite, and a continuation of such work will form the subject of a detailed
report in the future."

The Yield and Character of Pearls from Different Regions.

1 hiring the inspection of the Cheval Paar representative samples of mature oysters
were collected from the South-west, Mid-west and North-west beds respectively, and
from the 7th to the 9th of March Mr. Hornell was engaged superintending the

32

CEYLON PEARL OYSTER REPORT.

washing of these samples and the valuation of the pearls picked out. The results
showed that the South-west oysters were the richest in peai'ls, the North-west
following closely, with the Mid-west unmistakably much poorer in yield.
The following table will make clear the relative returns : —

Bed.

Quantity.

Total weight of
pearls.

Weight per 1000.

Valuation
per

Size of oysters in
millimetres.

Kalanchu.

Manchadi.

Kalanchu.

Manchadi.

1000.

South-west "1
Cheval J

North-west "1
Cheval J

Miil-west 1
Cheval /

12,000
7,200
6,000

7
2
2

13-25

17-00

0-88

Rs. ots.
12-76 36.01

7-91 33.75

6-81 20.25

76-20 x 69-20 x 32-34
7267 x 66-50x31-50
72-30x66-50x31-57

" The weight of pearls per 1000 oysters was very high in the case of those from the
South-west, which fact may be correlated with the size of these oysters being
considerably greater than that of the other two lots. In regard to the quality of the
pearls, the North-west oysters, while yielding a less weight, contained a much greater
proportion of 'cyst pearls' of excellent quality than those from the South-west,
which owed their greater proportionate weight of pearls to a noteworthy abundance
of ' muscle pearls. '

"External characteristics give the most meagre indication of what the pearl yield
may be — in regard to ' cyst pearls ' the abundance, or otherwise, is connected with
the factors which control the relative abundance of the pearl-inducing cestode and
those which conduce to its death during encystment in suitable localities within the
tissues — problems as yet obscure. The fact, however, is to be noted that dissection
showed the oysters from the North-west Cheval to be more extensively infested with
cestode cysts than those from the other districts, while the valuation sample showed
that cyst-pearls were much more abundant in these oysters than in those from either
the Mid-west or South-west Cheval sections. With regard to muscle pearls the
general fact stands out, deduced from my experience of two fisheries, that the more
vigorous and well grown are the oysters the more abundant are the muscle pearls."

Shell-pearls.

" At the close of the inspection, when the valuation samples of oysters had been
washed, I seized the unique opportunity thus afforded to examine over 30,000 pairs
of separated valves of the pearl oyster. I did this chiefly to see if any considerable
proportion of shell -pearls were induced by the irritation caused by perforation of the
shell by boring parasites — sponges, lamellibranch molluscs and annelids. The result

THE PEAEL FISHERY OF 1904.

33

Fi

negatived this idea and showed that these animals have little or no influence upon
pearl production ; the sponge, Cliona margaritifercs, causes occasional roughening, or
even rugged tuberculatum (fig. 10), at the insertion scar of
the adductor muscle, and the annelid, Polydora, may
produce an occasional discoloured internal excrescence,
usually sharp-pointed : more rarely a parasitic nematode,
free or encysted, is cemented to the nacre and sealed up in
a pearly sarcophagus.

" Bv far the larger numher of shell-pearls, fully 90 per
cent, of the whole number, are due to the attachment of
cyst- and muscle-pearls to the nacreous lining of the shell,
consequent upon fusion of the pearl-forming sac with the
epidermal layer outside, and rupture of these tissues caused
by the pressure of the growing pearl, thus placing the pearl
in actual contact with the nacre and converting its closed
pearl-secreting sac into a pouch or ampulla, the open neck directed outwards and in
continuity with the general epidermal layer upon the exterior of the mantle.

" Shell-pearls originating in this manner are at first pedunculated, and at this
stage may be detached readily. The peduncle tends, however, to become obliterated,
and eventually the pearl may be entirely lost in the substance of the shell nacre.
Such shell-pearls are in most cases affixed singly, without definite arrangement.
Examples may be located anywhere, the largest, however, being usually either in the
peripheral region of the nacre, or in the central region, corresponding with the lateral
surfaces of the visceral mass.

10. Inside of pearl oyster
shell, showing adductor im-
pression affected by Cliona
borings on the outside, x h

Fig. 11. Diagram showing comparative frequency
of cyst pearls in the various parts of the mantle.

Fig. 12. Diagram showing the positions most
frequently occupied by muscle pearls.

" Smaller attached pearls are not infrequently either singly or in a serial line of
from two to five along the pallial line, each pearl coinciding in position with a muscle

p

34 CEYLON PEARL OYSTER REPORT.

scar. Very rarely small ones may be seen along the edge of the adductor. Most
frequent of all are pearl masses of irregular or even botryoidal form attached at or
adjacent to the levator insertions. Decalcification of these pearls reveals no organic
nuclei, whereas the pearls irregularly disposed have cestode embryos as nuclei, exactly
as ' fine ' pearls have.

" Examination of these 30,000 shells gives, therefore, strong confirmatory evidence
in support of the. theory set forth by Professor Herdman and myself""" associating
such pearls as have no organic nuclei (cestode or other platyhelminthian embryos)
with the attachment surfaces of those muscles which are inserted on the shell,
especially the small and comparatively weak levator and pallial muscles. Figs. 11
and 12 show the distribution in the body of cyst pearls and muscle pearls respec-
tively, and fig. 12 is very instructive in showing the perfect way in which the
positions of the shell pearls there depicted coincide with the series of pallial muscle
scars.

" Phosphorescent and other Phenomena.

" On the evenings of April 9th, 10th, 11th, and 12th, when at anchor on the
Western Cheval, we beheld a repetition of the strange phosphorescent phenomenon
witnessed by Professor Herdman on March 13th, 1902 (see this Report, Part L,
1903, p. 81). The display began at about the same hour each evening, varying from
8.45 to 9.30 p.m., the sea being dotted with large isolated flashing lights that
pulsated and vanished ghost-like at regular and frequent intervals.

" The rhythmic and progressive blotting out and rekindling of these phosphorescent
fires was as if the sea were swept by regularly recurring search -light rays. The
beginning of each successive light- wave appeared to lie to the south ; the intervals
were approximately of two seconds' duration. All the four nights were dark, but the
weather was variable, one night tempestuous, another fairly calm, a moderately strong
swell j^revailing the whole time. The displays lasted approximately for an hour each
evening. The cause remains undetermined.

" Trichodesmian calm. — During several days in March vast quantities of the red-
brown scum of Trichodesmium erythrceum lay profuse and filthy upon the surface of
the sea. So abundant was this scum that twice did it extend the period of noontide
calm, acting precisely as a film of oil does. At midday when the first puffs of the
sea-breeze came out of the north-west, while the breeze was evident enough
otherwise, yet it failed for sometime to disturb the surface where lay this scum.
Where it was absent, or where the progress of the launch broke a lane through the
film, the wind raised ripples and wavelets which little by little spread and encroached
upon the Trichodesmian calm. This with the gradual freshening of the wind finally
scattered the particles, driving them downwards.

* ' Brit. Assoc. Report,' Southport, 190-j, p. 695.

THE PEARL FISHERY OF 1904. 35

" Surface drift. — Nothing can be done towards the solution of this question without
the employment of bottle-drifters, except to note the direction of the current from day
to day, when at anchor on the banks. This was attended to from 19th February to
19th April, the result being to show that during February the current set in the
main to the south and to the south-west in dependence upon the general direction of
the wind which at this period was mostly north-east. The following month, March,
distinguished for its alternate land and sea breezes, showed a to-and-fro motion of
the water upon the banks, the current altering with every change of wind and never
attaining any strength. In April, consequent upon the continuous south-west wind
and swell, the current ran strongly northwards, frequently at a rate of from 2 to 3
knots per hour."

Finally, Mr. Hornell has sent me a series of "Recommendations" in regard to
the further exploitation and conservation of the pearl banks. Some of his paragraphs
(h-al with matters treated of above, while others are better left over for discussion in
the Final Part of the Report. I give here the following headings of the chief points
in his recommendations : —

Synopsis of Recommendations.

(a.) Improved and extended inspection.

(/*.) The provision of a beacon on Aripu reef and the improvement of Kudiramalai
Tower.

(c.) Transplantation on an extensive scale by means of steamer dredging whenever
blank spaces occur upon the Cheval, coincident with the presence of young oysters in
the neighbourhood, within a radius of 20 miles.

(d.) Cultching to go on concurrently with transplantation. If these operations can
be carried out within the next twelve months, 500 tons of oyster shells and 300 tons
of Kalpitiya stone to be provided for this purpose.

(e.) The thinning out of overcrowded beds by means of the dredge.

( f.) The cleaning of the oyster banks by means of the dredge.

(;/.) Further investigation of the life-history of the pearl-inducing Cestode.

(A.) A series of drift-bottle experiments in conjunction with the Madras Presidenc}-
officials.

(/.) The establishment of a chank fishery on Government account and under
Government management in the neighbourhood of the Aripu pearl banks.

F 2

36

CEYLON PEARL OYSTER REPORT.

TABLE showing the Principal Facts relating to the Oyster-population on
the Pearl Banks in November, 1904 — to illustrate " The Present
Condition of the Pearl Banks" (p. 37).

Name of Paar.

Estimated

number

of oysters

on
the bank.

Valua-
tion per

1000
oysters.

Age

(November,

1904).

Condition.

Prospects.

South Cheval. . .
South-east Cheval .

40,000,000
23,600,000

Rs.
24.65
10.76J

years

2| to 3i

1 Healthy and vigorous

Should live through

two more seasons

without excessive

diminution in

[ numbers.

Mid-east and South-
central Cheval

1 13,750,000

13.21

years

2^ to 3^

("Healthy, but"!
< smothered in >
|_ young oysters J

The majority will
die oft' within the
next year owing
■I to the smothering
effects of the
myriads of young
oysters present.

South-west Cheval .

3,500,000

—

years

51

Healthy for their age

J" Too old to live be-
|_ yond 1905.

The remainder of
the Cheval

(Thousands!
[of millions}

—

months

u to n

Healthy ....

("Good. Should fur-
's nish fisheries in
[ future years.

North Modragam .
South Modragam .

4,700,000
21,000,000

17.86|

years

2fto3U

n ,, 3,}/

Unhealthy , shells and
bodies diseased ;
those on the south
smothered in
young oysters

- Dying off rapidly.

Periya Paar . . .

j Thousands]
[of millions)

—

months

H to 2i

Healthy ....

("Will probably dis-
J appear during the
j next South-west
(_ Monsoon.

Muttuvaratu . . .

20,000,000

3.16*

years
2§ to 5i

Stunted and diseased

Dying oft" gradually.

[ 37 ]

THE PRESENT CONDITION OF THE PEARL

BANKS.

[Based on Mr. Hornell's Inspection in November, 1904-]

Mr. Hornell left Colombo for the Pearl Banks on October 25th, 1904, with the
inspection barque " Rangasami Puravi " and the dredging steamer "Violet" — a
trawler recently acquired by the Government for pearl-fishery purposes. I have
received a detailed MS. report from him which has furnished the quotations and facts
that form the basis of the present section.

"The actual examination of the banks was begun on the morning of October 29th
and lasted with but one day's partial interruption, caused by heavy rain squalls, until
November 17th, when I brought the inspection to a close and went ashore [at
Marichchukaddi] to superintend the washing of the samples obtained."

This last inspection of the Cheval, Modragam, Periya and Muttuvaratu paars was
evidently unusually extensive and thorough for a November inspection, and
examination of the valuation records since the year 1854 shows that the number of
oysters (87,500) lifted by the divers was by far the largest ever collected.

The size of these samples affords a better opportunity than usual of ascertaining the
true average value of the oysters upon the different banks.

The result of the inspection is briefly as follows (see also Table opposite, p. 36) : —

1. Mature oysters, aged from 2f to 5|- years, were found in quantity upon the
North and South Modragam and Muttuvaratu paars, and upon the South, South-
east, South-central and Mid-east sections of the Cheval Paar, together with a small
and thinly covered patch remaining from last year's fishery (1904) upon the South-
west Cheval.

2. Very young oysters, aged from 1| to 2| months at the beginning of November,
cover the whole area of the Cheval excepting the South-west, South and South-east
sections, and are also found equally abundant on the Periya Paar and the South
Modragam Paar. If we estimate their age from the average size it is evident that a
fall of spat covering an immense extent of ground and quite unusual in numbers
must have occurred between the middle of August and the middle of September.
This is a most important occurrence which, though hoped for, was not expected and
could not have been foreseen. It greatly improves the fishery prospects for some
years to come and diminishes the need for any immediate transplantation from the
Periya to the Cheval Paar.

38

CEYLON PEARL OYSTER REPORT.

Proposed culture
areas recommended for
the Cheval Paar and
Periya Paar Kerrai.

Details of the Condition of Each Bank.

Ou the South and South-east sections of the Cheval Paar (see fig. 13) is a large
and continuous bed of excellently grown oysters from 2§ to 3^ years old. The area
covered with oysters Mr. Hornell computes at 6,705,000 square yards, but as the

bed extends in one direction beyond the limit surveyed we
may safely calculate upon a fishable area in this one locality
alone of fully 7,000,000 square yards. But even taking the
former figure and reckoning the population at an average of
from 16 to 26 per dive, it is estimated that this bed has a
total of 63,820,000 oysters. Mr. Hornell reports that "these
oysters are extremely well grown for their age and while they
contain but a small proportion of cyst-pearls and therefore
yield, at present, but few pearls of high individual value, they
give a remarkably heavy weight of muscle-pearls, so much so
that the valuation of the sample pearls ranges from Rs. 10.76
per 1000 oysters in the case of the South-east to as much as
Rs. 24.65 per 1000 in the case of those from the South section
— a fact accounted for by the ready sale which comparatively
low-class pearls meet with locally and in India."
Scarcely any young oysters are mixed with the mature individuals in this bed — a
fact of some importance when considering the rotation of fishing operations. This
point will be considered later on.

On the eastern part of the South-central and over the South-west quarter of the
Mid-east Cheval another extensive bed of oysters occurs, continuous along the
southern margin with the South Cheval bed. In area it contains about 2,300,000
square yards, bearing oysters averaging 15 to a dive, from which we estimate the
number of oysters present to be fully 13,750,000. The bulk of this ground was
fished, but imperfectly, in 1903, and, as a consequence, there are occasional patches of
oysters nearly 5|- years old, intermingled here and there with the younger generation,
2f to 3-g- years old, which forms the greater part of the bed (fig. 14). The presence
of a proportion of older oysters might be thought to make the quality and value of
the pearls from this locality higher than that of those from the South Cheval, but
actually the valuation is only Bs. 13.21 per 1000 as against Rs. 24.65 per 1000 in the
case of the latter bed. The difference seems due to a less vigorous growth upon the
Mid-east Cheval, a character correlated with inferior power of nacre-secretion and
consequent inferiority in the yield of pearls.

Over the whole of this bed of oysters, and extending west and north upon those
portions of both sections (South-central and Mid-east) which are bare of mature
oysters, myriads of the very young oysters before mentioned (now a few months old)
occupy every available point of hard ground. Where older oysters are present the

PRESENT CONDITION OF THE PEARL BANKS.

30

young ones crowd thickly upon the valves, ranging from 20 and 30 up to actually
(and frequently) counted totals of 96 to 99 upon each individual. Where cultch and
unoccupied outcropping rock occur every inch is crowded, the little oysters seeming
to shoulder each other in their fight for vantage ground. Mr. Hornell goes on to
state: "Towards the west, upon the sandy cultch-barren stretches of the South-
central ( 'heval, the struggle for existence among the young oysters is already intense;
clusters of individuals are grouped round nuclei of absurdly small and inadecpiate
dimensions, and often I have counted from 30 to 50 oysters in a bunch having for its
centre a fragment of stone or shell no larger than a pea in size. In other cases even
this may be wanting, the oysters adhering to one another."

The danger to these unstable clusters is obvious ; a strong current or under-tow
will sweep them hither and thither, and while this may have a certain value in
possilily bringing about the stocking of bare stretches of the bank elsewhere, the
dangers that these young oysters run of being swept into unsuitable localities far
outweigh the advantage to be gained by wider dispersion.

Kg. 1 1. Sketch-plan of the Cheval and Modragam paars, showing the distribution of pearl oysters in
November, 1904. Scale: Half an inch to one nautical mile. The broken line surrounds the areas
covered with spat a few mouths old, the dotted areas are those with oysters about 3 years old, the
solid black indicates those beds of oysters over 5 years old.

40 CEYLON PEARL OYSTER REPORT.

No other fishable oysters in any quantity are to be found on the Cheval Paar, with
the exception of an irregular and sparsely populated patch on the south-west corner
of the South-west Cheval — a remnant left from the last fishery. The quantity is
just under 3,500,000, reckoning the area at 1,150,000 square yards and the number of
oysters per dive as 9. No spat is here present, and as there is no admixture of a
younger generation, these oysters, which will be nearly 6 years old in March next,
should give a high valuation in pearls. They are too unevenly distributed to permit
of being dredged to any advantage. The native divers should, however, find the
patch remunerative, and it will probably be found advisable to begin next year's
fishery on this ground. The rest of this area (south-west) is blank, without even a
trace of the very young oysters found elsewhere (see fig. 14).

The whole of the remaining sections without exception — the Mid-west, the North-
west, the North-central, the Central, the North and the North-east — were everywhere
packed densely in the beginning of November with very young oysters of from 1|- to
1\ months of age, healthy in condition, vigorous in growth, and too abundant to
estimate in numbers. Indeed, so great is the profusion that Mr. Hornell reports
that " the remainder of the 2 \- to 3 -year-old oysters that abundantly stocked much of
this region in the early part of the present year are so masked by the covering of
spat as to be unrecognisable by sight. The divers cannot distinguish between rock
and old oysters ; both are submerged in this young brood of oysters, and till the
masses are brought to the surface the differences are not seen. Even if the quantity
of old oysters were considerable, fishing under such circumstances would be well nigh
impossible ; but, as it is, we are saved having to face this difficulty by the fact that
the North-west, Mid-west, and North-east sections have been ravaged by ray-fish
(Trygonidse) during the last few months, and the numbers of mature oysters so
thinned that these regions, for all practical purposes, may be considered blank so far
as concerns a fishery during the next 2 years. The prospects for an extensive fishery
4 years hence are on the other hand excellent ; by that time the present spat will be
old enough to be fished on those regions where growth progresses most rapidly."

On the North Modragam there is a small bed of 700,000 square yards in extent,
covered with oysters of 2§ to 3^ years old, ranging in density from 1G to 24 per dive.
This bed should yield at least 4,700,000 oysters if fished in March next. Spat is
absent from the whole of this section.

The South Modragam bears a much larger bed, fully 2,200,000 square yards in area.
At 24 oysters per dive the present population is estimated to be about 21,000,000.
Unfortunately for fishing operations, very young oysters (spat) are extremely
abundant, and may prove somewhat troublesome to the divers, as practically every
old oyster will have to be, at the least partially, stripped of its load of young.
Mr. Hornell recommends that " fishing operations should be hastened as much as
possible so far as this bank is concerned, as the health of the fishable generation is far
from satisfactory. The oysters are thin and the shells of a large proportion are so

PRESENT CONDITION OF THE PEARL BANKS.

41

diseased that it' not fished within the next 4 months T shall expect to find the bank
cleared of this generation in March, 1906. The age is the same as of those on the
North Modragam."

The only other bank where, from our previous observations, there was the remotest
possibility of finding fishable oysters is the Muttuvaratu Paar, lying off Dutch Bay
(see Map. Mu., fig. 15). Three days were devoted to the inspection of this, with the

Fig. 15. Sketch-map showing the principal pearl-oyster banks in the Gulf of Manaar. C, Chilaw Paar;
Ch., Cheval Paar; K., Dutch Modragam and Karativo paars; M., Modragam Paars ; Mu.,
Muttuvaratu Paar; P., Periya Paar; P.K., Periya Paar Kerrai.

result that Mr. Hornell estimates the number of oysters thereon, aged from 2§ to
5^ years, to be about 20,000,000. " All are stunted in outward appearance, the
shells both small and unhealthy, covered and distorted in many cases with a parasitic
growth of coral, while the substance is riddled and tunnelled by boring sponges,
molluscs, and worms. The soft parts are in like condition, miserably thin and sickly,
and many are obviously diseased." It is therefore not surprising to find that the
sample of 10,000 average individuals which was collected by the divers on November
11th yielded only a very small quantity of pearls, valued at Rs. 3.16i per 1000
oysters.

The valuation of the same bank of oysters in the previous November (1903) gave
Rs. 1.50 per 1000, so, as the value has more than doubled in the twelve months,
there is the possibility that a remnant might provide a small fishery in 1906 if
oysters be then scarce elsewhere. The only remaining bank inspected was the

G

42 CEYLON PEARL OYSTER REPORT.

Periya Paar. An examination of this remarkable region during the progress of the
fishery of 1904* showed the entire extent to be densely packed with untold millions
of young oysters from 1^ to 3 months old. At the present time but a few handfuls
of these survive ; the others have, as usual, disappeared, having been no doubt swept
from the bank into deep water or buried in sand during the turmoil of the south-west
monsoon. Their place is, however, now taken by a new generation of young oysters
of the same age as those so plentiful upon the Cheval Paar. They will be of
excellent size for transplantation in March next, if there be then need of such an
operation. If not utilised they must inevitably share the usual fate of oysters on
this bank, and will without doubt disappear during the next south-west monsoon.

Enemies of the Pearl Oyster.

Mr. Hornell reports that " Oyster-eating fishes are at present abundant on the
banks. Numbers of 'Kilati' (Batistes mitis, chiefly) and of ' Vellamin' (Lethrinus
nebulosus and other species) were caught by the crew whenever we anchored upon
spat-covered ground, and unmistakable evidence of their diet was given by the
stomach contents. In the stomach of one Balistes which I examined there was found
a minute pearl."

The quantity of young oysters is, however, sufficient to afford ample margin of
safety from such dangers, and with another 6 months' growth the shell will be
sufficiently strong to give protection against these particular enemies. " Greater
danger threatens from the huge ray called ' Walwadi tirikkai ' by the Tamil
fishermen (Rhinoptera javanica). This fish can destroy the strongest-shelled oyster,
and so great is its appetite and the capacity of its mouth and stomach that a
shoal is capable of destroying an entire bed of oysters within a few weeks or even
days, as witness the destruction of the fine bed of oysters on the Periya Paar Kerrai,
which was found to be in progress at the time of the fishery of 1903.f During the
recent inspection traces of widespread havoc from this cause were much in evidence
over the northern sections of the Cheval Paar, and this enemy is responsible for the
greater part of the reduction in numbers of the younger generation of oysters so
abundant on this area in March last."

The region of the bank in question is notably deficient in natural cultch, and till
this defect is remedied this and other similar parts of the Cheval and Modragam
paars must always be more or less subject to ravages and losses, even to the extent
of occasional total destruction of a bed and the consequent failure of expected
fisheries.

The boring sponge, Cliona margaritifercB, Dendy, is doing great harm to the oysters
of the Modragam paars. On these banks there is also an excessive abundance of

* See this volume, p. 28.

t See this Report. Part II. , p. 25.

PEESENT CONDITION OF THE PEARL BANKS. 43

shell-tunnelling annelid worms (Polydora), of boring molluscs (Lithodomus), and of
encrusting and enveloping sponges. This combination of parasites has induced in the
oysters a weakly condition which doubtless predisposes to disease. "As a conse-
quence the oysters here are sickly, the rate of mortality is increasing, and compara-
tively few ore likely to survive to 1906. Star-fishes are also abundant on this bank
and are aiding in the destruction of this bed of oysters."

On the Periya Paar, where very young oysters are plentiful, small carnivorous
gastropod molluscs (Sistrum and other genera) are unusually abundant, and are
causing, by boring through the shell, the death of considerable numbers of the young
oysters. On the Cheval Paar they are less numerous, and, speaking generally, there
is, saving for the great rays, no special or dangerous abundance of oyster enemies
upon this the most important bank.

Valuation Results.

The exceptionally great number of oysters collected at this inspection was made up
as follows : —

40,000 from the South Cheval Paar.

15,000

, Mid-east Cheval Paar.

6,000

, South-east Cheval Paar.

1,500 ,, North-west Cheval Paar

15,000

, South Modragam Paar.

10,000

, Muttuvaratu Paar.

87,500

Mr. Hornell reports that " One-half (20,000) of the South Cheval sample was
given to Mr. G. G. Dixon to be treated by his oyster-washing machine for
comparison with the other half, which was to be hand-washed and hand-picked in
the ordinary manner.* In addition, one bag containing 750 of the South-east Cheval
oysters was handed by mistake of the landing boatmen to Mr. Dixon, who will
account for it. All the rest of the samples which I collected were washed in the
ordinary way by the divers and munducks."

The result shows that the two highest valuations were yielded respectively by the
South Cheval and the South Modragam, the former being worth Ps. 24.65 per 1000
oysters, the latter Rs. 17.86^ per 1000. In the case of the former (higher) valuation
the result is due to the great weight of pearls present, in the latter to the
comparatively high value of the individual pearls. The one gives quantity, due to an
unusiial abundance of muscle-pearls, the other quality, owing to fine cyst-pearls being
present in a higher ratio than in the former case.

* See this volume, p. 6.
G 2

44

CEYLON PEAKL OYSTER REPORT.

Regarding the 40,000 oysters from the South Cheval Paar washed for comparison,
half by hand and half by Mr. Dixon in his machine, Mr. Hornell reports that " the
half washed by hand gave a valuation of Rs. 24.65 per 1000 oysters as against
Rs. 16.79 in the case of the machine-washed. The weight of pearls from the machine-
washed oysters was also markedly inferior, being but 7 manchadi per 1000 as against
a weight of a little over 11 ^j manchadi in the case of the hand-washed."

Mr. Hornell adds : " Lest it should be thought that the two lots might have been
taken from different localities, I wish to state that the whole 40,000 oysters were
lifted on the same days and from the same parts of the bank. They were originally
all piled on the deck of the inspection vessel, bagged under my supervision, and sent
ashore in one lot. The final division after being landed was carried out by
Mr. Saunders, Assistant Government Agent, Manaar, and by Mr. G. G. Dixon
conjointly, I not being present."

Fishery Prospects and Recommendations.

As the result of this inspection in November, 1904, Mr. Hornell reports that at
the present time there are available for fishing the following oysters, all yielding a
pearl valuation above Rs. 10 per 1000, the minimum at which it is considered
profitable to fish a bank, viz. : —

P;iar.

Estimated number
of oysters.

Value per
1000.

Age next March.

South Cheval

40,220,000
23,600,000
13,750,000
3,500,000
4,700,000
21,000,000

Rb.

24.65

10.764,

13.21

17.86*

years.
3 ' to 3i
3 „ 3|
3 „ 6
H „ 6
3 „ 3|
3 „ 3i *

Mid-east and South-central Cheval ....

South-west Cheval

North Modragam

Total

106,770,000

The highest recorded number of oysters landed at any one fishery has been
44,311,441 in 1891. At the utmost we could not hope to lift above 50,000,000 at
the next fishery and this would therefore leave rather more than half the number to
fish in 1906. Thus the number of fishable oysters appears amply sufficient to give
two large fisheries in 1905 and 1906 respectively and possibly a small one in 1907.
This can probably be brought about, if a careful watch be kept upon the banks and if
we take note of the various biological contingencies likely to menace the different
beds of oysters during the next 18 months. Mr. Hornell recommends that the
order in which the banks should be fished be as follows, provided the inspection of

PEESENT CONDITION OF THE PEARL HANKS. 45

February next shows the position of matters to be much the same as at present,
viz. : —

In 1905— (a.) 3, 500, 000 oysters on the South-west Cheval,

(b.) 21,000,000 „ „ South Modragam,

(c.) 4,700,000 „ „ North Modragam,

(d.) 13,750,000 „ „ Mid-east and South-central Cheval.

Total . . 42,950,000 oysters,

leaving for 1906 and 1907 the unmixed oysters now aged between 2f and 3^ years
upon the South and South-east Cheval.

The 40,000,000 oysters on the South Cheval are, with the exception ot the small
patch of 5^-year-olds on the South-west Cheval, the richest in pearls at the present
moment, and it might be thought that Mr. Hornell was assuming a grave responsi-
bility in advising the postponement for a whole year of a fishery iqjon this bank.
It must be remembered, however, that the section of the Cheval in question is very
reliable, and has scarcely ever disappointed the hopes of the authorities. There are
also the following special reasons for the course recommended : —

(<7.) The Modragam oysters are unhealthy and liable to die oft' in great numbers
within the next 12 months.

(b.) The South-west Cheval oysters, being nearly 6 years old, have all but reached
the span of oyster existence, and, though as yet fairly healthy, cannot be expected to
survive to 1006, when they would be nearly 7 years of age.

(e.) The oysters on the Mid-east and South-central Cheval are so densely covered
with very young oysters, a few months old, that, judging from the results of similar
competition seen repeatedly elsewhere during the last 3 years, the inevitable
consequence will be that in a year's time, or even less, the growth of such enormous
numbers of young will smother and kill oft' so large a proportion of the old that
the bed will cease to be one of mature or fishable oysters. Fishing on this bed
will, therefore, only be possible if carried out within the next 3 or 4 months.
If fished within this time the biilk of the young oysters could be returned to the sea,
and the bank will then be one of young oysters holding out good prospects of a fishery
in 1!)08 or 1909.

The above are arguments in favour of fishing certain banks first. Those against
immediate fishing on the South Cheval and the South-east Cheval are : —

The oysters on these two beds are at present in a healthier condition than
those on any of the others ; their growth is vigorous and they are not overcrowded
with young oysters. The oysters are clean and wholesome in appearance, food is
plentiful, and, so far as we know, no harmful agencies are present in a marked degree.
These oysters offer every prospect of being able to survive till 1906 — which cannot be
said of any other bed of mature individuals now known.

46 CEYLOISI PEARL OYSTEE REPORT.

In the case of the South-east Cheval there is also the obvious additional argument
against fishing in 1905, derived from the small pearl yield given by the present
valuation sample ; a year's growth would, no doubt, result in an enhanced pearl
value.

On the other hand, there is, of course, always a certain risk in leaving a fishable
bed of oysters unfished, and, once the biological facts given above have been stated, it
lies with the Government to decide what risk can be run and what course should be
taken. If the 40,000,000 oysters on the South Cheval Paar, or a considerable
number of them, can be fished in addition to the 43,000,000 which certainly ought to
lie secured first, there will, no doubt, be a large additional profit now — a present
certainty in place of the prospect of a possibly much greater result next year.

Summary of the Prospects of Fisheries from 1905-1909.

The following forecast assumes («) the adoption of the suggestion made above as to
the order in which the banks be fished ; (b) that extensive cultching operations be
undertaken next year to improve the purely sandy areas ; and (c) that no exceptional
catastrophe happens, such as an inroad of rays or abnormal weather conditions.

1905. — A fishery numerically as extensive as that of 1904, to be held on the North
and South Modragam paars, South-west Cheval Paar, Mid-east Cheval Paar, South-
central Cheval Paar. [Possibly also the South Cheval Paar.]

1906. — A fishery of large extent upon the South and South-east sections of the
Cheval Paar. [Unless the South be fished in 1905.]

1907. — A small fishery on the same grounds and possibly upon some small patches
elsewhere.

1908. — A portion of the Northern and Central thirds of the Cheval Paar now
covered with young oysters, if the pearl yield be then found sufficiently high to give
a remunerative return.

1909. — A fishery upon the remainder left from the preceding year's fishery.

That accounts for all the beds of oysters, old and young, now in sight. During
these five years other deposits of spat may fall upon reliable paars such as the Cheval.
If not, young oysters for future fisheries must be transplanted from the Periya Paar,
where they will no doubt be present in abundance from time to time.

Culture of the Banks.

In various parts of this Pteport Mr. Hornell and I have urged the necessity of
transplanting large quantities of young oysters by means of a dredging steamer, in
the event of no fall of spat occurring in the immediate future upon the unoccupied
sections of the Cheval Paar. Accordingly, on setting out upon the past inspection,
Mr. Hornell was authorised to employ the dredging steamer "Violet" for this

PRESENT CONDITION OF THE PEAEL HANKS. 47

purpose if, upon examination of the banks, he still considered this procedure necessary,
and provided the conditions on the Periya and Cheval paars respectively were found
suitable. Nature, however, proved unexpectedly generous ; the last spatting season
(August to September, 1904) must have been unusually bountiful, and the currents
favourable, as the result has been that all the great blanks on the Cheval Paar and
elsewhere have been filled up with young oysters in profusion.

The need for transplantation iu consequence of this quite exceptional spat-fall has
ceased to be urgent for the present season. The Government has been saved the
considerable outlay which transplantation on the scale contemplated would have
entailed. The dredging steamer consequently became available at the inspection for
fishing mature oysters, and means were thus provided to supply Mr. G. G. Dixon
with large numbers of oysters for experimental washing in the machine devised by
him for this purpose. It must not be thought, however, that the necessity for
transplanting has disappeared altogether. It is only postponed for a season, and
might become urgent again at the next inspection if, for example, it were found that
any catastrophe had occurred to the young oysters over any large section of the
Cheval Paar. The Inspector of the Pearl Banks should be authorised to transplant
from the Periya Paar whenever he may find it necessary.

The spat-fall has taken place equally upon sandy and upon rocky ground. That
upon the latter may for the present be safely left without special attention, but the
deposit upon the sandy areas requires careful nursing if it is to be brought to
maturity and yield a fishery. The prime necessity is extensive cultching operations,
the enriching of the surface of the sandy wastes of the South-central and North-east
Cheval especially with large quantities of fragmentary hard material in order to
furnish foothold to a few, at least, of the many millions of young oysters now
existing there in a very precarious condition. These young oysters have upon such
sections of the bank but few opportunities to make attachment to any fragments of a
size and weight sufficient to resist the strength of the bottom currents during even
moderately severe monsoon weather. They are liable at such times to be swept from
the banks, more especially during the first two years of existence, when the shells
are still light in comparison with their bulk, and when the small bunches into which
they are aggregated are of just the right form to permit of their being readily rolled
along the level stretches of the sandy areas till they finally perish.

Time and opportunity did not permit of the adoption during this last inspection of
active measures for the protection of these young oysters upon anything approaching
an adequate scale. All that Mr. Hornell could do was to take from the beach at
Marichc'hukaddi a quantity of nullipore balls (Lithothamnion, see fig. 16), dead coral and
broken calcrete ('" paar- rock ) and spread that hard material over a small portion of the
South Cheval area. He also directed the Master of the dredging steamer to instruct
his crew to break up all large masses of rock and coral that came up in the dredge
and to return the fragments to the sea. This is being done at present, and if carried

48 CEYLON PEAKL OYSTER REPORT.

out consistently will go some little way towards effecting what is desired. Much
more active measures must, however, be taken to ensure an immediate improvement
of the ground, and to save the spat now on the sand a great deal of cultching work
must be done during the next three months.

Fig. 16. Natural cultch (Lithothanmion) and, to the left, a similar Nullipore ball with a dozen

young pearl oysters attached.

The cultching operation which Mr. Hornell has recommended to the Ceylon
Government, and which I entirely approve of, is the disposal annually, for the next
few years, of a quantity amounting to not less than 500 tons of rubble, carefully
broken to a standard size of 3^- inches X 3-g- inches X 2^ inches, upon the South-
central and North-east sections of the Cheval Paar. If it be possible to get any
broken or waste tiles and bricks, such material is preferable to stone rubble and
should be utilised so far as obtainable. The cost delivered on the banks of the above
quantity should not exceed Rs. 1500, an insignificant sum compared with the
enormous return in the form of oysters saved from destruction which we should
reasonably expect to receive from this outlay. This quantity should, however, be
regarded as the minimum annual amount ; a much larger deposit of cultch would
probably well repay the increased expenditure.* The ravages of ray-fish have been
great during recent months upon some parts of the Cheval Paar, and it should be
mentioned that the efficient strewing of rubble over the ground is probably one of the
most feasible methods of stopping their depredations.

This section may appropriately end with a reference to the tabular statement on
p. 30, drawn up by Mr. Hornell to show the present condition of the beds of
oysters on the principal pearl banks.

* As this goes to the printer, I have received the information that the Ceylon Government has very
wisely decided upon the larger measure of cultivation. Rs. 5000 in place of Rs. 1500 has, I understand,
been placed upon the estimates to meet the expense of this year's cultching operations.

[ 49 J

FURTHER REPORT ON PARASITES

FOUND IN CONNECTION WITH THE

PEARL OYSTER FISHERY AT CEYLON.

BY
ARTHUR E. SHIPLEY, M.A., F.R.S.,

FELLOW AND TUTOR OF CHRIST'S COLLEGE, CAMBRIDGE,

JAMES HORNELL, F.L.S.,

MARINE BIOLOGIST TO THE CEYLON GOVERNMENT AND INSPECTOR OF PEARI. BANKS.

[With ONE PLATE.]

The following short Paper is based on material collected by Mr. Hornell largely
from Elasmobranchs which frequent the Pearl Beds of Ceylon. The specimens
unfortunately arrived too late for our account of them to be incorporated in the
article on the " Parasites of the Pearl Oyster," published in the second volume of
this Report, but as some of the forms are new and the Cestodes may possibly, though
perhaps not probably, be the parent form of the pearl-producing larvae, it has seemed
advisable to publish this further instalment. Unfortunately our further researches
throw no direct light on the problem of the 'provenance of the pearl-producing
parasite.

I. CESTODA.

Staurobothrium setiobatidis, n. gen. et sp. — Plate, figs. 1 and 2.

A considerable number of tapeworms, with large cruciform heads, were taken from
the intestine of the Bird or Cockle-eating Ray, Mtiohatis uarinari, Marcg., at
Marichchukaddi, the centre of the recent pearl fishery.

The head has the form of a cross with very short broad arms (Plate, fig. I). Each
arm ends in a sucker which, however, does not penetrate far into the arm, and whose
lumen is rather shallow. From the centre of the cross posteriorly arises the trunk,

H

50 CEYLON PEARL OYSTER REPORT.

whilst anteriorly the centre bears usually a slight papilla with rings round it ; the
distinctness and size of this latter varies, however, greatly with the state of
contraction of the specimen (Plate, fig. 2). In some specimens the head was swollen,
and then its longitudinal diameter equalled its greatest transverse diameter, but, as a
rule, the transverse axis is three times the length of the antero-posterior. There is no
armature. Also there is no neck, the strobilization commencing immediately behind
the head. The proglottides are at first many times as broad as long, but these
proportions alter as we pass backward, and at the posterior end of the body they are
perhaps one and a half times as long as they are broad. Each proglottis has a very
well marked salient posterior border, and their shape and relative size at different
regions of the body are well shown in fig. 1. Altogether there are about 100
proglottides, and the whole tapeworm measures on an average 1 "5 centims. in length,
and about 0"5 millim. to 07 millim. in breadth where it is broadest. These tape-
worms were, when preserved in formalin, stout and stiff, with well cut outlines.

In transverse sections the proglottides are circular, and the genital pore opens for,
at any rate, some consecutive proglottides on the same side, if one may speak of the
side of a cylindrical body. It was not possible to make out details of the sexual
organs beyond this fact, and that the uterus in the posterior proglottides is follicular.

As it has been impossible to find any figures or descriptions which correspond at all
closely with this animal, we have ventured to establish a new genus which we
propose to call Staurobothrium, and since it is found in the Bird Ray, I suggest the
specific name of cetiobatidis. The name Staurocephalus would better have fitted the
facts, but unfortunately it has been used before.

The formal definition is as follows : —

Staurobothrium, n. gen.

Cestode with large cruciform head, without hooks, genital pore lateral, no neck.

Staurobothrium cetiobatidis, n. sp. — Head without hooks or any armature, it consists
of four well marked arms projecting from a centre, like the arms of a Maltese cross ;
each arm ends in a shallow sucker ; anteriorly where the arms meet is a low annulated
papilla representing the rostrum, but, as said above, there are no hooks. There is no
neck. Each proglottis overhangs the one which succeeds it by salient angles forming
a funnel-shaped skirt. The genital opening is on one side. The uterus, when full of
ova, is follicular. The average length of the worm, with about 100 proglottides, is
15 millims., the average width from 0*5 millim. to 07 millim.

The systematic position of this Cestode is obscure. In structure the head super-
ficially resembles that of Ci/lindrojiJiorus, Dies., but it bears no hooks. Very little
change would be required in the architecture of the head to turn Staurobothrium
into the Calyptobothrium of Monticelli, but the terminal sucker of the latter is
absent in the new genus. In this respect it differs also from Linton's genus Crosso-
bothrium. On the whole I think we are justified in placing Staurobothrium in

FUETHEE REPORT ON PARASITES. 51

the family Phyllobothkiid.e, and not very far from the genus PlujUobothrium,
van Bex.

Tetragonocephalum trygonis, n. gen. et sp. — Plate, figs. 3, 4, 5, G, 7 and 8.

A number of very minute Cestodes were found in the alimentary canal of Trygon
walga. It frequently happens that most delicate tapeworms are found in the most
powerful and voracious Elasmobranchs, and nothing could he more delicate and fragile
than the Cestodes in question. The worms are fragile, and very limp when preserved
in formalin. They do not lie stiff and distinct as do those described above from
.Ettobatis ncvrincvri, but they are entangled together and form a mass like a knot of
chewed fine white cotton thread (Plate, fig. 3).

The head forms a distinct knob, hardly more than visible to the naked eye,
borne on the slender neck. Its diameter is some 0"03 millim. and its antero-posterior
axis is usually rather less than its diameter from side to side. In some specimens, as
in the one shown in fig. 3, the head is swollen, and its longitudinal axis is longer
than its transverse. It is a curious cushiony-looking head, consisting of two
distinct parts. Anteriorly there is a circular and rounded knob, about twice as broad
as it is long and resembling in shape the stones which are used in the Scottish sport
of curling. This probably corresponds with the rostellum of other forms. It is
quite unarmed (Plate, fig. 4). This rounded rostellum rests on a second region
like a crown upon a cushion. This second region is square in outline, and at each
of its corners it carries a small but distinct sucker, the orifice of which is minute.
From these suckers small papillae protrude, passing through their orifice. The
rostellum is separated from the square sucker-bearing portion by a thick basement
membrane, and it is traversed by many muscle bundles. Similar bundles are attached
to the hinder surface of this basement membrane and run down into the neck, where
they soon fade away.

The neck is short, and the narrow strictures separating the nascent proglottides
commence close behind the head. The constrictions between the proglottides always
remain slight ; the older proglottides somewhat resemble those of Dipylidium
cucumerinvm, but are less distinct and, except for the slightest possible thinning
between adjacent proglottides, the posterior part of the Cestode, after it has attained
a certain dimension, remains the same width throughout. At the same time, every
here and there there are constrictions which do not seem to correspond with the
divisions between proglottides. These are well shown in Plate, fig. 3.

The reproductive apertures are lateral, and the penis lies concealed in a spacious
recess (Plate, fig. 6). The pores irregularly alternate, some four being in the
left followed by one or two on the right, then a few on the left and again perhaps
four on the right. At its first appearance the uterus seems double, an anterior and
a posterior part lying one in front and one behind the genital pore ; the two are,
however, in communication by a narrow channel (Plate, figs. 5, G and 7). The

H 2

52 CEYLON PEARL OYSTER REPORT.

whole uterus in the ripe proglottides is thus somewhat dumb-bell-shaped, the narrow
part being pinched in by the cirrus bulb. The ova are slightly oval.

At first it might be thought that we had to do with a tapeworm of the same
genus as Linton's Lecanicephalus, but on referring to his description* and figures it is
evident that this can hardly be the case. Linton describes the head as " consisting
of two disciform plates," but he figures a round anterior part lying in the foremost
plate. The figure is very poor and we may be wrong in this interpretation, which,
however, is strengthened by his comparison of Lecanicephalum with the Discobotlirium
of van Benedkn, for in this genus the head is divided into three distinct parts.
The disks of Lecanicephalum are nearly circular, and the posterior bears four suckers,
but there is no mention of any papillae projecting from them. Van Beneden's genus
Discobotlirium found in Trygon pastinaca is figured, but not described by him.f It
has two distinct circular disks, corresponding with the rounded anterior part and the
first disk of Lecanicephalum, and then a thicker somewhat cruciform disc with very
salient angles which terminate in suckers. Our specimen has the anterior round
part — the curling stone as we have described it — resting on a quadrangular cushion
with suckers at the angles. On the whole it seems that we must describe it as a
new genus, and this we do under the name Tetragonocephalum, from the square
cushion which forms the larger part of the head.

Tetragonocephalum setiobatidis, n. sp. — Plate, figs. 9 and 10.

A single specimen of another Cestode of apparently the same genus was found
with those described a,bove in dEtiobatis narinari. Its length was 1"3 centims., and
its breath, which was remarkably uniform behind the head, was 0'5 millim. (Plate,
fig. 9). The head was three times this breadth and consisted of a rostellum, long
and conspicuous and unarmed, and of a swollen base, squarish in cross-section, with
four small suckers at the anterior angles (Plate, fig. 10). Posteriorly the basal
portion overlapped the anterior proglottides. There is no neck, but the proglottides
appear immediately after the head, at first very narrow but with marked constrictions ;
as they increase in size the posterior angle becomes salient, less so, perhaps, than in
Staurobothrium letiobatidis, but more so than in Tetragonocephalum trygonis. The
last three proglottides are twice the length of those which immediately precede them
and this growth is somewhat sudden.

The head, though it differs greatly in its proportions, resembles in essentials the
head of T. trygonis. The marked saliency of the posterior edge of the proglottides
separates off the species in question from the species which inhabits Trygon ivalga.
As there was but a single specimen, it did not seem advisable to cut it, and as it was
preserved in osmic it was not possible to make out anything of the internal anatomy.

The definition of this genus is as follows : —

* 'U.S.A. Commission of Fish and Fisheries,' 1SU1, p. 802.
f ' Mem. Ac. Belgique,' xxxviii., 1871.

FURTHER REPORT ON PARASITES. 53

Tetragonocephalum, 11. gen.

Head unarmed, consisting of an anterior knob-like portion arising from a cubical
liase ; t lie four posterior corners of the cubical base have minute suckers, each with
a papilla.

This genus would he a meinher of the family Lecanicephalidae, and would probably
come not very far from the genus Lecanicephalutn, Lint., in Braun's classification
given in Bronx's " Thierreich."

The species above described may he forniallv defined as follows:—

Tetragonocephalum trygonis, n. sp. — Fragile, minute, head hardly visible to the
naked eye ; length of body 2-4 centimetres : in section the head is circular and the
proglottides almost so, their greater diameter being 0'03 millim. ; head consisting of a
curling-stone-shaped anterior portion resting on a square cushion with suckers at
each corner from which a papilla protrudes ; neck short ; furrows between contiguous
segments very slight and in some places invisible ; reproductive pores lateral and
irregularly alternate ; genital recess large, and the remains of this constrict the
ripe uterus into a dumb-bell-shaped structure.

Tetragonocephalum CBtiobatidis, n. sp. — Minute head not much more than visible
to the naked eye; length of body and head 1"5 centims. ; head consists of an
elongated rostellum, unarmed, which projects freely from an almost cubical base, this
base is as long or longer than it is broad; at its anterior angles it bears four small
suckers ; there seems to be no, or at most a very short, neck ; the proglottides
overlap.

II. TREMATODA.

In the following account of two Trematodes, one from Batistes sp. and the other
from either a species of Carcharias or Rhinodon typicus, the " basking shark"* of
tropical waters, we are much indebted to Mr. Norman Maclaren both for notes on
the structure of the animals and for the drawing of tig. 11 on our Plate.

Distoma palleniscum, n. sp. — Plate, fig. 11.

This Trematode comes very near D. pattens,^ but differs from it in having certain
peculiarities which seem of specific rank. D. pattens was found in Chrysophrys
aurata, Cuv., by the authors mentioned in the footnote, and by Linton in Alutera

* There is some doubt as to the host represented by the " liasking-shark." The " basking-shark " of
the Indian Ocean, according to the books, is the rather rare Rhinodon typicus, but Professor Hekdman s
llection is that the term was applied by the sea-going men to a Carcharias. A drawing of one of these
sharks caught on the pearl banks has, however, been identified by Mr. BOULENGEK as Stegostoma tigrinum.

t RuDOLPHI, ' Entozoorum Synopsis,' 1819, pp. Ill and 410; Dujakdix, 'Hist. Nat. d. Helminthes,'
pp. 457 and 458; Diesim;, ' Syst. Helm.,' vol. 1, p. 348; StossicH, ' Saggio di una Fauna Elmint. di
Trieste,' p. 47 ; ' Ll.vioN, ' Proc. Nat. .Museum,' vol. 20, pp. ul'ti and 527, plate xlvii., figs. 8 and 9.

54 CEYLON PEARL OYSTER REPORT.

schcepfii. The present species is from the intestine of Batistes sp., the File- or Trigger-
fish found on the Ceylon pearl banks (Plate, fig. 11). The distoma is characterised
as follows : — Length 5 millims. ; perfectly smooth skin ; anterior sucker almost
globular, with a relatively small mouth ; the pharynx does not directly abut on the
anterior sucker ; oesophagus wide ; digestive cseca reach to posterior end ; yolk-
glands extraordinarily prominent, arranged in 2 rows, one on each side and behind
the testes ; the shell glands and receptaculum seminis not easy to distinguish, but
they lie anterior to the ovary ; the termination of the vas deferens is apparently
eversible and is probably used as a penis ; a definite penis is absent ; numerous
glands surround the terminal parts of both the male and the female ducts.

This species seems to differ from the Distoma pollens of Rudolphi in the facts
(i.) that the ventral sucker is not twice the size of the oral, (ii.) the aperture of the
ventral sucker is rounded and is not a transverse slit, (iii.) the ovary is not globular,
(iv.) the yolk-glands are prominent and arranged in two rows.

Distomum richiardii, Lopez.

Many specimens of this elegant little Trematode were taken from the perivisceral
cavity of Rhinodon typicus. The species has recently been described by Monticelli,*
who has found it in the body-cavity of Aeanthias vulgaris and more rarely in
Mustelus vulgaris and Myliobatis aquila, in the Bay of Naples.

III. NEMATODA.

Cheiracanthus spinosissimus, n. sp. — Plate, figs. 12 and 13.

Length 137 millims., breadth 0-45 millim. ; oesophagus ■£$, and the conical tail,
which is round quite at the end, is -^j of the whole body-length ; cuticle 0"022 millim.
thick, tough, transversely wrinkled ; head end rounded like the mouth of a Turkish
pipe (Plate, fig. 12) ; four tongue-like projections pointing backwards in the inter-
mediate areas ; 30 to 33 transverse rings, each consisting of some hundreds of spines ;
six lips around the mouth ; four neck glands of § to f the length of the oesophagus, in
front these are thin, with two swellings, but they thicken behind and contain spiral
muscles ; intestine dark brown ; the tail has on each side two prse-anal and four post-
anal papilla?, arranged as shown in fig. 13 ; the papilla? in outline resemble a skittle ;
cirri equal and 07 millim. long.

This Nematode was obtained from Myliobatis aquila, in the Gulf of Manaar.

This above definition has been furnished by Dr. VON Linstow, who kindly
undertook to investigate the Cheiracanthus. He adds that this new species differs
from C. unicinctus both in its head and in its neck glands.

* ' Zool. Jahrb. Syst.,' vi., Supplement, l6\)'o.

FURTHER REPORT ON PARASITES.

55

LIST OF PARASITES WITH THEIR HOSTS.

The species described in this and the previous Paper are as follows :-

PARASITE.

Cestoda :—

Small Cestode larvae
Tebrarhynchnts unionifactor

„ balistidis

„ pinna

„ minimus, v. Lixs

Staurobothrium ceiiobatidis
Tetragonocephahm trygonis
(eiwbalidv.

HOST.

from Margaritifera vulgaris,

„ Batistes mitts, II. wndulatus, and B. steUatvs.

n 11 ii ii

,, Toeniura melanospiks.

,, /Etiobatis narinari.

„ Trygtm walga.

„ JEHobatis narinari.

Trematoda : —

Mutt 'mi margaritifera
Musalia herdmani
Aspidogaster margaritifera'
Disfom Km palleniscum

,, richiardii

Margaritifera vulgaris.

Batistes sp.
lihinodm typicus.

Nkmatoda : —

Ascaris meleagrince
Cheiracantkus tmcinatus

sjnnosissimus

Margaritifera vulgaris.

larva, and Batistes mitis and
B. stellatus.

Myliobatis aguila.

56 CEYLON PEARL OYSTEli REPORT.

EXPLANATION OF PLATE.

Fig. 1. Staiirohothrivm (etiobaticUs. x 12.
,, 2. Anterior end of the same more highly magnified.
„ 3. Tetragonoc&phabi/m trygonis. x 24.

„ 4. Another head of the same, showing different dimensions, x about 75.
„ 5. Some mature proglottides of the same more highly magnified, showing the dumb-bell-shaped

uterus.
,, 6. Transverse section through the same, showing the penis, the genital recess, and the narrow part

of the uterus.
,, 7. Another transverse section, showing the opening of the genital recess to the exterior.
„ 8. A third transverse section nearer the end of a proglottis, showing the uterus taking up almost

all the space.
„ 9. Tetragonocephaht/m cetidbatidis. x about 20.
,. 10. The head of the same, x 75.
,, 11. Distomv/m paMeniscwm. Ez.bl., excretory vesicle ; Ov., ovary ; Sh.gl., shell glands; 7'., testis; 11.,

uterus; Y.g., yolk glands ; Y.d., yolk ducts ; ^, male, and $, female, reproductive pores.
,, 12. Cheiracanthus spinosisdmus, head end.
,, 13. The same, tail end.

CEYLON PEARL OYSTER REPORT

PARASITES PLATE

Abridge

Figs.] and 2 Stactrobothrium aetj:obatidis Figs 3-8. Ti cragonocephalum frygonis
Figs. 9 and LO, Tetragonocephalum aetiohatiih - Fig.11, Distomum palleni

Figs 1- and 13, Cheiraganthus spinosissimus

[CEYLON PEARL OYSTER FISHERIES— 1905— SUPPLEMENTARY REPORTS, No. XVIII.

REPORT

ON THH

SPONGES

COLLECTED BY

Professor HERDMAN, at CEYLON, in 1002.

BY

ARTHUR DENDY, D.Sc., F.L.S.,

pr.oFEsson of zoology in the south African college, cape town.

[With PLATES I. to XVI.]

INTRODUCTION.

Considering the frequent opportunities for collecting which have been afforded by the
Pearl Fisheries of Ceylon, it is remarkable that our knowledge of the Sponge-Fauna
of this locality, rich and varied as it is, should so long have remained in so backward
a condition. In 1887, Mr. Ridley and I (1)* called attention to this fact in our
Report on the Monaxonida collected by H.M.S. "Challenger," but as the "Challenger"
unfortunately did not visit Ceylon, the results of that memorable voyage did nothing
towards improving matters in this respect. In the same year, however, as that in
which our " Challenger " Report was published, I was able to report (2) also upon a
collection of sponges from Madras, made by Mr. Edgar Thurston, the Superintendent
of the Government Central Museum, and again in 1889 I reported (3) on a second
collection of sponges made by the same Zoologist in the Gulf of Manaar. In this
way considerable additions were made to our knowledge of the Sponge-Fauna of
Ceylon and Southern India, but Mr. Carter's papers (4, 5) on the sponges collected in
the Gulf of Manaar by Captain W. H. Cawne Warren still remained the principal
works on the subject ; a very remarkable fact, when we remember the extraordinarily

* The numbers in brackets refer to the List of Literature at the end of the Report.

I

58 CEYLON PEARL OYSTER REPORT.

small bulk of the material which he investigated and amongst which he found
altogether something like 70 species — a notable addition indeed to the 13 or so
previously described from the same neighbourhood by Esper (6), Haeckel (7),

BoWERBANK (8), and HOLDSWORTH (9).

The collection of sponges (146 species) made by Professor Herdman is, however, far
more extensive than any previously obtained in Ceylon waters, and includes not only a
large proportion of the species already described by the authors mentioned, but also a
considerable number (77 species) of new ones, some of which are of very great interest.
With the exception of the Calcarea, which are singularly few in number, and of the
Hexactinellida, which are, of course, a deep-water group and not likely to be met
with in the neighbourhood of the pearl banks, the collection contains representatives
of practically all the important sub-divisions of the Phylum. In dealing with so
large a mass of material, the cpiestion of classification has assumed a very formidable
aspect, and I have found it desirable to make certain modifications in this respect
which I hope may be regarded as improvements. These modifications will be duly
explained, and I hope justified, in the proper place.

Since, in spite of the enormous advances which have been made during the last
twenty years in our knowledge of the sponges, our ideas as to taxonomy are not yet
by any means in a settled condition, I have considered it desirable to give brief
diagnoses of the various sub-divisions with which I have to deal, in order that there
may be no mistake as to the sense in which I employ them. We are, undoubtedly,
progressing rapidly towards a satisfactory and natural classification of the group, but
the problem is an extremely difficult one, and every new collection of any considerable
extent, such as the present, must for a long time to come necessitate more or less
modification of existing views. The classification which I have adopted may at
present be regarded as the latest of these modifications, based principally upon the
writings of Sollas, Ridley, Lendenfeld, Topsent, and myself. Topsent, in
particular, has done much in the last few years both in arranging and classifying the
chaotic mass of information left by earlier writers, and in making notable additions to
our systematic knowledge of this difficult but interesting group.

In order to facilitate reference to the type specimens by future workers and to make
the systematic portion of this report as precise as possible, I have adopted the
system, already employed in my Catalogue of Non-Calcareous Sponges collected by
J. Bracebkidge Wilson, &c. (10), of giving at the end of the description of each
species the Register Number (R.N.), which I have attached to each specimen. This
method also has the great advantage of enabling one to refer readily to individual
specimens in the text.

At the conclusion of the Report I propose to give as complete a list as possible oi
the Sponge-Fauna of Ceylon and to discuss the problem of Geographical Distribution.
In the meantime I append a list of the principal memoirs dealing with the subject
arranged in chronological order.

SPONGES. 59

Special Bibliography of the Ceylon Sponge-Fauna.*

1798-1806. Esper.— " Fortsetzung tier Pflanzenthiere," Part 2.
1870. Ehlers. — " Die Esper'schen Spongien." (Tn this memoir the author
re-descrihes two Ceylon species.)

1872. Haeckel. — "Die Kalkschwumme." (This work includes descriptions oi

half a dozen species of Calcareous Sponges collected by Wright in Ceylon
waters.)

1873. Bowerbank. — "Report on a Collection of Sponges found at Ceylon by

E. W. H. Holdsworth, Esq.," ' Proc. Zool. Soc. Lond.,' 1873, p. 25.
(This paper contains descriptions of only four species. It was criticized
by Gray in ' Ann. and Mag. Nat. Hist,,' Ser. 4, vol. XII., p. 266.)
1873. Holdsworth. — "Note on the Occurrence of Xenospongia patelliformis,
Gray, on the Coast of Ceylon," 'Proc. Zool. Soc. Loud.,' 1873, p. 32.

1880. Carter. — "Report on Specimens dredged up from the Gulf of Manaar and

presented to the Liverpool Free Museum by Captain W. H. Cawne
Warren," 'Ann. and Mag. Nat. Hist,,' Ser. 5, vol. V., p. 437; vol. VI,
p. 35 and p. 129. (In this paper the author describes fifty-four species,
the great majority new, from material which he tells us would hardly
fill a quart measure, the largest of the specimens being not more than
three inches in its longest diameter.)

1881. Carter. — " Supplementary Report on Specimens dredged up from the Gull

of Manaar, &c," ' Ann. and Mag. Nat. Hist.,' Ser. 5, vol. VII, p. 3G1.
(A considerable number of species are added to the previous list.)
1884. Ridley. — "Zoological Collections of H.M.S. 'Alert, " British Museum.
(The author refers incidentally to the Sponge- Fauna of Ceylon.)

1887. Dendy. — "The Sponge-Fauna of Madras. A Report on a Collection of

Sponges obtained in the neighbourhood of Madras by Edgar Thurston,
Esq.," 'Ann. and Mag. Nat. Hist.,' Ser. 5, vol. XX., p. 153.

1888. Sollas. — " Report on the Tetractinellida collected by H.M.S. ' Challenger.' '

(In this work the author deals critically with the Tetractinellida described
by Carter from the Gulf of Manaar.)

1889. Dendy. — "Report on a Second Collection of Sponges from the Gulf of

Manaar." 'Ann. and Mag. Nat. Hist,,' Ser. 6, vol. III., p. 73. (The list
i >f the Ceylon Sponge-Fauna is herein considerably extended, and the
report contains records made by Mr. Thurston of the colours of many of
the species when alive.)
1889. Lendenfeld. — ■" Monograph of the Horny Sponges," 'Royal Society.'

* This list includes only works which deal directly with Ceylon species, and was inserted here on
account of its historical interest. In the course of the Report I shall have occasion to refer to many other
memoirs, and, in order to conform with the system adopted in the other Reports of this series and to save
space, I have decided to give a full list of the literature cited at the end of the work and to refer to it by
means of numbers.

I 2

60 CEYLON PEARL OYSTER REPORT.

SYSTEMATIC DESCRIPTION OF THE GENERA AND SPECIES
OF SPONGES IN PROFESSOR HERDMAN'S COLLECTION.

Phylum PORIFERA,

Class: NON-CALCAREA.

Porifera without calcareous spicules.

The non-calcareous sponges are divisible into four natural groups, for which
I propose to use the names Myxospongida, Triaxonida, Tetraxonida, and
Euceratosa respectively, and which may be conveniently regarded as of ordinal rank.

Order: MYXOSPONGIDA.

Non-calcarea which are primitively destitute of spicules and horny fibre ; with simple
canal system and usually large flagellate chambers.

In this order I include the genera Halisarca, Bajalus, Hexaddla, and Oscarella,
all of which appear to me to be nearly related to one another, so that it is hardly
necessary to divide them, as is usually done, into two families, Halisarcidse and
Oscarellidse. All of them are, I believe, primitive forms, and should therefore stand
at the bottom of the series of non-calcareous sponges.

It is cpiite possible that Lendenfeld's " Hexaceratina " may be closely related to
the " Halisarcidse," as supposed by that author ; but, if so, it appears to me that the
Halisarcida? are the more primitive forms, from which both Hexactinellida and
Hexaceratina have been derived. Lendenfeld, on the other hand, appears to regard
the Halisarcidse as being derived from the Hexaceratina by loss of horny fibres, and
in this opinion he is followed by Topsent (11), who observes " Depourvu a la fois de
spicules et de fibres conies, le genre Hexadella est vraiment le dernier chainon de la
chaine des Hexaceratina et doit occuper la place qu'on a parfois assignee, sans raison
valable, aux Halisarca."

The Myxospongida may therefore be regarded as representing a common starting-
point, from which have originated the Triaxonida, the Tetraxonida, and the
Euceratosa, and I cannot agree with Professor Minchin (12) in placing this order
at the end of the siliceous series instead of at the beginning. The primitive
character of the canal system argues strongly against the belief that they are forms
in which the skeleton has been lost by gradual reduction, and the case is obviously
quite different from that of Chondrosia, whose highly specialized canal-system,
strongly developed cortex, and evident relationship to Choiulrilla, afford good
grounds for regarding it as a reduced siliceous sponge in which the absence of
skeleton is a secondary and not a primary character.

SPONGES. 61

Hexadella, Topsent.

Myxospongida with large, sac-shaped flagellate chambers and a distinctly differen-
tiated, tough ectosome.

The description given by Topsent (loc. cit., p. 119) of this genus and of the two
species which he includes in it is evidently of a preliminary character and is unfor-
tunately without illustrations. His own diagnosis of the genus runs, " Hexaceratina
revetantes, molles, sans spicules ni fibres cornes ; un peu plus epais que dans les
genres voisins, l'ectosome jouit seul de quelque consistance et sert, dans une certaine
mesure, de squelette externe a la masse."

As the genus is as yet so little known, I shall postpone the discussion of its
relationship until I have described in some detail the anatomy of the Ceylon species.

Hexadella indica, n. sp. — Plate I., figs. 1-3.

Sponge thin, encrusting ; spreading extensively over masses of calcareous Polyzoa
and other organisms, but usually only about 1 millim. in thickness. Colour in spirit,
grey (in life, red ?). Surface smooth, glabrous, minutely reticulate. Vents few,
minute, scattered. Consistence (in spirit) rather tough and membranous (owing
chiefly to the ectosome), so that the sponge can be stripped off from the substratum
like a skin.

The ectosome (Plate I., fig. 1, Eel.) forms a tough pellicle about 0"08 millim. thick,
which can be stripped oft* from the underlying choanosome. The choanosome evidently
consists of a thin lamella (fig. 1, Ch.L., fig. 2), very much folded upon itself. The
lamella itself is little, if any, thicker than the ectosome, and contains a single layer of
large, sac-shaped flagellate chambers (figs. 1,2, F.C.), whose arrangement, of course,
follows the curvature of the lamella. The structure of the choanosome thus reminds
one very strongly of what occurs in OscareUa, but there is no extensive internal
portion free from flagellate chambers as in that genus.

Between the folds of the choanosomal lamella lie the primary inhalant (P.I.C) and
exhalant (P.E.C.) canals. The former expand at their outer ends into large, irregular
crypts (fig. 1, Cr. ), roofed over by the ectosome, which merges insensibly into the
choanosome at the points of contact. These crypts, resembling subdermal cavities,
may sometimes be seen from the surface, ramifying horizontally beneath the ectosome.
The primary exhalant canals open at their inner ends into larger exhalant channels
(fig. I, EEC).

» (wing to the development of the ectosome, the water does not enter the primary
inhalant canals directly, as in Oscarella, but by means of numerous well-developed
"chones" (fig. 1, Ch.), which penetrate the ectosome at right angles to the surface.
Each chone expands somewhat at its outer end in a trumpet-like manner, and is
covered over by a membrane (fig. 3, Mem.) which doubtless contains the inhalant
pores, but the pores are now all closed and cannot be recognised. I have, however,

(V2 CEYLON PEARL OYSTER REPORT.

seen indications, in the form of fine, slit-like canals in the closing membrane, of the
existence of several minute pores in the roof of each chone.

From the primary inhalant canals the water reaches the flagellate chambers by fine,
sometimes branching canaliculi of considerable length (fig. 2, I.Ca.). The chambers
themselves (figs. 1, 2, F.C.) are sac-shaped, unbranched (at any rate usually) and up
to about 0'098 millim. long, though usually less (say about 0"065 millim.). They
open into the primary exhalant canals either directly (eurypylous) or through short
exhalant canaliculi.

The mesoglcea, both of the ectosome and choanosome, is abundantly developed and
chiefly collenchymatous, with a profusion of small connective-tissue cells (fig. 2, C.T.C.)
with much-branched processes. The mesoglcea of the ectosome passes insensibly into
that of the choanosome at the points of contact ; the former, however, stains rather
more darkly than the latter, and is often distinctly, though very finely, fibrillated.

In the mesoglcea we also find embedded a large number of small, darkly staining,
spherical cells (figs. 1, 2, 3, Sy.A.), about O'OOti millim. in diameter, sometimes
apparently with small central nuclei. These cells are most abundant around the
various parts of the inhalant canal-system, and are congregated in immense numbers
in the inner portion of the ectosome. They sometimes appear to be broken up into
groups of smaller cells. They are in all probability symbiotic algte, such as are known
to occur frequently in sponges, a conclusion which is rendered almost certain by the
fact that, when treated with iodine and sulphuric acid, they give a distinct purple
coloration (although it must be admitted that when tested for starch with Schultzk's
solution, only negative results were obtained). These bodies are probably closely
comparable to the "cellules spheruleuses " to which Topsent (loc. cit., p. 120)
attributes the red colour of his Hexadella racovitzai, and which are also (in my
opinion) probably symbiotic algae. Unfortunately, I have no information as to the
colour of Hexadella indiea in life, but if, as seems just possible, it is identical with
Carter's Halisarca rubitingens, it is also red, and the colour may likewise be
attributable to the symbiotic algae.

This very interesting sponge appears to form a connecting link between the three
genera Oscarella, Halisarca, and Hexadella, to which latter genus I have referred it
only after long hesitation. It resembles the first in the distinctly folded character of
the choanosomal lamella and the arrangement of the flagellate chambers, which are,
however, much larger and usually (though not always) eurypylous, while a more
important distinction is introduced by the strong development of the ectosome. The
structure of the ectosome, the absence of the curious connective-tissue fibres, and the
unbranched character of the flagellate chambers, together with the less extensive
development of the mesoglcea, separate it from Halisarca. The mesoglcea appears,
however (as far as one can judge from Tofsent's description), to be a good deal more
extensively developed than in his species of Hexadella. Topsent says nothing about
the existence of a distinctly folded choanosome lamella, or of inhalant chones, in his

SPONGES. 63

species, and it is quite possible that a new genus may he necessary for the reception
of the < Vvlon sponge, hut this is a point which cannot be decided until we have more
information as t<> the European species.

Hexadella. iml><-<i may. as I have already suggested, possibly be identical with
( 'arter's Halisarca rxibitingens (5), which was described from dry material from the
Gulf of Manaar and referred to the genus Halisarca provisionally by its author.
Having examined one of Mr. Carter's original preparations of Halisarca ruhitimjcns,
now in my possession, I have come to the conclusion that we cannot say for certain
what that organism was, and that it will be best to abandon the name altogether.

R.N. 26, 46 (Gulf of Manaar), 254 (south of Adams Bridge, 5 fathoms).

Order: TRIAXONIDA (HEXACTINELLIDA).

Non-calcarea in which the skeleton is composed of siliceous spicules whose funda-
mental form is triaxonid and hexactinellid.

Owing, doubtless, to the collecting operations having been confined to comparatively
shallow water, there are no representatives of the Hexactinellida in the collection.
There are, however, a few specimens of the so-called Hexaceratina which may be
found ultimately to belong here. As, however, I do not consider that their
systematic position is yet definitely settled, I have decided to deal with them later
on, in connection with the other horny sponges (see below, p. '200).

Order: TETRAXONIDA.

Non-calcarea in which the fundamental form of the spicule is tetraxonid and
tetractinellid. The spicules may, however, be more or less reduced, and also to
a greater or less extent replaced by spongin or even sand.

It will be seen that I use the term Tetraxonida in a wider sense than that in
which it is employed, for example, by MlNCHIN, in his admirable article in
Laxkester's ' Zoology.' It appears to me necessary to do this in order to bring out
the most fundamental character of the group, viz., the primitively tetraxonid and
tetractinellid character of the spicule, in contrast with the triaxonid and hexactinellid
character of the spicule in the Triaxonida.* The term Tetractinellida may be
reserved for those non-Lithistid Tetraxonida which retain to a greater or less extent
the primitive spicule-form, a form which doubtless characterised the ancestors of all
the Tetraxonida, but has heen lost by reduction in many of the existing members of
the group.

The Lithistida have evidently developed along special lines ot their own, and

* The occurrence of triods in the most primitive family of the Tetraxonida (viz., Plakinidse) .suggests
a triaxonid (and triradiate) precursor of the tetraxonid and tetractinellid spicule. The triaxonid spicule
• it the llakinidse is, however, a fundamentally different form from that of the Triaxonida (Hexactinellida).

64 CEYLON PEARL OYSTER REPORT.

include both tetractinellid and monaxonellid forms, they may therefore be most con-
veniently regarded as constituting a separate " grade." By this exclusion of the
Lithistida, the term Tetractinellida is confined to Sollas's group " Choristida," a
sense in which it has already been employed by Lendenfeld in his work on the
Tetractinellida of the Adriatic (13).

Recent researches, however, have shown that the Tetractinellida, Lithistida,
Monaxonellida, and certain horny sponges which I propose to call Pseudoceratosa, are
so closely related to one another that they must certainly be included in one and the
same order. Indeed, it is difficult to draw a hard and fast line between these
sub-divisions, which merge very gradually into one another. It is certain that
the Monaxonellida have been derived (probably polyphyletically) from ancestral
Tetractinellida, and the Pseudoceratosa (or, at any rate, the great majority of them)
from ancestral Monaxonellida (probably polyphyletically), while the Lithistida may
also be of polyphyletic origin and are closely related to the Tetractinellida on the
one hand and to the Monaxonellida on the other.

The sharp separation of these four groups is, therefore, to a large extent artificial,
but must be retained as a matter of convenience till we know a good deal more about
their phylogeny. It must, however, be clearly recognised that the groups Tetracti-
nellida, Lithistida, Monaxonellida, and Pseudoceratosa simply represent stages in a
complex evolutionary series, commencing with the first-named group, and that each
of the later stages may possibly have been reached along more than one line
of descent.

Grade : TETRACTINELLIDA.

Tetraxonida in which the primitive tetraxonid and tetractinellid condition (or a
possibly still more primitive triaxonid and triradiate condition) is retained by
some at least of the spicules, while no desmas are developed.

Sub-order: HO MOS CLERO PH OR A.

Tetractinellida in which microscleres and megascleres are not yet sharply differentiated
from one another and no triaenes are developed.

I venture to propose the name Homosclerophora to replace the old name
Microsclerophora of Sollas. The latter is very misleading in that it does not
indicate the primitive undifferentiated chai-acter of the spicules as regards size, and
leaves it to be inferred that microscleres, as distinct from megascleres, are present.

Topsent (14), followed by Minchin (12), associates the Homosclerophora (Micro-
sclerophora) with the " Microtriaenosa " and the " Oligosilicina " in one group, to which
he gives the old name " Carnosa." This appears to me a most undesirable proceeding,
for while the Homosclerophora (especially the Plakinidae) are evidently primitive forms,
as Sollas (15) long since pointed out and as is clearly shown by their anatomical

SPONGES. 65

characters, the " Microtrisenosa " are what Mixchix terms " a heterogeneous collection
of sponges of divers affinities," and the " Oligosilicina " (" Chondrosidse," of Schulze)
are simple only by the reduction of the skeleton and not primitively. I therefore
consider that Topsent's revived "Carnosa" should he again abandoned, the
"Microtrisenosa" placed amongst the other trisene-bearing Tetractinellida, and the
"Oligosilicina" placed near the Tethyidse, to which they are evidently allied by
their corticate character, the general arrangement of their canal-system, and the
astrose microscleres of Chondrilla.

The Homosclerophora are thus left as the natural starting point of the Tetraxonid
series, a position already clearly indicated for them by Lexdexfeld in his work on
the Tetractinellida of the Adriatic. Topsext, it is true, fully recognises the artificial
character of his " Carnosa " and places them after his "Tetractinellida." Mixchix,
on the other hand, places them first, a diversity of opinion which clearly indicates
the impracticability of associating together such widely different forms as are included
in the group.

Owing to their primitive character, the Homosclerophora form a very important
and interesting sub-order, and it is here that we must seek for the origin, not only
of the various types of canal-system, but also of the almost innumerable types of
both micro- and megascleres met with amongst the Tetraxonida.

Family : PLAKINIDiE.

Homosclerophora in which no distinct cortex is developed.

This appears to be the most primitive family of the Tetraxonida. The genus
Plakinastrella, formerly included herein, but possessing more or less well-developed
trisenes, finds a more natural position amongst the Pachastrellida?.

Dercitopsis, n. gen.

Plakinidse with calthrops, triods and smooth oxea, but no candelabra. The oxea vary
greatly in size, and some of the smaller ones form a special dermal layer, in
which thev are commonly arranged at right angles to the surface.

It is, perhaps, doubtful whether this genus ought to be separated from Plakortis,
hut the species for which it is founded differs from the type species of Plakortis
(P. simplex) in three respects : — (1) Calthrops are present as well as triods ;
(2) some of the oxea are so large as to deserve the name of megascleres, although
they pass gradually into the smaller forms (microxea) ; (3) some of the microxea
form a special dermal layer, in which they are commonly arranged at right angles to
the surface.

In this genus we have a refutation of SoiXAs's statement that '" megascleres are
absent" from the so-called " Microsclerophora " (including the Plakinidse), and we

K

66 CEYLON PEAKL OYSTER REPORT.

see very clearly how oxeote megascleres may be derived directly from primitive
triaxon or tetraxon spicules by suppression of one or two actines (see Plate II.,
fig. 1). From the same source are derived microxea, and these, being frequently
curved, have probably given rise to the sigmata and toxa and, perhaps, other
curvilinear microscleres of higher groups. Increase in the number of actines of the
primitive spicule, on the other hand, accompanied by other modifications, has
probably given rise to the astrose series of microscleres. Thus Dercitopsis and its
allies may be regarded as representing at the present day the common ancestors of
both Astrophora and Sigmatophora.

Dercitopsis ceylonica, n. sp. — Plate II., fig. 1.

This species is represented in the collection by two small, flat pieces which have
probably been encrusting and may have formed parts of the same specimen. One is
about 22 millims. in length, 13 millims. in greatest breadth, and 7 millims. in greatest
thickness. The other is about 26 millims. in length, 16 millims. in greatest breadth,
and 7 millims. in greatest thickness. The surface is smooth but rather uneven.
Colour in spirit, dark slate grey on the outside, dull yellow internally. Vents
minute, scattered singly, with a tendency towards marginal arrangement or groujnng
on prominent parts. Inhalant pores dispersed abundantly over the surface, not in
sieves.

Skeleton very dense, for the most part quite confused and irregular, but with the
smallest microxea forming a thin dermal crust, in which they are commonly arranged
at right angles to the surface.

Spicules. — (1.) Calthrops and triods (Plate II., tig. 1, a-h); rays smooth, sharply and
gradually pointed, commonly about 0-033 millim. long by 0-005 millim. in maximum
diameter, but subject to considerable variation, and occasionally exhibiting monstrous
forms such as twins.

(2.) Oxea (fig. 1, i-6) ; varying enormously in size, but in such a perfectly
graduated series between the smallest and the largest that they cannot be divided
into mega- and microscleres. The dermal microxea measure only about 0-025 millim.
by 0-002 millim., while the oxea in the deeper parts of the sponge frequently measure
as much as 0'42 millim. by 0"012 millim. Both large and small oxea are smooth,
gradually and finely pointed at each end, and frequently centrotylote. They often
exhibit a curvature, which may be sigmoid or toxoid, and thus strongly support the
conclusion arrived at by Sollas (loc. cit., p. 109) to the effect that the toxa of
Dercitus are probably microxea which have acquired a curvilinear growth. The
same author has also pointed out that the spinose microrhabds of Dercitus originate as
smooth centrotylote microxea, which he regards as diactinal asters. From a similar
origin we get, in the case of Dercitopsis ceylonica, comparatively large oxeote
megascleres, which seems to indicate that this common type of spicule may also have
been derived from an aster. It also seems probable that the well-known sigmata, sy

SPONGES. 67

widely distributed in certain tetractinellid and monaxonid groups, may have originated
from such microxea as we find here, in other words, from diactinal asters.

The distinction between ectosorae and choanosome is not well-marked, and there
is no properly defined cortex. The ectosome is, however, fairly thick (about
0"27 millim.), and much more densely spicular than the choanosome; it contains
numerous small brown pigment cells (which are also met with in the outer part of the
choanosome). There is no fibrous tissue in the ectosome, and what little mesoglcea
there is between the densely packed spicules is probably collenchymatous. The
mesoglcea of the choanosome is very finely and uniformly granular.

The dermal pores lead into short inhalant canals which penetrate the ectosome
more or less vertically and, after uniting together to a greater or less extent, open
beneath the ectosome into spacious " crypts," from which the inhalant canals of the
choanosome take their origin. The flagellate chambers are rather large, pouch-
shaped, about 0"04 millim. in longer diameter; eurypylous, or with short, wide
exhalant canals.

R.N. 139, 235 (? Parts of the same specimen. From Station XLL, 12 miles off
Galle, 100 fathoms).

Sub-order: ASTROPHORA.
TetractinelUda with trisenes and with astrose microscleres ; without sigmata.

Family : I'AOHASTEELLIDyE.

Astrophora without long-shafted trisenes and without sterrasters. Calthrops, in
addition to short-shafted trisenes resembling calthrops, may be present.

Plakinastrella, Schulze.

Pachastrellidas with calthrops and (or) short-shafted trisenes and oxea for megascleres ;

and oxyasters and microxea for microscleres ; the microxea forming a special

dermal skeleton.
This genus occupies an intermediate position between the Plakinida? and the
Stellettida?, and probably indicates the first stage in the evolution of the long-shafted
trisene from the calthrops. The relationship to the Plakinicbe is clearly shown by the
form of the oxea, and the differentiation between megascleres and microscleres is not
yet by any means complete.

Plakinastrella intermedia, n. sp. — Plate I., fig. 4 ; Plate II., fig. 2.
Specimen (Plate I., fig. 4) irregularly branched and shortly stalked ; branches short,
thick, irregularly nodose, spreading more or less horizontally, rounded at the extremi-
ties. Surface smooth but uneven, minutely porous in parts. Texture firm and
incompressible, with many relatively large foreign bodies embedded in the substance

K. 2

68 CEYLON PEARL OYSTER REPORT.

of the sponge and sometimes projecting beyond the surface. Vents small and few, on
rounded ends of branches. Pores scattered in irregular groups ; many in each group,
easily visible under a pocket lens. Colour in spirit, greyish brown. Total height,
20 millims. ; greatest breadth, 26 millims.

The skeleton is for the most part very confused. The choanosome is densely
strewn with large oxea lying in every direction and occasionally aggregated in loose
bundles ; with these are mingled oxea of smaller size and a few tetract spicules,
while the interspaces are filled in with immense numbers of oxyasters, uniformly and
thickly scattered through the soft tissues. On the surface of the sponge is a dense
layer of small oxea, tangentially disposed, and for the most part lying so close
together as to touch one another and form a continuous crust, perforated by the
numerous circular inhalant pores where these occur. Beneath this dermal crust are
extended horizontally a number of the large oxea and the heads of the dichotriasnes.

Spicules. — (1.) Dichotrirenes (Plate II., fig. 2, a~d), with cladi extended beneath
the dermal crust and shaft projecting inwards at right angles to the surface. The
shaft is short, stout, and fairly gradually sharp-pointed, only about as long as the
radius of the cladome (or even shorter), measuring, say, about 0"37 millim. by 0-055
millim. The protocladi are short and stout, about 0'092 millim. by 0-055 millim. The
deuterocladi are about two and a half times as long as the protocladi, but variable ;
fairly gradually and sharply pointed, commonly slightly curved and often unequal.
Boiled-out preparations show various monstrous forms of this spicule, e.g., one with
one cladus unbranched, another with one deuterocladus itself branched, another with
the shaft branched, and also a number of very much slenderer forms of the same spicule
which are probably not fully developed. Although these dichotrisenes are essentially
characteristic of the sub-dermal skeleton, yet we find a good many branching spicules
scattered through the deeper parts of the sponge, which evidently belong to the same
type.

(2.) Oxea (Plate II., fig. 2, e-p). a, large ; more or less curved or even angulated
in the middle ; fusiform ; usually gradually and fairly sharply pointed ; size about
1*2 millims. by 0'037 millim. Occasionally a slightly curved stylote spicule may be
observed, with one end broadly rounded ; apparently derived from an oxeote by
suppression of one ray. b, small ; found chiefly in the dermal crust ; fusiform,
gradually and sharply pointed, slightly curved ; varying very much in size ;
averaging, sav, about 0"18 millim. by U'01 millim., but ranging from about a third of
these dimensions through intermediate sizes to the large forms.

(3.) Rather large oxyasters (Plate II., fig. 2, q-v), abundantly strewn through the
choanosome ; with usually about five long, slender and gradually sharp-pointed rays,
which are smooth and up to about 0"025 millim. in length ; there is no centrum.

Stained sections show that the ectosome is rather thin, apparently consisting only
of the dense spicular crust and with no fibrous tissue except a little around the
oscular tubes. The choanosome is gelatinous (collenchymatous). Both ectosome and

SPONRES. 69

choanosome contain numerous pigment cells full of brown granules, especially in the
walls of the canals. The material is not sufficiently well preserved to show the
characters of the flagellate chambers. The inhalant pores, about 0'13 millim. in
diameter, are the single openings of short cylindrical chones which penetrate the
spicular dermal crust and are provided with well-developed sphincters near its inner
limit ; but there are no extensive subcortical crypts.

(A fragment of a large, massive anatrisene was found in a boiled-out preparation of
the spicules of this sponge, but I could find no others, either in boiled-out preparations
or in sections. Should this form of spicule prove to be proper to the species, it would
necessitate its removal from the genus Plakinastrella).

R.N. 224 (from Station XLL, 12 miles off Galle, 100 fathoms).

Plakinastrella schulzei,* n. sp. — Plate II., fig. 3.

This well-characterised species is represented in the collection by an irregular
massive specimen (or fragment) which has been cut off at the base from its attachment.
The upper surface is strongly convex and is produced into irregular elevations. The
specimen contains many foreign organisms, both internally and encrusting the surface
(it is penetrated by the spiral shells of a species of Tenagodes (Siliquaria), a gastropod
which habitually lives in sponges). The colour is purplish grey externally (owing to
the presence of numerous granular pigment cells), yellowish within. The vents are
irregularly scattered, well-defined circular openings, which may be crateriform ; up to
about 3 millims. in diameter, but usually a good deal smaller. Inhalant pores (?) are
rather conspicuous on some parts of the surface, but irregularly scattered and varying
much in size. The specimen measures about 50 millims. in length, 30 millims. in
breadth, and 18 millims. in thickness.

The main skeleton is a very irregular interlacement of numerous large oxea and a
few short-shafted triaenes ; the spicules occasionally forming loose wisps. The
dermal skeleton is a dense feltwork of small oxea.

Spicules. — (1.) Short-shafted triaenes (Plate II., fig. 3, a-d), resembling calthrops
and subject to much variation in form ; commonly with three rays long and bifid and
the fourth short and simple, all fairly sharp-pointed; or all the rays may be simple ;
or one only may be bifid ; or one may be trifid, and doubtless other variations could
be found. Total diameter measured up to about 0"68 millim. from apex to apex of
rays, with primary rays about 0'046 millim. thick.

(2.) Oxea (Plate II., fig. 3, e-t). The large oxea of the main skeleton (e) are stout,
fusiform and gradually sharp-pointed at each end ; more or less curved and occasionally
slightly angulated at two points, but not centrotylote. They measure, when fully
developed, about 1*3 millims. by 0'055 millim. The small dermal oxea (r, s, t) are
commonly centrotylote, more or less sharply pointed at each end, and with the two
terminal thirds bent upon the middle third at slight angles in the same direction ;

* Named in honour of the most eminent of Spongologists, Professor F. E. SCHULZE.

70 CEYLON PEARL OYSTER REPORT.

they are quite smooth and measure about 0-35 millim. by 0-018 millim., but with
considerable variation in size. Between the large and small oxea thus described we
find so many intermediate in shape and size (f-q) that it is impossible to distinguish
them as megascleres and microscleres respectively.

(3.) Oxyasters (Plate II., fig. 3, v -x) with smooth, slender, sharp-pointed rays and
little or no centrum, are very abundant in the deeper parts of the sponge ; they
measure about 0'02 millim. in total diameter.

The irregular arrangement of the short-shafted trisenes (which presumably
originated phylogenetically at the surface of the sponge and acquired their
characteristic form in direct relation to that position), and the absence or scarcity of
typical calthrops, seem to indicate that this species is less primitive than the type of
the genus [P. copiosa, Schulze), and, in fact, somewhat degenerate.

B.N. 149 (from Station XLL, 12 miles off Galle, 100 tathoms).

Stceba, Sollas.

Thin, encrusting Pachastrellida?, with short-shafted triaenes, resembling calthrops,
for megascleres, and only spined microxea for microscleres.

In 1888 Sollas (loc. cit., p. 102) proposed the genus Stceba for Carter's Samus
simplex from the Gulf of Manaar, but Lendenfeld (13), followed by Topsent (14), has
merged the genus in Dei cit us. Inasmuch, however, as the type species of Detritus
(D. bucklandi, Bk. sp.) possesses toxa amongst its microscleres, while Stceba does not,
it seems to me desirable to retain the distinction between the two, especially as we
now know thi'ee more or less distinct species without toxa, viz., Stceba simplex
(= Samus simplex, Carter; Dercitus simplex, Topsent) ; Stceba plicata (= Coriicium
plicatum, Schmidt ; Calcabrina plicata, Sollas ; Dercitus plicatus, Lendenfeld and
Topsent), and Stceba extensa, n. sp.

Sollas's genus Calcabrina, founded for Schmidt's Corticium plicatum, must be
regarded merely as a synonym of Stceba.

Stceba extensa, n. sp. — Plate V., fig. 1.

Sponge thin, encrusting, spreading extensively over the surface of and into the
cavities of a mass of calcareous debris (Melobesia, Polyzoa, Coral, Worm-tubes, &c,
mixed together). Surface for the most part smooth and sub-glabrous, slightly rugose
in parts ; consistence tough and rather fleshy ; colour in spirit, pale grey. Vents and
pores not seen. The exposed part of the sponge forms an almost uninterrupted sheet
about 45 millims. by 33 millims. in extent, but of irregular outline.

The main skeleton consists of short-shafted trisenes scattered abundantly, but quite
irregularly through the choanosome, but very sparsely in the ectosome, and there is a
thin dermal crust of microxea.

Spicules. — (1.) Triaenes (Plate V., fig. 1); with short, stout, sharp-pointed shaft,
measuring about 0-136 millim. by 0-02 millim. when fully developed, with cladome

SPONGES. 71

about 0'2 millim. across; the cladi are usually extended nearly at right angles to
the shaft ; they are short and each usually divides into two short, sharp-pointed
branches; occasionally, however, the cladi are unbranched and sometimes they are
very irregular.

(_.) 8pined microxea (Plate V., fig. 1); very slender, straight or very slightly
curved, with spination minute but sharp and abundant and fairly uniform throughout
the length, excepting that there is frequently a constriction in the middle of the
spicule, in which the spination may be more or less wanting; size about 0"02 millim.
by 0*00133 millim. (excluding spines). These spicules occur very abundantly in the
dermal crust already mentioned and also scattered throughout the sponge.

The ectosome forms a cortex about 0"27 millim. thick, composed of large oval
cystenchymatous cells with fibrous tissue between them, the fibres, for the most part
at any rate, lying parallel with the surface. The cystenchyme cells measure about
0'06 millim. in longer diameter and are also abundant in the choanosome.

This species is obviously very closely related to the European Stceba plicata, on
the one hand, and to *S', simplex, from the Gulf of Manaar, on the other. From the
former, which has been fully re-clescribed by Topsent (14), it differs in its longer and
slenderer and more sharply-pointed microxea, and in the fact that the trisenes are
nearly all dichotria3ues. From the latter it differs in the smaller size of the trisenes
and the larger size of the microxea, and in the fact that the spines of the microxea
are not " most prominent towards the ends," though, as I have pointed out above,
they may be more or less absent from the middle of the spicule. It appears to be a
good deal more robust in growth than either species. It is quite possible, however,
that subsequent researches may make it possible to consider all three as mere
varieties of one species, but for the present it seems desirable that they should be
kept separate.

Topsent describes the ectosome of Stceba -plicata as being collenchymatous, but he
mentions and figures large " cellules spheruleuses," which are evidently closely similar
to the large cystenchyme cells of S. extertsa. The partly fibrous cortex of the latter,
almost devoid of megascleres, may possibly afford another means of specific
distinction.

In external appearance (in spirit at any rate) Stceba externa bears such a close
resemblance to the thin encrusting form of Chondrilla australiensis that it is difficult,
if not impossible, to distinguish the two without microscopical examination.

RN. 107 (Station LXVL, off Mutwal Island, March 19, 1902, 10 to 35 fathoms).

Family : STELLETTLLVE.
Astrophora with long-shafted trisenes, without calthrops and without sterrasters.

Myriastra, Sol-las.
Microsclere a euaster of one form only. Ectosome not a cortex.

72 CEYLON PEARL OYSTER REPORT.

Myriastra clavosa (Ridley).

1884, Stelletta clavosa, Ridley (16); 1888, Myriastra clavosa, Sollas (15).

I have no hesitation in referring to this species four small specimens from deep
water off Galle. The sjjecimens are approximately spherical in form and up to about

9 millims. in diameter. There is a single slightly depressed vent, which, in three of
the specimens at any rate, has a membranous margin. The colour is grey-yellow.
As this is an entirely new locality for the species, which has hitherto been obtained
from waters north of Australia, I propose to give the details of spiculation of one
specimen in justification of the identification.

Spicules. — (1.) Dichotrifenes ; shaft, 1'4 millim. by 0'025 millim., gradually and
finely pointed; chord, 0'6 millim. ; protocladi, 0'08 millim. ; deuterocladi, 0'2 millim.

(2.) Anatriasnes ; shaft, 2-3 millims. by 002 millim., very finely and gradually
pointed; chord, 0"1 millim.; cladi, 0"0f> millim., rather stout; of the shape figured
by Ridley.

(3.) Large oxea ; fusiform, gradually and sharply pointed, 2"2 millims. by 0-025
millim.

(4.) Cloacal oxea; commonly 0"2 millim. by 0'004 millim., may be rather larger.

(5.) Chiasters ; extremely minute, with no distinct centrum, and long, slender,
tylote rays varying in number; total diameter, 0*01 millim.

It will be seen from the above that the shaft of the dichotrisene is a good deal
shorter than in previously described specimens, but, considering the variation in this
respect, we cannot regard this difference as being of specific importance.

R.N. 131, 132, 227, 228 (all from off Galle and onwards up the West Coast of
Ceylon, depths up to 100 fathoms).

Myriastra tethyopsis (Carter).

1880, Stelletta tethyopsis, Carter (4) ; 1888, Myriastra (?) or Anthastra (?) tethyopsis,
Sollas (15).

There are four specimens of this very remarkable sponge in the collection. Sollas
regards the species as insufficiently characterised, but, as a matter of fact, Carter's
description is very good, and leaves no doubt as to the identification. The fact that
Sollas was unable to assign it definitely to any of his genera tends rather to show
the unsatisfactory nature of his system of classification than the insufficiency of the
original description.

The specimens (in spirit) are light grey in colour. The smallest is about

10 millims. and the largest about 42 millims. iii diameter. The shape is very
characteristically turbinate, more or less flattened below, and conical above. The
surface is harsh to the touch and finely granular, very minutely and rather sparsely
hispid. In the larger specimens a dense mass of long, hair-like spicules projects from
more or less of the lower surface. The larger specimens were evidently attached

SPONGES. 73

below. The apparent total absence of vents is a remarkable feature. Only in the
smallest specimen have T seen what looks like a small natural vent, and that is on
the flattened (presumably lower) surface. Doubtless the exhalant apertures are
completely concealed by contraction in preserved specimens. Owing to the enor-
mously strong development of the dense radiating skeleton, and the manner in which
the cladi of the triaeues come close up to the surface (giving it its granular character),
I have found it impossible to cut a tangential or surface section in the ordinary way
in order to look for pores in the dermal membrane. On attempting to cut such a
section the sponge splits up radially along the lines of the gigantic spicules.

The skeleton is remarkably strongly developed, the shafts of the huge oxea and
trieenes being arranged in dense bundles, which radiate from a central " nucleus" and
leave only narrow interspaces between them ; while the expanded cladi of the
dichotrieenes form a thin but dense dermal crust, in which innumerable very minute
asters also occur.

Spicules. — (1.) Dichotrisenes ; shaft about 8'0 millims. long by 0'074 millim. thick
a short way below the cladome, tapering very gradually to a narrow, long-drawn-out
but rounded point ; cladi short, stout, once-forked, expanded almost at right angles
to the shaft ; chord about 0'46 millim.

(2.) Protriaanes ; with long and very slender shafts and slender cladi about
0'05 millim. long, commonly projecting from the general surface. Much larger
protrieenes, with comparatively stout shafts and short stout cladi, occur in the
anchoring tufts on the lower surface of the adult sponge.

(3.) Anatriaenes ; with long slender shafts, and rather stout cladi up to about
O'l millim. long.

(4.) Oxea; associated in bundles with the shafts of the dichotriasnes ; fusiform,
gradually and fairly sharply pointed at each end; size about 8'0 millims. by 0'073
millim.

(5.) Asters; very numerous at and near the surface of the sponge, with small
centrum and numerous slender conical rays of equal length ; total diameter about
O'Ol millim. Such spicules are very rare in the deeper parts of the sponge, and
those which I have seen do not differ in any important respect from the asters of the
surface.

The smallest specimen possesses in one part, beneath the surface, in addition to
the asters, numerous minute refractive globules, which look as if they might be
siliceous.

Stained sections show that there is a very thick gelatinous (collenchymatous)
ectosome, excavated by numerous large, irregular cavities, from which occasionally
wide canals, provided with numerous transverse diaphragms, lead vertically inwards
through the choanosome. Small dermal pores are scattered between the cladi of the
dichotrisenes. The choanosome is not sharply differentiated from the ectosome, but
is finely granular. The flagellate chambers are oval and measure up to about

h

74 CEYLON PEARL OYSTER REPORT.

0"04 millim. in longer diameter; they are densely crowded together, and either
eurypylous or with very short, wide exhalant canaliculi.

This species is so peculiar that it might almost form the type of a new genus, but
I cannot fix upon characters which would separate it absolutely from Sollas's
Myriastra. The sponge, however, is not exactly small, nor are the oscules distinct,
nor are the pores in sieves, nor yet is the ectosome thin ; all of which features are
mentioned by Sollas as characters of Myriastra, though I should hardly consider
them as being of generic value myself.

B.N. 128, 128a, 128b, 130 (Station XLL, 12 miles off Galle, 100 fathoms).

Pilochrota, Sollas.
Microsclere a euaster of one form only. Ectosome differentiated to form a cortex.

Pilochrota haeckeli, Sollas. — Plate II., fig. 4.
1888, Pilochrota haeckeli, Sollas (15).

There are eight specimens in the collection which agree so closely with Sollas's
description as to leave no doubt of their specific identity with the " Challenger "
species, represented in the " Challenger " collection by a single specimen from the
Philippine Islands. The Ceylon specimens are irregularly spherical or oval in shape,
and the largest is about 16 millims. in maximum diameter. The colour in spirit is
grey. The vent (when visible) is single, small, and surrounded by a membranous lip,
minutely hispid owing to projecting oxea. The inhalant pores are arranged in small
sieves, which are irregularly scattered or situated in shallow meandering grooves.
The sponge has a strong tendency to attach itself to foreign objects by means of short,
root-like processes.

I subjoin the spicular measurements taken from one specimen (ll.N. 127), in
support of the identification : —

(1.) Orthotria3nes (Plate 11., fig. 4, a) ; shaft, 2-3 millims. by 0'074 millim.,
gradually and sharply pointed; cladi, 0'3 millim. long, stout, conical, sharp-pointed,
expanded almost at right angles to the shaft, slightly recurved.

(2.) Anatrisenes (Plate II., fig. 4, c) ; shaft, 2-8 millims. by 0*037 millim., long
and slender, finely pointed; cladi, strongly recurved, fairly stout, sharp-pointed,
0"13 millim. 1ou<j\

(3.) Somal oxea; fusiform, gradually and sharply pointed at each end, symmetrical,
often slightly curved; measuring about 2"3 millims. by 0'04 millim.

(4.) Cloacal oxea ; small and slender, about 0'2 millim. by O'OOo millim.

(5.) Chiasters, with very slender, tylote rays ; total diameter about 0"012 millim.,
not very abundant.

(There are also a few curiously abnormal trisenes with very short and stout shafts,

SPONGES. 75

as shown in fig. 4, b, d, e, but T think that these must be regarded as monstrous
forms of the orthotriaenes and anatriaenes, and not as having any taxonomic value.)

R.N. 127, 127a, 1276, 127c, 127d (all from deep water off Galle and onwards up
the West Coast); 197 (no special locality); 215 (deep water outside the pearl banks,
Gulf of Manaar); 225 (Station XLI., 12 miles off Galle, 100 fathoms).

Pilochrota hornelli, n. sp. — Plate II., fig. 5.

The single specimen is shaped like a somewhat elongated potato, with a fairly
smooth but rather uneven and finely granular surface, to which a number of foreign
bodies, such as shell-fragments, Foraminifera, &c, are attached. In its present
condition it measures about 45 millims. by 33 millims. by 33 millims., but a small
piece has been cut off from one end. The texture is somewhat spongy internally,
but with a firm, dense outer crust nearly 1 millim. thick in places. The colour is
light brown throughout. The inhalant pores are arranged in small groups (pore-
sieves) thickly and generally scattered over what was probably the lower half of the
sponge : they are also scattered singly on the upper part. There is a single large
vent near one end of the upper surface, elongated transversely to the long axis of the
sponge and with a narrow, thickened, smooth, fleshy margin. The vent measures
about 6 millims. by 2 millims., and is the opening of a short, wide oscular tube formed
by the union of a number of large exhalant canals.

The skeleton in the interior of the sponge is loose and irregular, composed
principally of large, scattered oxea, not arranged in definite fibres. Towards the
surface the spicules tend to collect together into fibres which end in dense brushes
composed almost entirely of ortho- and anatriaenes. These brushes separate the wide
subcortical crypts from one another, but their expanded outer ends form a continuous
spicular crust at the surface. The cladi of the orthotriaenes are for the most part
extended at the surface of the sponge, but do not project beyond it, those of the
anatriaenes lie somewhat deeper down in the cortex. There are also a large number
of orthotriaenes whose cladi lie beneath the subcortical crypts while their shafts
project inwards into the choanosome.

Spicules. — (1.) Orthotriaenes (Plate II., fig. 5, a-d) ; with stout shaft tapering
very gradually to a narrow, bluntly rounded or sharp apex, and with short, stout,
conical cladi extended almost at right angles to the shaft ; two typical examples from
a boiled-out preparation gave the following measurements: — (a.) Shaft, 1 "5 millims.
by 0-0:32 millim.; cladi, 0'123 millim. by 0-03 millim. (b.) Shaft, T017 millim. by
0-04 millim. ; cladi, 0115 millim. by 0"033 millim. In the dermal crust one or two
of the cladi not infrequently become bifurcate, but this takes place very irregularly,
and a good many monstrous forms occur.

(2.) Anatrirenes (Plate II., fig. 5, e) ; shaft long and fairly stout, not tapering to
hair-like dimensions, though, of course, much narrower at the proximal than at the
distal extremity ; bluntly or sharply pointed ; cladi stout, strongly recurved, sharply

l 2

76 CEYLON PEARL OYSTER REPORT.

pointed ; front of cladome flattened. A specimen from a boiled-out preparation,

somewhat longer than usual, gave the following measurements: — Shaft, 1'96 millim.

by 0-019 millim. ; cladi, 0-0656 millim. by 0'0i64 millim.

(3.) Oxea (Plate II., fig. 5,f); fairly stout, fusiform, gradually and usually fairly

sharply pointed, gently curved, approximately iso - actinate, measuring about

1-3 millims. by 0-032 millim.

(4.) Chiasters (Plate II., fig. 5, <j, h, i) ; with no centrum and very slender, slightly

roughened and distinctly tylote rays ; total diameter of spicule up to about

0-02 millim. ; abundant, especially in the deeper parts of the sponge.

The histological structure of the strongly developed cortex agrees very closely in

some respects with that described and figured by Sollas (loe. cit., p. 122) for his
Piloelirota pachydermata. The cortex itself, about 0'8 millim. in thickness, consists
chiefly of very dense fibrous tissue, in which the fibres interlace in various directions.
Imbedded in this tissue we find histological elements of two other kinds : — (l.) Large,
rounded, deeply staining bodies up to about 0'2 millim. in diameter, each made up of
an aggregation of much smaller bodies (cells ?), and the whole very finely and
uniformly granular. These remarkable bodies have a rather deep yellowish brown
colour in unstained preparations, and they form a single layer beneath the surface in
the outer part of the cortex. They are evidently homologous with what Sollas
terms " oval or round clusters of granule-cells" in P. pachydermata. Sometimes the
finely granular " cells" occur separately as well as aggregated in clusters, especially
in the margin of the vent. (2.) Numerous irregularly scattered, small groups of
faintly staining, homogeneous, spherical globules, varying in size in each group up to
about 0'005 millim. in diameter; abundant in the deeper parts of the cortex and also
in the choanosome. These bodies do not seem to have been observed in P. pachy-
dermata.

Beneath the fibrous cortex lie extensive sub-cortical crypts, separated from one
another by the radiating spicule bundles and a sparse development of ectosomal
collenchyma.

The inhalant pore-sieves overlie well-developed chones which penetrate the fibrous
cortex to reach the sub-cortical crypts, being provided with sphincters or diaphragms
where they join the latter ; the sphincters, however, do not lie quite so deeply as the
lower limit of the fibrous cortex. From the sub-cortical crypts originate numerous
fairly wide incurrent canals which penetrate the choanosome and sub-divide therein.

The choanosome is finely granular, and the flagellate chambers are approximately
spherical, about 0'02 millim. in diameter, and appai'ently eurypylous.

It is not impossible that this species may be identical with Sollas's Piloelirota
cingalensis, also from Galle, but the description of the latter (15), apparently based upon
a mere fragment, for the author says " Sponge (?)," is too imperfect for safe
recognition. Sollas, moreover, states that the cortex is thin, about 0'35 millim. in
thickness, and all the megascleres seem to have their shafts characteristically bluntly

SPONGES. 77

pointed or even strongylote, a condition to which there is only a tendency in our
specimen. Our species is also very closely related to Pilochrota pachydermata,
Sollas, from Tahiti, hut differs in certain respects, such as the form of the chiaster,
which is not tylote in /'. pachydermata.

R.N. 176 (collected in the lagoon inside the reef, Galle, shallow water, by
Mr. Hoknell, after whom, on Professor Herdman's suggestion, 1 have much pleasure
in naming the species).

Stelletta, Schmidt.

Microscleres euasters of two forms.

It might be advisable to sub-divide this genus according to whether or not a
fibrous cortex is present, as Sollas has done in similar cases. Stelletta herdmani
would then come under the first sub-division and S. vestigium under the second. I
do not, however, think it would lie desirable to take such a step until we know more
about the value of this character in classification.

Stelletta herdmani, n. sp. — Plate II., fig, 6.

Sponge quite irregular in shape ; sometimes vallate ; sometimes with occasional
digitiform or mammiform projections ; often very much mixed up with coarse
calcareous debris. Surface uneven, very harsh to the touch and striate owing to the
presence of the huge megascleres immediately beneath the thin dermal membrane ;
not hispid except where apparently worn. Vents few, scattered on prominent parts,
up to about 3 millims. in maximum diameter ; leading out of deep cloacal tubes,
which are of about the same diameter and lined by a sieve-membrane. Inhalant pores
scattered in the dermal membrane. Colour (external) ranging from pale yellowish to
slate-grey in the same specimen The largest specimen is about 55 millims. in
greatest diameter, but is a good deal damaged and much mixed up with foreign
matter.

Skeleton very dense and more or less confused ; consisting of irregular bundles of
huge oxea and trisenes, for the most part radiating towards the surface, but lying
tangentially beneath the dermal membrane and thus forming an ill-defined but dense
spicular cortex about 0'3G millim. thick.

Spicules. — (1.) Plagio- or protrieenes (Plate II., fig. G, a, b, c.) ; with short, stumpy,
conical cladi projecting more or less forwards, and stout shaft, usually somewhat
curved, more or less swollen at some distance below the cladi and tapering gradually
to a sharp point at the other end. Size somewhat variable, shaft up to about
1/2 millims. by 0"074 millim., with cladi 0-18 millim. long.

(2.) Oxea (Plate II., fig. 6, </.); fusiform, straight or slightly curved, tapering
gradually to a sharp point at each end; size about 2"8 millims. by 0"09 millim.

(3.) Chiasters (Plate II., fig. G, e-h) ; very abundant in the dermal membrane,

78 CEYLON PEARL OYSTER REPORT.

scarcer within ; with numerous rather short, truncate conical rays, and often with a
distinct hut small centrum ; total diameter about 0"012 millim. or less.

(4.) Oxyasters (Plate II., fig. 6, i-h) ; fairly numerous in the deeper parts, but may
be difficult to find ; with little or no centrum, and long, slender, conical, sharp-pointed,
smooth rays, varying in number; total diameter about 0-029 millim.

Owing to the great size and abundance of the megascleres it is impossible to cut
satisfactory sections by the ordinary paraffin method, but such as I have been able to
obtain have yielded some interesting results. The ectosome is remarkably developed
as a very thick gelatinous layer containing numerous stellate cells (collenchyma), and
a thin layer of fibrous tissue (about 0"08 millim. thick), situated about 2"5 millims.
beneath the surface and separating the gelatinous layer from the choanosome. The
arrangement of the main skeleton appears to be in no way correlated with the
differentiation between ectosome and choanosome, the large megascleres passing
through the fibrous layer from one to the other without distinction. The mesoglcea
of the choanosome ranges from collenchymatous to finely granular, and the flagellate
chambers are more or less spherical and about 0'024 millim. in diameter.

There are six specimens of this sponge in the collection. The arrangement of the
skeleton reminds one very strongly of Sollas's genus Stryphnus, but the absence
of amphiasters or sanidasters prevents us from placing the species in that genus
(according to Sollas's classification). The species also resembles to a considerable
extent Kirkpatrick's Stelletta (Astrella) horrens, from South Africa (17), but is
distinguished by its external form, by the arrangement of the pores and vents, and
by the details of spiculation.

R.N. 66 (Gulf of Manaar) ; 89, 137, 137a, 157, 203 (all from off Galle and onwards
up the West Coast, depths up to 100 fathoms).

Stelletta vestigium, n. sp. — Plate II., fig. 7.

Specimen irregular in shape, massive, encrusting, and containing many calcareous
foreign bodies. (In the same mass as the two specimens of Tethya lyncurium
var. a.) Maximum diameter about 23 millims. Colour in spirit, nearly black. Vents
and pores not seen. No cortex.

The skeleton is a confused reticulation of megascleres, mingled with foreign bodies
and sometimes collected into loose fibres ; with a thin dermal crust of asters, also
much mixed with foreign bodies.

Spicules. — (1.) Trisenes (Plate II., fig. 7, a-d) ; with cladi reduced to three, two, or
one mere spines or protuberances of varying size at the broad end of the spicule ; the
other end tapering gradually to a more or less sharp or irregular apex. These spicules
are of about the same dimensions as the oxea.

(2.) Oxea (Plate II., fig. 7, e.) ; rather slender, usually slightly curved; more or
less gradually sharp-pointed or irregularly ended ; measuring about 0-74 millim. by
0-016 millim.

SPONGES. 79

(3.) Spherasters (Plate II, fig. 7, f, y); with well-developed centrum and sharp,
conical rays ; total diameter about 0*024 millim., with rays about 0"008 millim. long- ;
most numerous in the dermal crust, but also abundant below. (A rather noteworthy
feature of these spicules, in many cases, in the boiled-out preparations, is the
conspicuous nature of the central canal in the rays, which gives the spicule a very
peculiar appearance.)

(4.) Oxyasters (Plate II., fig. 7, h, i) ; with very small centrum and a rather small
number of slender, finely-pointed, smooth rays ; total diameter up to about 0*024
millim. Abundant in the choanosome.

This species is particularly interesting on account of the extreme reduction of the
cladi of the trisenes, affording an absolute transition from the tetractinellid to the
monaxonellid condition. This vestigial condition of the cladi, an approach to which is
already seen in Stelletta Jierdmani, appears to be associated in this genus with the
irregularly scattered arrangement of the triaanes (compare the vestigial condition ot
the cladi of the dermal anatrisenes in Geodia peruneinata, described later on).

R.N. 200a (Station LXVIL, off Talaivillu Paar, 10-14 fathoms).

Ecioneina, Bowerbank.

The microscleres include microrhabds in addition to euasters ; the former are commonly
minutely spined or roughened, and usually form a dermal layer.

Ecionema carteri, n. sp. — Plate I., fig. 5 ; Plate III., fig. 1.

There are three specimens of this sponge in the collection. The largest and only
well-developed example (R.N. 175, Plate I., fig. 5) may be taken as the type of the
species. This specimen is massive and quite irregular in shape, with a number of
foreign bodies attached to it. The maximum diameter is about 35 millims. The
surface appears, in most places, distinctly porous, even to the naked eye, owing to the
presence of very numerous, thickly scattered, small pore-sieves (Plate I., fig. 5, 2>-s-)
containing the inhalant pores. It is also very minutely and sparsely hispid, at any
rate in parts. The vents (Plate I., fig. 5, o) are rather numerous, without prominent
margins ; circular or oval openings varying up to about 2 millims. in diameter ;
mostly congregated in a depression on what was apparently the upper part of the
sponge. They are the apertures of wide, diaphragm-bearing oscular tubes. The
texture of the sponge is fairly firm, but compressible. The colour in spirit is pale

grey.

Another specimen (B..N. 188) is almost spherical and only about 11 millims. in
diameter, without visible vents ; it is probably immature. The third (R.N. 259) is
an irregular fragment.

The main skeleton in the interior of the sponge is an irregular interlacement of
large oxea which, towards the surface, collect into loose radiating wisps, ending in

80 CEYLON PEARL OYSTER REPORT.

dense brushes of trisenes whose cladomes lie at varying depths beneath the dermal
membrane, but never very deeply. The dermal membrane is strengthened by a thin
crust of spinose microstrongyles, interrupted by the very numerous dermal pores.
The minutely hispid character of the surface is due to the presence of very slender,
hail-like anatriaenes projecting beyond the dermal membrane.

Spicules. — (1.) Plagiotrisenes (almost orthotriasnes) (Plate III., fig. 1, a, b) ; with
long shaft tapering very gradually to a sometimes almost filiform extremity, and
rather short, stout, sharply pointed or rather blunt cladi ; shaft about 1 "5 millims. by
0"033 millim., with cladi about 0-148 millim. by 0-027 millim.

(2.) Plagiotriaenes (almost protripenes) (Plate III., fig. 1, c, <•') ; with long, slender
shaft tapering to almost hair-like dimensions, and short conical cladi projecting
forwards at an angle of about 45°; shaft about 1'8 millims. by 0'0094 millim. ; cladi
about 0'028 millim. by 0-008 millim. Not very numerous.

(3.) Anatriaenes (Plate III., fig. 1, d, <!') ; with long, slender shaft and short, sharp-
pointed, strongly recurved cladi ; cladome somewhat flattened in front ; shaft and
cladi of about the same dimensions as in the preceding.

(4.) Very slender, hair-like anatriaenes with minute cladomes projecting beyond the
surface of the sponge. These spicules seem to be fairly numerous and characteristic,
but the cladomes are nearly always broken oft' in the preparations.

(5.) Oxea (Plate III., fig. 1, e) ; long and comparatively slender, usually slightly
curved ; fusiform ; gradually and usually sharply pointed at each end ; measuring,
when fully developed, about l-8 millims. by 0-037 millim.

(6.) Chiasters (Plate III., fig. 1, g); with little or no centrum and very slender,
cylindrical, very minutely roughened, tylote rays ; total diameter up to about
U-016 millim.

(7.) Microstrongyla (Plate III., fig. l,f) ; straight, rather slender and minutely
spined or roughened ; measuring up to about 0'02 millim. by 0'002 millim., most
abundant in the dermal membrane, but also plentiful in the choanosome.

There is a well-developed cortex about 0'25 millim. thick, composed of a mixture
ot chondrenchymatous, collenchymatous, and fibrous tissue, the first-named developed
chiefly towards the outside and the last towards the inside. The mesoglcea of the
choanosome is finely granular, but with an admixture of collenchymatous and
chondrenchymatous tissue. The flagellate chambers are approximately spherical,
about 0"02 millim. in diameter, and eurypylous. The dermal pore-sieves form the
thin roofs of wide chones, which pass vertically through the cortex and are continued
directly into the inhalant canals of the choanosome, without any specially differentiated
subcortical crypts.

The presence of a fibrous cortex, though not very strongly developed, seems to
indicate Sollas's genus Psammastra for this species, which is nearly related to three
sponges described by Mr. Carter under the names Stelletta geodides (18),* Stclletta
* Vide also Sollas, "Challenger" Tetractinellida, p. "200.

SPONOES. 81

badllifera (20) and Stelletta badllifera, var. robusta (21),* the first and third from
Australia and the second from the Mergui Archipelago. Sollas places the first in
his genus Psammastra, while he places the last in Edonema, which he distinguishes
from Psammastra by the absence of cortex, but Mr. Carter's specimens of Stelletta
bacilli/era, var. robusta, were dry. and it is extremely probable that they possessed
a fibrous cortex like that of Edonema carteri; the type of Stelletta bacilli/era from
the Mergui Archipelago may also have been corticate for anything we know to the
contrary, and it is almost certain that Carter's three sponges and the Ceylon species
belong to the same genus. Moreover, we do not even know whether Bowerbank's
original type of the genus Edonema (E. acervus) was corticate or not, a question
which it is impossible to decide without properly prepared sections of good
material.

U.N. 175 (Lagoon inside reef, Galle, shallow water) ; 188 (Muttuvaratu Paar, Gulf
of Manaar, 8 fathoms) ; 259 (Ceylon seas).

Ecionema laviniensis, n. sp. — Plate III., fig. 2.

The single specimen is somewhat finger-shaped, but flattened and slightly bifurcate
at one (? the upper) end. It measures about 30 millims. in length and 13 millims. in
breadth. The greater part of the surface is encrusted with coarse calcareous debris,
chiefly Foraminifera and shell-fragments ; where these bodies are absent there is a
tendency to be minutely and sparsely hispid. In places there is a minutely porous
appearance, easily visible under a pocket lens, due to the presence of small thickly
scattered pore-sieves. The vents are probably very minute and scattered — I have
only observed one. The colour in spirit is pale grey throughout, the texture firm and
very compact.

The main skeleton is a confused interlacement of large oxea, towards the surface
disposed more or less nearly at right angles to it and occasionally projecting beyond
it. At the surface there is a spicular cortex, about 0"13 millim. thick, composed of
microstrongyla with a thin layer of minute chiasters externally. Beneath the
spicular cortex lie the cladomes of the comparatively few triasnes, with their shafts
penetrating the deeper parts of the sponge at right angles. (It is, perhaps, worth
notin"- that in the deep groove at one end of the sponge, where it is beginning to
bifurcate, the spicular cortex is not developed and there are apparently no triaenes,
but the large oxea project much further beyond the surface than they do elsewhere.)
There are also a few slender anatriames and protrisenes (?) with cladomes projecting
far beyond the surface.

Spicules. — (1.) Dichotriames (Plate III., fig. 2, a, b); with stout shaft which
typically tapers very gradually to a fine point, but may be blunted; protocladi short
and stout ; deuterocladi about the same length or a little longer or shorter, conical,

* Vide also Sollas, "Challenger" Tetractinellida, p. 197.

M

82 CEYLON PEARL OYSTER REPORT.

typically sharp-pointed, but may be blunted. A typical example gave the following
measurements: shaft, l'O millim. by 0'0328 millim. ; protocladus, O'Oo millim. by
0'027 millim. ; deuterocladus, 0'065 millim. by 0"0164 millim.

(2.) Protrisenes?

(3.) Anatrkenes (Plate III., fig. 2, c, e'); with slender, often hair-like shaft and
small, variable, often abnormal cladome, perhaps characteristically flattened in front.
These spicules are evidently becoming vestigial and are so variable that it would be
useless to give measurements ; they are, however, still present in abundance.

(4.) Oxea (Plate III., fig. 2, d) large and usually stout; nearly always more or less
curved, often strongly so, sometimes crooked ; gradually and sharply pointed at each
end, or more or less blunted; measuring up to about 1'4 millims. by 0'049 millim.,
but very variable in thickness.

(5.) Chiasters, with short, blunt, sub-cylindrical rays (Plate III., fig. 2, e), total
diameter about 0'008 millim. ; very abundant in the dermal membrane, and often met
with in the choanosome, where they are commonly about half as large again and may
vary into oxyasters (?) with very slender rays (Plate III., fig. 2,f).

(6.) Microstrongyla (Plate III., fig. 2, g) almost cylindrical, usually with rounded
ends, but occasionally oxeote ; minutely spined all over ; usually slightly curved ;
sometimes centrotylote. A typical example measures about 0"1 millim. by 0-008 millim.,
but they are often much more slender. These spicules are found chiefly in the cortex,
but may also be met with occasionally in the deeper parts of the sponge.

(A few slender, smooth microxea (Plate III., fig. 2, h) also occur, possibly young
forms of other spicules. )

There is no distinct fibrous cortex, at any rate over the general surface, but a dense
spicular cortex formed as described above. In the deep depression at the end of the
sponge, however, where the spicular cortex is not developed, a certain amount of
fibrous tissue is present in its place, and it is cpuite possible that such is also present
concealed between the spicules of the spicular cortex elsewhere. (These facts serve
to emphasise the unsatisfactoriness of the fibrous cortex as a generic distinction
between Ecionema and Psammastra, to which attention has already been called
under the head of Ecionema carteri.) The spicular cortex contains an immense
number of rounded pigment cells filled with pigment granules of a rich brown
colour.

Scattered abundantly throughout the choanosome are rounded groups of minute
highly-refractive granules, staining with borax carmine, but in themselves practically
colourless. These at first sight look like flagellate chambers, each group being about
0*02 millim. in diameter, but I do not think that they are really of that nature.

The species is evidently closely related to Ecionema carteri and its allies, but is
well characterised by the dichotriames and by the comparatively large size of the
microstrongyles.

P.N. 265 (Station XLVI., off Mount Lavinia, 30 fathoms).

SPONGES. 83

Family: GEODIIDJ'..

Astrophora, in which the characteristic microsclere is a sterraster, funning a dense
cortical layer.

Geodia, Lamarck.
Geodiidse, in which the somal microsclere is a polyactinose aster.

Geodia perarinata, Bowekbank.

1873, Geodia perarmatus, BOWERRANK (22); 1880, Geodia perarmata, CARTER (4); 188S,
Geodia perarmata, Sollas (15).

There is an interesting series of specimens belonging to this species in the collection.
The smallest is only 13 millims. in diameter, and is approximately spherical in shape;
unfortunately, the surface of this specimen is considerably abraded, and I have not
been able to make out the arrangement of either inhalant or exhalant apertures.
The next in size, about 18 millims. in diameter, is also too much injured to show the
arrangement of the apertures satisfactorily. The next is more irregular in shape,
about 28 millims. in diameter, and flattened on one side so as to form an approxi-
mately circular oscular area about 14 millims. in diameter, but not definitely bounded.
This area is covered by a pore-bearing, cribriform membrane, with pores about
0-08 millim. in diameter, and few, if any, larger apertures. The remainder of the
surface of the sponge, outside the oscular area, is covered with a cribriform dermal
membrane containing thickly scattered inhalant pores of somewhat smaller diameter
than the exhalant pores. The specimen next in size is also irregular in shape, closely
resembling a potato, and about 45 millims. in diameter. The oscular area now forms
an ill-defined, but rather deep concavity on one side, whose floor is covered, as before,
by a cribriform membrane, hut with a few larger openings where the pores have
ajiparently become confluent and thus formed small " vents" ; the maximum diameter
of this basin-shaped oscular area, which is not quite circular, is about 17 millims.
The largest specimen is a very fine one, irregularly rounded in shape, and about
75 millims. in greatest diameter ; it has been attached below at only a few points to
calcareous debris. The depressed oscular area occupies the greater part of the
flattened upper surface of the sponge, and is about 38 millims. in greatest diameter,
and irregularly oval in shape. The floor of the depression is somewhat flattened, but
rises up in the middle; the margin is much better defined than in the smaller
specimens, but rounded off, except on one side, where it actually overhangs slightly.
The pore-bearing membrane presents the same appearance as in the last-described
specimen, but the small " vents," due to confluence of the pores, are more numerous ;
they are hardly 0'5 millim. in diameter. As before, the general surface of the sponge
is covered by a richly porous cribriform membrane. There is also in the collection an
irregularly-shaped sub-cylindrical specimen, about 36 millims. in length and
18 millims. in diameter, which does not show any distinct oscular area.

M 2

84 CEYLON PEARL OYSTER REPORT.

It appeal's, then, that the basin-like character of the oscular area, which was
figured by Bowerbank (22) in the type specimen, and which appears to form a charac-
teristic feature of the adult sponge, is not developed until comparatively late in life,
and it appears also that its development is due simply to more rapid groAvth of the
surrounding parts. According to the classification adopted by Sollas in his work
on the " Challenger " Tetractinellida, this mode of development would necessitate our
placing the species in the genus Cydonium. Sollas himself, however, places the
species in the genus Geodia, having been insufficiently acquainted with the characters
in question. Lendenfeld, in his work on the Tetractinellida of the Adriatic, merges
the genus Cydonium, in Geodia, and in this I must agree with him", for the present
instance demonstrates very clearly the impracticable character of a classification
which involves a study of the development, at any rate until our knowledge is much
more complete than at present.

The general surface of the sponge is smooth and, perhaps, even glabrous, where the
dermal membrane is intact, but from a few points on the lower surface, in the
immediate neighbourhood of the attachment, a few long, hair-like spicules (mostly
anatriasnes) project. The colour of the general surface (in spirit) is purplish brown,
mottled with dull yellow, the pigment being lodged in an immense number of
granular brown pigment cells which lie between the layer of sterrasters and the
dermal membrane. Internally the sponge is nearly white.

The cortical layer of sterrasters is about 1 '2 millim. thick in the largest specimen,
and between this layer and the dermal membrane lies a zone about 0-27 millim. thick,
in which the cladi of most of the triaenes are found. The manner in which the shafts
of these trisenes pierce the layer of sterrasters, so that their cladi come to lie
externally to it, is, as already pointed out by Bowerbank, very characteristic of the
species, and forms a conspicuous feature even in the smallest specimen which I have
seen.

The spiculation, as observed in the largest specimen, is as follows : — ■

(1.) Dichotrisenes ; shaft stout, about 3"2 millions, by 0"07 millim.; chord, 0'48
millim. to 0"64 millim.

(2.) Protriasnes ; scarce (seen only in another specimen); cladi about 0'075 millim.
long.

(3.) Anatriaanes ; shaft slender, about 4-2 millims. long; cladi about 0'07 millim.
long.

(4.) Somal oxea ; about 2'67 millims. by 0"042 millim.

(5.) Cortical oxea; about 0"3 millim. by 0'0083 millim.

(6.) Sterrasters; about 0'13 millim. in diameter.

(7.) Somal chiasters or spherasters ; about 0'009 millim. in diameter.

(8.) Choanosomal spherasters with spined rays; about 0-04 millim. in diameter.

The spherasters with spined rays are found just beneath the layer of sterrasters
and appear to be characteristic of the species. The fact that the spination of the

SPONGES. 85

rays is not figured by Bowerbank in his illustration, nor mentioned in the text,
makes me a little doubtful of the identification of the Ceylon species with that
originally described by Bowerbank (from an unknown locality). Carter, however,
made the identification in the first instance and Sollas has accepted it ; and the
differences, if they really exist, must be extremely slight.

K.N. 119, 121, 124, 125, 126, 135. (All from deep water, up to 100 fathoms, off
Calle and onwards up the West Coast of Ceylon.)

Geodia peruncinata, n. sp. — Plate III, figs. 3, 3a, 3b.

The single, specimen in the collection is, unfortunately, only a fragment, amounting
to probably somewhat more than half of a small spherical sponge, attached to a
small calcareous nodule. The diameter of the specimen is about 1 1 millims. The
colour in spirit is nearly white, both on the surface and in the interior. The cortex
is rather thin and very brittle, and the interior of the sponge is soft and friable.
It is impossible to make out the arrangement of the exhalant canal-system, but the
thin dermal membrane contains numerous small pores, presumably inhalant.

The layer of sterrasters (Plate III., fig. 3, e, e) is about 0-33 millim. thick, and is
separated from the dermal membrane by an interval of about 0'165 millim. occupied
by extensive sub-dermal cavities. The dermal membrane is supjwrted on the
cladomes of dichotrirenes (Plate III., fig. 3, a'), whose shafts pierce the layer of
sterrasters and the zone of sub-dermal cavities ; it is also strengthened and rendered
hispid by the presence of immense numbers of short anatrisenes which project from
the surface at various angles (Plate III., fig. 3, c, c). Some of the dichotrisenes
(Plate III., fig. 3, a) have their cladi extended beneath the layer of sterrasters, and
the remainder of the skeleton is arranged in the usual way, with radiate primary
lines.

Spicules. — (1.) Dichotrigenes (Plate III., fig. 3, a, a'); with stout shaft tapering
gradually to a narrow, but blunt apex, and short cladi ; the deuterocladi usually two
or three times as long as the protocladi, but variable and often unequal, gradually and
sharply pointed, slightly curved towards one another, or straight. Shaft measuring
up to about 2-4 millims. by 0"044 millim., with cladome 0'62 millim. in total diameter
and protocladi 0'038 millim. thick.

(2.) Protriaenes ; a lew stout protriaenes, more or less broken and sometimes with
four cladi, occur amongst the mass of spicules projecting from the base of the
sponge.

(3.) Anatriaenes (Plate III., fig. 3, b) ; with well-developed cladome and very long,
slender shaft. Shaft measured up to about 4 "9 millims. in length (and then broken
off), with a thickness of O'OIG millim. near the cladome, and cladi stout, conical,
and about 0"057 millim. long. (Projecting from the base of the sponge one finds
anatriaenes with long, hair-like shafts and slender, very sharply recurved cladi.)

(4.) Anatritenes (Plate III., fig. 3, c; fig. 3a); with short, slender, fusiform shaft

86 CEYLON PEARL OYSTER REPORT.

and verv small, often vestigial cladome, which may be reduced to a mere knob. The
attachment of the cladome to the shaft is so slender that the former is generally
broken off and the spicules then may easily be mistaken for the well-known dermal
oxea of other species of Geodia. The proximal end of the shaft may be oxeote or
rounded off, or even tylote. In a typical example the shaft measures about
0'21 millim. by O'OOG millim. The cladome varies so much in its degree of reduction
that it is useless to attempt exact measurements, but is generally not more than
0-008 millim. in total diameter. These spicules are found for the most part hispidating
the surface of the sponge, but a few occur beneath the layer of sterrasters. I have
seen one example which appears to be a reduced protriaene, with a single remaining
cladus, but the great majority are anatrisenes.

(5.) Oxea (Plate III., fig. 3, d) ; very long and comparatively slender, frequently
curved, sometimes very crooked ; fairly gradually and sharply pointed ; size variable,
e.g., 2"5 nhllims. by 0'029 millim.

(6.) Sterrasters (Plate III., fig. 3, e, e' ; fig. 3b, c) ; markedly oval in shape and
measuring about 0'12 millim. by 0D82 millim.

(7.) Small spherasters (or chiasters?) (Plate III., fig. 3b, a); with small centrum
and numerous almost cylindrical or slightly tylote rays about as long as the diameter
of the centrum ; total diameter of the spicule about 0'008 millim. Most abundant in
the dermal membrane, but plentiful also in the interior.

(8.) Comparatively large spherasters (Plate III., fig. 3, /'; fig. 3b, b); with large
centrum and very numerous short, conical rays, whose length is only about one-third
the diameter of the centrum or a little more. Total diameter when fully grown
about 0'024 millim. Sparsely scattered beneath the layer of sterrasters.

In the structure of the ectosome and the presence of dichotriasnes this species is
evidently closely related to Geodia perarmata. It is, however, distinguished by
three characters : (1) the absence of brown pigment cells in the outer part of the
cortex ; (2) the form of the larger spheraster, which, in G. perarmata, has (? always)
distinctly spined rays ; (3) the presence of the numerous short-shafted, cortical
anatriasnes,* which are replaced in G. perarmata by cortical oxea, though I am
inclined to think that a few of these oxea may still show a vestige of a cladome even
in G. perarmata, while it is quite possible that a few true cortical oxea may occur
amongst the reduced anatriaenes in G. peruncinata.

The presence of the vestigial anatriaenes is extremely interesting as indicating the
anatrirenal origin of the cortical oxea of Geodia. Soeeas (loc. cit., p. cxlvii) has
already observed, in his general remarks on the family Geodiidas, that "A second
finer hispidation is frequently produced by small oxeas, which are confined to the
cortex (cortical oxeas). Associated with these, in some few instances, are minute
anatrisenes, which much remind one of the cladose tylostyles described by Dendy and
Ridley in Proteleia sottasi." We have here an admirable example of the evolution

' Very similar small cortical anatrisenes occur in Lindgeen's Geodia arripiens from Cochin China (86).

SPONGES. 87

of a monaxonellid spicule from a tetractinellid form by subordination and final
suppression of three of the rays.

R.N. 223 (Station XLL. 12 miles oft'Galle, 100 fathoms).

Geodia areolata, Carter.

1880, Geodia areolata, C.vhtkr (4); 1888, Geodia areolata, Soi.i.as (15).

There are several specimens of this sponge in the collection. They are approximately
spherical in shape, and in form and colour closely resemble young specimens of
(i. <pera/rmata, but the largest sent to me is only 21 millims. in diameter. The surface
of the sponge is smooth and characteristically marked out into small stellate ai'eas of
a pale yellow colour, separated from one another by a reticulation of chestnut-brown
pigment. This pattern, however, is not recognisable everywhere. Numerous small
apertures appear irregularly scattered over some parts of the surface, but with a
tendency towards grouping. These frequently occupy the centres of the pale stellate
areas, and they are often guarded by a fringe of convergent dermal oxea. They are
frequently covered over by a reticulate dermal membrane. They are the openings of
chones, which pierce the layer of sterrasters, but whether they are inhalant or
exhalant, or both, I have been unable to decide.

The layer of sterrasters is about 0"46 minim, thick in the largest specimen. The
trisenes almost invariably extend their cladi beneath this layer, and do not pierce it
to reach the dermal membrane. The dermal oxea are abundant and arranged more
or less perpendicularly to the surface ; they extend from the outer surface of the layer
of sterrasters and project for a short distance beyond the dermal membrane. The
granular brown pigment cells, to which the sponge owes its characteristic colour, are
scattered in the ectosome outside the layer of sterrasters.

The spiculation is as follows : —

(1.) Orthotrisenes ; with shaft about 2 '37 millims. by 0"06 millim.

(2.) Anatriasnes ; with very long and slender shaft and cladi about 0*054 millim.
long.

(3.) Soma! oxea; about 2*0 millims. by 0'029 millim.

(4.) Cortical oxea ; about 0"2 millim. by 0'008 millim.

(5.) Sterrasters; about 0*08 millim. in diameter, oval or nearly spherical.

(6.) Somal chiasters or spherasters ; about 0'0083 millim. in diameter.

(7.) Choanosomal oxyasters or spherasters ; mostly with few long and slender,
smooth rays; about 0"0165 millim. in total diameter.

No typical protrisenes were seen, but only small plagiotrisenes, which may be young
forms of the large orthotrisenes.

In this case, as in that of G. perarmata, we can include the species in the genus
Geodia only by abandoning the restricted definition thereof given by Sollas.

The species, as already indicated, is at first sight easily confounded with
G. perarmata. The specimens of the two were mixed together in the collection, and

88 CEYLON PEARL OYSTER REPORT.

only a part of the total number was forwarded to me for Investigation, under the
impression that they were all specifically identical. I am therefore unable to say how
many of each were really collected by Professor Herdman and his assistants. Even
small specimens are, however, very easily distinguished by microscopical examination,
as will be evident from the above descriptions, and the species are not really even
closely related.

Mr. Carter received his material of this species from the Gulf of Manaar.

R.N. 120, 122, 123 (deep water off Galle, up to 100 fathoms).

Geodia ramodigitata, Carter.

1880, Geodia ramodigitata, Carter (4); 1888, Geodia (?) ramodigitata, .Soi.las (15).

There is one specimen in the collection which appears to be referable to this species.
It is of remarkably irregular shape, with one strongly curved, finger-like projection or
branch. Unfortunately the specimen is a good deal injured on one side, so that the
exact shape cannot be determined. The maximum diameter of the specimen in its
present condition is about 40 millims. The colour (in spirit) is pale grey, and a good
many foreign bodies are attached to the surface here and there. The vents are
minute, for the most part irregularly grouped over the ends of large exhalant canals,
with which they communicate by means of narrow chones penetrating the cortex.
The outer ends of these chones are usually covered over by a sieve-like dermal
membrane which sub-divides the vent into a group of comparatively small pores.
Inhalant pores irregularly scattered, perhaps sometimes in groups, but of the numerous
pore-sieves present in the dermal membrane it is extremely difficult to say which are
inhalant and which exhalant.

The layer of sterrasters is about 1 millim. thick and lies close up to the dermal
membrane, with the cladi of the trieenes spreading out beneath it. The dermal
membrane is very minutely hispid with projecting oxea.

The spiculation is as follows: — (1.) Orthotriames ; shaft stout, gradually and
finely pointed, about 2 '5 millims. by 0"08 millim; cladi stout, sharp pointed, up to
about 0"35 millim. long, extended nearly at right angles to the shaft. Sometimes
one or more of the cladi undergoes reduction or may even become obsolete, leaving the
spicule with two, one or no properly developed cladi, as the case may be.

(2.) Protrisenes ; shaft very long and slender, measured up to about 3 '7 millims.
by 0-023 millim., with sharp-pointed cladi about 0"18 millim. long.

(3.) Anatrisenes ; shaft very long and slender, measured up to 4-0 millims. by
0"018 millim.; with stout, recurved, sharp-pointed cladi about 0*09 millim. long;
not very numerous.

(4.) Somal oxea; gradually and sharply-pointed at each end, about 2'57 millims.
by 0"04 millim., often slightly curved.

(5.) Cortical oxea; about 0"2 millim. by 0"007 millim.

(G.) Sterrasters ; about 0"09 millim. by 0'075 millim.

SPONGES. 89

(7.) Chiasters ; extremely minute, with short, truncate or slightly tylote rays;
total diameter about 0"006 millim. ; very abundant in the dermal membrane.

(8.) Oxyasters; of unusually large size, with no distinct centrum, with large,
slender, smooth, sharp-pointed rays, rather few in number ; total diameter about
0'0G6 millim. ; very abundant in the deeper parts of the sponge.

There are two points in which the spiculation of Professor Herdhan's specimen, as
given above, diners considerably from that given by Carter (4) for the type of the
species. In the first place, Carter mentions no cortical oxea, but he remarks that
the surface is much worn, so that we need lay but little stress upon this difference. In
the second place, the internal oxyasters appear to be twice as large in our specimen as
in the type ; this also I do not consider a very important difference, as the internal
asters of Geoilia are subject to much variation in size. Their form, on the other
hand, agrees very closely in the two cases and appears to be characteristic. The
other spicules, also, with the exception of the sterraster, appear to be somewhat
larger in the present specimen ; but on the whole there can be very little doubt of
the correctness of the identification, especially when we remember that Mr. Carter's
type specimen came from the Gulf of Manaar.

R.N. 13G (Station XLL, 12 miles off Galle, 100 fathoms).

Sub-order: SIGMATOPHOEA.

Tetractinellida with trirenes ; with sigmata for microscleres (when present); without
asters.

Family : TETILLID^E.

Sigmatophora with well-developed protriaenes and with skeleton usually strongly
radiate in arrangement.

Tetilla, Schmidt.

Cortex absent or feebly developed ; no special cortical skeleton.

Tetilla hirsuta, Dexdy.

1889, Tetilla hirsuta, Dendy (3).
This remarkable species is represented in the collection by three specimens which
differ somewhat amongst themselves, but all agree in the possession of the highly
characteristic, cup-shaped, poriferous pits. In all the surface is more or less strongly
hispid from projecting spicules, and encrusted with a layer of dirt between, to which
the colour of the surface is due, this colour varying with the nature of the foreign
matter. Two of the specimens are more or less spherical, one (R.N. 129) being
attached to a mass of Nullipore ; the third (E.N. 241) is constricted and slightly
elongated below, where it has been torn off from its attachment. In one specimen
(R.N. 129) there are, in addition to the characteristic spicules, a fair number of quite
small oxea, varying much in size, and irregularly scattered in the choanosome. Such
small oxea do not appear to be present in the type, but, in view of their variability

N

90 CEYLON PEARL OYSTER REPORT.

in size and sporadic distribution, one can hardly regard their presence as constituting
a specific or even varietal distinction. It should be noted, however, that very
similar spicules appear to be characteristic of the next species (T. pocvlifera).

The arrangement of the pores and vents in this sponge is very peculiar and might
be thought by some authorities to deserve generic recognition. The former are
congregated in more or less deep, cup-shaped pits, guarded by very slender projecting
protrisenes and prodisenes ; these pits occur, mostly, at any rate, on the lower parts of
the sponge and may slope upwards. The vents (or better, cloacal cavities) are
represented by pits containing the openings of excurrent canals. As Sollas has
pointed out (15, \i. cxxv), a somewhat similar arrangement is found in his genus
Cinachyra, closely related to Tetilla, but that genus is also characterised by the
])resence of a thick cortex containing cortical oxea. In Tetilla hirsuta the cortex is
only feebly developed, more or less fibrous, but without special cortical spicules.

For further specific details I must refer to the original description, to which I may
add that the flagellate chambers are apparently eurypylous.

In the choanosome of R.N. 129 numerous minute and apparently siliceous spherules
were observed, up to about 0'004 millim. in diameter. The nature of these bodies
will be discussed under the head of Tetilla anomala, in which species they also occur.

R.N. 129 (deep water off Galle and onwards up West Coast); 177 (lagoon inside
the reef, Galle, shallow water); 241 (Ceylon seas).

Tetilla poculifera, n. sp. — Plate I., fig. 6 ; Plate III., fig. 4.

The type specimen (R.N. 230, Plate I., fig. 6) is somewhat pear-shaped, with the
broad end attached to a mass of calcareous organisms (Polyzoa, &c). The apex of
the sponge is occupied by a deep cup-shaped cloaca (Plate I., fig. 6, CI.), whose
margin is fringed with spicules, and whose floor is a sieve-membrane. The surface of
the sponge is uneven, irregularly hispid, and thinly encrusted with sand-grains.
Colour (in spirit) yellowish grey. Height of specimen 16 millims., transverse
diameter 14 millims. Diameter of cloacal aperture 2 "5 millims.

The skeleton consists chiefly of bands of large oxea radiating outwards from a
central "nucleus"; the ends of these bands commonly project slightly beyond the
surface.

Spicules. — (1.) Plagiotrirenes (Plate III., fig. 4, a, b) ; few in number, but
evidently proper to the sponge, though, perhaps, not specifically characteristic. A
typical example from a boiled-out preparation has a fairly stout shaft, tapering
gradually to a hair-like extremity, and about 1'3 millim. long by 0'0185 millim. thick
below the cladome ; cladi short and stout, about 0*13 millim. long, with apices fairly
sharp and slightly incurved. (I have also seen fragments of two large dichotriames,
but it is doubtful whether these are proper to the sponge.)

(2.) Protrisenes (Plate III., fig. 4, d, d') ; with long, slender shaft, and long,
slender, sharp-pointed cladi; conspicuous around the cloacal opening; size variable.

SPONGES. 91

(3.) Anatriienes (Plate III., fig. 4, c, <■') : of" ordinary form, with short cladi and
very long, slender, hair-like shaft. A specimen in a boiled-out preparation has cladi
about 0"037 millim. long and shaft about 2"8 millims. long.

(4.) Oxea of the main skeleton (Plate III., fig. 4, e) ; stout, straight, fusiform, and
tapering very gradually to the extremities, which may be sharply pointed, or rounded
off or irregular; size about 3'1 millims. by 0-046 millim.

(5.) Sigmata (Plate III., fig. 4, h) ; slender, more or less contort, measuring about
0"017 millim. from bend to bend; abundant.

(G.) Small scattered oxea (microxea) (Plate III., fig. A,f,g); straight or slightly
curved ; smooth, slender, fusiform, gradually and finely pointed at each end ; size
about 0'23 millim. by 0'005 millim. These spicules are scattered irregularly in
enormous numbers throughout both choanosome and ectosome ; in the latter position
they may form quite a dense layer at the surface of the sponge, in which most of
them lie tangentially. They may, perhaps, be regarded as microscleres.

Stained sections show that there is no cortex, and that the ectosome, though
fairly thick, is not sharply differentiated from the choanosome. The material is not
in a fit condition for minute histological investigation, but I have been able to make
out that the flagellate chambers are oval and about 0-029 millim. in greater diameter.
The inhalant pores are apparently scattered over the surface, and the inhalant canals
appear to originate in irregular lacunar subdermal cavities. The larger exhalant
canals are provided with numerous diaphragms, and terminate below the sieve-like
floor of the cloaca.

The most characteristic features of this species are the structure of the cloaca and
the presence of the very numerous smooth microxea. In Tetilla (?) australiensis
(Carter) minutely spined microxea, 0-21 millim. long, are present, and Sollas (15,
p. 43) observes of this species that the oxeote microscleres are almost unique amongst
the Tetillkke.* The case of T. hirsuta, described above, however, seems to show that
in the oxea of this genus (as in PlakinastreUa, &c.) it is impossible to draw a hard and
fast line between mega- and microscleres.

In addition to the type described above, there are in the collection two other
specimens which may be referred to the same species, viz., Pt.N. 189 and 205. Both
are somewhat imperfect. R.N. 205 has several cup-shaped (cloacal ?) cavities, with
sievedike floors, irregularly distributed ; the other is too imperfect to show the
character of the vents.

K.N. 189 (Muttuvaratu Paar, Gulf of Manaar) ; 205 (Gulf of Manaar) ; 230 (deep
water off Galle and onwards up West Coast of Ceylon).

Tetilla anomala, n. sp. — Plate III., fig. 5.

There are two specimens of this sponge in the collection, the larger of which
(R.N. 153) may be regarded as the type. It appears to be a fragment, amounting to

* (Jxeotu microscleres occur also in the genus Pwatetttla (vidt infrd).

92 CEYLON PEARL OYSTER REPORT.

nearly half, of au irregularly spherical sponge, ahout 35 millims. in greatest diameter.
The surface is rather uneven, minutely hispid, and thinly encrusted with sand.
Neither pores nor vents are visible externally. The colour, in spirit, is grey, the
texture firm and compact and only slightly compressible.

The skeleton consists chiefly of stout bands of oxeote spicules radiating from a
centrally (?) placed "nucleus" to the surface of the sponge. The outer portions of
these bands are often abundantly echinated by the cladi of the anatrisenes, lying
chiefly in the ectosome.

Spicules. — (1.) Protriaenes (Plate III., fig. 5, a) ; with rather short, stout cladi and
fairly stout shaft, tapering to hair-like fineness. Shaft about 276 millims. by
0-015 millim. (at the thickest); cladi about 0-054 millim. by 0'00625 millim. The
cladi may be slightly irregular, and I have seen one forked at the extremity. They
are sometimes much longer than in the specimen measured. These spicules occur in
positions similar to those of the anatriasnes, but are very scarce.

(2.) Anatriaenes (Plate III., fig. 5, b, b') ; with fairly stout, sharp-pointed cladi and
long hair-like shaft; length of shaft about 27 millims., of cladi about 0'058 millim.
Abundant.

(3.) Oxea (Plate III., fig. 5, c) ; stout, fusiform, gradually and finely pointed at
each end; size about 2 "8 millims. by 0'046 millim.

(4.) Sigmata (Plate III., fig. 5, d) ; slender, usually contort; very numerous,
especially in the walls of the inhalant canals; measuring up to about O'Ol millim.
from bend to bend.

(5.) Spherules ; smoothly rounded, but irregular in shape ; up to about 0"004 millim.
in diameter, but usually smaller. These bodies are enormously abundant in the
choanosome, but are practically absent from the ectosome. In the type specimen
they are thickly, but irregularly scattered; in R.N. 192 they are grouped in oval
clusters about O'll millim. in longer diameter.

Stained sections show a fairly thick ectosome pretty sharply differentiated from
the choanosome and, to some extent, fibrous, but composed chiefly of chondrenchyme
with very numerous granular cells. The thickness of this ectosome, which almost
amounts to a cortex, in the type specimen, is about 074 millim. It is penetrated
here and there by narrow inhalant canals, leading almost vertically inwards from the
surface. The flagellate chambers are oval or nearly spherical, up to about 0'033
millim. in diameter. Their mode of opening and the arrangement of the excurrent
canal system have not been made out. The mesoderm of the choanosome ranges from
collenchymatous to chondrenchymatous.

The extraordinary number of the siliceous spherules or globules in this sponge is
very remarkable. In R.N. 192 they are, as already stated, grouped together in oval
masses. They appeal' in this case to originate many together in special mother cells,
in which they first appear as very minute, highly refringent points. The oval groups
are probably still associated with the remains of the mother cells. The exact nature

SPONGES. 93

of these bodies and their taxonomic value are extremely doubtful. Sollas (15)
mentions that such bodies occur in the Tetillkke associated with sigmaspires,
hut in Oinachyrcc barbata, where they are said to be as much as 0"0535 millim. in
diameter, he regards them as " accessory or accidental " forms, and the only species
of Tetilla in which they are mentioned appears to be Tetilla arabica (Carter), where
they are of about the same size as in T. anomala. I myself have seen similar bodies
in one specimen of Tetilla hirsuta (ride swpra), but only very locally developed.
That the spherules are really siliceous appears to be tolerably certain, for they
appear abundantly in preparations of T. anomala which have been boiled out with
nitric acid. Whether they can be regarded as definite and characteristic spicules is,
however, another question, and it appears to me not impossible that they may be
simply reserves of siliceous material destined to be re-absorbed later on and used
for spicule formation. This view is strongly supported by their very sporadic
distribution.

There is possibly a close relationship between T. hirsuta, T. poculifera, and
T. anomala.

R.N. 153 (Station V., off Chilaw, 10 fathoms) ; 192 (Station LVIL, outside Dutch
Modragam Paar, 11^ to 3(5 fathoms).

Tetilla limicola, n. sp.— Plate I., fig. 7 ; Plate III., fig. 6.

Sponge (Plate I., fig. 7) somewhat fig-shaped, constricted below, broadly rounded
above ; may be slightly lobose ; may be laterally compressed ; anchored in the mud
by a great mass of fine silky spicules attached to the base. Surface glabrous, but
at the same time minutely and sparsely hispid above, more strongly hispid on the
sides ; also very minutely reticulate on the upper parts. Texture soft and spongy,
but very compact, i.e., without wide canals or cavities in the interior. Colour in
life, pink ; in spirit, grey. Vents (Plate I., fig. 7, o) of fair size, several, slit-like,
in deep or shallow depressions on the upper part of the sponge ; each leading into
a wide but shallow cloaca! cavity with an almost flat floor perforated by numerous
minute openings of very narrow exhalant canals. Inhalant pores scattered between
the surface tufts of spicules. A spirit specimen (exclusive of the root tuft) measures
about 43 millims. in height, 48 millims. in greatest breadth, and 31 millims. in
greatest thickness. The root-tuft (Plate I., fig. 7, r.t.) is nearly as large as the
specimen itself, and in its present condition consists of a mass of soft mud held
together by the extremely long and slender silky anchoring spicules, which
individually are scarcely visible to the naked eye.

The skeleton consists of the following parts : —

(1.) Loose longitudinal bands of very long, slender oxea run almost parallel to one
another throughout the body of the sponge. In a longitudinal section taken across
the greatest breadth of the sponge these fibres are seen to converge towards a point
situated at a short distance below the depressed apex of the sponge. Owing to the

94 CEYLON PEARL OYSTER REPORT.

constriction of the sponge below, the lower ends of these fibres approach the surface
almost without diverging from one another.

(2.) More or less mingled with the foregoing in the lower parts of the sponge are
long, silky bundles of the very long, sinuous, hair-like shafts of anatrisenes, whose
heads echinate the bundles at varying distances beneath the surface.

(3.) Dense surface brushes, composed partly of oxea which form the ends of the
long fibres and partly of protrisenes of hair-like thinness, whose cladomes project far
beyond the surface.

(4.) Numerous oxea, much shorter than those of the fibres, scattered irregularly
throughout the body of the sponge.

(5.) The root-tuft, composed of anatrisenes with extremely long and slender shafts,
irregularly matted together.

Spicules. — (f.) Protrisenes (Plate III., fig. 6, b.), (and diaenes) ; shaft and cladi
often of hair-like thinness ; cladi commonly of unequal length (one sometimes
suppressed entirely), projecting forwards at only a very small angle with the shaft.
In a perfect example in a boiled-out preparation the shaft is about 3"0 minims, long,
about 0'004 millim. thick just below the cladi, and tapering to hair-like fineness at
the other end. The longest of the cladi is about 0'05 millim. in length.

(2.) Anatriasnes (Plate III., fig. 6, a); with extremely long, hair-like shafts, so
long and slender that they appear to be invariably broken off even in boiled-out
preparations, and I am therefore unable to give the measurements. The cladi are
fairly stout, gradually sharp-pointed, recurved at a very acute angle to the shaft,
and about 0-04 millim. long by 0"004 millim. thick at the base. In boiled-out
preparations the shafts of these spicules stick together in silky wisps.

(3.) Oxea (Plate III., fig. 6, c.) ; long and slender, commonly slightly curved,
very gradually and finely pointed, varying greatly in size according to position,
measured up to about 2"1 miHims. by 0-01 millim.

(4.) Sigmata (Plate III., fig. 6, d) ; slender and commonly contort, measuring about
0"008 millim. from bend to bend. Very numerous.

The material is hardly sufficiently well preserved for minute histological investi-
gation, but the examination of sections prepared by the usual paraffin method shows
us the following features. There is no cortex and no distinct dermal membrane, the
ectosome not being sharply differentiated from the choanosome. The small inhalant
pores, scattered between the dermal brushes of spicules, lead directly into narrow,
elongated canals, which run inwards at right angles to the surface and unite below
the spicular brushes in larger inhalant canals which penetrate the deeper parts of the
sponge. The flagellate chambers are approximately spherical and about 0-025 millim.
in diameter ; they are probably eurypylous. The exhalant canals are all narrow,
and converge towards the flask-shaped cloaca! cavities already mentioned. Numerous
large ova, many of them having very prominent pseudopodia, are scattered through
the choanosome. These ova measure up to about 01 1 millim. in diameter. They

SPONGES. 95

have uniformly granular cytoplasm and large, well-defined nuclei, each typically with
a single darkly-staining nucleolus.

Of this remarkable sponge eight specimens were found, anchored in the soft mud
at the bottom of Tamblegam Lake (a large inlet from Trincomalee Bay) ; but only one
was forwarded to Cape Town for more minute investigation.

The species is evidently nearly related to Tetitta dactyloidea (Carter), which that
author records (20, 34, 43) from the south-east coast of Arabia, Bombay, and the
Mergui Archipelago. The principal difference apparently concerns the excurrent canal-
system. In T. dactyloidea tbere appears to be a single vent at the summit of the
sponge, and Mr. Carter observes that the terminal aperture divides into a number of
branches, which, sub-dividing, permeate the mass generally down to its base. In
T. limicola, as we have seen, the sponge is very compact throughout, and there are
no wide tubes in it, the excurrent canals being very narrow and opening by numerous
minute apertures in the floors of somewhat flask-shaped cloacae with slit-like vents on
the surface of the sponge. It is highly probable that this arrangement is a special
adaptation to the conditions of life, serving to hinder the entrance of the very fine
soft mud, in which it lives, into the interior of the sponge ; even as it is, a considerable
amount of mud may be seen in the excurrent canals just beneath the floor of the
cloaca.

R.N. 70 (one of eight specimens from Tamblegam Lake, Trincomalee).

Craniella, Schmidt.

The ectosome is differentiated into an inner fibrous layer, containing more or less
radially arranged cortical oxea, and an outer collenchymatous layer excavated by
large subdermal cavities.

Craniella elegans, n. sp. — Plate IV., fig. 1.

Sponge free, irregularly spherical. Surface covered with close-set conuli ; hispid
with brushes of spicules projecting from the conuli ; minutely reticulate between the
conuli. Vents not visible in the type specimen. Pores in the thin dermal membrane
stretched between the conuli. Colour, in spirit, purplish grey externally, yellow
internally. Greatest diameter of type (R.N. 193) about 20 millims. Consistence
firm and compact.

The main skeleton consists of stout bundles of spicules radiating from a central
" nucleus" and ending in the brushes which project from the surface conuli (Plate IV.,
fig. 1). These radiating bundles (r.b.) are separated from one another by fairly wide
intervals, and are composed of large oxea, with anatrioenes and protrisenes. The cladi
of the anatrieenes are, for the most part, situated in the outermost part of the
choanosome, just beneath the cortex, while those of the protrieenes mostly project
beyond the surface of the sponge in the hispidating brushes. The special cortical
skeleton is, as usual in Craniella, confined to the inner, fibrous layer of the cortex,

<)f> CEYLON PEARL OYSTER REPORT.

and is composed of stout oxea ; these are not arranged strictly radially, but are
inclined at various angles to the surface, with their outer extremities commonly
abutting at more or less acute angles against the stout bundles of the main skeleton
as the latter penetrate the cortex on their way to form the surface brushes.

Spicules. — (1.) Protrisenes (Plate IV., fig. 1, b) ; with long shaft, rather stout
distally but becoming setaceous proximally, and short, stout cladi of approximately
ecpial length. Shaft in measured specimen 4*78 millims. by 0*0368 millim. (near
cladome) ; cladi, 0-129 millim. long.

(2.) Anatrisenes (Plate IV., fig. I, a); with long and very slender, setaceous shafts,
and short, stout, sharp-pointed cladi ; shaft measured up to 4*87 millims. by
0*01 38 millim. (near cladome), with cladi 0"0736 millim. long. Very numerous.

(3.) Oxea of the radiating bands (somal oxea) ; stout, aniso-actinate ; more or less
gradually sharp-pointed distally, but drawn out into long setaceous filaments
proximally; size up to about 3"27 millims. by 0"046 millim.

(4.) Cortical oxea (Plate IV., fig. 1, c.o.) ; stout, fusiform, aniso-actinate, sharp-
pointed at each end, the proximal end narrower than the distal, but not setaceous ;
commonly slightly curved; size about l'O millim. by 0'0368 millim.

(5.) Sigmata ; slender, 0"01G6 millim. from bend to bend.

The cortex of the type specimen measures, in places, as much as 1*8 millims. thick,
and is very sharply differentiated into two layers, of which the outer appears to be
usually somewhat thicker than the inner. The outer layer is excavated by very large
subdermal (intra-cortical) cavities (Plate IV., fig. 1, i.c.c), which occupy nearly all
the space between the radiating spicule bundles of the main skeleton, with the thin
dermal membrane stretched over them. The dermal membrane (d.rn.), and also the
deeper parts of the outer layer of the cortex, between the subdermal cavities,
contain numerous small brown pigment granules grouped in more or less spherical
cells. The inner layer of the cortex (f.c.) is densely fibrous and contains no pigment
granules ; it is strengthened by the special oxea above described, and the fibres lie,
for the most part at any rate, parallel to the surface.

The flagellate chambers are about 0'02 millim. in diameter, spherical or oval,
eurypylous or with short, wide excurrent canaliculi. The choanosome is crowded
with large embryos (Plate IV., fig. 1, crnh.).

In addition to the type above described, there is in the collection another very
small specimen (H.N. 204) of pear-shaped form, with a single apical vent. The
maximum diameter of this specimen is only about 8-5 millims. In the structure of
the cortex and in the form and arrangement of the spicules, it agrees very well with
the type, but the megascleres are all much smaller. This difference, however, is
probably due to immaturity, the specimen being of very small size and containing no
embryos, while the type, as we have seen, is crowded with them.

R.N. 193 (Station LVIL, outside Dutch Modragam Paar, ll£ to 3G fathoms) ; 204
(Gulf of Manaar).

SPONGES. 97

Paratetilla, n. gen.

Tetillidse with a special layer of modified triaenes, resembling calthrops, lying at the
junction between the ectosome and choanosome (or in the ectosome).

The discovery by Professor Herdman of another species of Tetillid which shares
the remarkable skeletal peculiarities of Carter's Tethya merguiensis (21), appears to
justify the establishment of a new genus for the reception of the two. Paratetilla
merguiensis is, like our species, an inhabitant of the Indian Ocean (Mergui Archi-
pelago), but its area of distribution extends, according to Sollas (15), as far as Torres
Straits. The synonymy of this species is, however, somewhat involved, according to
Lindgren (86) and Thiele (87) it should be known as Tetilla bacca (Selexka), with
which, according to Thiele, Kieschxick's Tetilla ternatensis, T. amboinensis,
T. violacea, and T. rubra are again synonymous. If these identifications be correct
the range of the species must be yet further extended.

Paratetilla cineriformis,* n. sp. — Plate III., fig. 7.

This species is represented in the collection by four specimens, ranging in diameter
from about 12 millims. to about 41 millims. The shape is irregularly hemispherical.
All but the smallest have been injured below by tearing off from the attachment, the
base of the sponge having evidently been left behind. The smallest, which appears
to be fairly perfect, is cushion -shaped, flat beneath and convex above. The surface is
more or less uneven, and to some extent hispid, though not very markedly so ; it
may be very irregular. The consistence is rather soft and spongy. The colour
externally (in spirit) is almost black ; internally much j:>aler, purplish grey. The
arrangement of the vents and pores is very difficult to determine. A few small,
pocket-like depressions, irregularly scattered over the surface, probably represent
cloacal cavities ; their floors are perforated by the numerous minute apertures of
slender, exhalant (?) canals ; their external openings may be narrow and slit-like.
The inhalant pores are probably scattered irregularly over the surface, but I have not
been able to make them out distinctly. Some of the specimens are infested by
parasitic cirripedes, living beneath the surface in cavities whose external apertures
simulate vents.

The main skeleton consists of stout bands of sjjicules radiating from a large central
" nucleus." In addition to these there is the layer of modified triaenes lying at the
junction of ectosome and choanosome.

Spicules. — (1.) Modified triaenes (Plate III., fig. 7, a-d), lying at the junction of
ectosome and choanosome. These spicules are so irregular in size and shape that it
is almost impossible to describe them ; the following have been measured : (a.) with
four simple unbranched rays, one much shorter than the others, resembling a plagio-
trisene with shaft shorter than cladi ; length of longest ray 0-165 millim., with two

* The specific name is derived from the close resemblance which the sponge bears to a black cinder.

O

98 CEYLON PEARL OYSTER REPORT.

others a little shorter, and the fourth much shorter; (b.) with one very short
unbranched ray (shaft) and three much longer rays (cladi), of which two are forked ;
cladi of somewhat unequal length ; maximum diameter of cladome from tip to tip of
rays 0-276 millim. ; maximum diameter of cladi about 0"01G6 millim. ; (c.) similar to
(a.), but with cladi much curved ; the very short shaft, as is commonly the case,
rounded off at the extremity ; (d.) with only three rays, two almost in the same
straight line and the third, much shorter (? = shaft) at right angles to them ; length
of each of longer rays about 0'147 millim. ; of short ray about 0"055 millim. The
rays of these spicules are often irregularly bent or curved. They are probably, as
pointed out by Carter and Sollas in the case of P. merguiensis, modified " zone-
spicules " (plagiotrisenes).

(2.) Protrisenes (Plate III., fig. 7, e, e') ; few in number and very variable in
dimensions, with shaft and cladi all very slender or fairly stout, and cladome some-
times irregular. The cladome may or may not project beyond the surface.

(3.) Anatrisenes (Plate III., fig. 7, f, f')\ with short, stout cladi, and very long
slender shaft, often inflated in a bulbous manner at a short distance below the
cladome and tapering to hair-like dimensions at the other end. Length of shaft in
measured specimen 3 '3 millims. ; diameter between bulb and cladome 0"007 millim. ;
length of cladi 0"027 millim. The cladomes of these spicules sometimes project
beyond the surface of the sponge.

(4.) Oxea of the main skeleton (Plate III., fig. 7, g, (/, g1') ; stout, straight or
nearly so ; fusiform, with the ends gradually and finely pointed, or more or less
irregular; size about 2*9 millims. by 0-03G5 millim., but usually more slender.
Numerous much smaller oxea, apparently young forms, also occur.

(5.) Sigmata (Plate III., fig. 7, h) ; extremely slender, frequently, if not always,
contort, about G"014 millim. from bend to bend ; abundant.

In stained sections of the largest specimen the cortex measures up to about
0"28 millim. in thickness. It is composed of a compact tissue which appears to be
cystenchymatous rather than collenchymatous, and the cells contain numerous minute
brown pigment granules. The choanosome appears to be mostly collenchymatous and
contains brown pigment granules similar to those of the cortex, but fewer in number.
The flagellate chambers are oval or nearly spherical, up to about G"0249 millim. in
diameter.

This interesting sponge is evidently nearly related to Carter's Teihya merguiensis
(21) (Tetilla merguiensis of Sollas) from the Mergui Archipelago, off the coast of
Burmah, but differs to such an extent that there can hardly be any question of
specific identity. The characteristic inhalant pore-sieves and dermal reticulation of
Carter's species are not recognisable here ; the dermal microxea are absent, and the
modified " zone-spicules " appear to be much more irregular in shape.

R.N., 184, 187 (Stat. VI., Muttuvaratu Paar, Gulf of Manaar) ; 214 (deep water
outside pearl banks, Gulf of Manaar) ; 245 (Ceylon seas).

SPONGES. 99

Grade : LITHISTIDA.

Tetraxonida in which the megascleres form desraas, typically united with each other
to form a continuous skeleton.

'Hie relationships and classification of the Lithistida still require a great deal of
investigation before we can consider our knowledge of the group as by any means
satisfactory. They are. of course, usually considered as a sub-division of the
Tetractinellida, but inasmuch as many of them have only monaxonid spicules this
method of disposing of the group seems somewhat unjustifiable. There appear to
lie two logical alternatives. We may either regard the Lithistida as being of
polyphyletic origin and partition them to the best of our ability amongst other
groups, or we may regard them as representing another grade of evolution in a
special direction and retain the group provisionally as a matter of convenience,
without committing ourselves to an opinion as to whether it is a natural group or not.
In the present state of our knowledge the latter appears t<> me the wisest course
to adopt.

As regards the internal classification of the group, I have no doubt that important
modifications of the existing arrangement of genera, which we owe to Sollas, must
shortly be made. My own experience of the Lithistida is, however, so limited that
I prefer not to experiment in this direction, but will content myself with a few
critical remarks suggested by the investigation of certain species in Professor
Herdman's collection.

Discodermia, Bocage.

Lithistida with tetracrepid desmas and ectosomal discotriienes, and with microscleres
in the form of microxea or microstrongyla.

Discodermia emarginata, n. sp. — Plate IV., fig. 4.

The single specimen is of massive form with spreading base attached to a mass of
calcareous worm-tubes. The upper surface is strongly convex, rising into two low,
mammiform projections, from the broad apices of which (where the small vents are
situated) radiate shallow grooves, the outward indications of convergent exhalant
canals. Colour throughout (in spirit) pale yellow; texture compact and stony.
Diameter of massive central portion about 12 millims., with the basal portion
sj Heading a good deal further as a thin crust.

Skeleton composed of a close reticulation of firmly united tetracrepid desmas witli
a single layer of discotriames on the surface. Below the surface are a few slender
oxea, which become more numerous and are sometimes collected into loose fibres in the
neighbourhood of the vent.

Spicules: — (1.) Stout tetracrepid desmas (Plate IV., fig. 4, a-e), more or less
tuberculated. especially at the ends of the branches, which proliferate into numerous
rounded tubercles, by the interlocking of which the union of adjacent desmas is.

o -1

LOO CEYLON PEARL OYSTER REPORT.

effected. There are no transitional forms intermediate between these desmas and the
discotrisenes of the surface, such as Mr. Carter describes (4), and the desmas are
evidently not derived from discotrisenes. Each originates from a tetract crepis
(calthrops) by branching of the actines and secondary deposition of siliceous material.
The initial stages in the development of the desmas are to be found in abundance
just beneath the superficial layer of discotrisenes, where the skeleton is evidently
growing. The young crepis (Plate IV., fig. 4, a) is a regular tetract calthrops with
minutely roughened surface. In one of the youngest I have seen, the actines measure
about 0"04 millim. by O004 millim. One actine may be longer than the other three ;
they are sometimes conical and suddenly constricted at the base. Having attained a
length of about 0*08 millim., or it may be less, the actines swell out at their ends
and begin to branch irregularly (Plate VII., fig. 4, c). Presently the original minute
roughening of the surface is covered over by smooth secondary deposits of silica, and
the adult, much larger tuberculation appears, though the more central portions of the
spicule may often be nearly smooth when fully developed. The diameter of the
actines of the fully formed spicule (" epactines " of Sollas), between the centre and
the point where they commence to branch, is about 0"036 millim. This account of
the development of the desmas, while differing Avidely from that given by Carter,
agrees with the views expressed by Sollas in his " Challenger" report.

(2.) Discotrisenes (Plate IV., fig. 4,f,g); disk flat, about 0"23 millim. in diameter ;
outer surface beset with numerous minute, conical papillae ; margin approximately
circular, but minutely crenulated and usually strongly emarginate at one, two or three
places, so as to form semi-circular bays or indentations ; shaft well developed, stout,
conical, fairly sharp-pointed, up to about 0*147 millim. long. These spicules form a
single layer on the surface, with overlapping disks. The youngest stage I have
recognised in their development is shown in Plate IV., fig. ■!,/.

(3.) Monaxonid spicules (rhabdi) ; probably oxeote ; long and slender, and commonly
tapering very gradually to a fine point. These spicules are so long and slender and
so interwoven with the desmas that I have never succeeded in seeing both ends of
the same spicule in situ, and even in boiled-out preparations they are always broken.
They are particularly numerous in the neighbourhood of the vent, but do not, at any
rate usually, project beyond the surface. I have measured one up to a length of
about 1'2 millims., with a maximum diameter of about 0'006 millim.

(4.) Microrhabds (Plate IV, fig. 4, h) ; usually fusiform and strongylote, with very
minutely roughened surface; sometimes slightly curved; size about 0-016 millim.
by 0-0027 millim. These spicules are very abundant beneath the layer of disco-
trioenes ; and especially in the membrane which immediately surrounds the vent,
where they form a continuous layer, the discotrisenes being absent.

It is with some reluctance that I have been constrained to add another species of
Discodermia to the six already described by Mr. Carter (4, 5) from the Gulf of
Manaar (D. papillata, D. aspera, D. spinispirulifera, D. laividiscus, D. sinuosa, and

SPONGES. 101

D. sceptrettifera), but I cannot identify Professor Herdman's specimen, from deep
water oft" Galle, with any of these. The richness of the Ceylon seas in species of
Discodermia is very remarkable. It is strange that I have not been able to find any
ot Mr. Carter's species in Professor Herdman's collection, but they are apparently
all small forms, which would not be likely to attract the attention of a collector.
P.N. 234 (Station XLL, 12 miles off Galle, 100 fathoms).

Aciculites, Schmidt.

Lithistida with monocrepid desmas and rhabdi, the latter forming a special dermal
layer. Without microscleres.

Sollas, in his report on the " Challenger " Tetractinellida, places this genus in his
family Scleritodermidad in the "Sub-order" Hoplophora, but it seems to me that it
would fall more naturally amongst the Azoricidte (in the Sub-order Anoplia of
Sollas). The fact that some of the rhabdi form a dermal layer can hardly be of
sufficient importance to justify us in placing the genus not only in a different family,
but even in a different "sub-order" from the Azoricidae, to which it is naturally
allied by the form of the desma, the presence of rhabdi, and the absence of
microscleres.

Aciculites orientalis, n. sp. — Plate IV., fig. 3.

The single specimen is massive, compact, cushion-shaped, rather irregular, but with
rounded outlines, somewhat flattened above and attached below by a broad base to a
mass of calcareous debris. Height about 28 millims. ; greatest breadth, 38 millims.
Texture compact and stony, but with a comparatively soft dermal membrane, which,
on the upper surface of the sponge, is easily separable from the underlying part.
Vents numerous, minute (say about 0-2 millim. in diameter, but varying), scattered
over the upper surface.* Pores scattered. Surface, in parts at any rate, slightly
corrugated, with narrow, ramifying and meandering canals showing through the
dermal membrane. Colour in spirit grey, both internally and externally.

The main skeleton is a very close and compact reticulation of monocrepid desmas,
intermingled more or less abundantly with strongyla. In the dermal membrane the
strongyla are very abundant and lie for the most part tangentially, forming a more or
less continuous dermal skeleton. Beneath the dermal membrane, where the main
skeleton is growing, the desmas are found in various stages of development, not as yet
connected with one another.

Spicules. — (1.) Monocrepid desmas (Plate IV., fig. 3, a-d). The young spicule (a)
is an irregular, elongated rod, with minutely roughened surface, which soon begins to
branch. Tbe fully grown spicule (d) usually consists of a strongly curved main axis
with branches coming off chiefly on the convex side. These branches proliferate and

* One much larger opening looks as if it might be artificial.

102 CEYLON PEARL OYSTER REPORT.

cud in rather sharp, conical papillae. Similar papillae may occur on the main axis,
hut when this is strongly curved they are usually absent from the concave side. The
fully developed desma measures about 0*328 millim. in maximum length in a straight
line from point to point.

(2.) Strongyla (Plate IV.. fig. 3, <\/) ; slightly curved and a good deal broader at
one end than at the other. The broad end. which may he sliffhtlv tvlote, is covered
with minute spines, while the narrow end is smooth, or nearly so. though often with
a very few minute projections. Size, commonly about 0*328 millim. by 0'00'J millim.
(in the middle).

So far as 1 am aware, the only other species of Acicvlites hitherto described is the
type of the genus, Aciculites higginsii, Schmidt (44), from Havanna. ' >ur species is
evidently very closely related to the West Indian form, which has fortunately been
re-examined and described by Sollas (15, p. 347). In fact, the only specific
difference which the description of the latter has enabled me to detect lies in the fact
that in the ( leylon species the vents are not protected by tent-like arrangements of
radiating rhabdi as described by SoLLAS. Probably, however, other specific differences
will be found to exist in the form of the desmas, &c.

We have here an interesting case of apparently discontinuous generic distribution,
though the imperfect state of our knowledge of the sponge-fauna of intermediate
localities makes it possible that the discontinuity is apparent rather than real.

It may be noted that Tui'sent's Aciculites incrustans has now been recognised by
that author as belonging to a totally distinct genus, Desmanthus (14).

R.N. 150 (Ceylon seas).

Taprobane.:; n. gen.

Lithistida of plate-like or cup-shaped form, with minute sphinctrate apertures
abundantly scattered on each side of the plate ; with monocrepid, tuberculate
desmas and long, slender oxea ; without special ectosomal spicules and with
niicroscleres in the form of sigmata only.

We have here another proof of the artificial character of Sollas's classification of
the Lithistida. It will he seen from the diagnosis that the sponge upon which the
genus Taprobane is based might be regarded either as a Scleritodermid without
ectosomal spicules or as an Azoricid with sigmata : it is thus very closely related on
the one hand to Scleritoderma, and on the other to Azorica. We have already had
occasion to notice, in speaking of the genus Aciculites, that Sollas places
CScleritbderma and Azorica in different " sub-orders," his Hoplophora and Anoplia
respectively, but the discovery of Taprobane, combining characters of these two
groups, viz., the absence of special ectosomal spicules and the presence of microscleres,
alone seems sufficient to necessitate a revision of Sollas's scheme.

* So called from tin.' wl<l Greek name tor the Island of Ceylon.

SPONGES. 103

Taprobane herdmani, n. sp. Plate I., fig. 8 ; Plate [V., fig. 2.

Specimen (Plate L, tig. 8) consisting of a stout, erect plate or lamella, strongly
curved (almost into the shape of a cup). Surfaces more or less uneven, especially on
the outside (which bears numerous galls due to parasitic barnacles). The upper edge
is broadly rounded, and the base of attachment is slightly expanded, like the foot of
a cup. Both surfaces are slightly and unevenly hispid, the hispidation being almost
confined to the more depressed portions. Both surfaces are thickly studded with
minute, pustule-like, circular areas, hardly visible to the naked eye. These areas are
only about 0"3 millim. in diameter, and each hears the single aperture of an incurrenl
or excurrent canal surrounded by a well-developed sphincter membrane. The colour
(in spirit) is dirty grey : the texture compact and stony. Greatest height about
70 millims. ; thickness of lamella about 15 millims.

The main skeleton is a very dense and compact reticulation of desmas, together
with loose wisps or brushes of long, slender oxea running at light angles to the
surface, beyond which their ends project so as to give rise to the hispidation. The
oxea seem to disappear more or less completely from the deeper parts of the sponge,
while around the apertures of the canal svstem, at any rate on the outer surface,
they may project as a scanty fringe.

Spicules. — (1.) Desmas (Plate IV., fig. 2); monocrepid and rather sparingly
branched, the branches usually coming off almost exclusively from one side of the
main axis, which is usually more or less strongly curved. Main axis and branches
more or less abundantly ornamented with rounded tubercles ; branches (? always
terminating in conical papilla?, which may be bifid at the apex. The union of these
desmas is so close and compact that it seems almost impossible to isolate an adult
spicule for measurement : but the total length, measured in a straight line from point
to point, appears to be about 0"36 millim. The thickness of the main axis, exclusive
of tubercles, is about 0"028 millim.

(2.) Oxea; very long and slender, gradually and sharply pointed at each end;
commonly more or less curved or crooked; size variable; they are usually broken,
but two complete spicules measured about TO millim. by 0"008 millim. and l-88 millim.
by 0*008 millim. respectively.

(3.) Sigmata ; slender, contort ; total length in a straight line from bend to bend,
about 0 01 millim. : abundant throughout the sponge.

One of the most characteristic features of this species is the arrangement of the
apertures of the canal system. These appear to be identical in form and arrangement
on the two sides of the sponge, but we may conclude from the analogy of other plate-
like and cup-shaped sponges that the apertures on the outer side are inhalant pores,
and those on the inner side vents. Each aperture, whether pore or vent, lies in the
middle of a circular area, sharply defined by the sudden cessation of the reticulation
of desmas at its margin. This area is occupied by a very well developed, iris-like
diaphragm, with abundant circularly and perhaps also radially arranged myocytes,

104 CEYLON PEARL OYSTER REPORT.

the contraction of the former of which lias, in almost all cases, completely closed
the aperture. The diaphragm contains sigmata, but no other spicules. A similar
arrangement of the external apertures of the canal-system appears to be characteristic
both of Azorica and Scleritoderma ; indeed, our Ceylon species, except for the absence
of microstrongyles, resembles very closely indeed Schmidt's Scleritoderma packardi
from the Gulf of Mexico, as described and figured by Sollas in his work on the
" Challenger " Tetractinellida.

K.N. 40 (Stat. XV., Periya Paar, Gulf of Manaar, 9 fathoms).

Petromica, Topsent.
Lithistida of massive form ; with scattered pores and vents ; with thin dermal
membrane destitute of special skeleton ; with monocrepid desmas feebly united
or quite separate ; with monaxonid rhabdi often collected in fibres which may
terminate in surface conuli ; without microscleres.

This genus was founded by Topsent in 1898 (45) for a new Lithistid sponge from the
Azores, very closely related to Azorica. Topsent gives the following diagnosis : —
" Azoricidse massives, en forme de cones dresses, a surface conuleuse, a pores disperses,
a oscules membraneux, a ectosome devellope aspiculeux, a desmas peu ornes et
faiblement relies entre eux."

The discovery of a second species in Ceylon waters, closely related to the type of
the genus, forms a further justification for separating Petromica from the other
Azoricidee.

Petromica inassalis, n. sp. — Plate IV., fig. 5.

Sponge massive, may be attached by a broad base, may be compressed vertically
or horizontally. Vents usually numerous, rather small, but variable in size ;
scattered on the upper part of the sponge, each forming the termination of a
vertical oscular tube. Pores scattered. Surface uneven, variable, more or less
corrugated ; in places covered with a thin, reticulate, pore-bearing dermal membrane
overlying large sub-dermal cavities and supported on bundles of large monaxonid
spicules which sometimes terminate in conuli. Texture incompressible, but friable ;
colour (in spirit) yellowish grey throughout. The type specimen (Pt.N. 257) measures
about 37 millims. in height, the same in breadth, and 24 millims. in thickness.

The main skeleton is a reticulation (sometimes close and sometimes so loose that
the desmas do not touch one another) of much -branched monocrepid desmas, inter-
mingled with numerous large monaxonid spicules which are partly collected together
in coarse fibres (without being united together by any cementing substance).
These fibres, as already noticed, sometimes terminate in surface conuli. There is
no special dermal skeleton.

Spicules. — (1.) Monocrepid desmas (Plate IV., fig. 5, a) ; much and very irregularly
branched ; not tuberculate (or very slightly so) ; branches usually terminating in
short, blunt, conical points, or in curved flattened expansions pressed against other

SPONGES. 105

desmas. Maximum length from apex to apex of branches about 074 millim., with
main axis about 0-06 millim. thick.

(2.) Stout and slender monaxonid spicules (Plate IV., fig. 5, b-c) ; more or less
curved and variously ended, ranging from oxeote to strongylote and stylote ; size
about 0-98 millim. by 0*0369 millim., 1*2 millims. by 0'0246 millim., 1-28 millims. by
0"0328 millim., &c. ; very variable.

The most interesting feature of this sponge is the strong development of the
monaxonid spicules (rhabdi), which, in three out of the four specimens (the fourth
being dead and infested by another sponge\ play an equal part with the desmas in
the composition of the skeleton. This condition of the skeleton suggests that
certain Monaxonellid sponges usually associated with the Axinellidse may be derived
from Lithistid ancestors by the suppression of the desmas, or that certain Lithistids
with monocrepid desmas may possibly be derived from Monaxonellid ancestors.

Petromica massalis is evidently very closely related to Topsent's P. grimaldii, the
type of the genus, from the Azores. The principal difference appears to concern
the ornamentation of the desmas, which, in P. grimaldii, are ornamented at the
extremities with little conical tubercles, not pointed. These are absent or very feebly
developed in the Ceylon species, but subsequent researches may make it desirable to
unite the two.

R.N. 198 ; 216 (dead, and infested by another sponge ; deep water outside pearl
banks, Gulf of Manaar) ; 257 (type) ; 269 (young ; deep water oft' Galle and onwards
up West Coast of Ceylon).

Grade : MONAXONELLIDA.

Tetraxonida in which the primitive tetraxonid and tetractinellid coDdition of the
megascleres has been entirely lost and none but monaxonellid megascleres
remain. No desmas are developed.

With the recognition of the true nature of the monaxonellid sponges as reduced
Tetraxonida, it becomes once more desirable to modify the name of the group. The
old name, " Monactinellidre," which we owe to Zittel, was objectionable because
implying that the spicules are necessarily monactinal (one-rayed), while, as a matter
of fact, they may be also diactinal (two-rayed). Sollas therefore altered the name
to " Monaxonidse," which was afterwards altered to " Monaxonida," under which
name the group appears in the ' " Challenger" Reports' and other recent works.

The name " Monaxonida," however, becomes distinctly objectionable when used for
a mere sub-division of the order Tetraxonida, contrasted with the Tetractinellida ;
and I therefore propose to again modify the name of the group, choosing this time
the term " Monaxonellida," which is, I believe, both correct in meaning and in
harmony with the scheme of classification adopted.

The Monaxonellida form, however, an unnatural group of polyphyletic origin,
which we can only retain as a matter of convenience until we know more about their

p

106 CEYLON PEARL OYSTER REPORT.

phylogeny. They are evidently descended from more than one group of tetrac-
tinellid ancestors hy degeneration of the tetraxon megascleres, and have branched off
into an immense number of genera and species, the classification of which, as in the
case of all reduced forms, is extremely difficult.

We have seen that the two chief tetractinellid sub-orders are distinguished mainly
by the form of the microscleres, the " Astrophora " possessing some form of aster and
the " Sigmatophora " sigmata. This same distinction, broadly speaking, runs through
the Monaxonellida also, and may be used as the basis of their sub-division into
" Astromonaxonellida " and " Sigmatomonaxonellida," names which I now propose as
the most appropriate for the two great groups into which the " Monaxonellida " are
by general consent divided, and approximately equivalent to the "Clavulina" and
" Halichondrina " of Vosmaer and of Ridley and Dendy, and to the " Hadromerina "
and "Halichondrina" of Topsent, &c.

These two groups are sharply distinguished from one another. As far as I am
aware, there are only three cases on record of the occurrence of astrose microscleres
in association with sigmata or chelce, and all three are probably to be explained
as due to mixture of the spiculation of two distinct sponges — a very frequent
occurrence.*

We may regard these two great monaxonellid sub-orders as being descended from
the two corresponding tetractinellid sub-orders, though this is no doubt but a crude
way of looking at the problem, and it may well be that some of the forms which we
find most difficult to classify are descended directly from the Homosclerophora, and
others from the Lithistida,

The great difficulty in following out this system of classification to its logical
conclusions lies in the fact that in many cases the microscleres, as well as the tetract
megascleres, have entirely disappeared, apparently by degeneration, and we are then
dependent upon the much less trustworthy guidance of other characters. Thus the
genera Halichondria, Reniera, Axinella, Sube rites, &c, with a very simple spicu-
lation composed exclusively of monaxonid megascleres, are in reality more modified
forms than Gellius, Sigmaxinella, Ilymedesmia, Spirastrella, and so on, which have
not lost their microscleres.

Sub-Order: ASTROMONAXONELLIDA.

Monaxonellida in which the microsclere, when present, is some form of aster.

This sub-order is practically equivalent to the Hadromerina of Topsent, which

* The cases in question are Schmidt's Sceptrella regalis (23), p. 58 [ride also Ridley and Dendy (1),
p. lxii., footnote] ; Fristkdt's DesmaceMa peachii var. stellifera (24) ; and Topsent's Hymeraphia Umreti (25) ;
ii.ll very doubtful cases. The so-called amphiasters (or " birotulates ") of certain Desmacidonidse (e.g.,
Axoniderma, Iotrochota, Amphiastrella) are not true asters at all, but merely modified isochela?. [Further
discussion on this subject, necessitated by the appearance of Toi'.sknt'.s great work on the Sponges of the
Azores (62), will be found in the general remarks on the sub-order Sigmatomonaxonellida.]

SPONGES. 107

that author sub-divides into two sections, according to the prevailing form of
megasclere, viz., " (Jlavulida " and " Aciculida." This sub-division, however, appears
to me to be both unnecessary and unnatural, especially when its originator places the
genus Tethya, with its stylote megascleres, in the division " Aciculida," which is
characterized by oxeote megascleres.'" I therefore propose to divide the sub-order
immediately into families. Of these TorsENT (26) makes nine, of which five are
represented in the present collection.

In addition to these, I here include the family Ghondrosidse, which have completely
lost their megascleres, and, in the genus Choadrosia itself, their microscleres also.
Some of the old group Axinellidaa (e.g., Vibulinus, with astrose microscleres) must
likewise be included in this sub-order, and it may prove necessary to institute a new
family — Astraxinellidaa — for their reception.

On the other hand, I am inclined to think that some of Topsent's families will
prove to be superfluous. Thus I propose to abandon his " Coppatiidse " and
" Streptasteridse " in favour of the older " Epipolasidse " of Sollas.

Family: EPIPOLASIDSE.

Astromonaxonellida with oxeote megascleres, and usually euasters for microscleres.

This family, founded by Sollas (15) for the reception of the genera Amphius,
Coppatias, and Asteropus, is, at any rate as regards what we may consider to be its
typical representatives, evidently very closely related to the Stellettidaa, a fact which
has already been recognised by previous writers. In short, we may safely regard
these forms as reduced Stellettids which have lost their tetraxon megascleres, and
Sollas himself placed the family next to the Stellettidse as an appendix to the
Euastrosa, though apparently with some doubt.

Topsent divided the genera of Sollas's Epipolasidas between his own families
Coppatiidse and Streptasteridse. In the Coppathdae he also includes certain genera,
such as Spongosorites, which have no microscleres at all. Spongosorites is well
represented in the Sponge-Fauna of Ceylon, but, for reasons which will appear
subsequently, I find it desirable to remove it to the Axinellidse.

Coppatias, Sollas (15).

Epipolasidse with the skeleton composed of an irregular interlacement of oxea, and
microscleres in the form of euasters only.

This genus is almost synonymous with the genus Stelletthiopsis as employed by
Carter.

* It should be pointed out, however, that Sollas and ToPSENT regard the megascleres of Tethya as
modified oxea, which they term strongyloxea, but it is impossible to distinguish these from styli.

P 2

108 CEYLON PEARL OYSTER REPORT.

Coppatias reptans, n. sp. — -Plate V., fig. 2.

The single specimen is much elongated, irregularly sub-cylindrical, creeping over
and to a slight extent encrusted by calcareous debris. It has apparently lain
horizontally. The total length is about 11 centims., and the maximum thickness
about 14 millims. One end is much narrower and forms a free digitiform process,
strongly curved. The surface is slightly rugose, minutely conulose and minutely
and slightly hispid. The colour (in spirit) is dark purplish-grey, paler below and
internally. Vents small (mostly minute), numerous, scattered on the upper surface.
Inhalant pores abundantly scattered.

The main skeleton is a very confused reticulation of oxeote megascleres, here and
there collected into loose bundles. There is no special dermal layer of tangentially
disposed oxea, but the surface is rendered more or less hispid by the projecting points
of oxea which lie below it. There is, however, a thin dermal layer of densely crowded
asters.

Spicules. — (I.) Oxea (Plate V., fig. 2, a, b, c) slightly curved and gradually and
sharply pointed; varying a good deal in diameter; about TO millim. by 0'022 millim.
when fully grown.

(2.) Chiasters (Plate V., fig. 2, (/) ; very minute, with rather numerous, slender,
cylindrical rays and little or no distinct centrum ; total diameter about O'OOG millim.
(sometimes a little more) ; most abundant at the surface, but also plentiful in the
choanosome.

The ectosome is fairly thick, but is excavated by numerous irregular, spacious
sub-dermal cavities, into which the inhalant pores open and from which the inhalant
canals of the choanosome take their origin. It is clearly differentiated into two
layers : an inner, comparatively thin and densely fibrous layer, with the fibres mostly
lying parallel to the surface, and an outer, much thicker layer composed of collenchyma
with a considerable admixture of fibrous tissue, but with the fibres running irregularly
in all directions.

It is a noteworthy fact that the fibrous cells of the ectosome contain a large
proportion of the pigment to which the sponge owes its dark colour, arranged in them
in the form of minute spherical granules. Similar pigment granules also occur in
some of the ordinary stellate cells of the collenchyma, of which the fibre-cells are but
a slight modification.

The. outer part of the ectosome also contains immense numbers of large, spherical,
darkly staining cells, crowded together in large groups or loosely scattered. Similar
cells also occur very abundantly, scattered singly or grouped in dense masses, in the
choanosome. They remind one strongly of the symbiotic Algaj of Hexadella and the
corresponding cells of Asteropus haeckeli, and are probably of a similar nature. Their
immense numbers and their occurrence in such dense masses in both ectosome and
choanosome are alone enough to suggest that they are not true constituents of the
sponge-tissues.

SPONGES. 109

This species appears to be very closely related to Carter's Stellettinopsis (Coppatias)
iuberculata (18), an Australian species from which the Ceylon sponge differs chiefly
in its external form and reptant habit.

K.N. 242 (Stat. XXVII., Cod Bay, Tiincomalee, 5 fathoms).

Asteropus, SoLLAS (15).
Epipolasidse with two sorts of asters, oxyasters and sanidasters.

Asteropus haeckeli, n. sp. — Plate V., fig. 3.

The single specimen is a small, irregular crust which has probably been attached by
the base, to which fragments of calcareous debris are still adherent. The upper
surface rises up into a broad, low, mammiform projection, with a single, rather large
vent (about 2"5 millims. in diameter) at its apex. Surface smooth, but uneven and
harsh to the touch, owing to the presence of huge oxea lying tangentially beneath
the surface or projecting slightly beyond it. Colour (in spirit) pale grey. Texture
internally coarse and cavernous, owing to the large exhalant canals.* Inhalant pores
scattered. The specimen may be only a fragment of a much larger crust ; its
maximum breadth is about 25 millims., and its greatest height (where the vent is
situated) about 1 1 millims.

The skeleton is a very dense and very confused interlacement of huge oxea,
sometimes collected into very loose, ill-defined, coarse strands, which run towards
the surface.

Spicules. — (1.) Oxea (Plate V., fig. 3, a) ; usually large and stout, fusiform,
gradually and fairly sharply pointed at each end ; curved : varying much in size, up
to about 1*9 millims. by 0"065 millim. Numerous much shorter and more slender
forms also occur, which may be immature.

(2.) Sanidasters (Plate V., fig. 3, c, d); with straight, slender axis dividing into
two spines at each end, and with two irregular whorls of spines dividing the total
length into three approximately equal parts ; or with the axis irregularly spined, and
perhaps angulated, so that the whole closely resembles a Spirastrella microsclere ;
total length up to about O'OIG millim. These spicules are very abundant at the
surface and also common in the interior.

(3.) Oxvasters (Plate V., fig. 3, b) ; with few, slender, very slightly spined or
roughened (? sometimes smooth), sharp-pointed rays and no distinct centrum ; total
diameter measured up to about 0"04 millim. These spicules are found in the interior
of the sponge, but are scarce and easily overlooked.

The ectosome is very thick, and composed of a mixture of cystenchymatous and
collenchymatous tissue, slightly fibrous in places, and with numerous darkly staining,
oval cells scattered through it, which somewhat resemble the symbiotic Algas of
Hexadella, and are probably of a similar nature.

* The oscular tube is occupied by a Polychsete worm.

110 CEYLON PEARL OYSTER REPORT.

This species is closely related to Carter's " SieHcttiiwpsis simplex," from Australia
(Fremantle and Port Phillip Heads),* for which Sollas (15) established the genus
Asteropus, and which, with the doubtful exception of Schmidt's Stellcttinopsis
annulata, has, up to the present time, remained the only described species of the
genus. (Topsent has shown (28) that Lendenfeld's Asterojms incrustans is not
referable to the genus Asteropus at all.)

The occurrence of another species in Ceylon waters is extremely interesting and
affords a good illustration of the close relationship which exists between the Sponge-
Fauna of this region and that of Australia. That the two species are not identical
I have been able to convince myself by personal examination of a fragment of an
Australian specimen kindly forwarded to me some years ago by the authorities of the
British Museum (Natural History). The differences concern chiefly the arrangement
of the vents, the colour (tawny-brown in the Australian species), and the size of the
oxea (which are considerably more robust in the Ceylon species). Still, it is possible
that the future discovery of intermediate forms may justify us in uniting the two.

I have taken the liberty of naming this species in honour of the most distinguished
author of ' Die Kalkschwiimme.'

R.N. 219 (deep water outside pearl banks, Gulf of Manaar).

Cryptetethya, n. gen.

Epipolasidse in which the ectosome is differentiated into a thin, inner fibrous layer,
very dense, and a thick, outer, more or less gelatinous layer. The outer layer is
produced into more or less elongated, finger-like projections, between the bases
of which the more or less spherical body of the sponge is largely concealed by
agglutinated foreign bodies. The main skeleton is radially arranged, consisting
of large oxea extending through choanosome and ectosome alike. The microscleres
are euasters.

As regards its spiculation, this genus agrees very closely with Coppalias, and, like
the latter, is evidently very nearly related to Stelletta. In the structure of the
ectosome it makes a close approach to Stelletta herdmani, in which also the triasnes
have undergone considerable reduction. In fact, Cryptoteihya may be regarded as
derived from some such form as Stelletta herd man i by further reduction of the triaenes
and by the outgrowth of the ectosome into finger-like processes. In the spherical
form of the body, and the radial arrangement of the skeleton, it is also related to
Tethya, but perhaps its nearest ally is Sollas's genus Magog, established (15) for
the reception of Carter's Chondrilla saeciformis, from Mauritius, and included by
Sollas amongst the Tethyidse. In Cryptotethya, however, the oxea are not confined
to the choanosome as in Magog. Both genera appear to me to come most naturally
amongst the Epipolasidse, though they certainly seem to indicate the manner in

* Mr. Carter (27) also records the species from Hayti, but the Haytian form is very likely specifically
distinct from the Australian.

SPONGES. Ill

which the Tethyidse may have originated, through the Epipolasidse, from stellettid
ancestors.

Cryptotethya agglutinans, n. sp. — Plate V., figs. 4, 5.

The single specimen (Plate V., fig. 4) consists of a more or less spherical body from
which radiate irregular finger-like processes of varying shape, long or short,
cylindrical or flattened, and sometimes expanded at the free end. Between these
projections the surface of the sponge is for the most part concealed by a great quantity
of calcareous ddbris which firmly adheres to the sponge, including melobesiau nodules
of considerable size, worm-tubes, Foraminifera, &c, the whole forming an irregular
mass in the midst of which the body of the sponge is scarcely recognisable. Calcareous
debris may also be found in the interior of the sponge, even within the choanosome.
I have not succeeded in making out the arrangement of the vents and pores. Some
small openings on the ends of some of the projections resemble vents, but on close
examination are found to be merely the apertures of cavities inhabited by parasitic
barnacles. One at least of the shorter projections (fig. 4, x), however, contains
longitudinal canals which are evidently either inhalant or exhalant canals proper to
the sponge. The surface of the sponge, where exposed, is very uneven and very
harsh to the touch, owing to the projection of the large oxea, which readily break off'
in one's skin and thus make the sponge very unpleasant to handle. The colour of
the surface and of the thick outer layer of the ectosome (in spirit) is grey, of the thin
fibrous layer of the ectosome white, and of the choanosome nearly white. The
maximum diameter of the body of the sponge is about 40 millims., the length of the
longest projection about 25 millims.

In the body of the sponge the huge oxea of which the skeleton is composed are for
the most part arranged radially and without any distinction between choanosome and
ectosome, passing indifferently from one to the other through the dense fibrous layer.
Some of the oxea are associated in loose bands which spread out in brushes as they
approach the surface. In the projections the oxea naturally lie for the most part
longitudinally, but at the expanded end of a broad projection (containing longitudinal
canals) I have found a good many placed tangentially at or near the surface. The
asters are for the most part arranged (very abundantly) in a thin dermal layer,
through which the points of the oxea may project for a short distance.

S}ncules. — (1.) Oxea (Plate V., fig. 5, a, b) ; stout, fusiform, usually gradually and
fairly sharply pointed, but subject to some irregularity at the ends and occasionally
stylote. Size very variable, say about 2 5 millims. by 0'073 millim. when fully
grown, but often less.

(2.) (Jhiasters (Plate V., fig. 5, c), with little or no centrum and smooth, slender,
sub-cylindrical rays, ending bluntly, but not tylote ; total diameter commonly about
'CT012 millim. These spicules are most abundant at or near the surface of the sponge,
but a few precisely similar forms occur in the choanosome.

112 CEYLON PEARL OYSTER REPORT.

Owing, on the one hand, to the enormous quantity of foreign matter adhering to
the surface, and, on the other, to the great size of the megascleres, which interfere
greatly with the cutting of thin sections, the investigation of the canal system of this
sponge is attended with exceptional difficulties, and I have come to no satisfactory
conclusions on the subject.

One of the most striking features of the sponge is the dense fibrous layer of the
ectosome, which, when the sponge is cut in half, is conspicuous even to the naked eye
as a white layer about 0"4 millim. thick, dividing the body into inner and outer
portions, and forcibly calling to mind the similar layer of fibrous tissue in Stelletta
herdmani. This dense fibrous layer consists of bundles of fine fibres closely matted
together and running in all directions. It does not, perhaps, form quite the
innermost portion of the ectosome, for beneath it lies a thin gelatinous layer containing
subcortical crypts, from which the inhalant canals of the choanosome probably take
their origin.

The outer layer of the ectosome varies much in thickness. It is partly
collenchymatous, consisting of a clear gelatinous matrix with an immense number of
large granular stellate cells embedded in it, and partly fibrous, the fibrous condition
being apparently arrived at by elongation of similar cells in a direction parallel to the
surface. Roughly speaking, this fibrous tissue may be said to occur between two
layers of the collenchyma, but the two kinds of tissue are not sharply differentiated
from one another, and the fibrous layer is not nearly so well defined or so dense as
the inner fibrous layer already described. The outer layer of the ectosome alone
takes part in the formation of the finger-like projections on the surface of the sponge,
the inner fibrous layer not being continued into these.

The choanosome is rather compact and finely granular, but, owing perhaps to want
of penetration by the preserving medium, my sections do not enable me to make out
details of the histology or the arrangement of the flagellate chambers.

RN. 62 (Gulf of Manaar).

Family : TETHYID^.

Astromonaxonellida with stylote megascleres and euasters for microscleres ; with
strongly developed fibrous cortex and radially arranged skeleton.
I have discussed the probable origin of this family in speaking of the genus
Cryptotcthya.

Tethya, Lamarck.

More or less spherical Tethyidas, without highly specialised pore-bearing grooves and
without a sand-layer in the choanosome.

Tethya lyncurium, Lin.

There are in the collection a number of specimens of Tethya which have given me a
great deal of trouble as regards their collect nomenclature. 1 have finally decided to

SPONGES. 1 |:i

regard them as belonging to three varieties of the well-known and extremely variable
Tethya lyncurium of European waters. For convenience of reference we may dis-
tinguish the Ceylon varieties as a, b, and c respectively. .

Tethya lyncurium, Lin., var. a.

This variety is represented by two approximately spherical specimens growing side
by side on a mass of calcareous and other debris. The larger of the two is about
20 millims. in diameter, the other only a little less. The colour in spirit is dull
yellowish-grey. The surface is irregularly conulose and gemmiparous, not distinctly
tessellated, and to a considerable extent covered by adherent foreign matter. Each
has a single prominent and widely open vent, about 2 millims. in diameter, at the
summit of a thin-walled tubular projection. The pore-sieves between the conuli are
for the most part inconspicuous.

The cortex is very dense and more or less fibrous throughout, but the fibrous tissue
is most strongly developed in its deeper portion. The inhalant canals in the cortex-
are lacunar near the surface and constricted into definite canals deeper down.

The main skeleton consists of stout radiating bundles of megascleres, breaking up
into divergent brushes in the cortex. In the choanosome (but not in the cortex)
loose spicules of similar form are abundantly scattered (mostly lengthwise) between
the bundles.

The megascleres are rather slender, faintly tylote styli, of the ordinary Tethya
form. The feebly developed bead is narrower than the middle of the shaft, and the
apex is gradually and more or less sharply pointed. These spicules measure about
t-:l millims. by 0"02 millim. in the main fibres, but are much smaller in the surface
brushes, while between the main fibres, in the choanosome, the sizes are mixed.

The microscleres are of two forms only, spherasters and chiasters. The spherasters
are mostly found in the cortex, where they are rather sparingly scattered. The
centrum is fairly large, and they have sharp-pointed conical rays which may be
(rarely) spined or branched. The rays are about 12 in number, and nearly or quite
touch one another at their bases. Total diameter about 0"076 millim., with rays
0'02 millim. lono-. In the cortex the chiasters are most abundant at the surface and
in the walls of the inhalant canals ; they are also numerous in the choanosome. They
have no centrum and about from 6 to 9 distinctly tylote, rather slender rays. The
total diameter of the chiaster is about 0-012 millim. Those of the choanosome are
commonly six- rayed, and the rays are occasionally more elongated and i^oportionately
more slender than in the ordinary form, but they are almost always distinctly tylote,
and the whole spicule is never more than about 0-02 millim. in diameter.

I am convinced that the two specimens described above cannot be distinguished
more than varietally from the common European species. The chiasters are, it is
true, more distinctly tylote than is usually the case in T. lyncurium, but I cannot
regard this character as of specific value, for in a specimen from Budleigh Salterton,

I 1 4 CEYLON PEARL OYSTER REPORT.

in Mr. Carter's collection (now in my possession), the rays of the chiasters are
occasionally tylote, though not nearly so strongly as in the Ceylon specimens.*
R.N. 199, 200 (Stat, LXVIL, off Talaivillu Paar, 10 to 14 fath., Gulf of Manaar).

Tethya lyncurium, Lin., var. h.

There are five specimens in the collection which may be referred to this variety,
differing from the foregoing in the following respects : —

(1) There are no prominent vents.

(2) The surface is more or less distinctly tessellated.

(3) The spherasters are considerably larger and more numerous, and have more
numerous rays and larger centra. Total diameter up to about O'l millim., with rays
about 0'024 millim. long; sometimes they are a little smaller; usually they are
densely crowded in the cortex ; the number of rays appears to be usually about 25.

(4) The chiasters are more distinctly differentiated into two kinds, the rays of
those of the choanosome showing a strong tendency to lose their tylote character and
become strongylote or even oxeote, while at the same time elongating somewhat (but
not much) and being (? always) slightly roughened.

The largest of the five specimens is only about 21 millims. in diameter. The colour
in spirit is dull grey.

R.N. 180, 180a, 180b (all from Lagoon, Galle) ; 196 ; 211 (Gulf of Manaar).

Tethya lyncurium, Lin., var. c.

This variety is represented by a single specimen, about 24 millims. in maximum
diameter, with very strongly tessellated surface and one prominent vent. The cortex
is very lacunar between the polygonal, flattened tesserae.

As regards spiculation, this variety differs from the preceding only in the presence
of numerous well-developed oxyasters, chiefly in the choanosome. These spicules
have little or no centrum and usually six rays. The rays are rather slender and
sharp-pointed, sometimes slightly roughened, but very rarely spined or branched ;
usually straight or nearly so. Total diameter of the oxyaster about 0'04 millim.

This variety might be justifiably identified either with Selenka's Tethya maza or
with Percival Wright's T. seychellensis, but the two preceding varieties unite it so
closely with T. lyncurium that it seems to me quite unnecessary to make a specific
distinction.

R.N. 19 (Gulf of Manaar).

The occurrence of these three varieties of Tethya lyitcurium within the same limited
area is a very interesting fact, and their discovery is likely to be of great assistance
in elucidating the relationship of the so-called species of this extremely difficult genus.
In this connection it is interesting to note that Wilson (29) has lately recorded the
occurrence of both T. lyncurium and T. seychellensis in Porto Rico.

' Topsent, however, entertains (14) a different view ;is to the taxonomic value of the tylote character,

SPONGES, 115

Xenospongia, Gray.

Tetnyidse of discoidal form. Choauosome containing much Band in its deeper parts.
Inhalant pores localized in well-defined grooves, partly concentric and partly
radial in arrangement. Vents on slight prominences. Surface tuherculate.
Skeleton composed of bundles of styli ending in brushes, which project from the
surface tubercles and also form a slight marginal fringe. Microscleres euasters
of various forms and sizes,* forming a dense cortical layer, and also abundantly
scattered in the choauosome.

The genus Xenospongia was established by Gray (30) as far back as 1858 for the
reception of two remarkable sponges from Torres Straits, of which he gave a very
poor description accompanied by excellent figures of the external characters only.
He made no attempt to describe the spiculation, and, although this defect was
partially remedied by Carter in 1882, our knowledge of the sponge has remained
singularly incomplete up to the present day. The genus has only been recorded
once since its original discovery, viz., by Holdsworth (9) from the Ceylon pearl
banks in 1873. No description, however, has ever been published of Holdsworth's
specimen, and the identification with Gray's species, though doubtless correct, was
quite unsupported by evidence. Holdsworth's record and specimen (which appears
to have been forwarded to Gray for further investigation) both seem to have been
lost sight of by subsequent writers.

Under the circumstances, it is not remarkable that the systematic position of
Xenospongia has hitherto remained doubtful. Gray himself, in 1868 (31), made it
the type of a new family, for which he proposed the name " Xenospongiadas."
Carter, in 1875 (32), referred it to his group " Donatina," along with Tetliya
lyncurium, but expressed the opinion that it might ultimately have to come amongst
the " Suberitida." In 1882, however, this author (33) reverted to Gray's opinion,
and proposed " a group named Xenospongina = Xenospongiadae, Gray." Sollas, in
1888 (15), referred the genus doubtfully to the Tethyidse, pending further informa-
tion. Topsent, in 1898 (26), made a new departure by referring it to the
Spirastrellida?, though, as it seems to me, with very slight justification.

The minute investigation which I have been able to make of Professor Herdman's
well-preserved specimen proves beyond doubt that the opinion of Sollas is correct,
and that Xenospongia is very closely related to Tetliya itself, alike in spiculation,
skeleton arrangement, histology, and canal system, although the external form and
the arrangement of the inhalant pores are very peculiar. Even the surface tubercles
or conuli, and the somewhat pinkish colour in spirit, remind one strongly of the genus
Tetliya. In Tethya also one sometimes sees the beginning of the development of
pore-bearing grooves between the conuli, and the branching of the rays of the
asters, which takes place so extensively in Xenospongia, may also be observed
occasionally.

Q 2

L16 CEYLON PEARL OYSTER REPORT

Xenospongia patelliformis, Gray — -Plate VI.

1858, Xenospongia patelliformis, Gray (30) ; 1867, Xenospongia patelliformis, Gray (31) ;
1873, Xenospongia patelliformis, Holdsworth (9): 1875, Xenospongia patelliformis,
Carter (32); 1882, Xenospongia patelliformis. Carter (33); 1888, Xenospongia
patelliformis, Soi.i.as (15) : 1898, Xenospongia patelliformis, Topsent (26).

The single specimen in the collection (Plate VI., fig. 1) has the form of an almost
circular disk, ahout 28 millims. in diameter, with convex upper and slightly concave
lower surface and rather thin and slightly undulating margin. The lower surface is
formed by the agglutinated sand which makes up the greater part of the thickness of
the disk. The upper surface is covered with rather small, rounded tubercles or
conuli, thickly scattered at fairly regular intervals. These tubercles are about
I millim. in diameter, and are normally hispid from the projection of the ends (if
large spicules, now generally broken off short. The margin of the sponge is also
very shortly hispid from the same cause (tigs. 1, 2, m.f.). Just above the margin
two narrow grooves (tigs. 1, 2, m.p.g.) run round the disk, separated from one another
by an interval of about I "5 millim., across which they occasionally communicate with
one another by oblique connecting grooves. The width of the grooves varies up to
about 0'5 millim., and the floor is crossed at right angles by narrow parallel bands of
fibrous tissue of a whitish appearance, arranged very regularly at short intervals.
The inhalant pores are very minute and arranged in transverse rows between the
fibrous bands in the floor of the grooves (tig. 2). (In the specimens described by
Gray similar grooves radiate more or less abundantly from near the centre of the
disk towards the margin ; in our specimen these radiating grooves are very feebly
developed and recognizable in only a few places, fig. 1, >:j>.</.) The vents, now
more or less closed, are situated at the apices of three low, monticular elevations
near the centre of the disk (figs. 1, 2, o) ; these elevations are not very conspicuous
and are apparently formed each by the agglomeration of three or four of the surface
conuli.* (Dr. Gray appears to have mistaken the inhalant pores, or groups of pores,
for vents ; some of the prominent elevations figured by him on his larger specimen
probably bear the true vents, though one, at least, of these elevations appears to
contain a parasitic barnacle ; there appear to be no barnacles in our specimen, but a
parasitic worm occupied one of the larger exhalant canals.)

The colour of the sponge on the upper surface (in spirit) is pale yellowish grey, with
a faint pinkish tinge in places, reminding one much of Tetliya. The texture is Arm
and compact, leathery above, but the greater part of the thickness of the disk is
composed almost entirely of coarse sand (fig. 2, s.g.), firmly cemented together by the
tissues of the sponge. This sand is exposed only on the lower surface. The total
thickness of the disk in the middle is about 6 millims.

The main skeleton consists in the lower two-thirds of the thickness of the disk, or

* In addition to the vent-bearing elevations, there is one larger, wart-like protuberance containing some
hard foreign body (fig. 1, a).

SPONGES. 117

thereabouts, of the above-mentioned sand-grains, between which there are also
numerous styli, arranged, partly at any rate, in irregular bundles, and various forms
of asters. The upper third of the thickness of the disk is free from sand, and
includes a considerable thickness of the cboanosome as well as the cortex. Here the
main skeleton is formed by stout bundles of styli running vertically to the surface
and entering the tubercles, from which they project as dermal brushes (fig. 2, (/, 6).
These vertical bundles of styli are crossed at right angles by similar bundles which
lie near the inner limit of the sand-free layer of the cboanosome and terminate in the
feebly-developed spicular fringe (m.f.) at the margin of the disk.

The cortex (figs. 2, 3, cort.) is strengthened by an immense number of asters of
various shapes and sizes, forming a very dense skeleton, especially towards the surface.
Asters of various kinds are also very abundantly scattered through all parts of the
cboanosome.

Spicules. — (1.) Styli (Plate VI., tig. 4) ; long, slender, nearly straight, evenly
rounded oft* at the base and tapering very gradually to the apex. The size of these
spicules is so variable that it seems almost useless to give measurements. The largest
are so long that it is difficult to get them unbroken ; 17 millims. by 0"012 millim. is
perhaps a fair average size for the full-grown spicule, but considerably stouter (and
presumably also longer) examples frequently occur (broken) in boiled-out preparations,
while the styli which fringe the edges of the pore-bearing grooves, for example, are,
on the other hand, very much smaller than the measurements given above.

(2.) Euasters (Plate VI., rigs. 5, 6); enormously abundant and varying so greatly
in form and size that it is impossible to separate the different kinds sharply from one
another. The following may, however, be regarded as the principal types : —

(a.) Spherasters with very small centrum and long, conical, stout, sharply and
gradually pointed, often slightly curved actines, about 1 1 in number ; total diameter
of spicule about O'l millim., with centrum 0'02 millim. in diameter and actines
0'<»44 millim. long. These asters pass gradually on the one hand into smaller
oxyasters, and on the other into larger forms with very irregularly curved and more
or less (often much) branched actines; the most copiously branched forms appear to
be characteristic of the sandy layer of the choanosome, where they may attain a total
diameter of as much as 0-l 8 millim.

(6. ) Minute chiasters, with about 8 fairly stout, sub-cylindrical, tylote actines ;
total diameter about 0'008 millim.

(c.) Oxyasters or spherasters, with small centrum and about LI slender, conical,
oxeote, minutely spined actines ; total diameter about 0 "02 millim. The actines are
occasionally branched.

(d.) Similar to (&), but with tylote actines; this is perhaps the most uncommon
ton a of the aster.

After carefully removing the sand-grains from below, it is possible to prepare
microtome sections of the outer portion of the sponge, including the cortex and that

118 CEYLON PEARL OYSTER REPORT.

portion of the choanosome which is free from sand. The cortex (Plate VI., figs. 2, 3,
cort.) is about 0'13 millim. thick between the surface tubercles, but much thicker in
the tubercles themselves, which are composed exclusively of cortical tissue and
spicules. The greater part of the substance of the cortex is made up of asters of
various forms and their accompanying scleroblasts, but it also contains bands of fibrous
tissue running in various directions. Stout bands of dense fibrous tissue (figs. 2, S,f.b.)
also run vertically inwards from the cortex through the outer part of the choanosome
to the sandy layer, where they appear to assist in binding the sand-grains together.
The vertical spicule-bundles which run into the surface tubercles are also accompanied
by similar bands of fibrous tissue (fig. 2), and the same kind of tissue is also developed
in connection with the vents and the pore-bearing grooves.

A noteworthy feature of the cortex is the presence of the very numerous,
approximately spherical, vesicular scleroblasts (fig. 3, scl.), about O'OIG millim. in
diameter ; each resembles a cystenchyme cell and encloses one of the smaller asters,
the ends of whose actines abut against the thin limiting membrane of the cell, or
perhaps sometimes project beyond it. Similar scleroblasts may be observed in
the choanosome. The outermost part of the cortex is convposed of small-celled
chondrenchymatous tissue ; the inner part is more or less fibrous, and between the
two Ave find collenchyma with stellate connective-tissue cells.

The flagellate chambers (fig. 3, Jl.c.) are oval or nearly spherical, closely crowded
together in the choanosome and about 0'028 millim, in diameter where least contracted
by shrinkage. The state of preservation is not sufficiently good to enable me to make
out minute details very satisfactorily, but the chambers are apparently eurypylous.

The inhalant pores are, as I have already observed, minute openings in the floor of
the pore-bearing grooves (fig. 2, m.p.g.). They are very numerous, and 10 or more
may be indicated in a single transverse section of the groove. From each pore a very
narrow inhalant canal runs vertically inwards and opens, with its fellows, into a
system of irregular crypts which lie beneath the thick floor of the groove and from
which wider inhalant canals take their origin and run inwards to the choanosome, in
which they sub-divide into smaller branches. Stout bands of dense fibrous tissue run
across in the floor of the groove from side to side, between the transverse rows of
inhalant pores and pore-canals. Probably, by the contraction of these bands ot
fibrous tissue, the prominent lips of the groove can be brought together and the
groove thus closed.*

The main exhalant canals are only moderately wide and converge towards the vents.
Owing to the state of contraction I am unable to say whether there is a single large
vent or a group of small ones on each of the vent-bearing prominences. Sections
indicate that there may also be small vents between the conuli, and it appears as if
one exhalant canal sometimes opens through several small apertures.

* A very similar arrangement is found in a very different sponge, Esperella mwrrayi, as described in the
Report on the ' Challenger ! Monaxonida,

SPONGES. 119

The geographical distribution of Xenospongia patelliformis is, as pointed out by
Holdsworth (9), very interesting; the only records up to the present time being
Torres Straits and Ceylon. That the specimens met with in these two localities are
specifically identical, 1 have little doubt. Gray, in describing the species from
Torres Straits, gave no description of the spicules, and did not even mention the
occurrence of asters. Carter, to some extent, remedied this defect in 1882 (33). but
he only figured two forms of aster, and those not very characteristic. His
microscopical preparations, however, which are now in my possession and which were
presumably made from one of the Torres Straits specimens (for Holdsworth's
specimen and record from Ceylon appear to have been entirely lost sight of ever since
his note on the subject was published), show both large and small asters of various
forms, and the actines of the large ones may occasionally branch. Mr. Carter's
preparations are only teased ; had they been boiled out from all parts of the sponge,
they would probably have shown all the forms of aster described above from
Professor Herdman's example.

The external form is, of course, extremely characteristic, and so little does it look
like a sponge that the specimen described above was in the first instance placed
amongst the Clypeastroids.

R.N. 375. (Stat, I. First haul of trawl, off Negombo, 12 to 20 fathoms, January
31, 1902.)

Family : SPIEASTEELLID.E.

Astromonaxonellida usually of massive or encrusting form. Skeleton usually irregular
or reticulate, at any rate internally. Megascleres usually stylote or tylostylote.
Microscleres asters of various forms.

Hymedesmia, Bowerbank.

Thin encrusting Spirastrellidae whose spiculation consists of smooth tylostyli and
euasters of various forms (occasionally passing into spirasters).

Topsent has given an excellent account of the literary history of this genus in his
Monograph of the Sponges of France (14). A very large proportion of the known
species come from the Gulf of Manaar, and were described by Mr. Carter (4) in
1880 : viz., H. stellivarians, H. moorei, H. spinatostellifera, H. capitatostellifera, and
H. trigonostellata. Of these five species I have only met with one in Professor
Herdman's collection.

The genus is an extremely interesting one from the phylogenetic point of view, for
we see here, in the first place, the first appearance of the characteristic suberitid
" pin-head" spicule (tylostyle), and, in the second place, the evolution of the typical
spirastrellid microsclere from the euaster.

Hymedesmia forms a connecting link between Coppatias and Spirastrella,

120 CEYLON PEAEL OYSTER REPORT.

Coppatias, as we have already seen, is simply a reduced Stellettid in which the cladi
of the trisenes have become completely aborted, and the megascleres now consist
solelv of oxea. From the oxeote to the tvlostylote form appears at first sight a big
jump, but I think it is evident that the pin-headed type in Hymedesmia is simply the
result of the thinly encrusting habit. The megascleres in this genus are typically
arranged at right angles to the base of support, with their apices projecting outwards,
often beyond the surface of the sponge, an arrangement which is admirably adapted
both to support the soft tissues and to protect the sponge from the attacks of parasitic
Crustaceans or other enemies. In this position the proximal end of the megasclere,
pressed against the hard base of attachment, is unable to elongate in the normal
manner, and siliceous material, which would otherwise have been devoted to its
elongation, is accumulated in a swelling or knob — the " pin-head."

Such is, I believe, the origin of the spirastrellid and suberitid tylostyle, and an
exactly analogous process appears to have taken place in the echinating spicules of
the Ectyoninae, which are commonly swollen into a " head" at the end which abuts
against the skeleton fibre. The at first sight tempting idea that the head of the
tylostyle is the vestige of the cladome of a trisene is put completely out of court by
the fact that it lies at the wrong end of the spicule, viz., the proximal instead of the
distal end.

The evolution of the tvpical spirastrellid spiraster from the euaster has evidently
taken place bv elongation and bending of the centrum of the latter, and various
stages of the process may be seen in different species of Hymedesmia ; indeed, the two
forms of aster pass so insensibly into one another that it is impossible to base a
generic distinction upon their character alone. In the spirastrellid spiraster the rays
(actines) are commonly reduced to mere spines or minute blunt projections.

Hymedesmia stellivarians, Carter.

1S80, Hymedesmia stellivarians, Carter (4).

The single specimen forms a thin yellowish crust, attached to a small melobesian
nodule, in company with Paresperdla serratohamata, &c. In spite of the somewhat
meagre character of Mr. Carter's description and figures, 1 think there can be little
doubt of the correctness of the identification, especially as Mr. Carter's specimen
also came from Ceylon waters.

All three forms of aster mentioned by Mr. Carter are present, but they are not all
of the same size, as might, perhaps, be inferred from his description and figures.
Moreover, in our specimen the megascleres are usually a good deal longer in proportion
to their thickness than Mr. Carter's figure indicates, but they are also very variable
in dimensions.

Under the circumstances it seems desirable to give the following particulars as to
the spiculation : —

(I.) Tylostyli ; only slightly curved, if at all; with oval head at one end and

SPONGES. 121

gradually and more or less sharply pointed at the other; size variable, up to about
0-47 millim. by 0-01 millim.

("2.) Spherasters ; (a) comparatively large, with large centrum and stout, smooth,
conical, sharply pointed rays ; closely resembling the large spherasters of Tethya ;
total diameter about 0"02 millim. ; (b) of medium size, with moderately developed
centrum and stout, sub-cylindrical, very distinctly tylote rays (? heads sometimes
roughened), about as long as the diameter of the centrum ; total diameter of spicule
about 0'012 millim. ; (c) small, with very small centrum and comparatively long,
slender, sharp-pointed rays; total diameter about 0"008 millim. (possibly young forms
of one or both of the others).

R.N. 220b (deep water off Galle and onwards up West Coast of Ceylon).

Hymedesmia curvistellifera, n. sp.— Plate V., fig. 6.

Sponge thin, encrusting. (The single specimen has evidently been removed from
the surface of some Alcyonarian, the large calcareous spicules of which still adhere in
great numbers to the base of the sponge.) Surface uneven, corrugated, very slightly
hispid. Colour (in spirit) light brown ; texture rather friable. Vents and pores not
seen. Average thickness of specimen about 0-5 millim.

The main skeleton consists of loose fascicles of tylostyles, springing from the base
of attachment and running to the surface, where they spread out in loose brushes and
give rise to the more or less hispid character.

Sjnctdes. — (1.) Tylostyli (Plate V., fig. 6, a) ; straight, or nearly so, with large, oval
or sub-globular heads and rather slender, sub-fusiform shafts, gradually and usually
finely pointed at the apex. Size, when fully developed, about 0-38 millim. by 0"006
millim. Many much more slender forms, with proportionally larger heads, sometimes
pear-shaped, also occur ; these I take to be immature spicules.

(2.) Asters (Plate V., fig. 6, b, c, d) ; with strongly-curved centrum and stout,
conical, sharp-pointed, smooth spines (rays). The spines show a tendency in some cases
to arrange themselves in three groups, one group at each end of the centrum and one
in the middle, on the convex side of the curve. The middle of the concave side of the
centrum is free from spines, but the curvature is so great that this part of the
centrum is often concealed from view, and the spicule then resembles a spheraster.
The total diameter of the aster, when fully developed, is about 0-032 millim., but
numerous smaller (young) forms also occur. The asters are most abundant in the
dermal layer, where they form an almost continuous crust.

This species finds a near relative in Topsent's Hymedesmia tristellata (14), from
Banyuls and the Azores. Closely similar asters occur in that species, but their rays
are often covered with minute spines, which I have never seen in the Ceylonese form.
I cannot agree with Topsent's view that the " triple spherasters " (as he terms them)
have resulted from the concrescence of three centra covered with actines ; it appears
to me, on the other hand, that they have arisen by elongation of the centrum and

R

122 CEYLON PEARL OYSTER REPORT.

grouping of the actines, and that we have here the first stage in the evolution of the
typical Spirastrella microsclere. From this point of view it is of great interest to
compare Ridley's Spirastrella transitoria (16), from the Amirante Group. This is a
thinly encrusting species, with " spinispirular, extremely concentrated, composed of
only one entire bend," in fact, very closely resembling the aster of Hymedesmia
curvistellifera. As Ridley further observes, " in S. transitoria we have the spini-
spirular almost in the form of the stellate, with which Schmidt and

Carter consider it to be homologous."

S. transitoria is evidently closely related to S. curvistellifera, but differs in the
proportions of the spicules. Both species, together with H. tristellata, and possibly
some others, undoubtedly occupy a position intermediate between the typical species
of Hymedesmia on the one hand and of Spirastrella on the other, so that they might,
with almost equal propriety, be referred to either genus.

R.N. 320 (Ceylon seas).

Spirastrella, Schmidt.

Massive Spirastrellidae with styli or tylostyli for megascleres and spirasters for
microscleres ; the spirasters usually forming a dermal crust.

Spirastrella vagabunda, Ridley.

1884, Spirastrella vagabunda, Ridley (16).

There are in the collection a number of specimens which, while differing greatly in
external form, agree so closely in spiculation that I am obliged to regard them merely
as varieties of one and the same species, which appears to be identical with Ridley's
Spirastrella vagabunda. This species was originally described by Ridley from
Torres Straits, but at the same time he assigned to it certain specimens in the British
Museum Collection which came from Trincomalee and the Galle Coast, Ceylon, a
fact which, of course, strongly supports my identification of Professor Herdman's
specimens. The Trincomalee specimen referred to was very briefly described by
Carter (33) under the name " Suberites ? sp." Ridley, having examined the same
specimen, suggested (loc. cit.) that it should be distinguished as " S. vagabunda
var. trincomaliensis," on account of certain slight differences in spiculation. In
Mr. Carter's cabinet, now in my possession, there is a preparation, labelled in his
handwriting " Suberites trincomaliensis," which is evidently from the specimen
examined and described by him. Subsequently, in 1886, Mr. Carter (20) described
a sponge from the Mergui Archipelago under the name Suberites trincomaliensis,
identifying it with the Ceylonese form.

Spirastrella vagabunda thus appears to be a widely-distributed and variable
species. Ridley's Spirastrella congenera, from Torres Straits, is probably a mere
variety of the same, and perhaps, also, his S. punctulata, from Mozambique and
Mauritius. The principal characters of the species appear to be the dense, confused

SPONGES. 1:23

arrangement of the main skeleton, the usually stout and distinctly headed tylostyles,
and the slender, more or less elongated spirasters.

I propose to regard Professor HekDMAn's specimens as belonging to four varieties
of the species, which may he distinguished as trincomaUensis, tubulodigitata, fungoides,
and gallensis respectively.

Spirastrella vagabunda, var. trincomaUensis, Ridley.

1882, " Suberites, 1 ap., undescribed, Trincomalee," Carter (33); 1884, Spirastrella
vagabunda, var. trincomaUensis, RIDLEY (16) ; 1886, Suberites trincomaUensis,
Carter (20).

The single specimen in the collection evidently agrees very closely with the
specimen examined by Carter and Ridley. It consists of a massive base rising up
into a few short, stout, finger-shaped processes. The surface is sub-glabrous, but
slightly corrugated vertically ; not warty as described by Carter (this character
having probably been due to drying). The colour (in spirit) externally is nearly
black, with a greenish tinge, internally dark greenish -brown. Texture compact
throughout, but fleshy and fairly compressible, with little or no imbedded foreign
matter. The exhalant canals are (in spirit) extremely narrow and surrounded by
gelatinous tissue almost free from spicules ; they run vertically upwards through the
finger-shaped processes in considerable numbers and probably open by minute vents
(now nearly all closed) at the apices of these projections. Greatest height of specimen
5G millims. ; greatest breadth of massive base 52 millims. ; length of finger-shaped
projections about 23 millims.

The skeleton is a very dense and confused reticulation of megascleres, interlaced in
all directions. On the surface some of them form, in places at any rate, very poorly
developed surface brushes.

Spicules. — (1.) Styli and tylostyli ; moderately stout, usually more or less curved
or crooked ; apices, and heads when developed, very variable and apt to be irregular ;
size of fully grown spicule about 0"G2 millim. by 0-009 millim., but variable.

(2.) Spirasters ; not very abundant and varying in form from the ordinary zig-zag
to one with a simply but strongly curved axis with blunt projections on the convex
side; length about 0-012 millim. These spicules agree closely in form and size with
those in Mr. Carter's preparation, although a good deal shorter than in the type of
the species (according to Ridley's measurements).

R.N. 52 (Gulf of Manaar).

Spirastrella vagabunda, var. tubulodigitata, nov.

In this variety the sponge consists of hollow, finger-shaped processes or " fistula? "

rising from a sandy base to a total height of about 50 millims. Each process contains

several wide, longitudinal exhalant canals, separated from one another by narrow

partitions, and usually ends in a single conspicuous vent. The colour varies from

light to dark grey.

R 2

124 CEYLON PEARL OYSTER REPORT.

The skeleton arrangement and spiculation agree closely with those of the last-named
variety. The surface brushes of megascleres may or may not be well-developed in
different parts of the same specimen.

In one specimen (R.N. 246) a single finger-shaped process swells up at its free end
into an irregular nodular mass, and thus makes an approach to the variety Jungoides.

In 11. N. 218 the spirasters are rather longer and the megascleres have better
developed heads, and are perhaps usually somewhat stouter than in the type of the
variety.

RN. 154 (type of variety) ; 218 (deep water outside pearl banks, Gulf of Manaar) ;
246 ; 352 (Ceylon seas).

Spirastrella vagabunda, var. fungoides, no v.

The type of this variety is an extremely irregular, massive specimen, growing
amongst a quantity of nullipore and other calcareous debris. It is characterised
especially by the presence of definite porous areas, either forming irregular depressions
on the surface, like large pock-marks, or on the flattened tops of fungoid outgrowths.
These are probably inhalant pore-areas, for there are at least two fairly large vents
forming the outlets of wide oscular tubes ; one of these vents is situated ou a level
with the general surface of the sponge, and the other on a low mammiform projection.
The structure internally is somewhat cavernous, and the specimen contains a great
deal of imbedded foreign matter. The size of the entire mass is about 88 millims. by
47 millims. by 47 millims., but a large proportion of it consists of nullipore, &c. The
colour in spirit is pale yellowish -grey.

The main skeleton is dense and confused ; surface brushes are developed in the
pore-areas and, as usual, many at any rate of the spicules in these brushes are much
smaller than those of the main skeleton.

Spicules. — (1.) Tylostyli ; usually slightly curved; with stout fusiform shafts, well-
developed, oval heads and finely and evenly pointed apices ; size in main skeleton
about 0-5 millim. by 0*0167 millim.

(2.) Spirasters; mostly long, slender and zig-zag; occasionally up to as much as
0*048 millim. long, but usually much shorter. Rarely more than 0*002 millim. thick,
exclusive of spines ; occasionally nearly straight.

In the possession of the occasionally much elongated spirasters this variety
resembles Ridley and Dendy's Spirastrella solicla (1) from the Philippine Islands,
which should perhaps be regarded merely as another variety of S. vagabunda.

R.N. 54 (type of variety, Gulf of Manaar) ; 253 (Ceylon seas).

Spirastrella vagabunda, var. gallensis, nov.

The larger of the two fragments by which this variety is represented in the
collection is an irregularly cylindrical piece, about 52 millims. in length, and varying
in diameter from about 10 millims. to about 1U millims. The texture is compact and

SPONGES. 125

firm and the specimen contains a good deal of coarse sand imbedded in it. The colour
(in spirit) is pale yellowish-grey. The surface is smooth, hut rather uneven. No
vents visible. Pores scattered in small groups '.

The main skeleton is a very dense and confused reticulation of stout megascleres.
The preparation only shows very feebly developed surface brushes.

Spicules. — (1.) Tylostyli ; usually curved, with stout, fusiform or sub-fusiform
shafts and well-developed ovoid heads ; apex gradually and evenly and fairly sharply
pointed. Size when full grown about 0'5 millim. by 0'01'J millim.

(2.) Spirasters ; short and slender, closely resembling those of var. triucomalicusm ;
about 0*012 millim. long; not very abundant.

R.N. 178 (type of variety), 179 (both from Lagoon, Galle, June, 1902).

Spirastrella tentorioides, n. sp. — Plate V., fig. 7.

The single specimen bears a striking resemblance in external form to Tentorium
semisuberiles, consisting of a short columnar body ending above in a strongly convex
and sharply-defined pore- and vent-bearing area of darker colour* than the remainder
of the surface. The surface of the column is subglabrous and irregularly furrowed
longitudinally. On one side a much smaller column is given off as a vertical offshoot,
terminating above like the large one. The specimen is attached below to a mass of
calcareous debris by a broad base and narrows somewhat towards the apex of the
column. Total height about 24 millims. Diameter of the column in the middle
about 15 millims. Colour (in spirit) light grey. There are several wide exhalant
canals running vertically through the column, and probably several smallish vents at
the apex. Only one vent, however, is now visible, forming the outlet of the largest
canal, and measuring only about 1'5 millim. in diameter. The inhalant pores are
scattered between the surface brushes of spicules on the rounded apex of the column,
around the vents.

The main skeleton is a very dense, confused reticulation of megascleres, permeating
the whole of the soft tissues, close up to the walls of the canals, on the one hand, and
to the dermal surface on the other. In this reticulation the spicules lie in all
directions, but around the inhalant canals they are mostly placed lengthwise, with
their apices pointing upwards. Surface brushes are confined to the rounded summit
of the column, where they are well-developed.

Spicules.— (I.) Tylostyli (Plate V., fig. 7, a, h)\ straight or slightly curved, with
stout, sub-fusiform shafts gradually and sharply pointed at the apex, and well-
developed oval heads ; size when full grown about 0 6G millim. by 0'0164 millim., but
much smaller in the surface brushes.

(2.) Spirasters (Plate V., fig. 7, c-fj) ; varying much in shape and size, usually
slender, but sometimes stout; e.g., (a) short, slender, simply curved, with projections
(hardly spines) on the convex side ; length about 0*008 millim. ; (b) short, slender,
* The darker colour is due to the entanglement of dirt amongst the spitule-brushe=.

126 CEYLON PEARL OYSTER REPORT.

zig-zag, with projections on all sides; length about 0"016 millim., thickness about
0'002 millim. (excluding projections) ; (c) long, slender, zig-zag or crooked, with
small, sharp spines on all sides; size about 0"048 millim. by (P0027 millim. (excluding
spines) ; (d) short, stout, zig-zag, with stout conical spines on all sides ; size about
0"022 millim. by 0"004 millim. (excluding spines), with spines about 0"006 millim.
long. The spirasters are abundant and form a thin dermal crust.

The canal system of this sponge is somewhat remarkable and, like the external
form, reminds one of Tentorium. The wide exhalant canals, running vertically
upwards to the apex of the sponge, have already been noticed. The inhalant pores,
situated also at the summit of the sponge, lead into subdermal cavities from which
very narrow inhalant canals run vertically downwards, more or less parallel with the
oscular tubes. These canals unite together as they descend and, even in hand-cut
and unstained sections, can be easily traced by the brown colour of their walls.

Considering the extraordinary variation which the species of Spirastrella exhibit,
alike in external form and in the arrangement of pores and vents, I should hardly
have considered characters of this nature alone sufficient to justify the establishment
of a new one, but should have regarded this form as yet another variety of
Spirastrella vagabunda. We have here, however, a stout form of the spiraster
which is, perhaps, not represented in any of the varieties of that species, and this
fact, taken in conjunction with the other characters, seems to me to justify a specific
separation.

R.N. 239 (Ceylon seas).

Placospongia, Gray.

Spirastrellidse with a stony spicular axis and a similar cortex, both composed of
closely packed sterrospiroe ; with bundles of tylostyles radiating from the axis
towards the periphery. Cortex divided into polygonal areas by grooves con-
taining the inhalant and exhalant apertures.

Placospongia carinata (Bowerbank).
[For Literature and Synonymy, see Vosmaer and Vernhout (35).]
In view of the very recent publication of the elaborate monograph on the genus
Placospongia, by Vosmaer and Vernhout, it is unnecessary to say much about this
interesting species. The presence of numerous parenchymatous spirasters ("spini-
spirse"), and the fact that the dermal spicule is a microspire, justify the specific
identification according to the views of the authors quoted. I also agree with
Keller in placing the genus amongst the Spirastrellidse, it having been shown that
the "sterraster" of this sponge is a modified spiraster.

Carter (4) has described a species of Placospongia from the Gulf of Manaar which
he identified as P. melobesioides, and this identification will no doubt hold good, for
he expressly mentions the absence of spirasters (" spinispiruke"). He has also (5)

SPONGES. 127

recorded the same species from the vicinity of the Basse Rocks, off the south-east
coast of Ceylon. It appears, therefore, that the two species, melobcsioulex and
cannata, hoth occur in Ceylon waters.

R.N. 118 (Stat. V., off Chilaw, 10 fathoms).

Negombo, n. gen.

Spirastrellidre consisting (? always) of tubular processes (? arising from a common
base). Megascleres smooth styli ; microscleres sanidasters.

This genus may, perhaps, have arisen independently from some sanidastrose form
of Tetractinellid, but as regards its existing characters it is so closely related to
Spirastrella that it may be included in the same family. It is also quite possible
that its sanidaster may be merely a modified spiraster.

Negombo tenuistellata, n. sp. — Plate V., fig. 8.

Sponge consisting of a group of short, rather thin-walled tubes of very variable
diameter, growing up close together, side by side, and more or less fused with one
another laterally. Each tube ends above in a single widely-open vent, ranging in
diameter from about 3 millims. to about 8 millims. All the tubes are broken off' and
widely open below, so that it is impossible to say whether or not there was a basal
mass from which they sprung, but probably there was. The walls of the tubes contain
a great number of large sand-grains embedded rather sparsely in them. The colour
(in spirit) is pale yellowish-grey, translucent ; the texture rather soft and flexible, but
fairly tough. The tubes do not vary greatly in height, the height of the longest
being about 31 millims., while its width in the middle is about G millims. ; the tube
next to it is of about the same height, but as much as 12 millims. wide in the middle.
The walls of the tubes are scarcely 2 millims. thick in the middle, thinning out
somewhat towards the margin of the vent and thickening slightly towards the base.
The outer surface of the tube-wall is rough, with more or less embedded sand-grains,
and also, between the grains, irregularly reticulate with slightly-projecting ridges ; I
have not been able to find dermal pores in it. The inner surface of the tube-wall, on
the other hand, is covered by a kind of dermal membrane, strengthened by a reticulation
of megascleres, and bearing numerous small pores, resembling dermal pores but
presumably exhalant, in the interstices of this reticulation.

The main skeleton consists of long styli, not forming definite fibres but sometimes
collected into loose wisps. They mostly lie lengthwise in the thickness of the tube-
wall and are more abundant in the middle of its thickness than elsewhere. On the
inside of the tube-wall there is, as already indicated, a well-developed "dermal"
reticulation of styli, crossing one another singly, or in twos or threes, in all directions
parallel with the surface. On the outer surface of the tube-wall the dermal
membrane contains very numerous microscleres and the megascleres lie at a slightly
lower level. On both surfaces the dermal membrane is supported to some extent on

128 CEYLON PEARL OYSTER REPORT.

the outer ends of very loose, irregular wisps of styli which come oif from the dense
central portion of the main skeleton.

Spicules. — (1.) Styli (Plate V., fig. 8, a, b, c) ; rather long and rather slender,
subfusiform, with the base rather narrower than the middle and the apex fairly
gradually sharp-pointed ; usually slightly curved ; size about 0"54 millim. by
0-012 millim., but variable. The apices have a tendency to be irregular and
occasionally the spicule becomes oxeote.

(2.) Sanidasters (Plate V., fig. 8, d) ; each in the form of a very slender rod,
straight or slightly crooked, terminating at each end in a slight swelling or, perhaps,
a couple of small spines, and bearing very slender spines along its length, usually
most strongly developed in, or perhaps even confined to, a whorl on each side of the
middle of the spicule. Total length about 0'012 millim. ; maximum diameter,
including spines, about 0'004 millim. These spicules appear to be almost, if not
quite, confined to the dermal membrane on the outer surface of the sponge.

RN. 362 (Stat. I., hauls 1-4, January 31, 1902, off Negombo, 12 to 20 fathoms).

Family: CLIONID^.

Astromonaxonellida of boring habit ; forming excavations in the shells of Mollusca
and other calcareous bodies.

Cliona, Grant.

Clionidse of which the complete spiculation is composed of tylostyli, oxea, and
spirasters. One. or two of these forms of spicule may be absent by atrophy.
It will be seen that I have adopted Topsent's views (36, 37) as to this genus and
its systematic position, in preference to those expressed in the Eeport on the
" Challenger " Monaxonida.

Cliona margaritiferae, n. sp. — Plate V., fig. 9.

The specimens in the collection consist of pieces of the shell of Margaritifera
vulgaris (the Ceylon Pearl Oyster), abundantly excavated by the sponge. The
chambers which it makes are rounded or oval in form, more or less crowded together,
according to age, and connected with one another by narrow tunnels. From the
inner sides of the chambers slender, elongated, conical canals radiate at various angles
towards the inner surfiice of the shell. These outgrowths are shaped like spines, and,
as in certain other species which I shall refer to later on, give the chambers a very
characteristic appearance when viewed by transmitted light. From the outer side
of each chamber are given off usually about two cylindrical canals, which perforate
the outer layer of the shell at right angles to the surface and terminate in circular
vents or pore-areas (fig. 2, on p. 144, shows a shell excavated by this sponge).

SPONGES. 129

The walls of the excavations have a finely granulated or, under the microscope,
frothy appearance, due to the presence of innumerable shallow, conchoidal depressions,
which are in contact with one another all over the surfaces of the walls.

There is only room in the thickness of the shell for a single layer of chambers.
When these have attained their full size, their transverse diameter is about 1"5 millims.,
but they may be considerably elongated. The diameter of the cylindrical canals
which terminate on the outer surface of the shell varies up to about 0-65 millim.
The outer end of each is closed, completely or partially, hy a thin membrane containing
numerous micro- and megascleres. Where the membrane in question stretches com-
pletely across the end of the canal, I assume it to be an inhalant pore-sieve, though
the pores cannot now be seen ; such a membrane may be supported by bundles of
tylostyles converging towards the centre. The vents, on the other hand, are more
or less widely open and surrounded by a thin, membranous diaphragm, also supported
by tylostyles.

The body of the sponge itself (in spirit) forms for the most part a very thin lining
(of a pale yellowish-brown colour) to the chambers.

Spictdes. — (1.) Tylostyli (Plate V., fig. 9, a) ; straight, or nearly so, slender, very
gradually and sharply pointed, and with well-developed globular heads ; size about
0"25 millim. by 0-004 millim., with head about 0'006 millim. in diameter. These
spicules occur scattered generally, as well as in the neighbourhood of the vents and
pores.

(2.) Spirasters (Plate V., fig. 9, b, c, d) ; usually with about four angulations, fairly
stout, and abundantly but rather minutely spinous, size about 0"024 millim. by
0"004 millim., but variable. These spicules pass by transitional forms (Plate V.,

fig- 9> e> /> u) mto

(3.) Spined microxea (Plate V., fig. 9, h, 1-) ; usually bent or angulated in the middle,

and measuring about 0"06 millim. by 0-0027 millim., but variable. These forms are

extremely numerous in the deeper parts of the sponge. The gradual transition

between spirasters and microxea in this species is extremely interesting.

The excavations made by Cliona margaritiferce, even down to the spine-shaped
outgrowths of the chambers, are closely similar to those of Hancock's Cliona
spinosa (38) and Thoosa cactoides (38), but the spiculation is very different. Cliona
spinosa occurs in shells of Perna and Placuna, and Thoosa cactoides in shells of
Meleagrina margaritifera. Another distinct species, of closely similar form, occurs
in shells of Meleagrina albino,? and has been described by the same author (38)
under the name Cliona cervina ; the spiculation of this species makes a much closer
approach to that of C. margaritiferce, but the two appear to be quite distinct.

Topsext (37) has already described, under the name Cliona indica, a boring sponge
infesting a pearl oyster from Ceylon. This species is evidently nearly related to ours,
but the describer gives no information as to the character of the excavations or
apertures, and the spiculation differs so much from that of our species that it is

s

130 CEYLON PEARL OYSTER REPORT.

impossible to suppose that the two are identical. Both belong, however, to Topsent's
third division of the genus Cliona.

Thiele's Cliona concharum (39), from Japan, perhaps comes nearer to our species
than any other, at any rate so far as the spiculation is concerned, the differences in
this respect being so slight that subsequent researches may make it desirable to
consider the two as being only varietally distinct. In the Japanese form, however,
the chambers excavated by the sponge seem to be much smaller, while the spined
microxea are a good deal longer (0*09 millim.), than in that from Ceylon.

Carter's Cliona warren i (5), also from the Gulf of Manaar, on the other hand, is
a very different species., and, according to Topsent (37), is identical with the European
Cliona celata.

R.N. 261 (Gulf of Manaar — very abundant, and destructive, on the pearl banks).

Family : SUBERITf D,E.

Astromonaxonellida in which the megascleres are styli or tylostyli, and the microscleres
have completely disappeared.

Suberites, Nardo.

Suberitidaj of varying form, but without mammiform projections on the surface.
Spicules tylostylote (nearly always). Skeleton usually arranged radially, with
surface brushes of spicules smaller than those of the main skeleton.

The genus Suberites, at any rate so far as its typical species are concerned, for it
may possibly be of polyphyletic origin, is, as I have already indicated, probably
derived from Sjyirastrella by loss of the spirasters. Carter even admitted into the
genus certain species with spirasters.

It is remarkable that there is only one species of the genus, and of that only a
single specimen, in the present collection. In making my preliminary examination in
Liverpool (which had to be done very hastily), I identified one of the specimens as
Suberites inconstans, var. digitata, a form jjreviously described by me from near
Ceylon, but more careful examination subsequently revealed the presence of spirasters
and thus proved that the specimen was really a Spirastrella.

Curiously enough, Thiele (39) has come to the conclusion that Suberites inconstans
is in reality a Spirastrella in' which I have overlooked the spirasters, these being,
according to him, small and scarce. This is, of course, possible, but it is by no means
proved. Thiele has apparently never seen specimens from Ceylon or India, but
identifies certain specimens from Celebes with the species, under the name
Spirastrella inconstans. Unfortunately I am unable to re-examine the types of the
species here in South Africa, as the original specimens and preparations remained in
the British Museum.

Topsent has sub-divided the old genus Suberites into a number of separate genera,

SPONGES. 131

and there is a great deal to be said in favour of such a proceeding, though it is
doubtful whether all of his distinctions can be maintained. Suberites cruciatus, for
example, combines certain characters of Topsent's La xosv.be rites and Axosuberites
with peculiarities of its own, and I prefer to make use of the old generic name in this
instance.

Suberites cruciatus, n. sp. — Plate V., fig. 10.

Specimen consisting of a number of long, slender, flattened branches, springing from
a short pedicel of similar structure to themselves and branching in an almost
dichotomous manner, but with some of the branches fusing together again higher up.
Total height of specimen about !)1 millims. ; length of pedicel to first fork about
14 millims., breadth 3-5 millims., thickness 2 millims. : breadth of separate branches
about 2'5 millims.: thickness about 1'5 millims. Surface rather uneven, very
minutely hispid, and beset with very numerous small rounded translucent areas,
apparently pore-areas. Vents probably minute and scattered. Colour (in spirit) pale
brown ; consistence soft and very flexible.

The main skeleton consists of numerous loose bands of tylostyles running lengthwise
through the sponge ; with numerous similar spicules scattered between in a loose,
irregular reticulation. I have not detected any spongin cement. Towards the surface
this arrangement gives place to radiating brushes of tylostyles, whose apices project
slightlv beyond the dermal membrane. Between these brushes lie the fairly extensive
sub-dermal cavities.

Spicules. — Tylostyli (Plate V., fig. 10), of rather peculiar form. Usually straight,
slender, gradually and finely pointed at the apex : with heads usually elongated
transversely at a little distance from the base of the spicule, so as to form a cross.
This cruciate character is most pronounced in the youngest and slenderest spicules ;
in the mature forms the arms of the cross form rounded knobs projecting from the
shaft usually at a very slight distance from the base ; occasionally there are three of
these knobs instead of two. The full-sized spicules measure about 0"31 millim. by
0'005 millim., with head about 0'0093 millim. across.

The shape of the tylostyle in this curious little sponge reminds one of the
corresponding spicule in Carter's Hymedesmia spinatostellifera (4).

RN. 315 (Stat, L.V., outside Periya Paar, 24 fathoms).

Family: CHONDEOSIID^E.

Corticate Astromonaxonellida with complex canal system and small flagellate
chambers. Without megascleres.

These sponges appear to be Astromonaxonellida in which the megascleres (and in
the case of Chondrosia the microscleres also) have been lost by degeneration. Their
strongly developed cortex and complex canal system show that they are not primi-
tively simple forms like the Myxospongida, and, as the megascleres probably passed

s 2

132 CEYLON PEARL OYSTER REPORT.

through a monaxonellid condition before finally disappearing, we may include the family
in the monaxonellid rather than in the tetractinellid grade, though, perhaps, logically
speaking, it ought to occupy a distinct grade of its own. The form of the micro-
scleres (when present) and the corticate character suggest a close relationship with
the Tethyidse.

The family will always be historically interesting as having formed the subject ot
one of F. E. Schulze's classical memoirs (40).

Chondrilla, Schmidt.
Chondrosiidse with microscleres in the form of euasters of various kinds.

Chondrilla australiensis, Carter.

1873, Chondrilla australiensis, Carter (41) ; 1885, Chondrilla australiensis, Lendenfeld (42).

The specimens of this sponge form thin crusts of a greyish colour (in spirit) and
irregular outline, spreading over masses of calcareous debris. One large specimen is
about 80 millims. in greatest breadth. The surface is smooth and sometimes
glabrous ; it may be minutely reticulate when seen under a lens, and in one specimen
(R.N. 185) it is very minutely papillate, with the spherasters so thickly aggregated
in the papillge that they touch one another. The vents are minute and scattered.

Having in my possession several of Mr. Carter's own microscopical preparations of
this species, evidently from the original types, I have been able to make a direct
comparison and to assure myself of the correctness of the identification. In both
Mr. Carter's and Professor Herdman's specimens the spheraster, with smooth
conical rays, measures about 0"028 millim. in diameter, and the oxyaster, with
minutely spined rays, sometimes branched at the ends, measures nearly as much. In
a Ceylonese specimen I find that neither spicule is strictly confined to either the
cortex or the interior of the sponge, but while the spherasters are much more
abundant in the cortex, the oxyasters are much more abundant in the interior.

It is noteworthy that Mr. Carter has recorded (5) the occurrence of Chondrilla
nucula, the common European species, from the Gulf of Manaar. This species
appears to be of very wide distribution.

RN. 17 (Gulf of Manaar); 185 (Donnan's Paar) ; 251 ; 376 (encrusting a shell of
Margaritifera vulgaris, Cheval Paar).

Chondrilla australiensis, var. lobata, nov.

Tins variety is represented by two specimens (apparently obtained together) which
differ conspicuously from the thin, encrusting form above described, being massively
lobose, with comparatively large vents placed singly on the top of the lobes. The
base of attachment is constricted to a few narrow projections on the lower surface,
and the entire body of the sponge exhibits a swollen, tumid appearance. The vents
measure up to 2 millims. in diameter, and each is surrounded by a thick, membranous
margin, which, in its turn, is usually surrounded by a shallow groove. The surface

SPONGES. 133

is sub-glabrous, but uneven and very minutely reticulate. The colour below (in spirit)
is very pale grey, but on the upper parts of the lobes the grey is mottled with brown.
The texture is compact and fleshy ; compressible.

The two specimens are of about equal size ; the one selected for measurement is
about 42 millims. in length, 17 millims. in breadth, and 18 millims. in height. The
cortex is about 0-164 millim. thick.

The spicules agree closely in form with those of the thin, encrusting variety. The
spherasters are, however, rather smaller. They have, in both varieties, a strongly
marked tendency for their rays to be reduced to low warts or even, perhaps, to dis-
appear, so that in some cases the large centrum is left almost smooth, as in Keller's
C. globulifera, which is, however, specifically distinct.

A lobate variety of C. australiensis occurs also in Australia.

R.N. 286, 286a (deep water off Galle and onwards to Colombo. Hauls off Kaltura
and off Mount Lavinia, 20 to 30 fathoms, February 19, 1902).

Chondrosia, Nardo.
ChondrosiidaB in which all the spicules have completely disappeared.

Chondrosia reniformis, Nardo.

This well-known Mediterranean species is represented in the collection by two
specimens, both attached to the same fragment of a horny sponge. Each is roughly
hemispherical in form, with wide base spreading out into a broad, thin "stolon" on
one side. The surface is covered with low, irregular tubercles, which may be due in
part to contraction. The vents are small and difficult to make out, sometimes, at any
rate, on low mammiform projections. The structure of the dense fibrous cortex, with
its pigment cells, and that of the choanosome, agree very closely, so far as can be
made out in the material at my disposal, with Schulze's classical description (40).
The pigment cells, however, appear to be mostly in the inner part of the cortex,
instead of in the outer part as figured by Schtjlze ; but this is not an important
difference. Each specimen is about 13 millims. in diameter, and the colour (in spirit)
is mottled grey and brown on the surface, and pale greyish-yellow internally.

R.N. 226 (two specimens ; deep water off Galle and onwards up West Coast of
Ceylon).

Sub-order: SI GMATOMONAXONELLID A.

Monaxonellida in which the typical microscleres are sigmata, or forms derived
therefrom, normal astrose microscleres being absent.

The sponges which comprise this large sub-order may be regarded as descended
from the tetractinellid sub-order Sigmatophora (family Tetillidoe) by reduction of the
megascleres, in the same way that the Astromonaxonellida may be regarded as being
derived from the tetractinellid Astrophora.

134 CEYLON PEARL OYSTER REPORT.

In dealing with the Astromonaxonellida I have already had occasion to point out
that no reliable instance of the occurrence of true astrose microscleres in conjunction
with sigmatose forms has ever been recorded, a fact which argues very strongly in
favour of the primary cleavage of the Monaxonellida into two great groups corre-
sponding to the tetractinellid sub-orders Astrophora and Sigmatophora. Since that
part of my report was completed, however, and sent to England, Topsent's latest work
on the ' Sponges of the Azores ' (62), has been received here. In this work the author
describes two apparent Sigmatomonaxonellida in which asters were met with, viz.,
Yvesia alecto and Leptosastra constellate/, but in neither of these are other microscleres
present.

In Yvesia alecto the asters are, as Topsent himself shows, simply transformed
megascleres (spined oxea) with the spines greatly developed, so that they are clearly
of secondary origin. It may be pointed out, further, that an analogous transforma-
tion takes place in the echinating megascleres of the genus Cyamon * and, probably,
also in the genus Trihentrion, giving rise to spicules which simulate true asters. It
appears, therefore, that in certain Sigmatomonaxonellida astrose spicules have arisen
secondarily by transformation of spined megascleres. It is obvious, however, that
these cases do not affect the primary division of the Monaxonellida here adopted.

The case of Leptosastra coustellata offers a more serious difficulty, for here there is
no indication that the asters, which form a superficial crust, are other than true
astrose microscleres. It is, however, a suggestive fact that the sponge in which they
occur is an Ectyonine with spined styli. Two explanations appear to me possible :—
(1.) The asters may not belong to the same sponge as the megascleres. Considering
the well-known and frequent accidental admixture of the spicules of different species,
caused either by the sponges growing over one another or by the taking in of foreign
spicules in the same way that grains of sand are taken in, and especially in view of
the fact that only a single specimen of the sponge has been obtained ; although
Topsent has assured himself to the contrary, I venture to think that we may be here
dealing with a composite spiculation. (2.) It is not impossible that the asters, if
proper to the sponge, may be derived secondarily from the spined styli, in much the
same way as in the genus Cyamon. Their position at the surface of the sponge,
however, and their apparently normal form, are opposed to this view. However,
until we have further evidence before us, it is quite unnecessary to allow this isolated
case of a single specimen to make us alter our views on the classification of the
Monaxonellida.

Assuming;- then that the Sip'matomonaxonellida are derived from the tetractinellid
Sigmatophora, the question arises where are we to seek for the point of contact
between the two groups ? The answer to this question is easily given, for in the
genus Gellius we have a near approach to the massive species of Tetilla, such as
T. limicola. The replacement of the tetractinellid megascleres by oxea in a massive

* See later xmdeTjCyamon.

SPONGES. 135

Tetilla, accompanied by the complete loss of the radiate skeleton arrangement, both
of which changes are already partially accomplished in such forms as T. Umicola,
would give us a typical GeUius, and in GeUius I believe we have the starting point of
the entire sigmatomonaxonellid series.

From this starting point the Sigmatomonaxonellida have branched off in various
directions in the course of their evolution. Total loss of microscleres has given rise
to forms with the spiculation composed entirely of oxeote megascleres, constituting
the family Homorrhaphidse of Ridley and Dendy, and the strong development of
spongin cement has led to the evolution of the very large sub-family Chalininae,
from which in turn some of the so-called Ceratosa have been derived by total
suppression of the spicules and their replacement by horny fibre. The arrangement
of the megascleres in a characteristic plumose fashion, accompanied in most cases by
the replacement of many of the oxea by styli and the loss of microscleres, has
given rise to the family Axinellidaj. The development of a new type of microsclere —
the chela — by modification of the sigma, has given rise to the great family group
Desmacidonidse, within which the Ectyoninae have arisen by development of spined
echinating spicules.

In the Report on the " Challenger" Monaxonida, published 17 years ago, Mr. Ridley
and I proposed to divide the Sigmatomonaxonellida (= Halichondrina) into four
families, viz., Homorrhaphidse, Heterorrhaphidse, Desmacidonidse and Axinellidaj, an
arrangement which has been variously modified by subsequent writers, chief amongst
whom is Topsent. This author (59) has united our Homorrhaphida? and Heteror-
rhaphidse in one family which he terms " Haploscleridas," a proceeding which appears
to me justifiable in view of the obviously close relationship between the two, and
especially in view of the fact that the Renierinse and Chalininse have very probably
arisen independently from different though closely related genera, the Renierinse from
GeUius (and perhaps other genera), and the Chalininse from Gelliodes and Toxochdlina,
as well as, perhaps, in some cases from Renierinse.

Topsent has also proposed the name " Pceciloscleridai " in replacement of
" Desmacidonidaa " — a proceeding which appears quite unnecessary, although the
extent of his family is not quite the same as that of ours.

I therefore propose in this Report to sub-divide the Sigmatomonaxonellida amongst
three families, viz., Haploscleridaa (including the Homorrhaphidas and Heterorrhaphidaj
of Ridley and Dioidy), Desmacidonkke and Axinellkke. The Haplosclerkbe, there
can be little doubt, should stand hist, and indeed occupy the position of a parent
group from which the other two families have descended.

Family : HAPLO.SCLEIUD.E.

Sigmatomonaxonellida in which microscleres when present are usually in the form of
shnnata, or derivatives thereof, but never chelre. The skeleton is reticulate and
the fibre is typically not plumose. The megascleres are usually diactinal.

136 CEYLON PEARL OYSTER REPORT.

The most primitive sub-family of this group is undoubtedly the Gelliinte, from
which the Renierinee and Chalininae are clearly derived. The other sub-families may
be in part Desmacidonidse which have lost their chelae ; this I am able in the present
Report to demonstrate pretty clearly in the case of the Phlceodictyinae, which I have
accordingly removed. Topsent has already removed the Tedaniinse, Desmacellinae,
and Hamacanthinae as being Desmacidonkke without chela? ; this may be quite right,
but until clear evidence that they are descended from chela-bearing forms is
forthcoming, it seems to me equally justifiable to leave them in the parent group.
A Desmacella, for example, may very well be an Esperella which has lost its chelae,
but it seems at least equally probable that it has never reached the stage of
having any.

Sub-family : GELLIINiE.

Haploscleridas with diactinal megascleres, and sigmata or toxa or microxea for
microscleres.

Geilius, Gray.

Gelliinae with little or no spongin, the main skeleton being formed by a reticulation
of oxea.

Gellius fibulatus (Schmidt).

1862, Reniera fibulata, Schmidt (47); 1880, Reniera fibulifera, Carter (4); 1892, Gellius
fibulatus, Toi'sent (48).

There is in the collection a considerable quantity of this sponge growing amongst
the branches of a Floridean Alga. I identify it with the European species by direct
comparison with a preparation from a specimen from Budleigh Salterton in
Mr. Carter's cabinet. The differences in spiculation are very slight. In our specimen
the slightly curved and gradually sharp-pointed oxea measure about 0"184 millim.
by 0*007 millim., and the sigmata about 0-02 millim. from bend to bend.

R.N. 51 (Gulf of Manaar); 299; 348; 350 (all growing in association with
apparently the same kind of alga ; the last three probably fragments of one and the
same specimen).

Gellius angulatus (Bowerbank), var. canaliculata, nov. — Plate IX., fig. 7.

[For synonymy and references vide Ridley and Dendy (1)].

The single specimen is massive, rounded and slightly elongated, about 18 millims.
long by 12 millims. in transverse diameter. It was probably attached by one end, and
bears a group of vents at the other. The surface is even and smooth, conspicuously
veined by ramifying exhalant canals, which run towards the (upper ?) end of the
specimen and open there by means of the moderate-sized vents. Many of these
exhalant canals lie just beneath the surface and are covered over only by a thin,
translucent membrane, which easily gets rubbed off, leaving the canals as open

SPONGES. 137

grooves. Between the exhalant canals there is no separable dermal membrane, and
there are no conspicuous sub-dermal cavities. Texture between the canals very
compact, hut nol very hard and rather friable. Colour (in spirit) very pale grey.

The skeleton is a very close and pretty uniform reticulation of single oxea, crossing
one another in every direction. The dermal skeleton consists only of scattered oxea
placed tangentially.

Spicules. — (1.) Oxea (Plate IX., fig. 7, a) ; rather slender, slightly curved, usually
somewhat abruptly or even hastately pointed; about 0-25 millim. by 0-008 millim. to
0-01 millim.

(2.) Sigmata (Plate IX., fig. 7, h) ; slender, C-shaped, with shortly and sharply
incurved and sharply pointed ends ; often with a slight indication of enlargement or
angulation near the middle ; measuring about 01)28 millim. from bend to bend when
full grown.

(3.) Toxa (Plate IX., fig. 7, c); rather short and moderately stout; sharply
angulated in the middle and only slightly recurved at the extremities; length up to
about 0'044 millim.

This is a pretty little sponge, evidently very nearly related to the European
Gellius angulatus, from which it differs in the somewhat shorter oxea, the consider-
ably larger sigmata, and the much shorter toxa. It is possible also that the arrange-
ment of the exhalant canals may be characteristic. It is interesting as indicating
that sigmata and toxa are simply slightly different modifications of the same form of
microsclere.

P.N. 140 (deep water offGalle and onwards up West Coast of Ceylon).

Gelliodes, Ridley.

Gelliina? with much spongin, more or less completely enveloping or even replacing the
megascleres and forming distinct fibres. The microscleres are sigmata.

Gelliodes carnosa, Dendy — Plate VII. , fig. 5.
1889, Gelliodes carnosa, Dendy (3).

There is one fine specimen of this sponge in the collection, easily recognized by its
characteristic external form (Plate VII., fig. 5). The megascleres are much slenderer
than in the types and appear to be becoming vestigial, as in so many chalinine
sponges, being functionally replaced by the strongly developed horny fibre. The
sigmata are still numerous, about 0"02 millim. long, but very slender.

This species forms a conspicuous feature of the Ceylon Sponge-Fauna.

P.N. 69 (Stat. XXVII. , Cod Bay, Trincomalee, 5 fathoms; also Gulf of Manaar).

Gelliodes incrustans, n. sp. — Plate IX., fig. G.

Sponge thin, encrusting ; the single specimen growing over both valves of a Pecten
(which was evidently alive when collected). Maximum thickness about 7 millims.

T

138 CEYLON PEAEL OYSTER REPORT.

Surface smooth, but rather uneven. Vents numerous and conspicuous, circular, from
about 1 millim. to 2-5 millims. in diameter; mostly flush with the surface, hut
occasionally on low prominences. Colour (dry) dull greyish-brown, with a purplish
timre in places, which seems to indicate that it was purple in life. Texture (dry)
compressible and resilient, but rather stiff.

The main skeleton is a rectangular-meshed network of horny fibre cored by spicules,
in which the meshes vary greatly in size. The principal fibres are about 0'033 millim.
in diameter, and contain a multispicular core of small oxea which occupy only about
one-third (or less) of the total thickness of the fibre. The secondary fibres are more
slender and contain fewer spicules. The dermal skeleton (Plate IX., fig. G) is, for the
most part, a unispicular reticulation of oxea with comparatively little spongin ; some-
times one sees brushes of projecting oxea, but I am not sure how for these are proper
to the dermal skeleton.

Spicules. — (1.) Oxea (Plate IX., fig. 6, o) ; short, slender, slightly curved, sub-
fusiform, gradually and sharply pointed at each end ; size about 0-l millim. by
0'004 millim. A number of very slender, hair-like oxea, probably young or vestigial
forms, also occur.

(2.) Sigmata (Plate IX, fig. 6, s) ; very slender and hair-like, C-shaped, up to
about 0-02 millim. from bend to bend. Abundant.

This species is nearly related to Gellioden licheniformis (Lamarck),'" but differs, at
any rate from the " Challenger" specimen of that species, both in external form and
in the much smaller size of the spicules and more regular arrangement of the
skeleton.

R.N. 1 12 (Gulf of Manaar, dry).

Gelliodes petrosioides, n. sp. — Plate IX., fig. 3.

Sponge massive, depressed, cushion-shaped ; flattened below, where it has appa-
rently been attached by a broad base ; evenly rounded oft' and strongly convex above.
Surface coarsely granular, not hispid. Vents not visible. Pores numerous, scattered
in the thin dermal membrane which roofs over the numerous small, rounded sub-
dermal cavities. Colour (in spirit) pale yellowish-grey. Texture compact ; hard and
almost stony. Greatest diameter of specimen, which is irregularly rounded in outline,
about 24 millims.

The main skeleton is a very dense but quite irregular reticulation of very coarse,
stout, densely spicular fibre, with a great many loose megascleres scattered between.
The stout fibres have a thickness of about (P164 millim., and probably contain a
certain amount of spongin, which, however, is not visible in ordinary sections. The
dermal skeleton cannot he sharply distinguished from the main skeleton, and consists
of an irregular reticulation of coarse spicular fibre, the interspaces in which are
occupied by the thin, pore-hearing dermal membrane.

* Vide Riih.f.y and Dendy, ' "Challenger" Mon.axoirida,' p. 48.

SPONGES. 139

Spicules. — (1.) Short, stout, fusiform, slightly curved oxea (Plate IX., fig. 3, a, b, c),
usually sharply and fairly gradually pointed at each end (often becoming strongylote
or stylote) ; measuring about 0"25 millini. by 0"017 millim., but varying a good deal
in thickness.

(2.) Sigmata (Plate IX., fig. 3, d) ; slender, C-shaped or contort, about 0'022 millini.
from bend to bend ; very numerous.

Were it not for the presence of the sigmata, this species would be a typical Petrosia.
It appears to be very nearly related to Topsent's Gelliodes fayalensis* from the
Straits of Pico-Fayal, but is distinguished by the absence of the large oscula, and,
perhaps, by other characters.

Pi.N. 146 (deep water off Galle and onwards up West Coast of Ceylon).

Gelliodes petrosioides, var. fibrosa, now

I propose this name, at any rate provisionally, for a single small, much-damaged
specimen of irregular shape and cavernous structure, with large exhalant canals and
vents (?) and very soft, fibrous texture ; agreeing very closely with the type of the
species in spiculation and in the structure of the main skeleton fibres, but with the
fibres better defined and the meshes of the reticulation mostly very wide and -not
filled up by scattered spicules. The soft texture of the specimen, which is in
striking contrast with the hardness and density of the type, is due simply to this
greater laxity in the skeleton arrangement.

Except for the presence of the sigmata, this variety closely resembles a Pachy-
chalina. It may ultimately, when better specimens are forthcoming, have to be
considered as a distinct species.

R.N. 272 (deep water off Galle and onwards up West Coast of Ceylon).

Toxochalina, Ridley.

Gelliinse with much spongin, more or less completely enveloping or even replacing the
megascleres, and forming distinct fibres. The microscleres are toxa.

This genus differs from Gelliodes only in the replacement of the sigmata by toxa,
and in view of the occurrence of both these forms in the same species of Gellius
(vide under Gelliits anr/nlatus, var. canaliculata), it may be questioned whether the
two should be kept distinct. They are interesting as forming an obvious starting
point in the evolution of the great sub-family Chalininee.

Toxochalina robusta, Ridley.

1S84, Toxochalina robusta, Ridley (16).

There are a number of specimens of this sponge in the collection, agreeing closely,
both as regards external form and skeletal characters, with Ridley's description of
the tvpe from Port Jackson. It is perhaps noteworthy, however, that the megascleres

* Toi'sknt (48), [i. 78.
T 2

I ID CEYLON PEARL OYSTER REPORT.

are abundant between the stout horny fibres, as well as (sometimes) in their axes.
The species has also been recorded from off Bahia (" Challenger ").
R.N. 8, <J, 38 (all three from Gulf of Manaar) ; 351 (Ceylon seas).

Toxochalina robusta, var. ridleyi, nov. — Plate IX., fig. 2.

The type specimen is very irregular in shape ; massive and angular, with a slight
tendency towards branching. It has evidently been attached by a broad base to one
valve of a Lamellibranch shell (1 Margaritifera vulgaris), the impress of which is still
clearly visible. The surface is smooth and sub-glabrous, but uneven and very
distinctly granular when viewed under a lens. The vents are rather numerous (live),
about 4 millims. in diameter, with very prominent margins ; each is the opening of a
wide, deep oscular tube of the same diameter as itself. The texture (in spirit) is com-
pressible and resilient, but stiff and tough. Colour, pale brown. The specimen is
about G6 millims. long, 38 millims. broad, and 31 millims. high.

The main skeleton is a reticulation of horny fibre, with very few and slender
spicules (Plate IX., fig. 2). The primary fibres are very stout, sometimes as much as
0"164 millim. in diameter, but very variable ; typically they run at right angles to
the- surface and are united by short secondaries to form rectangular meshes, but the
network often becomes very irregular and the size of the meshes is very variable.
The secondary fibres are usually, but not always, more slender than the primaries.
The primaries are cored by a multispicular axis of slender oxea arranged in a plumose
manner (as in typical Axinellidas), but all entirely enveloped in spongin to such an
extent that the entire column of spicules only occupies one-third (or less) of the
thickness of the fibre. The secondary fibres contain only a few isolated spicules
arranged uniserially.

The dermal skeleton consists of a rather close-meshed reticulation of rather slender,
unispicular horny fibre. From the nodes of this reticulation brushes of oxeote spicules
project vertically. In certain places this dermal skeleton appears to become many
layers deep, and the vertical brushes of oxea are continued inwards as more or less
plumose columns enveloped in spongin and connected by numerous unispicular cross-
fibres, so as to form a close skeleton network beneath the surface, very consjucuous
in vertical section, and strongly contrasted with the much coarser, more widely
meshed and less abundantly spicular main skeleton below it.

Spicules. — (1.) Oxea (Plate IX., fig. 2, o) ; short and rather slender, slightly curved,
fairly gradually sharp-pointed at each end ; measuring about 0'08 millim. by 0"004
millim. near the surface, but usually smaller, and especially more slender, in the fibres
of the main skeleton. These spicules — at any rate, in the main skeleton — are
evidently becoming vestigial.

(2.) Toxa (Plate IX., fig. 2. t) ; slender, more or less strongly curved in the middle,
very slightly re-curved at the apices ; sometimes slightly roughened in the middle,
gradually sharp pointed at the ends; size varying up to about 0-08 millim. (in a

SPONGES. 141

straight Hue from end to end) by 0-002 niillim. (in the middle) ; very abundantly
distributed through the soft tissues in association with scattered oxea.
R.N. L09 (Gulf of Manaar. Type); 306 (Stat. XXIV., off Trincomalee).

Strongylophora, n. gen.

Gelliime with the skeleton composed of a reticulation of strongyla of various sizes,

partly collected in fibres, but with little (if any) spongin. With microscleres in

the form of smooth microxea, chiefly found in the dermal membrane.

This is a remarkable genus of somewhat doubtful systematic position. The

presence <>f the fusiform (and often angulated) dermal microxea suggests an affinity

with the Homosclerophora and Pachastrellidse amongst the Tetractinellida, rather

than with the Sigmatophora. We cannot, however, lay very much stress upon this

character when we remember the generalized character of microxea and the fact that

such spicules also occur in the Ectyonine genus Fusifer (vide Dendy, 10).

Strongylophora durissima, n. sp. — Plate IX., tig. 1.

Sponge massive, irregular ; may be depressed and cakedike or subcylindrical and
slightly ramose. Surface very uneven, sometimes with angular grooves and ridges,
giving it a curious crumpled appearance ; minutely and uniformly granular. Vents
few, scattered ; circular and often rather large, up to about 4 millims. in diameter ;
each the opening of a wide cylindrical oscular tube which runs vertically inwards for
a considerable distance. Inhalant pores minute, abundantly scattered in the meshes
of the dermal reticulation. Colour (in spirit) greyish-brown throughout. Texture
hard and stony, but brittle, and rather cavernous internally owing to the presence of
the numerous canals of varying diameter. The larger of the two specimens measures
about 45 millims. in maximum diameter.

The main skeleton is an irregular but fairly close-meshed and, towards the surface,
sub-rectangular reticulation of more or less stout spicular fibre composed of closely
packed strongyla, with numerous loose strongyla scattered between the fibres. The
dermal skeleton (Plate IX., fig. 1) is a reticulation of mostly large and single
strongyla ; their ends come in contact with one another, many together, at the
principal nodes of the reticulation, from which they radiate, and at each of these
nodes there is also a little heap of very short strongyla. The presence of these nodal
heaps gives the characteristic granular appearance to the surface of the sponge.

Spicules. — (1.) Strongyla (Plate IX., fig. 1, «.); usually stout, more or less curved
or angulated in the middle ; evenly rounded off at each end ; ranging in size from
about 0*026 niillim. by 0"006 niillim. (or perhaps even less) to about 0'26 millini. by
0'02 millini. (A few long and very slender spicules, oxeote and strongylote, occur in
the interior of the sponge ; they are probably abnormal forms of the strongyla, with
which they are connected by intermediates.)

('2.) Microxea (Plate IX., fig. I, •?>(.)'; fusiform and usually angulated in the middle,

142 CEYLON PEARL OYSTER REPORT.

from which they taper gradually to a very sharp point at each end ; size fairly
uniform, ahout 0-028 millim. by 0'002 millim. Very abundant in the thin transparent
dermal membrane in the meshes of the dermal skeleton ; also found less frequently in
the interior of the sponge.

R.N. 156; 244 (Ceylon seas).

Sub-Family : EENTERIN/E.

Haploscleridae in which the microscleres have entirely disappeared and the skeleton
consists of a reticulation of oxeote megascleres with little or no spongin.
Tli is sub-family is apparently derived from the Gelliinse by loss of microscleres.
Inasmuch, however, as the microscleres constitute the most important guides to the
classification of monaxonellid sponges, their total loss may in certain cases leave one
in considerable doubt as to the true systematic position of the species concerned. The
close relationship of Reniera to Gellius I take to be fully established by the form of
the megascleres and by their arrangement. In the genus Halichondria, on the other
hand, there appears to me to be less certainty, and the long, slender, slightly curved
form and confused arrangement of the oxea suggest a possible origin from some
astromonaxonellid genus, such as Coppatias, by loss of the astrose microscleres. It is
impossible in the present state of our knowledge to decide this question definitely,
but it is quite likely that, as regards the genera Reniera and Halichondria, we are
dealing with a case of convergent evolution rather than of close genetic relationship.

Reniera, Nakdo.

Renierinae in which the skeleton is composed of a close reticulation of usually single
megascleres, each forming one side of a rectangular, triangular or polygonal
mesh. Spicules short, oxeote or strongylote, usually united together at the ends
only by spongin cement.

Reniera implexa, Schmidt.

1868, Reniera implexa, Schmidt (50); 1887, Reniera implexa, Ridley and Dendy (1).

1 identify with this species a single small specimen consisting of a few irregularly
branched tubes, mostly about 5 millims. in diameter and widely open at the end.
The surface has a minutely reticulate or porous appearance. The colour (in spirit) is
brownish-yellow, and the consistence very soft, compressible and tender.

The skeleton is a rather irregular, triangular-meshed reticulation, for the most part
of single spicules, with occasional loose plurispicular bands feebly developed.

The spicules are slender, slightly curved oxea, gradually sharp-pointed at each end
and measuring up to about 0'13G millim. by 0'004 millim., usually perhaps a
little less.

This species has been recorded from the Adriatic by Schmidt, and from the Azores
by Ridley and Dendy (1) and Topsent (62).

R.N. 201 (Stat. LXIV., south of Modragam Paar, 5 fathoms, March 17, 1902).

SPONGES. 143

Reniera pigmentifera, n. sp. —Plate TX., fig. 10.

This species is represented in the collection by a large number of small fragments
which may perhaps represent only a single specimen. The external form appears to
have been more or less flabellate, with rounded margin. The vents are about
2 millims. in diameter and appear to have been scattered singly along the margin (and
elsewhere ?). The surface lias a porous appearance to the naked eye, and is very
minutely hispid. The colour (in spirit) is dark brown throughout, sometimes with a
purplish tint, and the texture is very soft and crumbling.

The skeleton is an irregular " Isodictyal" network of short spicules, sometimes con-
nected together at their ends by spongin cement ; primary fibres, from one to about three
spicules in thickness, are recognisable in places. There is no distinct dermal skeleton.

Spicules. — (1.) Oxea (Plate IX., fig. 10, a) ; slightly curved and gradually sharp
pointed at each end ; measuring about (M44 millim. by 0'007 millim., but often much
more slender.

(2.) Strongyla (Plate IX., fig. 10, h) ; stout, very slightly curved, broadly rounded
off at each end; variable in length, up to about 0"12 millim. by 0"009 millim. ; may
be shorter and at the same time somewhat stouter.

(3.) Styli (Plate IX., fig. 10, c, c/) ; short and stout, very similar in size and shape
to the strongyla, but pointed at one end. Of course, intermediate forms of spicules
also occur.

A remarkable feature of this sponge is the immense number of granular, brown or
purple-coloured pigment cells which it contains. These cells are rounded in outline
and about 0"002 millim. in diameter, and are thickly scattered all through the sponge.

P.N. 290 (numerous fragments. Jokkenpiddi Paar, 10 fathoms).

Reniera zoologica, n. sp. — Plate IX., fig. 8.

The single specimen appears to be half of a pear-shaped sponge which has been
torn in two longitudinally. It has probably been fixed by the narrower end, and
bears a single rather large vent (?) opening out of a wide oscular tube near the broad
upper end. The surface is encrusted with large Foraminifera and sand-grains. The
colour (in spirit) is pale grey, and the texture (internally) crumb-of-bread-like.
Total height of specimen 34 millims., greatest breadth about 20 millims.

The main skeleton is an irregular network of spicules, partly arranged singly in an
" Isodictyal " manner and partly collected in irregular multispicular bands. No
spongin cement is recognisable in my preparation.

The dermal skeleton is a dense but thin layer of oxea, lying very close together
and crossing one another in all directions parallel with the surface.

Spicules. — Oxea (Plate IX., fig. 8); moderately stout, slightly curved, gradually
sharp pointed at each end; size when fully developed about 0*18 millim. by
0-007 millim. Numerous very slender forms, probably young, also occur.

R.N. 262 (Gulf of Manaar).

144

CEYLON PEARL OYSTEE REPOKT.

Reniera, sp. ?

A small massive specimen with a couple of short, mammiform, vent-bearing
projections. Surface smooth. Colour (in spirit) pale yellow; texture compact and
firm, but brittle.

Main skeleton a compact, irregular, " Isodietyal " reticulation of spicules without
any fibres. Dermal skeleton a dense but thin layer of tangentially placed oxea
crossing one another in all directions parallel with the surface.

Spicules. — Oxea ; slightly curved, usually gradually sharp-pointed at each end ;
occasionally strongylote ; size about 0"164 millim. by 0"008 millim.

This species may possibly be identical with one of the numerous imperfectly known
European species of Reniera (Isodictya).

Tfc.N. 232 (deep water off Galle and onwards up West Coast of Ceylon).

Petrosia, Vosmaer.
Renierinae usually of hard or even stony texture, owing to the density of the skeleton,
which is composed of an irregular reticulation of oxeote or strongylote megascleres
(usually short and thick), packed close together, sometimes in stout fibres.
Those species of this genus which have a more or less fibrous skeleton make a close
approach to the genus Pachychalina,

Petrosia testudinaria (Lamarck).

[For literature and synonymy vide Dendy (3).]

This handsome sponge (see text-figure 1) has been recorded from Queensland (Ridley)

■$?**!%>*. .ifSsSc*

Fig. 1. Petrosia testudinaria (Lamk.), C4ulf of Manaar;
reduced one-half,

Fig. 2. Pearl-oyster shell honeycombed by
cliniiK marguriHfertB Dendy (see p. 128).

SPONGES. 14.i

and from the Mergui Archipelago (Carter, Dendy), as well as from the Gulf of
Manaar (Dendy).

R.N. 42 (Stat. LX1., Gulf of Manaar) ; 238 ? (deep water off Galle : fragments).

Petrosia similis, Ridley and Dendy.

1887, Petrosia similis, Ridley and Dendy (1).

There are in the collection a number of irregular, massive or more or less lohate
specimens, with numerous rather large, scattered vents, which I refer to this species.
The spicules measure about 0'22 millim. by 0"014 millim. The character of their ends
varies greatly, from strongylote to apiculate and sharp-pointed.

The species was originally obtained by the " Challenger" from south of the Cape of
( J< >od Hope and between Kerguelen and Heard Islands.

R.N. 12, Gl (both from Gulf of Manaar) ; 289; 327 (Ceylon seas).

Petrosia similis, var. delicatula, now

This variety is distinguished from the typical form of the species by its more
delicate texture and smaller vents, which are usually more or less blocked up by a
strong development of gelatinous tissue. The spiculation differs little, if at all, from
that of the types.

R.N. 84, 133, 276 (all from deep water off Galle and onwards up West Coast).

Petrosia similis, var. halichondrioides, nov.

In external appearance this variety closely resembles P. similis, var. delicatula.
The single specimen is strongly compressed, with the rather small but conspicuous and
widely open vents placed on prominent ridges. The texture (in spirit) is firm and
compact, but brittle. The skeleton is very confused and dense, without distinct
fibres, and the spicules are much more slender and Halichondria-like than in the
typical form, being gently curved oxea, for the most part gradually sharp-pointed at
each end, and measuring, say, about 0"2 millim. bv 0'006 millim.

R.N. 79 (Gulf of Manaar).

Petrosia densissima, n. sp. — Plate IX., fig. 9.

Sponge massive, sub-conical, attached by the broad base ; with rather irregular surface
bearing feebly developed, meandering grooves. Surface minutely granular, without
distinct dermal membrane, slightly sandv. Vents very small and scattered Colour
(in spirit) rather dark greyish-brown on the surface ; pale yellowish-grev internally.
Texture extremely compact and hard ; stony ; incompressible. The larger of the two
specimens (R.N. 138a) measures al>out 40 millims. in height by 42 millims. in greatest
breadth.

The skeleton is extraordinarily dense and compact, consisting of a sub-rectangular-
meshed reticulation of very stout spicular fibres, in which both primary and secondary
fibres are about as thick as the width of the meshes between them, say about

u

146 CEYLON PEARL OYSTER REPORT.

(VI 3 millim. The fibres are compact, but the arrangement of the spicules in them is
very confused and there is no visible spongin. Many spicules occur scattered
irregularly between the fibres, so that the whole skeleton forms an almost solid mass
of spicules. There is no special dermal skeleton.

Spicules. — Stout, fusiform oxea (Plate IX., fig. 9); slightly curved and usually
sharply and fairly gradually pointed ; size when fully developed about 0"24 millim. by
0"02 millim., but with numerous smaller forms which are presumably young.

This species is evidently very closely related to Thiele's Petrosia imperforata from
Celebes (39).

R.N. 138, 138a (both from deep water off Galle and onwards up West Coast of
Ceylon).

Halichondria, Fleming.

Renierinse in which the skeleton consists of a confused reticulation of long and slender
oxea (or strongyla) with little or no spongin ; the spicules sometimes associated
in ill-defined bands or fibres.

I have already indicated the doubt which exists as to the true relationship of this
genus. Possibly it is, as at present understood, of polyphyletic origin, including
species derived from several ancestral forms by loss of microscleres.

Halichondria panicea, Johnston.

[For literature and synonymy vide Ridley and Denuy (1) and Dendy (2).j
This widely distributed species is represented in the collection by two well
differentiated form-varieties, so that it seems desirable to distinguish them by varietal
names. In both varieties many of the more superficial oxea are arranged more or less
at right angles to the surface, with their apices projecting to a greater or less extent,
thus making an approach to the genus Trachyopsis. In both the full-grown spicules
measure up to about l'O millim. in length and are of the usual Halichondria type.
Their arrangement in the interior of the sponge is quite irregular and confused.

Halichondria panicea, Johnston, var. megalorhaphis, Carter.
1881, Amorphina megalorhaphis, Carter (5).
This variety is irregularly encrusting, growing out into lobose or digitiform
processes, and with small scattered vents.

R.N. 87, 231, 248 (all from deep water off Galle and onwards up West Coast).

Halichondria panicea, Johnston, var. heinispherica, nov.
Tins variety is massive and compact, more or less hemispherical or cushion-shaped,
with vents usually arranged in conspicuous groups on the convex upper surface. It
attains a considerable size, the largest specimen measuring about 100 millims. in
greatest diameter and about 40 millims. in thickness in the middle.

SPONGES. 147

R.N. 67 (Gulf of Manaar); 96 (Gulf of Mannar, dry); 141 ?, 142 (both from deep
water off Galle and onwards up West Coast) ; 249 (Stat. XV., Periya Paar).

Trachyopsis, n. gen.

Renierina? in which the main skeleton is composed of a dense, irregular network of
oxea, while the surface is protected by similar (or perhaps more slender) spicules
arranged in dense vertical brushes, which support the pore-bearing dermal
membrane.

This genus is of somewhat doubtful systematic position ; in certain features it
recalls the genera Trachya and Spongosoritcs, and it differs from typical Renierhw
in the replacement of the reticulate dermal skeleton characteristic of that group by
radially arranged brushes of oxea.

Trachyopsis halichondrioides, n. sp. — Plate X., fig. 10.

Sponge massive (or thickly encrusting ?) ; upper surface slightly convex, rising up
at irregular intervals into a few short, thick-walled, cylindrical, tubular processes,
each terminated by a single circular vent. General surface smooth and almost
glabrous, but uneven ; very minutely reticulate as seen under a lens. Inhalant pores
minute and scattered. Colour in spirit, pale yellowish-grey. Texture of body hard
and compact, penetrated by numerous narrow vertical canals. Greatest diameter of
specimen, which is irregular in outline, 44 millims. ; thickness about the middle
15 millims. (but the specimen has evidently been cut off from its base). Height of
largest projection about 8 millims.; diameter in the middle about 5 '5 millims.;
diameter of the vent at its apex 2'5 millims.

The main skeleton is an extremely dense and very irregular reticulation of stout
oxea, with a tendency to arrange themselves in ill-defined tracts running towards the
surface. Immediately beneath the surface the oxea, here perhaps somewhat more
slender than usual, are arranged in dense brushes perpendicularly to the dermal
membrane, beyond which their apices may project very slightly.

Spicules. — Oxea (Plate X., fig. 10); short, usually stout, sub-fusiform, gently
curved or (often) biangulate, fairly gradually and sharply jjointed at each end ; size,
when fully developed, about 0-64 millim. by 0'0328 millim. ; frequently more slender.

R.N. 147 (deep water off Galle and onwards up West Coast of Ceylon).

Sub-family : CHALININ.K.

Haplosclerida? without microscleres and with diactinal megascleres. Skeleton a

network of more or less strongly developed horny fibre cored by megascleres.

It is highly probable that this sub-family is of polyphyletic origin, being derived

from several genera of Gelliinre and Renierina? by loss of microscleres and strong

development of spongin. Some species have probably been derived from Toxochalina

0 2

I4« CEYLON PEARL OYSTER REPORT.

and Gelliodes simply by loss of microscleres, the horny fibre being already strongly
developed in those genera, while others have probably arisen from Reniera and
Petrosia simply by strong development of the horny fibre, the microscleres having been
already lost.

The excessive development of spongin appears to have taken place independently
in many genera, and this fact, coupled with the loss of the characteristic microscleres
and the uniform character of the megascleres, renders it extremely difficult to arrive at
a natural classification of the Chalininae.* The subject is, however, much too complex
to he discussed here at length, especially as there are not a very large number of
species in the collection. For our present purposes it will suffice to make use of the
established genera, Pachychalina, Chalina, Ceraochalina. and Siphonochalina,
without committing ourselves to an expression of opinion as to their genetic relation-
ships. Owing to their degenerate character, it is impossible to define even these in
such a way as to distinguish them quite sharply from one another.

Pachychalina, Schmidt.
Chalininee of various external form, lobose or digitate, not tubular ; with stout
skeleton fibres, containing very numerous well developed spicules arranged
multiserially.

Pachychalina subcylindrica, n. sp. — Plate X., figs. 1, 2.
Sponge elongated, rather slender, irregularly cylindrical or angular, probably
branched and repent. Surface fairly smooth but uneven, with a very few coarse
aculeations ; minutely reticulate to the naked eye. Vents fairly large (about
2'25 millims in diameter), irregularly scattered, with slightly prominent margins.
Colour (in spirit) light brown. Texture compressible, resilient, rather coarsely fibrous,
but somewhat fragile. The largest piece measures about 95 millims. in length, with a
very variable thickness up to about 9 millims.

The main skeleton is a sub-rectangularly meshed network of very stout multispicular
fibre, about 0*066 millim. in diameter ; with meshes varying greatly in size, and with
numerous spicules scattered irregularly between the fibres. The fibres themselves
contain a very large number of spicules, but no visible spongin. The dermal skeleton
(Plate X., fig. 1) is an irregular, polygonal-meshed network of similar coarse multi-
spicular fibre.

Spicules. — Oxea (Plate X., fig. 2) ; more or less curved or angulated ; when fully
developed stout and very sharply pointed at each end ; measuring about 0'14 millim.
by 0"008 millim. Numerous slender forms also occur, probably immature.

In the feeble development of the spongin this species occupies an intermediate
position between the genus Petrosia and the more typical Chalinina?.

P.N. 292 ; 360 (Stat, II., north of Negombo, 9 fathoms).

* Compare LENDENFELD (51) and Dexdy (63). In the paper referred to I have explained the reasons
why I cannot accept LeNDENFELP'S classification of the Chalinina'.

SPONGES. 149

Pachychalina delicatula, Dendy.

1889, Pachychalina delicatula, Dkxdy (3).

With this species I identity three specimens, all characterized by their great
softness and delicacy of texture, but all containing a good deal of sand.

K.N. 55, -26-1 (both from Gulf of Manaar) ; 364 (Stat. 1., hauls I 4. January 31,
L902, Colombo to Negombo, 12 to 20 fathoms).

Pachychalina brevispiculifera, n. sp. — Plate X., tig. 7.

The single specimen is compressed, digitate to flabellate (presumably erect),
branching and anastomosing. The branches or fronds are sometimes narrow and
sometimes broad, but always greatly flattened, and only about 6 millims. in thickness.
The surface is beset with small conical aculeations, which form the principal nodes in
a very strongly developed dermal skeletal reticulation. Vents rather large (about
4 millims. in diameter) but very shallow ; numerous, but confined almost or cpuite
entirely to one of the flattened sides of the frond or branch. Colour (in the dry state)
light brown ; texture coarsely fibrous, compressible, resilient, fragile. The single
specimen measures about 120 millims. in height by 110 millims. in greatest width.

The main skeleton is a very coarse, sub-rectangularly or irregularly meshed network
of very stout horny fibre almost filled with well developed and very abundant spicules
arranged multiserially in all the fibres. The primary fibres, running lengthwise
through the branches, measure up to about 0'164 millim. in thickness, and the
secondaries are sometimes nearly as stout, though usually a good deal slenderer.
Sometimes two or more primary fibres run close together side by side, connected with
one another at frequent intervals by numerous very short secondaries. The dermal
skeleton is a coarse, irregular reticulation of similar fibre, varying greatly in thickness
and with meshes of varying diameter. The fibres, especially those of the dermal
skeleton, are occasionally echinated by projecting oxea.

Spicules. — Oxea (Plate X., fig. 7) ; slightly curved; sub-fusiform; short, stout and
sharp-pointed at each end; measuring about 0-l millim. by 0'0055 millim.

R.N. 110 (Gulf of Manaar, dry).

Pachychalina spinilamella, Dendy — Plate VII., fig. 4.
1889, Pachychalina spinilamella, Dendy (3).

I identify with this species a number of specimens of somewhat variable external
form, but all characterised by their strongly conulose surface and with closely similar
skeleton arrangement. A fairly typical example is represented in Plate VII. , fig. 4.
R.N. 14. 296 and 326 are characterised by their more slender, irregularly branching
form and smaller vents.

U.N. 14. 25, 94 (Periya Paar, &c., Gulf of Manaar) ; 172, 296, 326 (Ceylon seas).

I 50 CEYLON PEAEL OYSTER REPORT.

Chalina, Grant.

Chalininse of various external form; not tubular. Skeleton reticulation typically
rectangular; fibres usually slender, with much spongin and few but usually well
developed spicules.

Chalina subarmigera (Ridley) — Plate X., fig. 5.

1884, Cladochalina subarmigera, Ridley (16); 1887, Chalinopora subarmigera, Lrnden-
FELD (51); 1898, Chalina subarmigera, Lindgren (86).

This species is represented in the collection by two specimens, which agree very
well with Ridley's original description, except that the margins of the vents are
slightly prominent and there are fewer spicules in the fibres of the main skeleton.
The characteristic external form is shown in Plate X., fig. 5.

The species was obtained by the "Alert" in Torres Straits and at Albany Island
(north coast of Australia), and has also been recorded by Lexdrnfeld from Port
Jackson, and by Lindgren from the Coast of Cochin China.

R.N. 116 (Gulf of Manaar, dry) : 288 (Ceylon seas).

Chalina obtusispiculifera, n. sp. — Plate X., fig. 9.

Sponge elongated, slender, cylindrical ; may be irregularly branched (? erect or
repent). Surface even, very minutely hispid in its present condition. Texture (in
spirit) soft and resilient, but fairly tough. Colour pale yellowish-brown. Vents and
pores not seen. The largest specimen (R.N. 370) is about 130 millims. long by
3 "5 millims. in diameter.

The skeleton is a well- developed reticulation of pale-coloured horny fibre cored by
strongyla. The principal fibres run lengthwise through the sponge, branching as they
go, and the branches curve outwards towards the surface ; they are connected by
short secondary fibres to form an irregular network. The principal fibres are about
0*04 millim. in diameter and contain many spicules arranged multiserially as well as
much spongin extending well beyond the spicular core. The secondary fibres are
only about half as thick and contain fewer spicules. There is no specially
differentiated dermal skeleton, unless we consider the outermost secondary fibres of
the main skeleton as such (Plate X., fig. 9).

Spicules. — Cylindrical strongyla (Plate X., fig. 9) ; broadly rounded off at each end,
never pointed; nearly straight; measuring up to about 0-12 millim. by 0-007 millim.,
lint frequently nmch more slender.

This species is easily recognised by its external form and blunt cylindrical spicules.
Both specimens are more or less washed out and contain numerous foreign spicules
and other debris. In R.N. 285 none of the spicules appear to attain as great a
thickness as that given above for the type.

R.N. 285 (deep water off Galle and onwards up West Coast); 370 (deep water
outside pearl banks, Gulf of Manaar).

o

SPONGES. 151

Chalina clathrata, n. sp. — Plate X., fig. 3.

Sponge massively encrusting, the single specimen being attached to the valve of a
Pinna ; clathrous ; with very uneven surface proliferating into numerous small, blunt
outgrowths. Vents numerous and large (up to about 8 millims. in diameter), scattered
singly, each at the end of a short tubular projection and forming the termination of a
wide cylindrical oscular tube. Surface minutely reticulate. Texture very delicate,
soft, compressible, resilient : colour (after drying) light yellowish-brown, with a tinge
f purple. The single specimen measures about 220 millims. in maximum diameter.

The main skeleton is a sub-rectangularly or irregularly meshed network of very
pale-coloured horny fibre. The fibre varies greatly in diameter (averaging, say, about
0'025 millim.) and contains very few spicules, arranged for the most part uniserially
and absent altogether in places. The dermal skeleton is a close, polygonally or
rectangularly meshed network of horny fibre cored by uniserially arranged spicules ;
the fibres being about 0'0165 millim. in diameter and the meshes about 0"1 millim. in
diameter.

Spicules. — Very slender, usually slightly curved oxea or strongyla, measuring about
0'112 millim. by 0-002 millim., occurring in and between the fibres and often reduced
to vestiges (Plate X., fig. 3).

With its large prominent vents and deep oscular tubes this species makes an
approach to the genus Siphonochalina, while its massive (though clathrous) form
recalls Ridley's Acervochalina.

R.N. 102 (Gulf of Manaar, dry).

Chalina cymaeformis (Esper ?).

1 1798-1806, Spongia cymaeformis, Esper (6); ? 1870, Spongia cymaeformis, Ehleks (58).

Sponge shortly stipitate, bushily lamellar or frondose or sub-digitate. Lamellae
about 9 millims. thick, with broadly rounded margins. Surface rather uneven,
minutely conulose, especially where the dermal membrane has been rubbed oft'. Vents
small (about 2 millims. in diameter), more or less abundantly scattered, chiefly on the
inner surfaces of the lamellse. Inhalant pores scattered in the dermal membrane.
Texture (in spirit) soft and compressible, but very tough and resilient ; rather woolly.
Colour brown. The most typical specimen (R.N. 16) is about 67 millims. high by
90 millims. in greatest breadth, with a stalk about 22 millims. high and 18 millims.
thick. It bears a very close resemblance to the figure of Chalina palmata, given by
Ridley and Dendy (1), as well as to Esper's figure of his Spjongia cymceformis
(Plate 69).

The main skeleton consists of what, at first sight, look like rather slender, ill-defined,
plurispicular fibres running towards the surface at irregular intervals and branching
as they go, connected by still less well-defined secondary fibres from one to about four
spicules broad. No spongin is at first sight visible, but closer examination shows that
a very large quantity is really present in the fibres, more or less completely imbedding

152 CEYLON PEAKL OYSTEE REPOKT.

the spicules; owing to its very pale colour and great transparency, however, it
readily escapes ohservation. Numerous spicules occur scattered irregularly between
the fibres. There is no special dermal skeleton, the dermal membrane being practically
destitute of spicules.

Spicules. — Slender, slightly or rather strongly curved oxea ; more or less gradually
sharp-pointed at the ends; size variable, say about 0'12 niillim. by 0"003 millim.

It is probable that this species is identical with Esper's Spongia cymoeformis (from
Ceylon). It also appears to be nearly related to Chalina palmata from European.
Indian and Australian waters ; differing, however, in the absence of the dermal
skeleton reticulation. In habit the species reminds one very much of some species of
A.chieJla, a resemblance which appears from Ehlers' description to have been
increased in the case of Esper's specimen by the presence of styli mingled with
the oxea.

RN. 16 (Gulf of Manaar) ; 349 (Stat. LV., west of Periya Paar, 20 fathoms).

Ceraochalina, Lendenfeld.

Chalininse of various external form ; not tubular. Texture hard, owing to the great
thickness of the skeleton fibres, in which the spongin is very strongly developed
and the spicules much reduced in size and sometimes also in number.

Ceraochalina retiarmata, n. sp. — Plate X., fig. 4.

The single specimen forms an erect, thin lamella, attached below by a constricted
base ; sub-dividing into flattened branches and giving off irregular digitiform
processes almost exclusively in one plane ; the branches to a slight extent anastomosing
with one another. Surface glabrous, minutely granular under a lens; rather uneven.
Vents minute, about O'u millim. in diameter, mostly arranged uniserially on the
narrow margins. Inhalant pores scattered in the meshes of the dermal reticulation.
Colour (in spirit) rather dark brown ; texture compressible, resilient, tough and
fibrous. Total height of specimen about 33 millims. ; greatest breadth about
66 millims. ; thickness of lamella about 3 '5 millims.

The main skeleton is a network of stout horny fibre, sparingly cored by very
slender vestigial oxea. The primary fibres are about O'l niillim. in diameter and
curve upwards and outwards towards the surface, branching as they go. They are
connected together by short secondaries about 0-05 millim. in diameter and containing
fewer spicules. There is also a system of tertiary fibres, much more slender (from
about U"008 millim. to about 0'024 millim. in diameter) and containing from one to
about four rows of well-developed oxea imbedded in spongin. These tertiary fibres
form an irregular network which seems to bear no relation to the rest of the main
skeleton, except that its fibres are attached frequently to those of the latter.

The dermal skeleton is very strongly developed, forming a close polygonal-meshed
reticulation of horny fibres cored by usually two or three rowrs of well developed oxea

SPONGES. L53

and echinated abundantly by similar oxea projecting from the fibre singly or in small
groups. The diameter of the dermal fibre is about 0'024 millim. ; of the meshes
between the fibres about O'lG millim., but variable.

Spicules. — Rather short, slightly curved, gradually sharp-pointed oxea (Plate X.,
fig. 4), measuring about 0"084 millim. by 0*004 millim. in the dermal skeleton and in
the tertiary fibres of the main skeleton, but becoming more or less vestigial in the
other fibres.

R.N. 342 (Stat. V., ofFChilaw, 10 fathoms).

Ceraochalina reticutis, u. sp. — Plate X., fig. 8.

The type specimen (R.N. 58) forms an agglomeration of short, irregular, sub-
cylindrical or angular branches, slightly anastomosing with one another and branching
with great irregularity. The branches vary greatly in diameter, from about 4 millims.
to about 11 millims., and have a slightly nodose appearance. The entire mass
measures about 80 millims. in greatest breadth. Vents small (about 1 millim. in
diameter), but conspicuous and with slightly projecting margins ; scattered abundantly
and sometimes in ill-defined longitudinal series. The dermal membrane is parchment-
like, and under a pocket lens appears very finely and regularly reticulate in triangular
meshes, which are the coarser meshes of the dermal skeleton composed of dark brown
spongin fibre. Texture (in spirit) compressihle, very resilient, tough ; colour dark
brown.

The main skeleton is a fairly regular, sub-rectangularly meshed network of strong
horny fibre. The principal fibres run longitudinally through the branches, sub-
dividing as they go and curving outwards to the surface. They measure up to about
0-08 millim. in diameter, and contain a considerable number of slender vestigial
spicules, irregularly and multiserially arranged. The secondary fibres are about as
thick as the primaries, but contain fewer spicules (which are also vestigial). The
meshes of the main skeleton reticulation vary a good deal in size ; averaging, say,
about 0*33 millim. in diameter.

The dermal skeleton (Plate X., fig. 8) is a very well developed, close, polygonal-
meshed reticulation of horny fibre containing only a very few slender spicules scattered
here and there. The fibres of which this reticulation is made up are of two principal
sizes : («) stout, about 0'03 millim. to 0"05 millim. in diameter, radiating from the
ends of the primary fibres of the main skeleton and forming the coarser triangular-
meshed reticulation visible under a pocket lens; (b) more slender, hut very variable
in diameter, forming a very close-meshed but irregular reticulation in the meshes of
the coarser reticulation. A large number of well developed spicules may be irregu-
larly scattered in the dermal membrane outside the horny fibres of the dermal
skeleton.

Spicules.— Oxea ; varying greatly in degree of development : in the horny fibres
very slender and vestigial (Plate X., fig. 8, v.s.), but often well developed in the soft

x

i;,4 CEYLON PEARL OYSTER REPORT.

tissues between the fibres, especially in the dermal membrane (Plate X., fig. 8, o).
When fully developed they are slightly curved and somewhat hastately sharp -pointed,
measuring about 0-072 millim. by 0*00265 millim.

R.N. 58 (Gulf of Manaar) ; 321 ; ? 346 (distinguished from the type chiefly by the
numerous well developed oxea arranged uniserially in the slenderer fibres of the
dermal skeleton and multiserially in the stouter fibres ; while the spicules themselves
are occasionally strongylote).

Ceraochalina multiformis, Lendenfeld, var. manaarensis, Dendy — Plate VII., fig. 2.
1889, Pachychalina multiformis, var. manaarensis, Dendy (3).

I identify with this variety a single specimen attached to a pearl oyster, which is
represented in Plate VII., fig. 2. A feature which I omitted to mention in my
original description of the variety is the echination of the fibres of the dermal
skeleton by projecting oxea, singly or in groups.

The species was recorded by Lendenfeld (51) from Australia and New Zealand.

K.N. 98 (Gulf of Manaar, dry).

Ceraochalina ceylonica, n. sp. — Plate VII., fig. 3 : Plate X., fig. 6.

Sponge (Plate VII., fig. 3) massive, irregular, sometimes clathrous, with a slight
tendency to become lobose or digitate. Surface strongly and copiously aculeated
by sharp-pointed conuli about 4 minims, in height ; minutely fibro-reticulate between
the conuli. Vents numerous, scattered, about 4 millims. or 5 millims. in diameter ;
the openings of deep, cylindrical oscular tubes. Colour (in spirit) pale brown ;
texture firm and tough, but compressible and resilient. The largest specimen
(K.N. 5, figured) measures 200 millims. by 1G0 millims. in horizontal dimensions, by
90 millims. in height.

The main skeleton is a very strongly developed, irregular or rectangularly meshed
network of stout, horny fibre, containing usually a large number of small oxea
scattered irregularly throughout the spongin substance, but almost always more or
less jjarallel with the long axis of the fibre. Occasionally (K.N. 5) the spicules are
much less strongly developed and may be completely absent from some of the fibres.
Numerous spicules also occur scattered between the fibres. The thickness of the
fibres is variable, say about 0*08 millim. for the primaries and not much less for the
secondaries

The dermal skeleton is a polygonally meshed network of fibre containing a very
large proportion of spongin and a good many spicules. The latter are, for the most
part, imbedded in the spongin substance more or less longitudinally; at frequent
intervals, however, little groups of oxea, or single spicules, project more or less at
right angles from the fibre in an echinating manner, and thus give a rather
characteristic appearance to the dermal skeleton. The meshes of the dermal

SPONGES. 155

reticulation are about 0*2 millim. in diameter, and the fibres from about 0'008 millhn.
upwards.

Spicules. — Slender oxea (Plate X., fig. 6), slightly curved and more or less gradually
and sharply pointed : size about 0*088 millim. by 0"003 millim., but subject to a good
deal of variation, and frequently, if not usually, more slender.

R.N. 4. 5, 50, 108, 113 (all from Gulf of Manaar).

Siphonochalina, Schmidt.

( ihalininse of tubular form. Tubes smooth, both inside and out, usually narrow ; each
with a large circular vent at the summit.

Siphonochalina communis (Carter), var. tenuispiculata, nov. — Plate VII., fig. 1.

This variety (Plate VII., fig. 1) agrees very closely in external form with the
specimens of Siphonochalina communis described by Carter (5) and myself (3) from
the Gulf of Manaar. Mr. Carter gave no measurements of the spicules in the case
of the type of the species, but the specimens in Professor Herdman's collection
differ rather strikingly from« that collected by Mr. Thurston and described by
mvself, in that the spicules are very much more slender and very much more
numerous in both primary and secondary fibres and in the fibres of the dermal
skeleton. In fact, the spicules, though very abundant in all the fibres and occurring
throughout the entire, or almost the entire, thickness of each fibre, are so slender as
to be almost vestigial, measuring about 0-072 millim. by O'OOl millim.

In skeletal characters, although the spicules are more slender, this variety agrees
much more closely with my Siphonochalina crassifibra from the same locality (3)
than with Carter's S. communis; differing from S. crassifibra chiefly in the smaller
size of the tubes. Thus it has the external form of S. communis combined with the
skeletal characters of S. crassifibra, and I therefore propose to regard all three forms
as mere varieties of one and the same species.

Professor Herdmax informs me that the sponge in life had a violet-pink colour.

Ridley (16) records the species (under Carter's name Tabvlodigitus communis)
from Port Jackson, Australia, and also from Kurrachee.

R.N. 6, 7, 117 (dry, figured; all from Gulf of Manaar, Stat. II., 8 fathoms).

Sub-family : DESMACELLIN^E.
Haploscleridns with monactinal megascleres. Microscleres various.

Desmacella, Schmidt.
Desmacellinre with reticulate skeleton composed of styli or tylostyli. Microscleres
sigmata, toxa and trichodragmata variously combined.

Desmacella tubulata, n. sp. — Plate IX., fig. 4.
Sponge consisting of cylindrical, tubular processes, more or less widely open above

x 2

L5(j CEYLON PEAEL OYSTER REPORT.

(? always) and (sometimes, at any rate) united together below. (Possibly they may
have been attached to a common body, but only fragmentary tubes are present in the
collection.) Diameter of individual tubes about (! millims. ; thickness of tube-wall
about 1 "25 millims. Outer surface slightly granular and very minutely hispid. Inner
surface with numerous very minute openings of exhalant canals. Colour (in spirit)
pale greyish -yellow. Texture very soft, compressible, fragile.

The main skeleton is a very irregular but close reticulation of slender stvli, either
isolated or in loose bundles : with no visible spongin cement. There is no special
dermal skeleton.

Spicules. — (1.) Long slender styli (Plate IX., fig. 4, a, h) ; slightly curved or bent,
broadly and evenly rounded off at the base, sharply and more or less gradually
pointed at the apex ; size about 0*28 millim. by 0'005 millini.

(2.) Trichodragmata (Plate IX., fig. 4, d); extraordinarily abundant, especially
beneath the outer surface, and very variable in size, sometimes forming wisp-like
fibres, sometimes breaking up into separate microxea (Plate IX., fig. 4, e) ; varying in
length from about 0-02 millim. upwards, and always very slender.

(3.) Sigmata (Plate IX., fig. 4, c) ; also extraordinarily abundant ; slender, com-
monly much contort, also C-shaped ; occasionally in small bundles (sigmodragmata) ;
length from bend to bend variable, say about 0'02 millim.

This appears to be a very well characterized species, and I know of no other which
comes very near it. The immense number of microscleres is very remarkable.

P.N. 209 (Gulf of Manaar); 324.

Sub-family : HETEKOXYLVE.

Haploscleridie with a dense cortex composed of radially arranged megascleres.
Megascleres smooth and spined oxea. Microscleres present or absent.
1 propose this sub-family for the reception of the genera Heteroxya, Toi*sent, and
Acanthoxifer, n. gen., the former of which is, apparently with very slight justification,
placed by its founder amongst the Tethyida?.

Acanthoxifer, n. gen.

Heteroxyinse with a dense spicular cortex broken up into polygonal plates by pore-
hearing grooves. Main skeleton a confused reticulation of oxea. Cortical
skeleton composed chiefly of dense brushes of oxea arranged at right angles to
the surface. Megascleres smooth and spined oxea. Microscleres trichodragmata.

This remarkable genus is evidently nearly related to Topsent'r Heteroxya (45), but
differs in several respects, notably in the presence of trichodragmata" which indicates
that the true position both of Acanthoxifer and Heteroxya is amongst the Haplo-
scleridse and not amongst the Tethyida;, where TorsENT has placed Heteroxya. The
presence of the spined oxeote megascleres suggests a possible relationship to the
Spongillinse,

SPONGES. 157

The breaking up of the cortex into polygonal plates or nodules by pore-bearing
(and '. vent-bearing) grooves remind one forcibly of the genus Placospongia amongst
Spirastrellidse, but the resemblance is entirely superficial.

Acanthoxifer ceylonensis. n. sp. — Plate IX., fig. 5.

Sponge massively encrusting, irregular, with flattened or convex upper surface and
broad base of attachment. Surface very minutely hispid or granular, uneven, nodular
or tubercular, the nodules or tubercles being very low and roundedly polygonal in
shape, separated by grooves of varying distinctness. Generally the grooves are broad
and shallow; sometimes they are narrow, with prominent margins. Diameter of
nodules, say, about 4 millims., but variable and irregular. Colour (in spirit) light
brown. Texture compact, fleshy, but with much calcareous debris embedded. Pores
and vents not recognisable externally. The largest specimen is cake-shaped, about
42 millims. in horizontal diameter, and up to 17 millims. in thickness.

The main skeleton is a quite confused, lax reticulation of very abundant, long, oxeote
megascleres, occasionally collected together into loose, ill-defined bands, but without
any distinct fibre, and with a little spongin becoming visible in stained sections. The
cortical skeleton is very strongly developed and consists of dense brushes of oxea
placed side by side and lying at right angles to the surface, beyond which their apices
project (but are now nearly all broken off). These brushes are backed up internally
by a very dense, irregular reticulation of oxea. The cortical skeleton is absent
beneath the grooves which separate the surface nodules. Thus there is a separate
section of the cortical skeleton for each nodule or tubercle, thinning out as it
approaches the grooves in such a manner as to become strongly convex on the inner
aspect.

Spicule*. — (I.) Smooth oxea (Plate IX., fig. 5, a) ; long and rather slender, slightly
curved or bent, variously and often irregularly ended, sometimes stylote (Plate IX.,
fig. 5, J>). Size variable, say about 0'74 millim. by 0"012 millim. Abundant in the
main skeleton and occasionally met with in the surface brushes.

(2.) Spined oxea (Plate IX., fig. 5, c) ; usually slightly and symmetrically curved
(or angulated) in the middle ; tapering fairly gradually to a sharp point at each end ;
provided with numerous very minute, sharp spines, most abundantly developed
towards the two ends; size about 0'38 millim. by 0'008 millim. Characteristic of the
cortical skeleton, but also common in the interior of the sponge.

(3.) Trichodragmata (Plate IX., rig. 5, d) ; oblong bundles of extremely slender
rhaphides, which do not usually become dissociated even on boiling with nitric acid.
Size usually about O'Olfi millim. by 0'004 millim. : occasionally much longer.

Stained sections show that the ectosome (between the spicules) is chiefly collenchy-
matous, sometimes with a tendency to become fibrous near the surface. In both
ectosome and choanosome are an immense number of minute granules of a pale

158 CEYLON PEARL OYSTER REPORT.

yellowish colour, aggregated in rounded masses of very varying size. These may he
symbiotic al^ae.

R.N. 213, 217 (both from deep water outside pearl hanks, Gulf of Manaar) ; 247.

Family : DESMAOtnONTP/F,

Sigmatomonaxonellida in which some of the microscleres are cheki? (except when
these have been lost by degeneration).

The presence of microscleres in the form of chelae constitutes a natural character by
which the Desmacidoniche are, as a rule, easily distinguished from all other sponges.
Unfortunately, however, the chelae are verv apt to disappear by degeneration,
especially in the sub-families Ectyonina? and Phlo?odictvinae, and we have then to
depend upon other characters — such as the presence of spined echinating styli — for
guidance in classification. That the chela, one of the most remarkable forms of
microsclere known to us, has originated by modification of the sigma, there can, I
think, be little douht. In the Report on the " Challenger" Monaxonida* we showed
that in ontogeny the chela arises from a sigmoid form (Esperetta mammiformis) and
that sigmata and chelae must therefore be grouped in the same category. Nor are
intermediate forms of adult spicules unknown to us, such as the curious " bipocilli " of
the genus Iophon, especially those of Iophon chelifer, Ridley and Dendy (1), and,
most notable of all, the bidentate sigmata of Topsent's Gellivs bidens (64).

We are, therefore, justified in regarding the Desmacidonidae as derived from the
Haploscleridae by modification of the sigmoid microscleres into chelae, though why this
modification should have taken place is verv hard to understand. It is extremely
difficult to see how the verv peculiar and highly specialized chelate form of microsclere
can he of any special advantage to its possessor, and we have here one of those
numerous cases in which, so far as we can see at present, the theory of natural
selection signally fails to account for the facts.

I have already pointed out that the sub-family Phloeodictyinae must be transferred
to the Desmacidonidae on account of the presence of chelate microscleres in the
genus Histoderma, certain species of which are obviously very closely related to
Phlceodictyon and Oceanapia. For the purposes of this Report the three sub-families
Esperellinae, Phlceodictyinae and Ectyoninae will be sufficient. TorsENT's sub-family
Dendoricinae appears to me to be an unnatural group which cannot he maintained, the
differentiation of the ectosomal megascleres being far too general and widespread a
character to be utilised as distinguishing the sub-family, whose members fall very
naturally in one or other of the remaining sub-families. The sub-family Buharinse,
proposed by the same author, has, I am glad to see, heen again abandoned by him in
his latest work (62).

* Page xx.

SPONGES. 15(J

SUB-FAMILV : ESrEIlELLIN.E.

DesinacidonkLe without echinating spicules, and without fistular outgrowths of the
sponge body.

Esperella, Vosmaer.

Esperellinse of various external form, usually massive, lobose or ramose. Skeleton
usually fibrous, often with much spongin. Megascleres monactinal, stylote or
tylostylote. Characteristic microscleres palmate anisochelre, to which may be
added smooth sigmata, toxa, trichodragmata and small isocheke* in various
combinations.

Esperella parishii (Bowerbank ?), Ridley.

[For synonymy and literature ride Ridley and Dendy (1), p. 65.]

There is one small, thinly encrusting specimen of this sponge in the collection,
growing on a calcareous nodule in association with Paresperclla serratohamctta and
Hymedesmia stellivarians. The specimen agrees very closely in spiculation with the
description given by Ridley (16), except that the megascleres are a little larger and
the trichodragmata contain many more spicules.

The species has been hitherto recorded from the Straits of Malacca (Bowerbank) ;
Port Darwin, Australia (Ridley), and the Philippine Islands ("Challenger"). Re-
examination of the " Challenger " specimen, however, has convinced me that it does
uot belong to the same species as those described by Ridley from Australia and
found by Professor Herdman in Ceylon waters, as it possesses toxa. It must be
remembered that Bowerbank (49) originally described toxa as forming part of the
spiculation, but Ridley regarded these as foreign elements. It is probable that we
have here a confusion between two species.

It appears not improbable that the " thin fragment" recorded by Carter (4) from
the Gulf of Manaar, under the name Esperia tunicata. Sdt., may be specifically
identical with our specimen.

R.N. 220a (deep water off Galle and onwards up West Coast of Ceylon).

Esperella plumosa (Carter).

1882, Esperia plumosa, Cahtku (33) ; 188G, Esperia plumosa, Cakteu (20).
Sponge irregularly frondose or digitate ; clathrous. Surface irregularly cactiform
or conulose, covered by a very well-developed, stellately reticulate dermal membrane.
Vents not seen; pores scattered in the meshes of the dermal reticulation. Texture
(in spirit) coarsely fibrous, compressible, resilient, fragile. Colour grey. The largest
fragment measures about 62 millims. in height by 41 millims. in greatest breadth.

The main skeleton is an irregular reticulation of coarse, stout, multispicular fibre
without visible spongin; the main fibres running lengthwise and branching and

* Possibly young forms of the anisoehelse.

160 CEYLON PEARL OYSTER REPORT.

anastomosing with one another. The fibres are not very sharply defined and
numerous megascleres occur scattered between them. The dermal skeleton is a very
well -developed, triangular-meshed, stellate reticulation of stout, multispicular fibre
(about 0'05 millim. in diameter).

Spicules. — (1.) Styli ; fairly stout, often slightly crooked, with fairly well-developed
oval heads and slightly constricted necks ; sharply and rather abruptly pointed at the
apex ; size about 0-3 millim. by 0'009 millim.

(2.) Broad palmate anisochelse, very similar to those of Esperella simonis;* about
0'048 millim. long; frequently in rosettes.

(3.) Minute palmate isocheke ; numerous, about 0'012 millim. long.

(4.) Large, stout sigmata, C-shaped and contort, with abruptly recurved and very
sharply pointed ends ; size about 0-08 millim. from bend to bend by 0"006 millim.
thick in the middle.

(5.) Slender toxa with gently rounded curves; up to about 0*08 millim. long by
0'002 millim. thick in the middle : often in sheaves (toxodragmata) when young.

I have been able to satisfy myself by personal examination of Mr. Carter's type
preparation of his Esperia plumosa (now in my possession) of the specific identity of
the Ceylon form with the sponge recorded bv Carter from Mauritius, and subse-
quently from the Mergui Archipelago also. As Mr. Carter's descriptions are very
scanty, I have thought it desirable to give the above details concerning what is
evidently a widely distributed and characteristic species in the Indian Ocean,
distinguished by an exceptionally varied and beautiful spiculation.

II. N. 298, 328, 361 (three fragments amongst a large number; all from Ceylon
seas).

Esperella crassissima, n. sp. — -Plate XL, fig. 6.

The single specimen is rounded, cushion-shaped, attached by the base to a mass of
calcareous debris. The upper surface is strongly convex and bears a single large
prominent vent about 3 millims. in diameter. Surface with a more or less strongly
marked reticulate appearance, due to the coarse sub- dermal network of very stout
spicular fibre, the oval meshes of which are normally covered over by a thin pore-
bearing membrane, now mostly rubbed off. Pores scattered in the meshes of the
dermal skeleton. Texture (in spirit) rather hard but slightly compressible and
resilient ; coarsely fibrous. Colour pale grey. Maximum diameter of specimen about
24 millims.

The main skeleton is a very well-developed reticulation of very stout, compact
spicular fibre up to about 0-33 millim. in diameter, composed of very numerous closely
packed spicules and without visible spongin. The meshes of this reticulation vary
greatly in size and shape; perhaps 1*3 millims. would be a fair average diameter.
Immediately beneath the surface the main skeleton passes into the more compact

* Vide (1), Plate XV., fig. 16.

SPONGES. n; |

sub-dermal reticulation, with rounded meshes averaging about 0*5 millim. in diameter,
separated by fibres of about the same diameter (more or less). Over this lies a true
dermal reticulation composed mostly of loosely scattered spicules (sometimes aogre
gated in slender fibres) lying tangentially in the pore-bearing dermal membrane.

Spicules.— (I.) Tylostyli (Plate XI., fig. 6, a) ; slightly curved, rather stout; with
well developed oval heads narrower than the middle of the shaft, from which they
are separated by a well marked constriction ; gradually or rather abruptly sharp
pointed at the apex. Size about 0'49 millim. by 0-01 G millim.

(2.) Large palmate anisochelse of the ordinary form (Plate XL, fig. (J, b, c),
measuring about G"06 millim. by 0-024 millim. Very numerous, mostly in groups
which look like disorganised rosettes ; found chiefly just beneath the surface. In the
dermal membrane numerous small palmate isochelse occur scattered singly ; these are
about 0-012 millim. long and are probably young forms.

(3.) Sigmata (Plate XL, fig. G, d) ; slender, C-shaped and contort, up to about
0"036 millim. long; numerous.

(4.) Trichodragmata (Plate XL, fig. G, e) ; short, compact ; about CH)2 millim. by
0-006 millim. Very abundant.

This species is very closely related to Espcrella fusca, Ridley and Dendy (1),
obtained by the "Challenger*' off Bahia ; it differs, however, in the more strongly
developed main skeleton and consequently greater hardness of texture, and in the
sharply pointed character of the tylostyles ; possibly also in colour.

R.N. 240 (Ceylon seas).

Esperella tenuispiculata. n. sp.

Sponge irregularly massive, with a tendency to grow out into rounded lobes or
short, thick branches. Surface very uneven, covered over by a soft dermal membrane.
Vents few, small, scattered. Pores scattered (perhaps in irregular groups) in the
dermal membrane. Texture (in spirit) soft and spongy, but intensely gritty from the
presence of an immense quantity of coarse sand, chiefly in the interior of the sponge.
Colour varying from grey to brown. The largest specimen is about 63 millims. long
by 36 millims. in greatest breadth.

The skeleton is to a large extent replaced by the abundant sand-grains, which may
he held together in very irregular bands by spongin cement. Between and in
association with these we have loose, wispy bands of styli running towards the
surface, or simply scattered styli. There is no dermal skeleton.

Sjtu-nh's. — (1.) Tylostyli, very much reduced ; straight, slender; with distinct oval
heads, constricted necks and fairly gradually sharp-pointed apices ; size about
0-21 millim. by 0'004 millim., but often more slender.

(2.) Small palmate anisochelee, about 0'02 millim. long; rather scarce but constant.

(3.) Slender sigmata, C-shaped and contort, about 0"036 millim. from bend to bend :
sometimes very abundant,

Y

162 CEYLON PEARL OYSTER REPORT.

In its intensely arenaceous habit and the consequent reduction of the proper skeleton,
this species resembles Esperella arenicola, Ridley and Bendy (1), and E. crassa,
Dendy (10), both from Bass Straits. Tt differs from the former, however, in the
absence of trichodragmata, and from the latter in the presence of sigmata.

R.N. 293 ; 305 ; 334 ; 344 (Ceylon seas).

Paresperella, n. gen.

Encrusting or massive Esperellinae, with megascleres in the form of tylostyli or styli ;
with microscleres in the form of palmate anisochehe and serrated sigmata, to
which others (such as toxa) may be added.

The existence of several esperelline species with serrated sigmata justifies the
erection of a new genus, of which the type will be Carter's Esperia, serratohamata,
I am not aware that this very extraordinary form of spicule has hitherto been met
with in any other genns.

There are probably at least three species of the genus Paresperella in the
neighbourhood of Ceylon, for in an indeterminable sponge encrusting a calcareous
nodule from " Deep water off Galle and onwards " I have found (as a foreign body)
a huge serrated sigma (Plate XI., fig. 3) like those of Paresperella serratohamata, but
far too large to be referred to that species. It measures 0'3936 millim. from bend to
bend and 0'0146 millim. thick in the middle of the shaft, while, according to
Mr. Carter's measurements, the corresponding spicule in P. serratohamata, though
large, measures only 0"1 millim. by 0'0052 millim. Curiously enough, P. serratohamata
was also first known from a single spicule.

Lindgren (86) has described, under the name Esperella macrosigma, a species of
Paresperella in which the serrated sigmata attain still more enormous dimensions,
measuring 0-48 millim. in length by 0"024 millim. in diameter. This species comes
from the straits of Korea, and it is quite likely that it occurs also in Ceylon waters
and may be represented by the single spicule above described.

The genus is, of course, very closely related to Esperella, from which it has
evidently been derived.

Paresperella serratohamata (Carter) — -Plate XL, fig. 2.
l.sSO, Esperia serratohamata, Carter (4).
A minute specimen of this remarkable sponge occurs on a calcareous nodule in
association with Hymedesmia stelhvarians and Esperella pafishii. The spiculation
agrees very well with Mr. Carter's description, but the apices of the tylostyles are
uniformly mucronate (Plate XL, fig. 2, a). This character is not mentioned in
Mr. Carter's description, and the figure which he gives of the tylostyle is on too
small a scale to afford satisfactory evidence.

Lam BE (85) records this species from Vancouver, but I am doubtful, from the

SPONGES. 163

measurements which he gives of the spicules, whether his identification is correct ;
we have probably here yet another species of the genus.

U.N. 220c (deep water off Galle and onwards up the West Coast of Ceylon).

Paresperella bidentata, n. sp. — Plate XL, tig. I.

Sponge intensely arenaceous, the single specimen consisting of a friable mass of
coarse yellow sand, held together and permeated hy the soft sponge-tissues. Somewhat
cavernous internally. Surface uneven, with a soft dermal membrane visible in places.
The single specimen is an irregular massive fragment about 25 millims. in maximum
diameter.

The main skeleton of the sponge, between the closely aggregated sand-grains,
consists of slender megascleres, mostly loosely scattered, but occasionally collected in
fairly stout multispicular fibres. In the dermal membrane there is a distinct but
very loose and wide-meshed reticulation of slender spicular fibre, three or four spicules
wide, supported here and there on very loose sub-dermal brushes belonging to the
main skeleton.

Spicules. — (1.) Tylostyli (Plate XL, fig. 1, a); long, slender, straight or nearly so,
with oval heads, about equal in diameter to the middle of the shaft, and slightly
constricted necks ; typically with the apex slightly enlarged, truncated, and provided
with two (sometimes three ?) minute conical teeth placed side by side on the truncated
end (with their long axes parallel with the long axis of the spicule). The enlarge-
ment from which these teeth project contains a diverticulum of the central canal of
the spicule. These spicules are best developed in the dermal membrane (and in the
lining membrane of the large canals ?). In the dermal membrane they form the
reticulation of spicular fibre mentioned above, and measure about 0"27 millim. by
0"0053 millim. In the deeper parts of the sponge, between the sand-grains, they are
more slender and their apices sometimes appear to end in simple, long-drawn-out
points.

(2.) Palmate anisocheke (Plate XL, fig. 1, c, d), about 0'028 millim. long; mostly
in rosettes near the surface, very abundant and similar to those of P. serratohamata.

(3.) Sigmata (Plate XL, fig. 1, b, b) ; long and slender, usually contort, with sharply
incurved apices and (? always) with more or less distinct teeth on the outer side of
each bend, similar to those on the sigmata of -P. serratohamata, but much less strongly
developed. Size about 0'052 millim. from bend to bend by 0-002 millim. thick in the
middle.

This species differs from P. serratohamata in the presence of two teeth, instead of
one, at the apex of the tylostyle ; in the much more slender and less distinctly toothed
or serrate character of the sigmata, and in the absence of toxa ; as well as in the
arenaceous habit, which has doubtless caused considerable reduction in the proper
skeleton. I know of no other case where one can trace such a close and evident
relationship between an arenaceous sponge with reduced skeleton and a non-arenaceous

Y 2

L64 CEYLON PEARL OYSTER REPORT.

congener. The bidentate character of the apex <>t the tylostyle is very remarkable,
especially in comparison with the mucronate character of the corresponding spicule in
P. serraboha/niata.

R.N. 2G3a (Gulf of Manaar).

Iotrochota, Ridley.
Esperellinse of massive, ramose, or flabellate form, and usually of dark purple or brown
colour. Skeleton reticulate. Megascleres styli, to which diaetinal forms may
be added, especially in the more superficial parts of the sponge. Typical micro-
scleres birotulate (apparently modified isocheke).
In previous papers I have used the term " Amphiaster" for the birotulate spicules
of this and certain other genera of Esperellime. There can, however, be little doubt
that these microscleres are modified isochelre, and do not belong to the astrose series
at all. They are connected with the more typical isochelse by the peculiar isocheke
of the genera Choudrocladia and Axoniderma, and may conveniently be termed
" birotulate isochelse."

Iotrochota purpurea (Bowerbank), Ridley.

1875, Halichondria purpurea, Bowerbank (49) ; 1884, Iotrochota purpurea, RlDLEY (16).
The Ceylon variety of this species is characterized by its dark brown instead of
purple colour in spirit.* It is also easily distinguished from Iotrochota baculifera by
this character as well as by the replacement of the tylote megascleres by slender styli,
commonly arranged in radiating tufts at the surface.

The specimens are for the most part irregularly cylindrical in shape, may be much
branched ; and the surface is covered irregularly with small, sharp conuli. The styli
of the main skeleton are usually sharp-pointed and very variable in size. The
birotulate isocheke are numerous and about 0'02 millim. long. Not infrequently, the
straight, slender styli in the radiating tufts at the surface of the sponge exhibit an
inflation not far from the middle of the shaft, causing them to resemble a pipette.
The fibres of the main skeleton reticulation are stout and contain much spongin as
well as very many spicules.

The original type of the species was described by Bowerbank from the Straits of
Malacca. Ridley recorded it from Torres Straits, Albany Island, Port Molle, and
the Amirante Group. The West Indian Iotrochota birotulata (Higgin) (53) and the
Southern Australian Iotrochota coccinea (Carter)"!" are very possibly identical with
this species, or only varietally distinct.

R.N. 258; 297; 309; 343; 354; 360 (Stat. V., off Chilaw, 10 fathoms, and
elsewhere in Ceylon seas. Some are mere fragments).

* It should, however, be noted, that a portion of the collection was preserved in formalin and trans-
ferred to spirit afterwards. I do not know which specimens were treated in this way, nor do I know
what effect such treatment may have had upon the colour. t Vide Dendy (10).

SPONGES. 165

Iotrochota baculifera, Ridley.

L884, Iotrochota baculifera, Ridley (16) ; 1887, Iotrochota baculifera, Ridley, var. flabel-
lata, Dexdy (2).

There are in the collection a number of dark purple-coloured specimens, mostly
small or mere fragments, which I identify with this species. The flahellate habit,
characterizing my variety Jlabellata, is not recognisable in most of them, and it seems
hardly worth while to retain a special varietal name for the Ceylon form. In a
boiled-out preparation of R.N. 1G4 I find the styli frequently much stouter than the
measurements which I gave from Mr. Thurston's specimens, often measuring about
0-2 millim. by 0*0 12 millim., but very variable in diameter, while the tylota measure
about 0'246 millim. by 0'007 millim., and the birotulate isocheke about 0"02 millim.
long. It will be seen that these measurements agree very closely with those given 1 > v
Ridley for the type specimen from Port Darwin.

Thiele (39) has recorded the species also from Celebes.

R.N. 47 (Gulf of Manaar) ; 164; 317 ; 322 ; 331 (all from Ceylon seas).

Sub-family : PHLCEODICTYDSME.

Desniacidonidae in which the sponge body is provided with fistular outgrowths, and.
usually at any rate, with a spicular rind or cortex.

In this sub-family I include the genera Phlceodictyon, Oceanapia, Histoderma,
Sideroderma and AmpMastrella, which appear to me to form a very natural group.

The microscleres are usually more or less completely suppressed.

Phlceodictyon, Carter.

Phloeodictyinse with oxeote or strongylote megascleres and no microscleres at all.

Lundbeck (88) has shown pretty conclusively that Carters name "Phlceodictyon"
must be revived for this genus, the type species of " Rhizochalina" described by
Schmidt, being true Chalininse.

Phlceodictyon fistulosum (Bowereank).

1S73, Desmacidon fistulosa, Bowerbank (22) : 1880, Desmacidon Jeffreysii, Carter (4) :
1884, Rhizochalina fistulosa, Ridley (16) ; 1888, Rhizochalina fistulosa, Ridley and
Dexdy (1); 1897, Oceanapia fistulosa, Topsent (83); 1904, Phlceodictyon fistulosum,

TOPSENT (62).

Mr. Carter recorded this species from the Gulf of Manaar under the name
Desmacidon Jeffreysii. It differs, however, from the British Oceanapia jeffreysi, as
Mr. Carter himself pointed out, in the absence of sigmata. In this respect it agrees
with the Australian Phlceodictyon fistulosum, with which I have no hesitation in
identifying it.

In a footnote to the account of this species in the Report on the "Challenger"

166 CEYLON PEARL OYSTER REPORT.

Monaxonida, Mr. Ridley and I suggested that Rhizockalina and Oceanapia should
lit' united in one genus, and in a subsequent paper (10) I have carried out this
suggestion. Our reason for this proposal was that at one of the " Challenger " stations
(188, off New Guinea) specimens with and without sigmata appear to occur together,
and are so closely similar in external appearance that they cannot he distinguished
otherwise than microscopically. It seems equally reasonable, however, to suppose
that the two genera actually occur together in this locality, or that there has been
some confusion in the sorting out of the specimens. We can hardly suppose that the
same species sometimes occurs with, and sometimes without, sigmata. I therefore
propose to return to the arrangement originally adopted in the " Challenger" Report,
substituting, for the reason above given, the name " Phlceodictyon" for " Rhizochalina."

Neither Mr. Carter nor I have been able to find sigmata in the Ceylon specimens,
though I have myself examined about half-a-dozen microscopically.

One of Professor Herdman's specimens has the body, which varies much in shape
and is often very irregular, up to 48 milliins. in maximum diameter. As pointed out
by Mr. Carter, the fistular processes are long and may be ramified. They sometimes
appear to be naturally closed at the extremity (except, perhaps, for small pores), and
sometimes open. In nearly all cases they have been broken off short.

R.N. 81, 90, 237 (all from deep water off Galle and onwards up West Coast ot
Ceylon); 260; 363 (fistula? only, Stat. I., hauls 1-4, January 31, 1902, Colombo to
Negombo, 12 to 20 fathoms). Also other unnumbered specimens and loose fistula'.

Histoderma, Carter.

Phlceodictyimt' with usually diactinal megascleres and microscleres in the form of
isochelas, to which others may be added.

Au examination of Mr. Carter's own preparations of his Phlceodictyon singa-
porense (54) lias revealed the presence of small palmate isochelse. I therefore propose
to place this species, together with a closely related Ceylonese form, in the genus
Histoderma.

The presence of isochelas in this genus is of great interest and necessitates the
inclusion of the Phlceodictyime amongst the Desmacidonida?, the two species referred
to forming an important connecting link between Phlceodictyon and Oceanapia on
the one hand and Desmacidon on the other.

Histoderma vesiculatum, n. sp. — Plate XL, figs. 8, 9.

This remarkable sponge is represented in the collection by two fragments of about
equal size and closely similar appearance. Each consists of an elongated cylindrical
body inflated at irregular intervals to form oval vesicles. In one specimen there are
two distinct vesicles connected by a short cylindrical piece, and from the larger of the
two, which is broken across at the end, another short cylindrical piece is given oft' as

SPONGES 167

a branch. The slender cylindrical portions, and the vesicles too. are now more or less
hollow, but they contain the remains of the coarse skeleton network and a certain
amount of decomposed soft tissue. Thus, in the present condition of the specimens,
the outer wall of the sponge forms a thin but firm shell enclosing a wide cavity in
which lie the remains of the choanosome. How far the sponge was hollow in life
cannot now he ascertained. The colour (in spirit) is dark brown throughout, with
a slight purplish tinge on the surface. The surface is smooth, but has a very
characteristic tessellated appearance, caused by the very stout sub-dermal reticulation
of spicular fibre showing through the thin dermal layer. Each specimen has a total
length of about 27 millims. The largest vesicle measures about 11 millims. by
!> millims., and the slender cylindrical portions are about 4 millims. in diameter
(sometimes rather less).

The main skeleton forms a very wide-meshed reticulation of very stout spicular
fibre, occupying the interior of the sponge ; the meshes being very irregular in shape
and size. The diameter of the fibre, which appears to be composed entirely of very
closely packed megascleres, is about 0'165 millim. There is a sub-dermal reticulation
of similar fibre with roundedly polygonal meshes, the meshes being about 1 millim. in
diameter. This supports the dermal skeleton (Plate XL, fig. 8), which is very well
developed and extremely beautiful, consisting of a single layer of strongyla of various
lengths lying tangentially side by side as closely packed together as possible, with a
few much larger oxea, or sometimes styli, intermingled with them.

Spicules. — (1.) Oxea (Plate XL, fig. 8, o, o, u) ; rather long and only moderately
stout ; slightly curved, usually sharply but sometimes rather abruptly pointed at each
end; size about 0*3 millim. by 0-009 millim., but -variable. Abundant in the fibres of
the skeleton and scattered in the soft tissues between, where numerous very slender,
hair-like forms, possibly young, also occur.

(2.) Strongyla (Plate XL, fig. 8, s, s) ; cylindrical, more or less curved, and evenly
rounded off at each end; size extremely variable, from 0'04 millim. by (V005 millim.
to 0"1 millim. by 0'007 millim., or perhaps more. Characteristic of the dermal skeleton,
but also found in the spicular fibres and soft tissues of the choanosome.

(3.) Palmate isochelre (Plate XL, fig. 9), about 0-016 millim. long ; very similar to
those of Desmacidon conulosa, as figured in the Report on the "Challenger"
Monaxonida ; numerous.

(4.) I have also seen one "birotulate" of about the same length as the ordinary
isochelaB ; with very slender and apparently straight shaft and short umbrella-like
ends with minutely dentate margins. Whether or not this forms a constant element
in the spiculation I am not prepared to say, but its presence is very interesting in
view of the normal occurrence of such spicules in the genus Amphiastrella.

This species differs from Histoderma singaporense (Carter) chiefly in external
form, but also in the more slender character of the oxea.

Lixix;i?ex, however, has described (86), under the name Rhizochalina singaporensis

I lis CEYLON PEARL OYSTEE REPORT.

(Carter), a sponge from the China Sen, in which he also discovered isochelse. This
specimen seems to resemble the Ceylon form very closely, and it is possible that
future investigations may show them all to he specifically identical.
li.N. 212 (deep water outside pearl hanks, Gulf of Manaar).

Sub-family : E0TY0NIN7E.

Desrnacidonidse in which some of the megascleres take the form of spined styli,
originally developed as echinating spicules of the skeleton fibre or projecting at
right angles from the substratum.

The members of this sub-family are usually easily recognised by their spined styli,
but, as in the case of perhaps every type of spicule, the spined styli may he lost
again by degeneration. Such degeneration has obviously taken place in the genus
Raspailia, which, until recently, has been confounded with the Axinellidse. As a
rule, microscleres are present in the form of isochelae (to which toxa are frequentlv
added), but these also may be lost.

Myxilla, Schmidt.

Ectvoninn?, usually of massive, irregular form, in which the megascleres are spined
styli, which may or may not be echinating, and variously ended diactinal forms
which typically belong to the dermal skeleton. The typical microscleres are
tridentate isochelna, to which other forms may be added.

This genus in a certain sense occupies, as pointed out by Mr. Ridley and myself
in the Report on the " Challenger" Monaxonida, a position intermediate between the
sub-families Esperellina? and Ectyoninae, including both species with definite echinating
spicules and species in which such spicules cannot be distinguished from those of the
ordinary skeleton reticulation. It appears probable, from their form, that the spined
styli originated in the first instance as echinating spicules, i.e., spicules in which the
growth of one end (the base) became arrested by pressure against a firm substratum,
while the apex projected freely into the surrounding soft tissues and thus formed an
internal defence against the attacks of parasites. In some species these echinating
spicules appear to have passed into the main skeleton and form a reticulation with
one another.* It appears desirable, therefore, to regard the presence of spined styli (or
tylostyli) as constituting the leading feature of the Ectyoninse, without insisting too
strongly upon their actual arrangement.

In the genus Myxilla itself it is sometimes impossible, owing to the irregular
character of the skeleton reticulation, to distinguish sharply between echinating and
non-echinating styli. and, as I have pointed out before (10), I cannot, therefore, agree
with Topsent (62, &c.) and Hanitsch (46) in retaining Gray's genus Dendoryx for
species in which no special echinating spicules are recognisable.

* Compare Rhabderemia indica and its relationship to Uic thin encrusting forms of the same genus. Sec
iilso below, under Mi/xilla tenuissinia,

SPONGES. 169

This important genus is represented in the present collection by only two species,
but both of these are of exceptional interest owing to their Unusual habit, the one
being a typical sand-sponge and the other a very thinly encrusting form.

Myxilla arenaria, n. sp.

Sponge massive, irregular ; intensely and coarsely sandy, with a distinct translucent
dermal membrane visible in places. Pores and vents (I). Greatest diameter of
type (R.N. 263) 30 milhms. Colour (in spirit) light brown (the colour of the sand)
or grey ; texture very friable.

Main skeleton composed almost entirely of sand, with slender strongyla scattered
between the sand-grains and also collected in loose, wispy fibres ; and with small
spined tylostyli or styli very sparsely echinating the sand-grains. Dermal skeleton
composed of slender strongyla scattered irregularly in the dermal membrane.

Spicules. — (1.) Strongyla, perhaps sometimes faintly tylote ; straight, slender,
smooth; measuring about 0*146 millim. by 0'002 millim.

(2.) Spined tylostyli or styli ; straight ; gradually and very sharply pointed at the
apex ; rather sparsely and irregularly spined, chiefly at and near the base, which is
generally distinctly tylote. Size variable, say about 0'056 millim. by 0"004 millim.
(near the head, including spines).

(3.) Tridentate isochelos, with small teeth and fairly stout, strongly curved shaft ;
length about 0"014 millim. ; not very abundant.

(4.) Slender sigmata ; C-shaped and contort ; measuring about 0-032 millim. from
bend to bend, but rather variable in size.

As usual in sand-sponges, the proper skeleton is reduced in accordance with the
arenaceous habit. In external appearance the sponge closely resembles Paresperella
bidentata, and might easily be mistaken for that species.

R.N. 2G3 (Gulf of Manaar) ; 266 (pearl banks off Aripu).

Myxilla tenuissima, n. sp. — Plate XI., fig. 5.

Sponge extremely thin, encrusting. The single specimen forms a small crust of a
pale yellowish colour growing on a mass of calcareous worm-tubes in association with
Discodermia emarcjinata. Vents and pores not seen.

The skeleton consists partly of slender tylota, scattered irregularly and in loose
wisps, and partly of spined tylostyles projecting vertically from the substratum,
upon which their bases rest while their apices point outwards.

Spicules: — (l.) Tylota (Plate XL, fig. 5, a) ; smooth, slender, straight or nearly
so, with rather feebly developed oval heads and commonly with the shaft slightly
inflated (tylote) in many places at irregular intervals. Size about 0-f 48 millim. by
0-002 millim.

(2.) Spined tylostyli (Plate XL, fig. 5, b) ; cpiite straight and abundantly spined all
over ; with spherical head not much wider than the shaft, which tapers gradually

z

170 CEYLON PEARL OYSTER REPORT.

from the head to a fine, sharp apex. Spines rather small, conical and sharp pointed ;
on the head rather longer than elsewhere and arranged in a radiating manner, so
that the head, when viewed end on, resembles a spheraster. Size variable, up to
about 0'08 millim. in length by 0-008 millim. in thickness where the shaft joins the
head (exclusive of spines). These spicules are very numerous and, though quite
separate from one another, are placed pretty close together.

(3.) Stout tridentate isocheke (Plate XL, fig. 5, c, d) of the usual Myxilla type,
with rather strongly curved shaft ; length about 0'032 millim. ; very numerous.

This is a Myxilla with the encrusting habit of a Hymedesmia, and with corre-
spondingly arranged skeleton. In spiculation it closely resembles Topsent's Dendoryx
certa (48), but differs in skeleton arrangement, the spined styli forming a network in
the latter species. In fact, Myxilla tenuissima bears much the same relationship to
Myxilla {Dendoryx) certa that Rhabderemia pusilla does to Rhabderemia indica.

R.N. 234a (on same mass as R.N. 234, " deep water off Galle and onwards").

Clathria, Schmidt.

Ectyonime of various habit, frequently clathrous ; skeleton a reticulation of fibre,
usually with much spongin, containing smooth styli and echinated by spined
styli. Typical microscleres small palmate isocheke, to which toxa are frequently
added.

I have already (10) indicated the necessity for abandoning the genus lihaphi-
dophlus of Ehlers, which differs from Clathria only in the possession of a dermal crust
of radially disposed styli. The degree of development of this crust varies to such
an extent, however, that it is quite impossible to base a generic distinction on this
character alone.

Clathria frondifera (Bowerbank).

1875, Halichondria frondifera, Boweubank (49); 1884, Clathria frondifera, Ridley (16);
1889, Clathria corallitincta, Dendy (3).

I do not think it is possible to keep the Ceylonese form, described by me under the
name Clathria corallitincta, separate from the common Australian and Indian Ocean
species, originally described by Bowerbank from the Straits of Malacca and Gaspar
Straits under the name Halichondria frondifera, and subsequently re-investigated by
Ridley.

The external form of the sponge is very characteristic, and one of Professor
Herdman's specimens (R.N. 44) bears a remarkably close resemblance to a specimen
from the Mascarene group figured by Ridley (loc. cit., Plate liii., fig. j). The
spiculation, however, seems to be somewhat variable, and I must add to the
description which I gave of Mr. Thurston's specimens the following particulars
concerning those in the present collection.

SPONGES. 171

The palmate isocheke may be abundant. Smooth, slender toxa of very various
length are also met with. These are best developed in R.N. 85 and 281, where
they attain a length of about 0-184 millim., being almost straight from end to end
except for a small bow in the middle. The stout styli of the main skeleton vary
greatly in thickness. The slender styli or tylostyli may form a fairly distinct dermal
skeleton, in which they are either irregularly scattered or arranged in more or less
definite radiating brushes. The bases of these spicules are sometimes minutely
spined (R.N. 169).

One of the specimens (R.N. 97) is attached to a pearl oyster (Margaritifera
vulgaris).

RN. 44, 97, 268 (all Gulf of Manaar) ; 85, 169, 281 (all deep water off Galle and
onwards up West Coast of Ceylon).

Clathria indica, Dendy.

This common species (Dendy, 3) is represented in the collection by four specimens.
R.N. 27, 106 (both from Gulf of Manaar, Stat, IX., &c.) ; 166 ; 291 (Ceylon seas).

Clathria spiculosa (Dendy).

1889, Rhaphidophlus spiculosus, Dendy (3).
There is one small specimen of this sponge which makes a close approach to the
original types both in external form and spiculation.
R.N. 335 (Ceylon seas).

Clathria spiculosa, var. ramosa, nov.

This variety agrees closely with the types of the species in skeleton arrangement
and spiculation, but differs conspicuously in its elongated, slender, sub-cylindrical,
irregularly branching external form. It thus makes a very close approach to
Ridley's Echincmema gracilis from the Mascarene Islands (16), but the megascleres
in the Mascarene species are much stouter.

R.N. 168 ; 308 ; 310 ; 333 ; 368 (deep water outside pearl banks, Gulf of Manaar).

Clathria spiculosa, var. tessellata, nov. — Plate VIII., fig. 2.

This variety again differs from the types of the species only in external features,
which are, however, very characteristic. The sponge consists of more or less erect,
thick, irregular lamellas, with the margin rising up here and there into short,
digitiform processes. The vents are minute and few, marginal or on the ends of the
processes. The most striking character of the variety, however, is the tessellated
appearance of the surface, due to the presence of a close, polygonal-meshed reticulation
of (presumably) pore-bearing grooves ; the meshes of the reticulation are about
1 millim. to 1'5 millims. in diameter. The general appearance of the sponge is well
shown in the figure. One specimen (R.N. 92a) measures about 70 millims. in greatest

Z 2

172 CEYLON PEARL OYSTER REPORT.

breadth and the same in height. The colour (in spirit) is dark grey, and the texture
characteristically firm and fleshy, and very tough.
R.N. 92; 92A (Gulf of Manaar).

Raspailia, Nardo.

Ectyoninas of elongated, slender, branching habit. Skeleton composed of a dense
central axis of spicular fibre containing much spongin, from which bundles or
tufts of spicules radiate to the surface. Smooth monactinal (sometimes diactinal)
megascleres are present, and also (normally) spined echinating styli. No micro-
scleres.
As already indicated, the members of this genus bear a strong superficial

resemblance to Axinellidoe, but it is evident from the presence of the spined

echinating styli (though these may be vestigial) that they are really highly modified

Ectyoninas.

Raspailia thurstoni, Dendy.
There is one specimen of this sponge (Dendy, 2, 1887) in the collection.
R.N. 39 (Stat. LX., outside Muttuvaratu Paar, 30 fathoms)

Raspailia fruticosa, Dendy, var. tenuiramosa, nov. — Plate VIII. , fig. 5.
1887, Raspailia fruticosa, Dendy (2).

There are in the collection two specimens, one dry and one in spirit, which differ
from the type of the species only in the much more slender and much shorter
branches, and in the greater density of the skeleton, which is less distinctly reticulate.
The branches are usually only from 2 millims. to 3 millims. in diameter, and very
numerous, so that the whole sponge is even more fruticose than the type. The
largest specimen (dry) is about 55 millims. in total height and 78 millims. in greatest
breadth ; it is represented in Plate VIII., fig. 5.

R.N. 100 (dry, Gulf of Manaar) ; 173.

Raspailia hornelli, n. sp. — Plate XL, fig. 7.
The single specimen is erect, arborescent, branching in one plane. The branching
at first sight appears to be dichotomous, but in reality the two branches present are
both given off from the same side of a main axis which in no way differs from them in
appearance, except that the part below the first branch, which we may consider as
representing the stalk, is more rounded in transverse section and slightly expanded
below at the attachment, while the branches themselves are slightly flattened. Each
branch is very much narrowed at the end to form a slender, elongated apex. The
surface is rather coarsely granular, with numerous minute apertures between the
granules ; slightly hispid in places. Colour (in spirit) dark brown. Texture tough,
compressible, resilient, Total height about 99 millims. Length of stem below the

SPONGES. 173

first branch 33 millims., diameter about 5 millims. Length of longest (lowest)
branch about 64 millims. ; greater transverse diameter of branches when fully
developed about 6 millims.

The skeleton consists of a network of stout horny fibre, irregularly cored by smooth
megascleres and fairly abundantly echinated by spined styli. There is a well marked
axial condensation of the reticulation, in which most of the larger spicules are to be
found densely packed together lengthwise in the stout horny fibre. From this axial
portion more slender fibres run obliquely outwards to the surface, where each one ends
(typically) in a projecting brush composed of a very few long, stout styli, surrounded
at the base by numerous much shorter and very slender styli (these often appear to
be absent). The outwardly-directed horny fibres are irregularly cored by usually
slender megascleres and abundantly echinated towards their outer ends by spined
styli. They are connected together to form a network by short secondary fibres of
similar structure, but more slender. There are also many of the smooth megascleres
scattered through the soft tissues apparently without any relation to the fibres at all

Spicules. — (1.) Smooth styli (Plate XL, fig. 7, a); long and comparatively stout,
evenly rounded off at the base, gradually and sharply pointed at the apex ; usually
slightly curved towards the base; size about 0"65 millim. by 0-018 millim. ; found
chiefly in the axial condensation (but also scattered in the soft tissues outside it) and
in the surface brushes.

(2.) Smooth styli (Plate XL, fig. 7, b, c) ; very slender, straight or nearly so,
evenly rounded off at the base and very finely and gradually pointed at the apex ;
size, say about 0'24 millim. by 0-0027 millim., but sometimes longer ; characteristic
of the surface brushes, but also occurring frequently between the axis and the surface ;
perhaps sometimes oxeote.

(3.) Smooth oxea (Plate XL, fig. 7, (/, e, f, g) ; rather short and fairly stout,
fusiform, more or less curved or angulated in the middle, gradually sharp pointed at
each end; size very variable, commonly about 0'2 millim. by 0-012 millim., but may
be as much as 0-4(3 millim. by 0-014 millim. ; abundant, scattered irregularly outside
the horny fibres, and within them.

(4.) Strongyla (Plate XL, fig. 7, //, h) ; very short and comparatively stout, straight,
equal-ended and broadly rounded off at each end ; size very variable, say about
0-2 millim. by 0'024 millim. ; sometimes pretty numerous in the axial condensation
and occasionally met with outside it, but very erratic in occurrence.

(5.) Spined styli or sub-tylostyli (Plate XL, fig. 7, k, I, m) ; rather short, straight,
tapering gradually from base to apex, which, though narrow, is bluntly pointed ;
covered all over with very minute, sharp spines, which are most abundant at the base
and apex ; size about 0"08 millim. by 0'008 millim. (at the base, including spines).

Numerous small, spherical, granular, brown pigment cells, about O'Ol millim. in
diameter, are scattered through the outer part of the soft tissues.

This species, though evidently closely related to the European Raspailia {Dictyo-

174 CEYLON PEARL OYSTER REPORT.

cylindrus) hispida, is readily distinguished by the details of its spiculation. I have
much pleasure in naming it after Mr. Hornell, to whose energy this collection owes,
I am told, so much of its value.

R.N. 59 (Stat. IV., off Karkopani, Gulf of Manaar).

Agelas, Duchassaing and Michelotti.

Ectyoninae in which the skeleton is composed of horny fibre echinated by verticillately
spined styli ; with no other spicules.

For synonymy and references I must refer to the Report on the " Challenger "
Monaxonida (p. 163).

Agelas mauritiana (Carter).

1883, Ectyon mauritianus, Carter (54) ; 1887, Agelas mauritianus, Ridley and Dendy (1);
190.3, Agelas cavernosa, Thiele (87).

This species is represented by a tubular fragment 31 millims. long and up to about
16 millims. in diameter ; widely open at both ends and with the wall of the tube only
about 3 '5 millims. thick. It is of rather dark brown colour and firm, resilient
consistence.

The skeleton is a close, irregular network of stout, amber-coloured horny fibre
echinated very abundantly by the characteristic verticillately spined styli, which
measure about 0-176 millim. by 0'016 millim. (including spines).

This beautiful species has hitherto been recorded from Mauritius (Carter), off
Tristan da Cunha (?) (" Challenger "), and from Ternate (Thiele).

R.N. 358 (Ceylon seas).

Agelas ceylonica, n. sp. — Plate XII., fig. 9.

The type specimen* consists of a few slender, anastomosing, sub-cylindrical branches,
arising from an irregular, proliferous basal crust attached to a calcareous nodule. The
branches are only about 2 millims. in diameter, and their surface is irregular and
sometimes minutely hispid. Texture (in spirit) compressible, resilient, fibrous and
fairly tough. Colour brown. Vents minute and scattered on the branches.

The skeleton is an irregular network of pale-coloured horny fibre about 0'03 millim.
in diameter, abundantly echinated by the spined styli, which are occasionally also
found embedded lengthwise in the fibre, two or three side by side.

Spicules. — Verticillately spined styli (Plate XII., fig. 9), straight or slightly curved ;
the spines mostly small and conical, especially about the middle of the spicule ; those
at the base irregularly arranged, larger, and often hook-shaped, with the sharp points
curved towards the apex of the spicule ; apex sharp-pointed and free from spines for
a short distance. Size variable, but characteristically long, say about 0"24 millim. by

* A second specimen closely resembles the type,

SPONGES. 175

0'02 millim. at the base (including spines). Although they are verticillately spined,
the annulation of these spicules is not nearly so distinct as in Agelas mauritiana.
R.N. 304 ; 312 (type). (Both from Ceylon seas.)

Echinodictyum, Ridley.

Ectyoninse with reticulate main skeleton composed of spicular fibre containing smooth
oxea and echinated by spined styli. Smooth styli may also be present.
Microscleres usually absent.

Echinodictyum clathratum, n. sp. — Plate XL, fig. 4.

The single specimen forms a sub-spherical, clathrous mass of rather thin, short,
flattened trabecule, echinated on the outer surface by slender conuli. Vents and
pores not seen. Colour (in spirit) pale greyish-yellow ; texture soft and resilient,
fairly tough. Total diameter about 20 millims.

The main skeleton is a very irregular network of fairly stout spicular fibre ; the
size and shape of the meshes and the thickness of the fibres varying greatly. The
fibres contain a very large number of oxeote megascleres closely packed together side
by side, while the spongin cement which unites them is inconspicuous on account of
its very pale colour. The fibres are echinated more or less abundantly by spined styli
which project from them almost or quite at right angles. Numerous loose megascleres
(oxea and styli) are scattered between the fibres. There is a dermal skeleton of well
developed brushes of very slender styli ; the brushes are well separated from one
another, and each exhibits a beautiful radiate arrangement of its component spicules.
These brushes appear to be confined to the outer surface of the sponge as a whole,
and are absent from the surfaces of the inner trabeculse.

Spicules. — (1.) Oxea (Plate XL, fig. 4, c, d) ; smooth, slender, usually more or less
sharply bent at or near the middle, gradually and sharply pointed at each end ; size
very variable, commonly about 0-25 millim. by 0-006 millim. In the fibres and
scattered between them, forming the greater part of the skeleton.

(2.) Smooth styli (Plate XL, fig. 4, a, a) ; very long and fairly stout, tapering
gradually from the evenly rounded base (which may be very faintly tylote) to the
sharply pointed apex ; nearly straight but generally slightly curved towards the base ;
size variable, say about 1'26 millims. by 0'012 millim. These spicules lie parallel to
the surface in the thickness of the sponge-lamella, without any definite arrangement.

(3.) Smooth styli (Plate XL, fig. 4, h) ; very slender, almost hair-like; straight or
gently curved ; evenly rounded at the base, very gradually and finely pointed at the
apex ; size about 0-34 niillim. by 0"002 millim. In the dermal brushes.

(4.) Spined tylostyli (Plate XL, fig. 4, e, f) ; rather long and slender, straight,
tapering gradually from the slightly developed head to the narrow, fairly sharp-
pointed apex. Spines small but abundant all over, especially on the head. Size about
O'l millim. by 0'006 millim. where the shaft joins the head.

176 CEYLON PEARL OYSTER REPORT.

I have also seen three small isochelse about 0-018 millim. long, and two or three
siemata, one of which measured about 0"05 millim. from bend to bend; but these
spicules are so scarce that I am doubtful whether they are proper to the sponge.

This is an interesting species, having (if we leave out of account the doubtful
microscleres) the spiculation of an Echinodictyum combined with the habit of an
Echinoclaihria. It is evidently nearly related to Tuiele's Echinodictyum cavernomm
from Celebes (39), but differs in its pale colour and in the presence of the large styli.

R.N. 325 (Ceylon seas).

Aulospongus, Norman.

Massive Ectyoninaa with plumose columnar skeleton, comprising both smooth and
spined styli in the spiculation. Without microscleres.

In my ' Report on a Second Collection of Sponges from the Gulf of Manaar ' (3) I
expressed the opinion that Norman's genus Aulospongus (55) was unnecessary, and
that the species for which it was established might be included in the genus AxineLla.
Further consideration has, however, induced me to alter my views on this question,
and to consider the presence of the spined styli as constituting a sufficient generic
distinction from Axinella, and, indeed, necessitating the removal of the genus to the
Ectyoninse.

This genus is evidently closely related to Raspailia, and constitutes one of the
apparent connecting links between the Ectyoninae and Axinellidae. Thiele (39) has
referred to the genus Raspailia two or three Japanese species, one, at least, of which
{Raspailia (?) villosa) should perhaps be included in Aulospongus. The Australian
Raspailia cacticutis (vide Dendy, 10) may also possibly belong here, and likewise
Mr. Carter's Dictyocylindrus sessilis from the Gulf of Manaar (4).

Aulospongus tubulatus (Bowerbank).

1873, Haliphysema tubulatum, Bowerbank (8) ; 1878, Aulospongus tubulatus, Norman (55);
1889, Axinella tubulata, Dendy (3).

There are a number of specimens of this sponge in Professor

Herdman's collection. The species (fig. 3) forms one of the

most characteristic elements in the Sponge-Fauna of Ceylon,

and is of especial biological interest as affording an example

of symbiosis, or perhaps commensalism, between a Sponge

and an Annelid. I have nothing to add to the account which

I gave in my ' Report on a Second Collection of Sponges

from the Gulf of Manaar ' (3).

Vi&J"'77''S 'r'Tl R-N. 1, 2, 3 (and others not numbered ; all from Gulf of

(Bowerb.) ; from Gull ot ' ' v

Manaar; „at. >ize. Manaar (Stats. V., IX., XV., LV., LXL, LXVIIL).

SPONGES. 177

Acarnus, Gray.

Ectvoninae in which the megascleres may be stylote, oxeote, tylote and cladotylote,
the latter forming the characteristic " grapnel-spicules " which typically echinate
the skeleton fibre. Microscleres may be present in the form of palmate isochelae
and toxa.

Acarnus ternatus. Ridley. — Plate VIIL, fig. 4.

There is in the collection a remarkably fine rlry specimen of this species (R.N. 105,
Plate VIIL, fig. 4), of proliferous and thickly flabellate habit, with rather large vents
(about 8 millims. in diameter) on the summits of the lobes. The specimen measures
270 millims. in greatest breadth by 150 millims. in height, and its colour is brown.
There are also a couple of small specimens in spirit, the best of which is sub-cylindrical
in form and irregularly branched.

The spongin of the skeleton fibres is very strongly developed but very pale-
coloured.

Spicules. — -(1.) Styli, measuring about 0"3 milbm. by 0"0164 millim.

(2.) Cladotylota (grapnels), with usually three large, strongly recurved, sharp
teeth ; shaft about 0'21 millim. by 0"012 millim.

(3.) Slender tylota with slightly spined heads ; about 0"22 millim. by 0"0035 millim.
in the middle.

(4.1 Oxea ; long and very slender, gradually sharp-pointed at each end, may be
angulated at or near the middle ; say about 074 millim. hy 0"004 millim.

(5.) Toxa; say about 0"152 millim. by 0*008 millim., but often more slender.

(6.) Palmate isochelse, about 0'02 millim. long.

The slender oxea occur irregularly distributed outside the horny fibres. Ridley (16)
makes no mention of them in his description, but they might be overlooked or
regarded as accidental. Keller, on the other hand, describes and figures similar
spicules in his Acarnus wolffgangi from the Red Sea (61). The latter species may
possibly prove to be a mere synonym of A. ternatus.

The species has hitherto been recorded from Torres Straits (Ridley), Bombay (?)
(Ridley), Amirante Islands (Ridley), and Tahiti (" Challenger").

R.N. 105 (dry, Gulf of Manaar) ; 313; 329 (Ceylon seas).

Cyamon, Gray (emend.).

Ectyoninee in which the principal megascleres are smooth styli and tylostyli ; the

echinating spicules have a radiate form, and there are no microscleres.

In 1867, Gray (31) proposed this genus for Bowerbank's Dictyocylindrus vickersii,

of which a single spicule was figured and described in the ' Monograph of British

Spongiadee.'* It was not until 1879 that this species was first really described by

* Vol. i., p, 267, fig. 23L
2 A

178 CEYLON PEARL OYSTER REPORT.

Carter (56), who, however, did not adopt Gray's genus, but adhered to Bowerbank's
original name.

In 1880, Carter (4) also described two species from the Gulf of Manaar, which he
named Microciona quadriradiata and M. quinqueradiata respectively, and called
attention to the resemblance of the echinating spicules in these species to those of
" Diclyocylindrus vickersii." It appears to me that all these three species must be
included in one genus, distinguished as in the diagnosis given above, for which we
may retain Gray's name Cyamon.

This genus appears to be nearly related to Trikentrion, but is distinguished from it
by the absence of oxeote megascleres (compare Carter, 56).

The peculiar radiate echinating spicule is probably derived from a spined echinating
stylus by great enlargement of three or four of the basal spines. In this way it may
come to resemble a tetractinellid spicule or an aster. It is extremely interesting to
compare with these forms the equally peculiar echinating " grapnel " spicule of
Acarnus, in which a tetractinellid form may also be arrived at secondarily, but by
enlargement of spines at the apex instead of at the base of the spicule.

Cyamon quinqueradiatum (Carter).

1880, Microciona quinqiieradiata, Carter (4).

The single specimen in the collection forms a sub-circular crust, about 1 1 millims. in
diameter and 3 millims. thick in the middle, attached to a mass of nullipore.* The
surface is uneven and irregularly conulose, and there is in parts a distinct, thin
dermal membrane. In parts also the surface is sparsely hispid from the projection of
some of the large spicules. The texture is soft and compressible, internally somewhat
cavernous, and the colour (in spirit) is pale yellowish-brown.

The skeleton consists chiefly of tylostyles and styles of various sizes, the latter (at
any rate usually) longer than the former. These spicules sometimes have their bases
resting on the substratum and sometimes they are arranged in short plumose columns ;
some have their apices projecting for a considerable distance beyond the surface.
The echinating spicules are rather scarce and local in their distribution, mostly to be
found amongst the bases of the other spicules ; though plentiful here and there, they
might easily be overlooked except in fortunate preparations.

The spiculation agrees fairly closely with Carter's description, but I think he has
laid too much stress upon the distinction of the various forms of styli and tylostyli.
In our specimen these pass gradually into one another; the slender " acuate " is not
recognisable as a distinct type (it may be simply a young form), while the tylostyli
have larger heads than appears from Carter's description and figure. There may be
either three or four smooth rays in the echinating spicule (in addition to the spined
ray), but it agrees very closely with Carter's description and figure.

R.N. 270 (deep water off Galle and onwards up West Coast).

* In association with a small specimen of Petromica massalis, R.N. 269.

SPONGES. 17 it

Plocamia, Schmidt.*

Ectyoninse of varying form ; may be encrusting or erect and branched. The
characteristic spicules are dumb-bell-shaped or sausage-shaped megascleres, in
addition to which styli or tylostyli of various forms (some of which are typically
echinating) may be present. The microscleres are isochelse and (at any rate
usually) toxa.

Plocamia nianaarensis (Carter) — Plate VIII., fig. 1.

1880, Dictyocylindrus nianaarensis, Carter (4) ; 1 881, Dirrhopalum nianaarense, Ridley (60).

There are several specimens of this curious sponge in the collection. Mr. Carter
figured and described both external form and spiculation, and Mr. Ridley added
valuable information on the latter point and on the question of systematic position.
It is therefore unnecessary for me to do much more than refer to Plate VIII. , fig. 1,
which represents a much finer specimen than any hitherto obtained. The colour of
the sponge (in spirit) varies from pale yellowish-grey to almost black,t and the
branching is extremely irregular. One specimen (R.N. 367) shows very clearly how
the main stem may be attached to the substratum by an expanded base.

Lambe (85) lias recorded this species from California, but I think that his
identification is probably erroneous.

R.N. 41 (Gulf of Manaar); 76 (Stat. XXXIII., 18 fathoms); 107 (off Galle,
February 13, 1902) ; 278 (deep water off Galle and onwards up West Coast of Ceylon) ;
367 (deep water outside pearl banks) ; 374 (Stat. LX., 20-30 fathoms, Gulf of Manaar).

Bubaris, Gray.

Ectyoninse of usually encrusting habit. Skeleton consisting of an inner mass of
diactinal (vermicular) spicules, from which large styli or tylostyli project
vertically outwards. Without microscleres.

This genus was founded by Gray (31) for Bowerbank's Hymeraphia vermiculata,
the type of the genus Hymeraphia (H. stellifera) being a totally different sponge.
Topsent (59, 62) has already adopted the genus.

It is probable that there are a large number of encrusting sponges, with skeleton
arrangement very similar to that of Bubaris, which are in reality not closely related
to one another, but derived from several groups by adaptation to an encrusting habit
(compare Hymedesmia amongst Astromonaxonellida). In such cases we must judge
of their systematic position rather by the form of the spicules than by the mere
arrangement of the skeleton.

The possible relationship of the genus Bubaris to the Axinellidae is indicated in
the Report of the " Challenger" Monaxonida (p. 182).

* For literature, &c, vide Ridley and Dendy (1).
t See footnote under lotrochota pwpwrea.
2 A 2

180 CEYLON PEAEL OYSTER REPORT.

Bubaris eruca (Carter).

1880, Hymerhaphia eruca, Carter (4); 1894, Rhabderemia eruca, Topsent (59).

There are three specimens of this very curious encrusting sponge in the collection ;
it has been well described and figured by Carter (loc. at.), who himself pointed out
its close resemblance to Bowerbank's Hymeraphia vermicidata, which is the type
species of the genus Bubaris.

Topsent (62) has recently proposed to include this species in his genus Monocre-
pidium, distinguished from Bubaris by the tuberculation of the diactinal megascleres.
In Bubaris eruca the spicules in question are annulated rather than tuberculated.

R.N. 183 (Stat. XLIIL, off Kaltura, February 19, 1902, depth 22 fathoms) ; 239a,
240a (both from deep water off Galle and onwards up West Coast).

Rhabderemia, Topsent.

Encrusting or massive Ectyoninse in which the principal megascleres are styli with
strongly curved base, shaped like a hockey stick, and the principal microscleres
are contorted sigmata. There are no chelae and the echinating spicules appear
to be greatly reduced or absent.

Topsent (48) established this genus in 1892, originally for the reception of Carter's
Microciona pusilla and M. intexta (57), and a new species described by himself under
the name Rhabderemia guernei. R. pusilla, being mentioned first, may be taken as
the type of the genus. R. guernei and the new species which I am about to describe
under the name Rhabderemia indica, agree so closely with R. pusilla that there can
be no reasonable doubt of their generic identity. R. intexta differs more from the
typical species, but may still, I think, be included in the genus. It is distinguished
from the others by the possession of spined megascleres and the absence of small
slender styli. There is in Mr. Carter's cabinet, however, a preparation labelled by
him " Microciona ininutula," which is intermediate in spiculation, having large bent
styli minutely spined at the apex and also very slender styli slightly roughened at
the base. It is further characterized by having toxa amongst the microscleres, and
is evidently quite distinct from Rhabderemia (Microciona) pusilla, with which
Mr. Carter has apparently confounded it.*

It is probable that some other species hitherto referred to the genus Microciona
will also have to be included under Rhabderemia ; on the other hand, as I have shown
in dealing with Bubaris eruca, this species has been erroneously included in the genus
Rhabderemia by Topsent.

Rhabderemia indica, n. sp. — Plate XII., fig. 10.
line specimen encrusts and almost completely envelopes some large fragments of
shell, and attains on one side of the shell a thickness of 7 or 8 millims. The shape
of the entire specimen with its enclosed shell-fragments is massive and irregular.

* The name pusilla was intended by Carter to be minutula, vide Carter (57 and 4).

SPONGES. 181

The surface is uneven but fairly smooth ; granular ; with a distinct, translucent
dermal membrane in places. Vents probably small and scattered. Texture rather
soft, spongy and friable. Colour (in spirit) dull grey. Greatest diameter of the entire
mass about 48 millims.

The skeleton is a close-meshed, very irregular reticulation of megascleres, many ot
which are collected together into loose nmlti-spicular bands running perpendicularly
to the surface. These primary fibres, if we may so call them, are connected together
by still looser and more irregular secondary bands, and the whole is confused by
immense numbers of irregularly scattered megascleres. There is no special dermal
skeleton and I have detected no spongin.

Spicules. — (1.) Styli, " rhabdostyles" of Topsent (Plate XII., fig. 10, a, b, c) ; base
evenly rounded off, not tylote ; basal part of shaft sharply bent at an angle to the
remainder, like the end of a hockey stick, occasionally somewhat spirally curved ;
remainder of shaft straight or nearly so ; gradually or somewhat hastately pointed at
the apex ; size fairly uniform, about 0"24 millim. by 0'006 millim., the bent basal
portion being about 0"018 millim. long. These make up the main skeleton.

(2.) Very small, slender styli (Plate XII., fig. 10, d) ; straight or nearly so;
tapering gradually from rounded base to finely pointed apex ; with very slightly
roughened surface; size about 0"044 millim. by 0-002 millim. at the base. Scattered
very abundantly through the soft tissues as microscleres, and very uniform in size. I
am inclined to think that these spicules are vestigial echinating styli.

(3.) Sigmata (Plate XII., fig. 10, e) ; very much contort, slender, often twisted
into a kind of half-knot in the middle, sharply pointed at each end (when one end
appears bluntly rounded, or even knobbed, it is probably due either to its having been
broken short or to fore-shortening in perspective). The greatest length, measured in
a straight line from bend to bend, is only about 0'012 millim., but if the spicule were
straightened out it would measure at least twice as much. Very abundant.

This species is distinguished from Rhabderemia pusilla by the greater length of the
bent styli and the much smaller size and roughened surface of the minute styli;
from R. guernei by the absence of the peculiar microscleres which Topsent terms
" thraustoxes," and by the smaller size of the megascleres and the roughening of the
minute styli ; from R. intexta by the presence of the minute styli, the smoothness of
the large megascleres and the form of the sigmata. It differs from all in its much
more robust growth, which constitutes perhaps its most noteworthy feature.

R.N. 341 (Ceylon seas).

Family: AXINELLID^E.

Sigmatomonaxonellida in which the microscleres have usually been entirely lost by
degeneration; the megascleres are usually, in part or entirely, stylote ; the
skeleton arrangement is usually, but not always, plumose ; and there are no
spined echinating styli.

182 CEYLON PEARL OYSTER REPORT.

Owing to their loss of microscleres and want of other well-marked characters, this
family is one of the most unsatisfactory with which we have to deal. The plumose
arrangement of the megascleres in the main skeleton cannot be regarded as exclusively
diagnostic, for it is met with also in Ectyoninse and even in Chalininse, and too great
reliance upon this character has led to the inclusion of forms amongst the Axinellidse
which certainly should not be included in that family. I have already indicated that
those so-called Axinellids in which true asters have been observed (e.g., Vibulinus)
should be placed amongst the Astromonaxonellida. The genus Raspailia, similarly,
has been removed from the Axinellidae to the Ectyoninse, but we are still left with a
somewhat heterogeneous collection of sponges which it is extremely difficult to define,
and which will probably be subjected to considerable re-arrangement in the future.
A curiously constant feature in this group is the irregularity exhibited by the ends of
the megascleres and the manner in which the stylote and oxeote forms tend to pass
into one another.

It is highly probable that the group, even as here restricted, is of polyphyletic
origin.

Spongosorites, Topsent (emend.).

Axinellidse with the main skeleton composed of an irregular but dense reticulation of
large oxea or styli, and dermal skeleton composed of a thin layer of very much
smaller oxea lying tangentially, and in close contact with the main skeleton ;
oxea typically biangulate.

Topsent (14, &c.) places the genus Spongosorites in his family Ooppatiidse, assuming
that it has lost the characteristic microscleres, and calling attention to the apparent
tetractinellid affinity indicated by the biangulate oxea. It appears to me, on the
other hand, to come more naturally amongst the Axinellids, being not distantly
removed from Leueophlceus and Ciocalypta, as indicated by the new species described
below, in one of which we perhaps find a clue to the evolution of the curious finger-
shaped " processes " of the latter genus. In Ciocalypta tyleri, var. aberrant, we also
sometimes find biangulate oxea.

The genus Spongosorites may also be nearly related to Thiele's Dactylella (39).

Spongosorites topsenti, n. sp. — Plate XII., fig. 1.

There are six specimens of this curious sponge in the collection, differing so mucn
amongst themselves in external form that it seems desirable to give a short account
of each.

R.N. 152 is a depressed cake-shaped sponge of nearly circular outline, with convex
upper and almost flat lower surface, the latter having been evidently attached to the
substratum all over. Greatest diameter 36 millims., thickness in the middle
13 millims. The upper surface is somewhat corrugated, with shallow grooves
(indicating underlying canals) radiating in a stellate manner from low, mound-like

SPONGES. 183

projections, one at least of which bears at its summit a small group of vents through
which the radiating canals open, while another hears no visible apertures at all. The
upper surface also bears one very small digitiform process, about 5 millims. long and
2 millims. in diameter, with do visible opening, and there may have been more of
these processes, now broken off. The texture is hard and compact, almost stony, and
the colour (in spirit) is dull brownish-grey.

R.N. 68 is similar in general features, but the upper surface is much more strongly
convex ; the broad, mound-like projections are rather better developed, but still few
in number (about three, grouped in the middle of the upper surface), and one bears a
conspicuous vent. The slender digitiform processes without visible openings are more
numerous than in R.N. 152. Greatest diameter of specimen 34 millims.

R.N. 182 is a small hemispherical specimen only about 15 millims. in diameter,
with the convex upper surface produced in the middle into a single stout digitiform
process, about 11 millims. long by (5 millims. thick, but of irregular shape. This
process contains wide canals which enter it from the body of the sponge. There are
apparently none of the slender digitiform processes. I have seen no vents, but the
specimen is somewhat damaged.

R.N. 202 is closely similar to the last, but rather larger, with a single stout
digitiform process in the middle, containing longitudinal canals, but now broken
short.

R.N. 134 is a small specimen about 2(5 millims. in diameter, with slightly conulose
surface and two very short but stout finger-like processes.

R.N. 77 is much larger than any of the preceding, and the base, instead of being
flat below, has grown partially round several loose calcareous nodules. It is strongly
convex and more or less conulose and corrugated above, and the upper surface also
bears two or three well- developed, stout, erect digitiform processes, very irregularly
distributed. These processes contain more or less well-developed longitudinal canals,
but there are no visible vents. On the general surface of the sponge the usually low
conuli are in places elongated to form slender projections about 4 millims. long, and
with strongly hispid surface. The base of this specimen is about 78 millims. long by
35 millims. in greatest breadth. One of the larger processes is about 26 millims.
high by 15 millims. broad at the base, tapering upwards almost to a point.

From the above descriptions it would appear that the sponge is normally cushion-
or cake-shaped, with the upper surface conulose or rugose, and sometimes bearing
processes of two kinds, viz., (l) large processes which appear to contain the exhalant
canals and to be produced by elongation of low, mammiform, vent-bearing projections,
with the conspicuous vents probably replaced by a cribriform dermal membrane ; and
(2) small slender processes produced by elongation of the surface conuli, and apparently
tearing no special relation to the canal system.

The main skeleton in the body of the sponge consists of a very dense and confused
reticulation of scattered spicules lying in all directions and varying greatly in form

184 CEYLON PEARL OYSTER REPORT.

and size, stout and slender mixed up together. At the surface, lying immediately on
the main skeleton, there is a thin dermal layer of short and rather slender oxea,
through which the apices of the large underlying spicules frequently project. In the
processes of both kinds, large and small, the spicules of the main skeleton for the
most part run lengthwise, and there is the same irregular dermal reticulation of small
oxea. There are none of those characteristic radiating pillars of spicules, supporting
the dermal membrane over large sub-dermal cavities, which we find in a typical
Ciocalypta.

Spicules. — (1.) Large and very stout (Plate XII., fig. 1, a-k), typically fusiform
oxea, but variously ended, sometimes stylote or strongylote, always more or less
curved, sometimes biangulate, sometimes very crooked and irregular ; size about
11 niillims. by 0"065 millim., but variable.

(2.) Short, slender oxea (Plate XII., fig. 1, I) ; usually biangulate, gradually and
sharply pointed at each end, symmetrical ; measuring about 0"18 millim. by 0'008
millim., but variable. These forms occur chiefly, but not entirely, in the dermal
skeleton ; while between these and the largest spicules numerous intermediate forms
(Plate XII., fig. 1, m) may be observed.

The above account of the spiculation is taken from R.N. 152, but the spiculation of
the other specimens does not differ in any important respect.

The species may be distinguished from its European congener, S. placenta (the type
of the genus), by the development of the finger-like outgrowths (which, however, may
be not always present) and by details of spiculation, such as the absence of the central
inflation of the oxea.

R.N. 68 (Gulf of Manaar); 77 (Stat. XLV., off Pantura, 25 fathoms) ; 134, 202
(deep water off Galle and onwards up West Coast); 152; 182 (Stat, XLIIL, off
Kaltura, February 19, 1902).

Spongosorites (?) lamellata, n. sp. — Plate XII., fig. 2.

Sponge irregular, compressed, lamello-digitate ; only 3 millims. or 4 millims. in
average thickness, with a maximum length of about 45 millims. ? Erect. Surface
granular in appearance and minutely hispid, fairly smooth but uneven ; margins
rounded. Colour (in spirit) pale greyish -yellow.

The main skeleton consists of a dense and very irregular reticulation of styli, mostly
scattered singly, but occasionally collected in loose strands. The dermal skeleton is
formed of a thin layer of small slender oxea, lying tangentially to the surface and
very irregularly scattered.

Spicules. — (1.) Styli (Plate XII., fig. 2, a, b, c) ; usually fairly stout and slightly
curved or crooked (especially towards the base), narrowing slightly to the base, which
is evenly rounded, and tapering gradually to the apex, which is sharply pointed ; size,
when fully grown, about 0'95 millim. by 0-02 millim., but numerous smaller and
more slender forms occur which are apparently young.

SPONGES. 185

(2.) Oxea (Plate XII.. fig. 2, d, e) ; slender, sub-fusiform, slightly curved or
angulated once or twice, sharply pointed at each end; size very variable, averaging,
say, about 0"2 millim. by 0*0055 millhn. These spicules occur chiefly at the surface
of the sponge, but are also found in the interior.

This species differs widely from Spongosorites topsenti, not only in external form
but also in the (? invariably) stylote character of the larger spicules.

R.N. 236 (deep water off Galle and onwards up West Coast of Ceylon).

Spongosorites (?) lapidiformis, n. sp. — Plate XII., fig. 3.

The three specimens by which this species is represented in the collection are quite
irregular in shape ; massive, and everywhere evenly rounded off, like water-worn
pebbles, with no recognisable points of attachment or differentiation of surfaces. The
surface is granular and very shortly hispid ; harsh to the touch ; occasionally veined
by underlying ramifying canals. Vents few, small, sometimes hardly recognisable.
Texture hard and compact, without separable dermal membrane. Colour (in spirit)
pale wax-yellow. The largest specimen measures about 29 millims. in maximum
diameter.

The skeleton is a very dense, irregular reticulation of megascleres, partly collected
together in ill-defined bands. This reticulation extends right up to the surface of
the sponge, and there is no special dermal skeleton, and apparently no spongin.

Spicules. — (1.) Very stout, fusiform oxea (Plate XII., fig. 3, a, b); slightly curved
and gradually and sharply pointed at each end ; measuring, say, about 0-87 millim.
by 0-0495 millim. ; pretty frequently becoming stylote (Plate XII., fig. 3, c) by
rounding off of one end, and rarely even strongylote. These oxea are connected by
intermediate forms (Plate XII., fig. 3, d) with

(2.) Slender oxea (Plate XII., fig. 3, e) ; slightly curved, gradually sharp-pointed at
each end; measuring, say, about 0-54 millim. by 0"012 millim. ; irregularly inter-
mingled with the large oxea, and perhaps only young forms thereof.

This species is of very doubtful systematic position. Had it not been for the
presence of so many stylote spicules, I should probably have referred it to Hali-
chondvia. It differs from typical species of Spongosorites in the absence of a special
dermal layer of small oxea.

R.N. 143, 144, 145 (all from deep water oft' Galle and onwards up West Coast).

Hymeniacidon, Bowerbajntk (emend.).

Axinellidse of massive habit. Skeleton reticulate, composed of spicular fibre usually
containing a good deal of spongin ; with no special dermal skeleton. Megascleres
styli or sub-tylostyli ; no microscleres.

I have pointed out on a previous occasion (10) that Lexdenfeld's genus Stylotella
(with which Topsent's Stylinvs is admittedly synonymous) is not distinguishable

2 B

186 CEYLON PEARL OYSTKR REPORT.

from Hymeniacidon. The position of the genus amongst the Axinellidse is, of course,
open to question, but it seems to come at least as naturally here as anywhere else.

Hymeniacidon petrosioides, n. sp. — Plate XII., tig. 4.

The type specimen (R.N. 151) is massive, cushion-shaped, strongly convex above
and irregularly concave below, having evidently been attached to the substratum at a
few points only. Upper surface fairly even but coarsely granular. Vents (probably)
few, small and scattered. Texture very hard and compact ; surface harsh to the
touch. Colour (in spirit) dull grey, with a purplish tinge here and there. Greatest
breadth about 47 millims. ; maximum thickness about 19 millims. There is another
much smaller specimen of irregularly massive form.

The skeleton is a dense, close-meshed reticulation of short, stout styli, in which one
can readily distinguish stout, multispiculous main fibres running at right angles to
the surface at distances of about one spicule's length from one another, and connected
crosswise by isolated spicules and bundles of spicules, with other similar spicules
irregularly scattered in the soft tissues.

Spicules. — Short, stout, more or less curved or bent styli (Plate XII., fig. 4),
broadly rounded off at the base (occasionally slightly tylote) and gradually sharp
pointed at the apex ; fairly uniform in size, measuring about 0"39 millim. by
0*022 millim. (There are apparently no oxea. )

This species appears to be nearly related to Topsent's Stylinos jullieni from the
Atlantic (48), but the spicules are much larger. In external appearance it bears a very
close resemblance to llirinacophora durissima, with which it may easily be confounded
until microscopically examined, and, indeed, I am inclined to think that these species
of Hymeniacidon are really closely related to the massive species of Thrinaco2)hora.
We may also have here, so to speak, a point of contact between the Axinellidse and
Desmacidonidse, indicated both by the form and arrangement of the megascleres
and the presence of trichodragmata in the last-named genus. The spicular fibre in
Hymeniacidon is not plumose, or, at most, very feebly so, but it is impossible to draw
a hard and fast line between the plumose type of fibre characteristic of the Axinellidse
and the non-plumose type characteristic of the Desmacidonidse, &c.
R.N. 151 ; 316 (Ceylon seas).

Thrinacophora, Ridley and Dendy.

Axinellidse with typically plumose skeleton and with microscleres in the form of
trichodragmata.

Thrinacophora agariciformis, n. sp.— Plate XII., fig. (J.
Sponge consisting (usually, at any rate) of a short, thick stalk (which has evidently
been attached below), supporting a thick, rounded, cushion-shaped body which is
irregularly depressed above so as to form a more or less shallow, very thick-walled

SPONGES. 1 87

cup : or the top of the sponge-body may be flattened, with several irregular, shallow
depressions. Surface minutely and uniformly conulose ; the conuli barely 0"5 millim.
in diameter and separated from one another by deep but narrow, meandering grooves.
In the depression at the top the conuli may be covered over by a thin translucent
membrane containing small circular vents. Consistence compressible, resilient.
Colour (in spirit) pale yellowish-grey. A typical specimen (R.N. 163, not quite the
largest) gave the following measurements : total height 25 millims. ; length of stalk
8 millims. : diameter of stalk 13 millims. ; longer diameter of body 31 millims. ;
shorter diameter 21 millims.

The skeleton is arranged in a typical axinellid manner, consisting chieflv of plumose
columns of spicules running at right angles to the surface and ending in the conuli.
These columns lie pretty close together and are connected with one another crosswise
by occasional groups of spicules, or by single spicules, running across at right angles
from one to the other ; there are also numerous irregularly scattered spicules in the
interspaces. The spicules are cemented together in the columns, and sometimes also
in the cross connections, by a large amount of very pale-coloured spongin. There is
no special dermal skeleton, but the surface may be rendered slightly hispid by the
projection of the terminal spicules of the columns.

Spicules. — (1.) Short styli (Plate XII., fig. 6, b, c) ; fairly stout and more or less
curved, especiallv towards the base, which is broadly rounded off; with gradually and
finelv pointed apex ; size about 0-268 millim. by O'Ol millim., but variable.

(2.) Oxea (Plate XII., fig. 6, d, e) ; usually gently and evenly curved and sharply
pointed, but often irregularly ended ; of about the same size as the short styli, and
very numerous.

(3.) Very long and slender, setaceous styli (Plate XII., fig. 6, a) (sometimes oxea),
measuring, say, about 075 millim. by 0-008 millim., but variable. These spicules are
found lying lengthwise in the interspaces between the plumose columns.

(4.) Trichodragmata (Plate XII., fig. 6, /) ; bundles of short and very slender,
hair-like microscleres, slightly curved, the whole bundle measuring, say, about
0'032 millim. by 0-005 millim., but variable. These occur scattered quite irregularly
between the spicular columns.

This beautiful and well-characterised little sponge is represented in the collection
by seven specimens. It differs widely from any of its known congeners in external
form, and spirit specimens are easily mistaken, at first sight, for young examples of
Phakellia donnani, though readily distinguished on closer inspection by their paler
colour, more finelv conulose surface, &c. It is apparently common and may be looked
upon as one of the most characteristic species of the Ceylon Sponge-Fauna.

R.N. 160a ; 160b ; 163 ; 163a ; 314 ; 332 ; 356 (all from Ceylon seas).

Thrinacophora durissima, n. sp. — Plate XII., fig. 5.

Sponge sessile, cushion-shaped, very strongly convex above and somewhat con-

2 b 2

188 CEYLON PEARL OYSTER REPORT.

tracted below, but attached by a broad base. Surface even, but granular or minutely
conulose, the conuli being in part covered over by a translucent dermal membrane ;
very minutely hispid. Vents small, few, scattered ; surrounded by feebly developed
grooves arranged in a somewhat stellate fashion. Colour (in spirit) pale yellowish-
grey. Texture compact, very hard ; surface harsh to the touch ; internal structure
radially columnar. Greatest diameter about 23 millims.

The skeleton is composed chiefly of close-set, rather irregular, plumose columns of
megascleres, running at right angles to the surface and ending in the small conuli or
granules. These columns are connected by numerous spicules, isolated or in loose
bands, which run across at right angles from one to another, while numerous
megascleres are also scattered irregularly in the soft tissues.

Spicules. — (1.) Oxea (Plate XII., fig. 5, b) ; short, fairly stout, fusiform, gently and
symmetrically curved, gradually sharp-pointed at each end ; size up to about
0*39 millim. by 0-024 millim., but usually somewhat less, especially in diameter.

(2.) Styli (Plate XII. , fig. 5, a) ; with broadly rounded base and gradually sharp-
pointed apex ; usually more or less curved towards the base ; dimensions about the
same as those of the oxea. Perhaps not quite so abundant as the oxea.

(3.) Trichodragmata (Plate XII., fig. 5, c, d) ; short, stout bundles of very slender,
hair-like spicules, the whole bundle having, as usual, a faint brownish colour ;
dimensions of the entire bundle about 0'02 millim. by 0'008 millim. ; abundant
towards the surface of the sponge.

This curious little sponge is evidently closely related to Topsent's Thrinaeophora
spissa (48) from the North Atlantic, and forms an interesting connecting link between
that species (which has only oxeote megascleres, apparently arranged in a hali-
chondrioid rather than an axinellid fashion) and the more typical species of the
genus.

K.N. 355 (Ceylon seas).

Axinella, Schmidt.

Axinellidse of varying habit, but not flabellate. With plumose skeleton composed of
smooth styli or oxea and no microscleres.

Axinella labyrinthica, Dendy.

There are three specimens of this sponge in. the collection. The species is easily
recognisable by its external appearance, and I have nothing to add to my former
description, Dendy (3), 1889.

R.N. 33, 103 (both from Gulf of Manaar) ; 357 (Ceylon seas).

Axinella manus, n. sp. — Plate XII. , fig. 8.

Sponge erect, stipitate, branched in a somewhat palmate manner, but with the
branches coming off at different levels and curving slightly towards one another (so
as to suggest a hand holding a ball). Stem cylindrical, slightly expanded below,

SPONGES. 1 89

about 25 rnillims. long and 9 millims. in diameter. Brandies about as thick as the
stem, slightly flattened, short (usually about 30 millims. long), rather few in number,
terminating in abrupt, conical apices. The vents are small openings in the floors of
stellately arranged or longitudinal grooves, which give a characteristic appearance to
the sponge, and are chiefly placed on the inner surface of the branches, but also
occasionally on the outer surface and on the stem. Surface between the vent-bearing
grooves granular or minutely conulose. Texture compressible, resilient but tough,
with the stem a good deal harder than the branches. Colour (in spirit) rather light
grey. Total height of specimen 84 millims.

The skeleton is rather loose and irregular, consisting (in the branches) of plumose
columns radiating outwards to the surface and with many spicules irregularly
scattered between ; the whole becoming quite confused towards the middle of the
branch, but without any special axial condensation.

Spicules. — (1.) Rather short and fairly stout styli (Plate XII., fig. 8) ; more or less
curved towards the base, which is broadly rounded off; gradually and sharply pointed
at the apex; size about 0"295 millim. by 0*016 millim., but often more slender.

(2.) Oxea ; almost symmetrically curved and gradually sharp pointed at each end;
of about the same dimensions as the styli ; abundant.

This species, in the arrangement of the vents and in the skeletal characters, makes
a close approach to Phakelha donnani and P. symmetrica, and demonstrates very
clearly the impossibility of distinguishing sharply between the genera Phakellia and
Axinella.

R.N. 53 (Gulf of Manaar).

Axinella tenuidigitata, n. sp. — Plate XIII. , fig. 4.

The single specimen is a small massive sponge of short, thick, irregularly cylindrical
form, attached by a broad base below and strongly convex on the upper surface, from
which a number of slender, elongated, finger-like processes are given off. Surface
uneven and irregularly hispid, especially on the digitiform processes ; in part minutely
and irregularly conulose and in part covered by a distinct, sub-glabrous, translucent
dermal membrane. The digitiform processes are solid and they may unite with one
another. Vents apparently small and scattered between the processes. Colour (in
spirit) pale wax-yellow ; texture hard and compact. Height of body about
20 millims., diameter about 15 millims. ; length of processes, of which there are about
half a dozen, about 11 millims., with a diameter of not much more than 1 millim.

The skeleton in the body of the sponge consists of an irregular reticulation of long
styli, which, as they approach the surface, arrange themselves in loose, irregular,
plumose columns. The digitiform processes are composed each almost entirely of a
dense axis of similar spicules closely crowded together and placed longitudinally,
with a few spicules projecting outwards beyond the surface, so as to give rise to its
hispid character.

190 CEYLON PEARL OYSTER REPORT.

Spicules. — Apparently all stylote and all long (Plate XIIL, fig. 4), but varying
much in thickness ; usually only very slightly curved (towards the base), but some-
times crooked ; broadly rounded off at the base, which may be somewhat narrower
than the middle part of the spicule, and gradually and sharply pointed at the apex.
The stouter forms measure about 1"18 millim. by 0-0328 millim. In the digitiform
processes they are a good deal more slender, and slender forms also occur intermingled
with the stout ones in the body of the sponge.

R.N. 202a (deep water off Galle and onwards up West Coast of Ceylon).

Axinella halichondrioides, n. sp. — Plate XII., fig. 7.

Sponge encrusting, extended horizontally. Upper surface somewhat convex and
rather uneven, with small monticular elevations, each bearing a single vent, scattered
at fairly regular intervals ; granular (minutely conulose) between the elevations, the
conuli being normally covered over by a thin, transparent dermal membrane. Colour
(in spirit) light brown ; texture firm and compact, columnar in vertical section.
Greatest breadth of specimen about 64 millims. ; thickness in the middle about
16 millims. Height of vent-bearing projections up to about 2 millims. Diameter of
vents about 1 millim. Distance between vents about 9 millims.

The skeleton consists chiefly of very stout but rather loose and irregular and only
slightly plumose columns of spicules running vertically to the surface and ending in
loose brushes in the small surface conuli. These columns contain a very large number
of spicules and are connected together by short, loose bands of spicules running across
the interspaces at right angles. There are also a large number of spicules irregularly
and loosely scattered through the soft tissues.

Spicules. — Mostly oxeote (Plate XII., fig. 7, a) ; gently and uniformly curved,
gradually and sharply pointed at each end; size about 0'31 millim. by O'Ol millim.
A few styli of about the same size also occur (Plate XII., fig. 7, b, c).

This species resembles pretty closely Thiele's Axinella incmstans (39) from Japan,
but its spiculation shows it to be distinct.

R.N. 75 (outside pearl banks, Gulf of Manaar).

Phakellia, Bowerbank.

Axinellidffi of compressed, fiabellate (or cup-like) form, usually with vents on one of
the flat surfaces and inhalant pores on the other. Without microscleres.

Phakellia donnani (Bowerbank).

1873, Isodictya donnani, Bowerbank (8); 1887, Axinella donnani, Dendy (2).

There are a dozen specimens of this characteristic species in the collection, in various

stages of growth. The spiculation, as usual in the Axinellidse, is somewhat variable,

and oxeote as well as stylote spicules occur. In view of the cup-shaped (or sometimes

fiabellate) form (see fig. 4) it seems desirable to remove the species from the genus

SPONOES. 191

Ixinella and put it iu Phakellia, if indeed the distinction between these two genera
is to be maintained.

Fig. 4. Phakellia donncmi (Bowerb.), from Gulf of Manaar, nat. size. A, flabellate ; B, cup-shaped form.

RN. 10, 15, 20, 21, 22, 23 (all from Gulf of Manaar); 160; 160a; 160b; 160c ;
181, 181a (the last two from Stat, XLIIL, off Kaltura, depth 22 fathoms,
February 19, 1902).

Phakellia symmetrica, n. sp. — Plate XIII., fig. 3.

The single specimen is a short-stalked, flabellate sponge, the somewhat compressed
stalk widening out not very suddenly into a single vertical expansion with evenly
rounded margin and without any proliferation. The two surfaces are exactly alike ;
minutely conulose with small circular openings (? inhalant pores) everywhere between
the conuli, and with numerous stellate vents scattered at moderately wide intervals.
Towards the margin the surface becomes longitudinally grooved rather than conulose,
and there are also a few marginal vents. Colour (in spirit) dull yellowish-grey ;
texture compressible, resilient, tough. Total height of specimen about 55 millims. ;
length of stalk about 16 millims. ; greatest breadth of frond about 41 millims. ;
thickness of frond in the middle about 8 millims.

The skeleton is composed of moderately stout, slightly plumose fibres curving
upwards and outwards to the surface (where they terminate in the small conuli) and
connected together by short, irregular cross fibres, so as to form an ill-defined
reticulation with many spicules scattered irregularly in the interspaces. The skeleton
fibres contain much spongin, not very conspicuous, however, on account of its pale
colour.

Spicules. — Styli (occasionally oxeote) of two principal varieties : («.) Comparatively
short and stout (Plate XIII., fig. 3, a, b) ; more or less curved towards the base,
which is evenly rounded off, and gradually and sharply pointed at the apex ; size
about 0'23 millim. by 0"009 millim. (b.) Comparatively long and slender (Plate XIII.,
fig. 3, d), slightly curved towards the base, which is evenly rounded off, and gradually
and finely pointed at the apex ; size about 0"5 millim. by 0-008 millim.

This pretty little sponge may prove to be merely a variety of Phakellia donnani,
connected with the typical form by the flabellate variety figured in my paper on the
Sponge-Fauna of Madras (2). For the present, however, it is perhaps better to keep
the two distinct,

R.N. 159 (Muttuvaratu Paar, 8 fathoms).

192 CEYLON PEARL OYSTER REPORT.

Phakellia ceylonensis, n. sp. — Plate VIII., fig. 3 ; Plate XIII., fig. 5.

The single specimen (Plate VIII., fig. 3) is shortly stipitate, erect, thinly flahellate
and very proliferous. The branching and anastomosing vertical lamellae of which it
is composed all terminate at about the same level in thin, sinuous margins. The
lamellae are scarcely 3 millims. in thickness and have a tendency to become perforated
by larger and smaller apertures. The two surfaces of the lamella are not distinguishable ;
each is finely conulose (granular) and minutely and slightly hispid. Vents not
recognisable. Texture tough and resilient ; colour (in spirit) greyish-brown. Total
height only about 41 millims., but with a maximum breadth of about 90 millims.

The skeleton is dense, composed of plumose columns radiating upwards and outwards
into the small surface conuli, and merging internally into an irregular but fairly dense
reticulation of spicules. There is a large development of very pale-coloured spongin.

Spicules. — (1.) Styli ; slightly curved or bent, evenly rounded off at the base,
usually very gradually and finely pointed at the apex ; of two principal sizes, but
very variable : («.) Comparatively short and stout (Plate XIII. , fig. 5, a), say about
0-2 millim. by 0'008 millim. (/>.) Long and slender (Plate XIII., fig. 5, b), say about
0*44 millim. by 0'005 millim.

(2.) Oxea (Plate XIII., fig. 5, c, (/, e) ; subject to much the same variations in size
as the styli, and variously ended.

This species is nearly related to Carter's Phakellia jiabellata from Australia {vide
Dendy, 10), but for the present at any rate it may be regarded as specifically
distinct.

R.N. 34 (Gulf of Manaar).

Phakellia crassistylifera, n. sp. — Plate XIII., fig. 6.

The single specimen is a small, irregular, proliferously lamellar and slightly
clathrous sponge, without recognisable point of attachment. The surface is granular
and minutely hispid, and there are no visible vents. Texture hard, tough, resilient.
Colour (in spirit) pale wax-yellow. Height (?) 31 millims. ; greatest breadth about
19 millims. ; thickness of lamellae variable, say about 2 millims.

The skeleton is a very dense, close and irregular reticulation of very stout fibre,
composed of a large quantity of almost colourless spongin, in which numerous usually
stout styli are more or less completely embedded. The primary fibres are stouter
than the secondaries, and may have a somewhat plumose character ; but the whole
reticulation is so confused, and the spongin, though very abundant, so pale in colour,
that at first sight the entire skeleton looks like a dense, irregular network of thickly
scattered styli.

Spicules. — Styli (Plate XIII. , fig. 6); usually stout and comparatively short and
more or less curved ; broadly rounded off at the base, which is not narrowed, and
usually sharply and gradually pointed at the apex ; but the apex is occasionally
bluntly rounded off, so that the spicule becomes strongylote with unequal ends. Size

SPONGES. 193

commonly about 0"46 millim. by 0*0285 millim., but very variable in diameter; often
much more slender than the measurement given, and occasionally a good deal stouter ;
sometimes rather longer.
R.N. 256 (Ceylon seas).

Acanthella, Schmidt.

Axinellidaj of usually flabellate form and more or less cartilaginous consistence. With
more or less strongly aculeate or conulose surface. With no microscleres.

Acanthella carteri, Dendy. — Plate VIII. , fig. 6.

There are several specimens of this sponge in the collection, and as the external
form has not yet been figured, Professor Herdman has kindly had one of them
photographed for this report (Plate VIII., fig. 6). There is a strong development of
very pale-coloured spongin fibre associated with the spicules, which I omitted to
mention in my original description, Dendy, 3 (1889).

Keller's Acanthella aurantiaca (61), from the Red Sea, comes very near to this
species, if it be not identical with it.

RN. 11, 36, 49 (all from Gulf of Manaar, Stats. II, IV, V.); ? 301 (perhaps young);
? 336 (perhaps young).

Acanthella nabelliformis, Keller.

The single specimen is thinly fiabellate ; ? stipitate (the stalk may have been broken
off, the specimen being somewhat damaged and worn) ; both surfaces beset Avith
rather sharp longitudinal ridges, showing a tendency to break up into conuli ; with
deep, semi-cylindrical, longitudinal grooves between the ridges. The floor of the
grooves is formed by a rather thick, translucent dermal membrane, containing no
spicules and easily stripping off; in this membrane are scattered numerous small
vents. Colour (in spirit) dark greyish-brown. Texture tough, compressible, resilient.
Height of the single lamella of which the specimen consists 37 millims. ; breadth
51 millims. ; thickness (including ridges) about 5 millims.

The skeleton is a reticulation of stout spicular fibre, containing much spongin and
comparatively few, though still very numerous, spicules. The main fibres curve
upwards and outwards to the surface, and are united by irregular cross-fibres. The
arrangement of the spicules, both in the fibres and between them, is very irregular
and confused. The main fibres are often somewhat plumose.

Spicules. — Oxea ; straight or curved, slender, gradually and sharply pointed at
each end ; size fairly uniform, up to about 0*3 millim. by 0"0065 millim. ; occasionally
stylote.

The soft tissues are densely charged with small, round, granular brown cells,
probably pigment cells.

This species is chiefly characterised by its oxeote spicules and by the distinctly reticu-

2 c

194 CEYLON PEARL OYSTER REPORT.

late skeleton with its strong development of spongin. It appears to form a connecting
link between the genera Acanihella and Phakellia.

I have little doubt of the specific identity of the Ceylon form with that from the
Red Sea, where it is, according to Keller, one of the most abundant and charac-
teristic forms on the reefs. Such slight differences as I have observed will be
sufficiently evident by comparison of the description given above with that given
by Keller (61), in 1889. Keller tells us that the colour in life is blue.

R.N. 31 (Gulf of Manaar).

Auletta, Schmidt.
Axinellidee of tubular form ; without microscleres.

Auletta lyrata (Esper).

1798-1806, Spongia lyrata, Esper (6) ; 1870, Raspaigella lyrata, Ehlers (58) ; 1889, Auletta
aurantiaca, Dkxdy (3).

There are a number of specimens of this species in the collection, which show it to
be an extremely variable one, both as regards external form and spiculation. The
re-discovery of the typical flabellate form, agreeing closely with Esper' s figs. 1 and 2
(Plate 67), enables me to identify my Auletta aurantiaca with Esper' s Spongia
lyrata, which was also obtained from Ceylon.

Typical examples may be described as follows : —

Sponge erect, flabellate, shortly stalked. Lamella thick, slightly proliferous, with
broadly rounded margin bearing a row of small, sometimes sphinctrate vents, which
are the outlets of vertical oscular tubes. Surfaces of lamella rather minutely conulose
or rugose, and slightly hispid. Texture (in spirit) compressible and resilient, but
tough ; colour yellowish-grey. One specimen (R.N. 345) has a total height of about
31 millims., the lamella is about 42 millims. in breadth, and 7 millims. in thickness.
The vents are rather less than 1 millim. in diameter.

The main skeleton consists of wispy bands of long, slender spicules united together
by abundant very pale-coloured spongin, running upwards and outwards to the
surface conuli (from which their terminal spicules project more or less) and connected
with one another by occasional short cross-fibres of one spicule's length, forming an
ill-defined, rectangular-meshed network of spicular fibre. In these fibres the spicules
lie approximately parallel to one another, there being none of the typical axinellid
arrangement except in the surface brushes. Between the fibres, which have a very
loose, irregular appearance, numerous isolated spicules are scattered.

Spicules. — (1.) Styli ; long and very slender, nearly straight; evenly rounded oil'
at the base and fairly sharply pointed at the apex ; size about 0"41 millim. by 0'005
millim., but variable.

(2.) Slender oxea; variously ended ; mostly a good deal shorter than the styli.

A second specimen (K.N. 174) differs in having the margin of the sponge thinner,

SPONGES. 195

the vents fewer and somewhat larger (with a tendency to occupy separate prominences
of the margin), and the spicules somewhat stouter.

In the arrangement of the vents, this typical form of the species resembles
Phakdlia tumida, from Australia (10), but the genus P/iakellia, if retained, should
probably be restricted to species with the vents arranged on one or both of the flat
surfaces of the lamella. It differs from my "Auletta aurantiaca" in the more slender
form of the spicules, as well as in the distinctly flabellate character of the sponge, whieh
may be regarded as formed from a number of Auletta tubes fused together side by side.

R.N. 174; 345 (Stat. LXVIIL, Gulf of Manaar, 10 fathoms).

Auletta lyrata, var. glomerata, nov.

In this variety the sponge consists of an irregular agglomeration of short, finger-
like processes, more or less united together laterally, and each containing a longi-
tudinal oscular tube terminating at the apex of the process in a sphinctrate vent. In
other respects, including spiculation, this variety resembles the preceding, and the
whole mass is attached to the substratum by a short stalk.

This variety makes a near approach to Thiele's Auletta halichondrioidcs, from
Japan (39), while in external form it closely resembles fig. 3 of Esper's Plate 67.

R.N. 170 (pearl banks off Aripu) ; 194 (south of Dutch Modragam Paar) ; 303.

Auletta lyrata, var. crassispiculata, nov.

I propose this name for three irregularly branched, loosely bushy specimens,
characterised mainly by the large size of the spicules, which, however, vary so much
that it is almost impossible to express the difference by measurements. The tubular
branches of which the sponge is composed are short and rather slender.

R.N. 43, 60 (both from Gulf of Manaar) ; 161 (deep water off Galle and onwards
up West Goast).

Auletta lyrata, var. brevispiculata, nov.

This variety is represented in the collection by one very fine specimen of much
larger size than usual and consisting of a great mass of branching and anastomosing
tubes, each ending in a sphinctrate vent now more or less completely closed by the
membranous sphincter. The whole mass is attached to a very short, stout stem,
and the shape and size of the individual branches do not differ from those of the
specimens of " Auletta aurantiaca " originally described by me. The total height of
the specimen, however, is 122 millims., and the greatest breadth about 73 millims.
The branching is very irregular.

The spicules commonly measure about 0-35 millim. by 0*0164 millim.

RN. 45 (Gulf of Manaar).

Auletta elongata, n. sp. — Plate XIII. , fig. 7.
The type specimen (R.N. 73) consists of a bunch of seven elongated, sub-cylindrical
tubes, branching out from each other and from the short, thick stem by which they

2 c 2

196 CEYLON PEARL OYSTER REPORT.

are attached to the substratum. All the tubes grow vertically upwards, close
together and parallel with one another, and they do not vary greatly in length.
Each terminates in a wide vent, which may be more or less closed by a membranous
sphincter. Colour in spirit, dull yellowish -grey. Texture of tubes compressible,
resilient, stiff; stem hard and tough. Total' height of specimen 90 millims. ; length
of stem about 18 millims. ; diameter of stem about 10 millims. ; length of longest
flnbranched tube about 52 millims. ; diameter of tube about 8 millims. The thickness
of the wall of the tube is about 2 millims., and the inner surface of the wall bears the
numerous small apertures of the exhalant canals, while the outer surface is granular
and porous in appearance and minutely hispid.

The skeleton consists chiefly of very stout bands of spicular fibre, which run longi-
tudinally through the inner half of the tube-wall, branching and anastomosing with
one another in a quite irregular manner. From these stout fibres very short, irregular,
loose, somewhat plumose columns of long, slender spicules run almost vertically
outwards to the surface of the sponge, beyond which the ajnces of some of them
project ; the distance between the longitudinal fibres and the outer surface being
only about one spicule's length. No spongin is recognisable in ordinary unstained
sections.

Spicules. — Very variable in form and thickness, the stoutest being found for the
most part in the coarse longitudinal fibres, while more slender ones radiate thence to
the surface. The following may be regarded as the chief varieties : —

(1.) Styli (Plate XIII., fig. 7, a, b) ; fairly stout or slender, slightly curved, evenly
rounded off at the base, bluntly or sharply pointed at the apex ; size about 0"83 millim.
by 0"022 millim. ; passing into

(2.) Oxea (Plate XIII., fig. 7, c), of about the same dimensions, but more or less
sharply pointed at each end.

(3.) Strongyla (Plate XIII. , fig. 7, (/, e) ; more or less crooked, often very much so ;
size, say, about 1*2 millims. by 0"022 millim.

More slender forms of all occur, and the slenderer styli may be nearly as long as
the strongyla or much shorter than the stout styli whose measurement is above given.

This species is evidently nearly related to Auletta lyrata. differing chiefly in the
arrangement of the skeleton and the length of the tubes.

R.N. 73 (outside pearl banks, Gulf of Manaar) ; 148, 283 (fragment, both from
deep water oft Galle and onwards up West Coast of Ceylon).

Leucophloeus, Carter."1''

Axinellidre of massive habit, often clathrous. Skeleton reticulate, composed of stout
multispicular fibres with little if any spongin ; with a well-developed dermal
skeleton composed of a reticulation of spicule-bundles or a crust of tangentially
placed spicules. Megascleres typically stylote, sometimes oxeote. No microscleres.

* Viib' Carter (54), p. 323.

SPONGES. 197

This genus, which was never diagnosed by its author, is an extremely difficult one
to deal with. In our Report on the " Challenger " Monaxonida, Mr. Ridley and I
decided to suppress it as a synonym of Hymeniacidon, but it has lately been revived
by Thiele (39) as a distinct genus, and I am prepared to follow him in this respect.
I cannot quite see, however, why Thiele should, in the same work, have established
another genus (Arnorphilla) for the reception of closely similar forms (especially as he
gives no definite generic diagnoses), unless it be on account of the comparatively small
size of the spicules.

If we decide to separate Leucophlccus from Hymeniacidon, as I think we must, it
must be mainly on the ground of the presence in the former of a strongly developed
dermal skeleton (composed, mostly at any rate, of tangentially placed spicules) which
often forms a white crust in dry specimens.

Our knowledge of these sponges, however, is still very deficient, and their systematic
position very doubtful. I retain the genus amongst the Axinellidse on account of the
presence of stylote megascleres and the apparent relationship to Ciocalypta as pointed
out by Mr. Carter. Leucopthhcm foetidus, with its long oxeote megascleres, perhaps
comes nearer to Halichondria than any other species, and may indicate a close
relationship with the Renierinpe.

Leucophloeus foetidus (Dendy).

1889, Hymeniacidon (?) foetida, Dendy (3); 1897, Amorphinopsis fcetida, TorsEXT (83);
1898, Ciocalypta foetida, Lindgren (86).

There is one good specimen of this sponge in the collection, differing from the type
as originally described in its partially trabecular and clathrous external form, and in
the pale yellowish (not blackish) grey colour of the surface (in spirit).

I have already pointed out the resemblance which this species bears to Ridley and
Dendy's Hymeniacidon (?) subacerata, and it is not impossible that the presence of
the small projecting styli in the dermal membrane of both these species may ultimately
prove to be of generic import.

R.N. 63 (Gulf of Manaar).

Ciocalypta, Bowerbank.

Axinellidfe provided with elongated, digitiform processes springing from a massive

body. In the processes the skeleton is arranged in a plumose manner with a

dense central axis. There is a thin dermal membrane supported on spicular

columns and usually strengthened by a reticulation of tangentially placed spicules,

overlying extensive sub-dermal cavities. There are no microscleres.

Ciocalypta tyleri, Bowerbank.

There is in the collection one specimen which I must refer to the typical form of

this species (22). This specimen (R.N. 29), which I examined hastily in Liverpool,

has unfortunately not been sent out to me, as it appeared to be identical with another

which was sent instead (R.N. 29a), and the only preparation which I have of it (made

198 CEYLON PEARL OYSTER REPORT.

in Liverpool) shows none of the very large spicules characteristic of the variety
manaarensis previously described by me (3), though it is not impossible that such
may occur in other parts of the sponge.
R.N. 29 (Gulf of Manaar).

Ciocalypta tyleri, var. aberrans, nov.

The single specimen consists of a flattened, cushion-shaped body of oval outline,
with a number (now four, but apparently one at least has been broken off) of slender,
upright, digitiform processes springing from the upper surface. These processes are
long and conical, tapering gradually almost to a point, and with no visible vents ;
their walls, however, are pierced by numerous small dermal pores. The surface of
the sponge between the roots of the processes is rather uneven and somewhat rugose,
with branching canals radiating from the roots of the processes beneath the surface,
through which they are plainly visible ; these canals are continuous with the large
canals which run lengthwise through the digitiform processes, about four in each.
The general surface of the sponge is very minutely hispid ; the walls of the
processes have a minutely reticulate appearance and may also be very slightly hispid.
The texture of the basal part of the sponge (which has been cut off below) is fairly
compact, except for the numerous cylindrical canals, some of which run almost
vertically downwards from the bases of the digitiform processes. The colour of the
sponge (in spirit) is dull brownish -grey. The longer diameter of the base measures
about 45 millims., the shorter diameter about 24 millims. ; the thickness in the middle
of the base (now) is about 13 millims. The digitiform processes are about 29 millims.
long and 6 millims. in diameter at the base, from which they taper gradually to the
narrow, blunt apex.

The skeleton in the body of the sponge is a dense, irregular reticulation of loosely
scattered spicules, chiefly oxeote, of various sizes. There is no special dermal skeleton
and no extensive sub-dermal cavities, and the spicules at the surface do not differ in
size, form, or arrangement, from those below. In the digitiform processes there is a
central axis surrounded by about four longitudinal canals, separated from one another
by rather thin longitudinal septa radiating from the central axis. Both axis and
septa are crowded with spicules similar to those in the body of the sponge, but mostly
arranged lengthwise. From the more peripheral portions of the septa radiate loose
brushes of, for the most part, more slender and shorter oxea, which spread out beneath
the dermal membrane and support it on their apices, which may project slightly
beyond the surface. Although the more peripherally placed spicules of each brush
are inclined very obliquely to the surface, there is no true dermal reticulation, but the
dermal membrane may contain irregularly scattered spicules of various sizes.

S/ncules. — Oxea of various dimensions ; usually gently curved, symmetrical,
gradually and fairly sharply pointed at each end ; sometimes biangulate instead of
simply curved ; occasionally stylote with broadly rounded base. Size varying up to

SPONGES. 199

about 0m'(9 millitn. by 0*025 millim. ; sizes indiscriminately mixed in the body of the
sponge, partially sorted out as described above in the digitiform processes, measuring
in the surface brushes usually only about 0-377 millim. by 0"0082 millim.

This variety differs from the typical form of the species in the absence of the special
dermal skeleton of slender oxea placed tangentiallv.

R.N. -29a (Stat. XLIIL, off Kaltura, 22 fathoms).

Collocalypta, n. gen.

Axinellidaj consisting of a basal crust, from which isolated digitiform processes spring
vertically upwards. With a thick, collenchymatous ectosome which, in the
digitiform processes, is penetrated by wide longitudinal canals communicating
with the exterior by groups of small canals ending in dermal pores. Skeleton
consisting in the basal portion of erect plumose columns of megascleres ; in the
digitiform processes of an axial core of spicular fibre from which loose bands of
spicvdes radiate outwards between the longitudinal canals to surface conuli.
Spicules more or less cemented together by spongin. Without microscleres.

This remarkable genus bears a very striking superficial resemblance to Ciocalypta,
from which, however, it differs widely in the presence of the thick, collenchymatous
ectosome and the typically axinellid (plumose) character of the main skeleton, as well
as in the total absence of dermal skeleton (compare, however, Ciocahjpta tyleri, var.
aberrant), and the presence of abundant spongin cement.

Collocalypta digitata, n. sp. — Plate VII., fig. 6 ; Plate XIIL, figs. 1, 2.

Sponge (Plate VII., fig. 6) consisting of a flat, wide-spreading, encrusting base,
about 8 millims. thick, from which arise erect digitiform processes widely separated
from one another by irregular intervals. These processes vary much in size, the
largest in my possession is about 53 millims. high by 8 millims. in diameter in the
middle. They usually taper to a sharp apex and are, as a rule at any rate,
unbranched. The surface of the basal crust, between the processes, is smooth or
nearly so, but it may be slightly hispid, and it is covered by a good deal of foreign
matter in places. The digitiform processes have a distinctly conulose and, at the
same time, longitudinally corrugated surface. They bear no visible vents, but
numerous minute pores (now all closed). The colour of the sponge (in spirit) is pale
grey ; the consistence tough and fleshy.

The skeleton in the basal crust consists of stout, erect, plumose columns of spicular
fibre, closely crowded together. The spicules are arranged in the typical axinellid
manner, with their outer ends projecting obliquely upwards and outwards, while their
inner portions are cemented together by a considerable amount of spongin. Each
column is continued through the ectosome as a loose tuft of more slender and longer
spicules than those which compose its deeper portion, and the apices of these spicules
commonly project beyond the surface. There is no dermal skeleton. In the digitiform

200 CEYLON PEARL OYSTER REPORT.

processes we find a very dense, stout axial core of spicules more or less cemented
together by spongin, from which numerous loose bundles of spicules radiate obliquely
outwards and upwards into the surface conuli, beyond which their apices may project.
Here, again, there is no dermal skeleton.

Spicules. — Oxea (Plate XIII., fig. 1), of various shapes and sizes according to
situation ; (a.) in the columns of the basal skeleton, stout, sub-fusiform, slightly
curved, irregularly ended, variable in size, say about 0'5G millim. by 0-03 millim. ;
(b.) in the ectosome of the base, comparatively long and slender, slightly curved,
irregularly ended, measuring, say, about 0-88 millim. by 0-02 millim. ; (c.) in the
digitiform processes the spicules are mostly of the long and slender type just
described, but occasionally stout and comparatively short forms occur.

One of the most characteristic features of this sponge is the thick gelatinous
ectosome (Plate XIII., fig. 2, ect.), composed of collenchyma, with a hyaline, or some-
times finely granular, matrix containing an immense number of large, stellate,
connective-tissue cells. In the basal crust this ectosome is about 0-G5 millim. thick.
In the digitiform processes it forms an even thicker layer around the dense central
axis of spicular fibre (Plate XIII., fig. 2), but it is penetrated by large longitudinal
canals (I.e.), about half a dozen in number, and varying in diameter up to about
2 '2 5 millims. These canals are separated from one another by radially arranged
longitudinal septa (sept,), in which the bundles of spicules run out from the central
axis to the surface conuli. The ectosome on the outer sides of the longitudinal canals
is reduced to a membrane of varying thickness, containing no spicules and penetrated
by short, narrow canals which place the great longitudinal canals in communication
with the exterior. These short canals are arranged in groups at wide intervals, and
those of each group unite together into a single larger canal before opening into the
longitudinal canal. The dermal pores are doubtless arranged in corresponding
groups, probably with one pore at the end of each of the smaller canals, but they are
now all closed (their position is shown in Plate XIII. , fig. 2, d.p.). The fact that the
smaller canals unite together as they pass inwards to the great longitudinal canals
seems to indicate that this is an inhalant system. The arrangement of the exhalant
system I have not succeeded in making out, and the choanosome is not sufficiently
well preserved to enable me to give any details with regard to the flagellate
chambers.

The digitiform processes may contain a good deal of sand outside the spicular axis.

R.N. 74, 74a (outside pearl banks, Gulf of Manaar).

Order 4 : EUCERATOSA.

Non-calcareous sponges without siliceous spicules, but with a skeleton consisting of
horny fibres developed independently, i.e., not in relation to any pre-existing
spicular skeleton. (The skeleton is sometimes replaced or supplemented to a
greater or less extent by foreign bodies.)

SPONGES. 201

The study of the very interesting series of twenty-two species, by which the horny
sponges are represented in the present collection, has caused me greatly to modify
my views as to the phylogeny of this group. Hitherto I have, in common with
certain other writers on the subject, been in the habit of regarding it as a group of
polvphyletic origin, derived probably from several distinct groups of monaxonellid
sponges by substitution of spougin for spicules. This view I now believe to be true
only for a very limited number of horny sponges, which might be distinguished from
the true Ceratosa (or " Euceratosa," as I propose to term them) under the name
" Pseudoceratosa," until such time as our increased knowledge shall enable us to
assign them to their proper systematic positions. This is already possible in some
cases, as, for example, in certain species of the Chalinine genus Siphonochalina
(Spinosella), concerning which I observed as far back as 1887, in my memoir on the
West Indian Chalinine Sponges (75) : —

"Here we can trace in different species of the same genus the gradual degeneration
and disappearance of the spicules until we come down to forms like Spinosella
maxima, mini (Plate LXL), and Spinosella plicifera, D. and M. (Plate LVIIL,
fig. 5 ; Plate LX., fig. 1), which sometimes still contain traces of the spicules
imbedded in the horny fibre, and apparently on the verge of disappearance, while at
other times they contain no spicules whatever ; and yet the specimens with spicules
and those without are specifically indistinguishable."

As regards the great majority of the horny sponges, however. I feel convinced that
they form a natural and compact group, in which it is almost impossible to separate
even the genera from one another by hard and fast lines. Thus I am in close agree-
ment with Polejaeff, who summarizes (74) his own observations on the classification
of the group as follows : — " 117/// tJte exception of the genera Darwinella, IantheUa,
and Psammopemma, all genera are devoid of any properties separating them absolutely
from one another."

Lendenfeld (66) has endeavoured to show that the Ceratosa are divisible into
two great groups of very different phylogenetic origin, viz., " Monoceratina " and
" Hexaceratina." It would not be difficult to expend a very large amount of
criticism upon his system, but, without going into detail, I must remark that this
main sub-division appears to me to be wholly erroneous, and that the connection
between these two groups is so close that it is quite impossible to separate them from
one another ; while, instead of the " Hexaceratina " being derived from the Hex-
actinellida and the " Monoceratina " from the Monaxonellida, as Lendenfeld would
have us believe, it appears to me tolerably certain that the majority of the
"Monoceratina"^ are descended from ancestral "Hexaceratina," and the latter in
turn from MyxosjDongida.

Some justification of my views concerning the phylogeny of the Euceratosa will,

* It must be remembered that Lendenfeld's " Monoceratina " include both Pseudoceratosa and
Euceratosa ; the latter alone are here referred to.

2 D

202 CEYLON PEARL OYSTER REPORT.

I hope, appear in the course of the subsequent pages ; in the meantime I may give
the following summary of the conclusions at which I have arrived : —

In the first place it is pretty obvious that the Aplysillicbe (constituting a large part
of the so-called " Hexaceratina ") form the starting point of the evolutionary series
within the order. The primitive character of such genera as AplysiUa and
Darwinella is clearly indicated by the simple canal-system, the large sac-shaped
flagellate chambers, and the very simple skeleton of branched spongin-fibres, supple-
mented in Darwinella by detached spicules of spongin. The presence of these
so-called spicules at first sight seems to lend colour to Lendenfeld's views as to the
relationship between the Aplysillidse and Hexactinellida. It is very difficult to see,
however, how the horny spicules in question can have anything to do with the
siliceous spicules of the Hexactinellida ; their shape is extremely variable and they
are probably best regarded simply as isolated portions of the general spongin
skeleton, secreted by groups of spongoblasts which, for some unknown reason, have
become isolated from their fellows.

Altogether the Aplysillidse agree very closely in structure with the Myxospongida,
especially with the genera Halisarca and Hexadella, and it is not impossible that
the curious fibres of Halisarca may represent a rudiment of a horny skeleton. That
the spongin skeleton in the Aplysillidse has been developed quite independently of
that of the Monaxonellida, and with no relation to a pre-existing siliceous skeleton,
admits, I think, of little doubt. The character of the skeleton, consisting in the
simplest cases of a thin basal lamina of spongin, from which slightly branched fibres
spring vertically upwards and end in surface coiruli, without anastomosing with one
another to form a network, is quite different from what we find in typical horny
Monaxonellida, in which the spongin is originally deposited as a cement which binds
together the spicules of a reticulate skeleton, and in which, consequently, when the
spicules disappear, the spongin is left in the form of a network of horny fibres. A
very similar network of horny fibres appears, however, to have been independently
evolved in the higher Euceratosa.

This difference in the arrangement of the horny skeleton — in the one case in the
form of a network and in the other case in the form of separate tree-like fibres — has
given occasion to Minchin (12) to divide his " Grade" Keratosa into two orders, viz.,
" Dictyoceratina (= Monoceratina, Ldf.)" and " Dendroceratina (= Hexaceratina,
Ldp., para)," the latter group including only the family Aplysillidse. Minchin's
distinction cannot, however, be maintained as a basis of classification, for, as
Lendenfeld himself recognised, there are undoubted Aplysillids (e.g., Denclrilla
elegans, Lendenfeld) which possess a reticulate skeleton, and in the present report
I propose the new genus Megalopastas for such forms, of which two species occur in
Ceylon waters.

The importance of the genus Megalopastas lies in the fact that it forms a
connecting link between the Aplysillidse and Spongeliidse, and thus completely breaks

SPONGES. 203

down the distinction hetween " Monoceratina" and "Hexaceratina," or "Dictyoceratina"
and " Dendroceratina." The Spongeliidse, like the Aplysillidaa, have large sac-shaped
flagellate chambers, simple canal-system, and clear transparent ground-substance.
In Tact, they differ from the Aplysillidaa only in their reticulate skeleton and in their
habit of taking foreign bodies into the fibre. The so-called " pith " in the fibre is
also less obvious, but this is an extremely variable character, and one upon which we
cannot place very much reliance for purposes of classification. In Mcgalopastas
pulvillus, for example, one and the same section may show great differences in this
respect, some fibres showing a strongly marked pith, differentiated by its darker
colour, and others apparently having no pith at all (Plate XV., fig. 3), the difference
apparently depending upon differences in local conditions at the time when the fibre
is growing, which give rise to a more or less distinct lamination analogous to the
annual rings in a tree trunk.

Moreover, when we remember that Schulze has described (71), under the name
Spongelia spinifera, a species in which the arenaceous fibres do not form a network
at all, but are arranged in a tree-like manner, as in the genus Aplysilla, we see at
once that the distinction between the Spongeliidai and Aplysillidee is purely arbitrary,
though, as a matter of convenience, it may, perhaps, still be maintained. From the
Spongeliidse the transition to the Spongiidas, by complication of the canal-system,
reduction in the size of the flagellate chambers and granulation of the ground
substance between them, is very simple.

I therefore conclude that the Euceratosa are a natural group descended from the
Myxospongidae, that their evolution starts with the Aplysillidaa and ends with the
Spongiidae, between which the Spongeliidse occupy an intermediate position, and that
the reticulate skeleton of the higher types has been independently evolved from a
more primitive dendritic skeleton.

Family: APLYSILLID.E.

Euceratosa with a dendritic or reticulate skeleton comjmsed of spongin-fibres
containing a more or less distinct pith, but usually without foreign inclusions ;
sometimes also with isolated spicules of spongin ; with a lacunar canal-system
and large sac-shaped* flagellate chambers opening by wide mouths direct into
wide exhalant lacuna?.

Darwinella, Muller.

Aplysillidaa with a dendritic skeleton and with isolated spicules of spongin.

Four species of this remarkable genus have been described, viz., D. aurea,
Muller (67) ; D. australiensis, Carter (18) ; D. joymxi, Topsent (89) ; and
D. simplex, Topsent (84) ; but it appears to me somewhat doubtful whether they
should all be regarded as specifically distinct from one another.

2 D 2

204 CEYLON PEARL OYSTER REPORT.

Darwinella simplex, Topsent, 1892 (84, vide also 62)— Plate XV., figs. 1, 2.

The single specimen forms a thin crust extending over a considerable area on the
surface of a branching tube which has apparently belonged to some annelid worni.
The surface of the sponge is glabrous and covered with sharp-pointed conuli, about
1 to 1*5 minims, in height and some 2 to 3 millims. distant from one another. The
maximum thickness of the crust is only about 2 millims. The dermal membrane is
minutely reticulate and lifted up in a tent-like manner on the ends of the vertical
skeleton fibres to form the surface conuli. The colour (in spirit) is dark purple,* and
the texture very soft and tender. The vents are inconspicuous, only one small one
having been detected. The inhalant pores are abundantly grouped in pore-sieves,
which occupy the oval or rounded meshes of the dermal reticulation.

The skeleton consists, in the first place, of sparingly and irregularly branched,
pithed horny fibres (Plate XV., figs. 1, 2) of the usual Darwinella type, which rise
more or less vertically from the base of the sponge and terminate in rounded apices
in the surface conuli. These fibres have a diameter of about 0-lG5 millim. near the
base, diminishing to about half as much in the conuli. The base of the fibre is
expanded into a thin plate of spongin, doubtless attached to the substratum ; the
fibres themselves do not appear to form any anastomoses. The wall of the fibre
is only about O'Ol millim. thick, and the interior is generally occupied by a much-
branched filamentous fungus (?), composed of rows of short cells, which more or less
completely replaces the pith in the older parts of the fibre (Plate XV., fig 2). In the
younger parts of the fibre (Plate XV., fig. 1) the pith exhibits the characteristic
thimble-shaped layers described and figured by Lendenfeld in D. aurea.

In the second place we have horny spicules of the usual Darwinella type, but
rather sparingly developed and, so far as I have been able to ascertain, all of the
triradiate form. They are irregularly scattered through the soft tissues, and I have
seen no union between them. The rays are long, slender and tapering ; say about
0'5 millim. long by about 0'025 millim. thick near the base.

The canal-system and general anatomy agree very closely with the description and
figures given by Lendenfeld (66) for D. aurea.

The Ceylon specimen agrees very well with Topsent's descriptions of the species,
but the horny spicules (in the single specimen available) do not appear to attain so
large a size. The species has hitherto been recorded only from the Mediterranean
and the Azores.

R.N. 302 (Ceylon seas).

Megalopastas, n. gen.
Aplysillidse with an entirely reticulate skeleton and without spongin spicules.

As I have already pointed out, Lendenfeld (66) includes in his genus Dendrilla
both species (e.g., D. rosea) without and species (e.g., D. elegans) with a reticulate

* Vide footnote under Iotrochota purpurea.

SPONGES. 205

skeleton. It is very doubtful, however, whether the genus, as constituted by
Lendenfeld, is separable from the older Aptysilla. The type species of Dendrilla
appears to be a form with a dendritic skeleton, viz., D. rosea, and if we are to
separate the species with reticulate skeleton, we can, for etymological reasons, hardly
employ the name Dendrilla for them.

In my report on Mr. Thurston's Second Collection of Sponges from the Gulf of
Manaar (3) I attempted to avoid the necessity for erecting a new genus by employing
Bowerbank's old name Spongionella for what must now be regarded as the type of
the genus Megalopastas, viz., Megalopastas nigra. I must admit, however, that
Bowerbank's type of the genus Spongionella, viz., S. pulehella, is probably not an
Aplysillid at all (according to Lendenfeld it is a "Leiosella"), while, on the other
hand, Bowerbank (8) also applied the name Spongionella to another totally different
sponge, viz., Phgllospongia {Spongionella) holdsworthi, and Spongionella is retained
by Lendenfeld (66) as a sub-genus of Phyllospongia. I therefore now revert to
what was my original intention in 1889, and propose the new genus Megalopastas,
the name being chosen in allusion to the large size of the flagellate chambers.

The anatomical characters of the genus are shown in Plate XV., fig. 4, which
represents, somewhat diagrammatically, a vertical section of an Australian species,
M. elegans (Lendenfeld's Dendrilla. elegans), of which I happen to have much better
preserved material than of the Ceylon species. Excepting that in M. elegans the
ectosome is thicker and the outermost secondary fibres of the skeleton do not lie so
near the surface, so that there is no " dermal skeleton," the figure would serve almost
equally well for either of the Ceylon species. My preparations of M. elegans contain
numerous embryos, enclosed in spherical endothelial cajisules, and mostly in the stage
represented in the figure, consisting of a solid inner mass of cells surrounded by an
outer layer modified near one end to form a pigment ring. This embryo agrees
pretty closely with those described by Schulze in Spongelia (71) and Euspo}igia (72),
and its occurrence perhaps tends to show the correctness of my views as to the close
relationship of the so-called " Hexaceratina " to the other Euceratosa.

Megalopastas nigra (Dendy). — Plate XIV., fig. 7 ; Plate XV, figs. 5-8.
1889, Spongionella nigra, Dendy (3).
This very remarkable species was originally discovered by Mr. Thurston and is
represented in Professor Herdman's collection by two specimens. The sponge
(Plate XIV., fig. 7) is sessile, and consists of a number of vertical lamellae, branching
and anastomosing with one another, often in a very complex manner. The largest
specimen I have seen was about 250 millims. high and the same in breadth, with
lamellae about 5 millims. thick. The colour of the living sponge is black (Thurston),
when dry, dull black, and in spirit rather lighter, blackish-grey. Texture (in spirit)
very compressible and resilient ; moderately tough. Surface granulated, the granules
being really minute, close-set conuli. Vents abundantly scattered, but almost or

•20(5 CEYLON PEARL OYSTER REPORT.

quite confined to one surface of each lamella. The vents are compound, each con-
sisting of an aggregation of several smaller ones, the entire group only from 1 millim.
to 2 millims. in diameter.

The main skeleton (Plate XV., fig. 5) is a rectangularly meshed network of very
distinct primary and secondary fibres, the primaries running vertically to the surface
and the secondaries crossing them more or less at right angles. The primary fibres
average about 0"049 millim. in diameter, and the secondaries about half as much.
The outermost secondary fibres form a well-developed dermal or sub-dermal skeleton
(Plate XV., fig. 5, d.s. ; fig. 6), in the form of an irregular network with fairly wide
polygonal meshes, the fibres averaging about 0'02 millim. in diameter.

There are many more minute conuli on the surface than there are of the stout
primary fibres, and those which do not contain the apices of such fibres are supported
by short fibres which spring vertically from the tangential fibres of the dermal
skeleton (Plate XV., figs. 5, 8).

All the fibres of the skeleton are composed of pale-coloured spongin without any
foreign enclosures. They are not distinctly " pithed " except at the growing apices
(Plate XV., fig. 7), where the usual thimble-shaped layers of spongin are added one
on top of the other, as in other Aplysillidte, but even here the " pith " is not distinctly
differentiated.

The ectosome forms a thin dermal membrane containing the inhalant pores, and
the choanosome is very delicate and gelatinous. The canal-system agrees closely with
that of Aplysilla, as described and figured by Schulze (70). It is lacunar, and the
flagellate chambers are sac-shaped and large, averaging when full-grown about
0"07 millim. in diameter. They are not placed very close together and they open
directly into the excurrent lacunas, without special exhalant canaliculi.

R.N. 71, 161a (Pearl banks, Gulf of Manaar — not uncommon).

Megalopastas pulvillus, n. sp. — Plate XV., fig. 3.

The single specimen has the form of a small flattened cushion, slightly convex
above and (has been) attached by a broad flat base below. The outline of the
specimen is irregularly rounded. The upper surface bears several small, compound
vents, each about 2 millims. in total diameter; it also appears granular from the
presence of numerous minute, slender, sharp-pointed conuli containing the ends of the
primary fibres. The inhalant pores are conspicuous under the microscope in small
groups in the thin, translucent dermal membrane. Texture (in spirit) firm, but
compressible and resilient ; colour, pale yellowish -grey. Diameter of specimen about
18 millims. ; thickness in the middle about 5 millims.

The skeleton (Plate XV., fig. 3) is a partly rectangular- and partly polygonal-
meshed network of pale amber-coloured homy fibre, in which the primary fibres are
very clearly differentiated, radiating towards the surface and terminating in long,
slender apices in the surface conuli. The primary fibres not infrequently branch,

SPONGES. 207

while at the base of the sponge they are seen to originate in a thin horizontal spongin-
lamella, which evidently forms the means of attachment to the substratum. The
primary fibres are about 0'09G millim. in diameter, the secondaries are usually a good
deal more slender, say about 0"04 millim. in diameter, but variable. Both primary
and secondary fibres frequently exhibit a very distinct pith of variable thickness,
which evidently simply represents the older part of the fibre surrounded by fresh
accretions of spongin. Even in old parts of the primary fibres the old slender apices
may frequently be observed thus imbedded in the new growth (Plate XV., fig. 3, ap.).
The principal secondary fibres run across at right angles between the primaries, but
numerous others run in various directions. The outermost secondary fibres form a
pretty definite dermal skeleton with polygonal meshes of very variable size. Here
and there short, vertical, gradually tapering branches arise from this network and
enter some of the surface conuli. These branches evidently form the commencements
of new primary fibres, as in M. nigra.

The ectosome is feebly developed, forming a thin dermal membrane containing the
inhalant pores and overlying the sub-dermal cavities. The canal-system is lacunar.
The flagellate chambers are large and sac-shaped, about 0-088 millim. in longer
diameter ; they open by wide mouths direct into wide exhalant lacunas. The ground
substance between them is very feebly developed, gelatinous-looking and broken up
into trabecular by the smaller inhalant lacunar in the characteristic aplysillid fashion.
The larger exhalant canals converge towards the compound vents, where they open
to the exterior.

It will be seen from the above description that this species agrees closely in
skeletal characters and canal-system with M. nigra, but it differs widely in external
form and colour (in spirit). The fact that the specimen contains large ova and
embryos in endothelial capsules likewise seems to indicate that it is not merely a
young form of M. nigra. Owing to the much greater distinctness of the pith in the
horny fibres the species is a more typical aplysillid than its congenei*.

K.N. 191 (Muttuvaratu Paar, Gulf of Manaar).

Family : SPONGELIID^.

Euceratosa with a (usually) reticulate skeleton of horny fibres without distinct pith,
but containing foreign bodies ; or with a skeleton composed of foreign bodies
united together by little if any spongin. With lacunar canal-system and large,
sac-shaped flagellate chambers opening directly by wide mouths into wide
exhalant lacunar.
This family may be retained as a matter of convenience, but it is, as I have

already indicated, logically impossible to separate it sharply from the Aplysillidar, for

the genus Megalopastas, on the one hand, and Schulze's Spongelia spinifera, on the

other, are strictly intermediate between the two groups.

208 CEYLON PEARL OYSTER REPORT.

Spongelia, Nardo.

Spongeliidse with a skeleton composed of distinct, but more or less arenif'erous, horny
fibres.

Spongelia fragilis, Montagu, var. ramosa, Schulze.

1879, Spongelia pallescens, sub-species fragilis, var. rarnosa, SCHULZE (71); 1889, Spongelia
fragilis, var. irregularis, Lendenfeld (66), pars.

There is a single, partly macerated specimen of this variety in the collection. The
external form appears to have been loboso-digitate. The skeleton network is very
coarse, with rectangular meshes, and densely charged throughout with sand. The
soft tissues are very densely charged with chains of algae, probably Oseillana
sjiongelue. The large sac-shaped flagellate chambers measure up to about 0"1 millim.
in longer diameter.

Under the name Spongelia fragilis, var. irregularis, Lendenfeld has already (66)
recorded a sponge from Ceylon which probably belongs to this variety, as he himself
admits that his var. irregularis partly corresponds with Schulze's var. ramosa. The
species, at any rate, appears to be cosmopolitan, if not the variety also.

R.N. 307 (Ceylon seas).

Spongelia elastica, var. lobosa, Schulze.

1879, Spongelia pallescens, sub-species elastica, var. lobosa, Schut.ze (71); 1889, Spongelia
elastica, var, lobosa, Lendenfeld (66).

There is one specimen of this variety in the collection, consisting of a massive,
columnar and slightly clathrous basal portion, giving off numerous short, irregular,
digitiform processes above. The surface is covered with acute conuli, larger and
further apart on the lower than on the upper portions of the sponge, varying from
about 0"5 millim. to 2'5 millims. in height. Between the conuli is stretched the
usual reticulate dermal membrane, and the lines of the dermal reticulation are
areniferous. The texture (in spirit) is soft and elastic and the colour pale greyish-
yellow. The specimen measures about 58 millims. in height by 40 millims. in
maximum diameter. The apices of the branches are commonly occupied by parasitic
barnacles, each enclosed in a cavity which opens to the exterior by a small terminal
aperture resembling a vent.

The primary fibres of the skeleton, ending in the conuli, are abundantly charged
with foreign matter, chiefly sponge-spicules, while the connecting fibres are almost
free from foreign matter, and form a moderately close network. The canal-system
and histology agree closely with Schulze's classical description. The large, sac-
shaped flagellate chambers, about 0"08 millim. in maximum diameter, are imbedded in
a sparsely developed gelatinous ground substance. They have numerous prosopyles
and open directly by wide mouths into the exhalant canals.

SPONGES. 209

This variety is well known in the Mediterranean and, according to Lendenfeld.
occurs also in the North Atlantic and in Australian seas.

R.N. 1G5 (deep water oft' Galle and onwards up West Coast of Ceylon).

Spongelia elastica, var. crassa, nov. — Plate XIV., fig. 4.

This variety differs from var. lobosa in the much coarser main fibres, filled with
much larger foreign particles (sand grains), and in the more areniferous character of
the connecting fihres, which, however, are still occasionally quite free from sand. In
external appearance (Plate XIV., fig. 4) the single specimen closely resembles the
specimen of var. lobosa described above, but it may readily be distinguished by its
much more rigid and coarsely arenaceous character. The very stout main fibres
contain sand grains of extremely various size. The network of connecting fibres is
very irregularly developed and sometimes is absent over wide areas. The sponge is
infested by numerous small cha3topod worms imbedded in the soft tissues.

R.N. 35 (GulfofManaar).

Spongelia incrustata, n. sp.

Sponge compressed, irregularly lobose, proliferous. Surface with small acute
conuli irregularly scattered over it at varying intervals, and only about 1 millim. in
height ; minutely reticulate over large areas, while over areas quite as large the
dermal reticulation is completely obliterated by the sand cortex. Vents rather small
and mostly on the margins of the lobes. Texture (in spirit) rather soft, compressible,
flaccid, cavernous internally. Colour yellowish-grey throughout. The largest
specimen is about 100 millims. in maximum diameter, and the lobes are generally
not more than 5 millims. or 6 millims. thick, though variable.

The skeleton consists in the first place of the dense arenaceous cortex, which is
well developed everywhere, except in the thin dermal membrane of the pore-areas
which lie between the meshes of the dermal reticulation (where j>resent). This cortex
is about 0-5 millim. thick. Internally the skeleton also consists chiefly of sand-
grains, arranged in irregular tracts rather than in well-defined fibres, but often held
together by spongin cement. Between these tracts are wide areas free from sand.

The canal system is that of a typical Spongelia. The sub-dermal cavities, under-
lying the pore-sieves, are large, and the whole canal-system is lacunar in a high
degree. The flagellate chambers are sac-shaped and up to about 0'08 millim. in
longer diameter, opening by wide mouths into the exhalant canals and provided with
numerous prosopyles. The mesoglcea between them is very scantily developed, clear
and gelatinous, but with numerous stellate connective-tissue cells. The walls of the
larger exhalant canals contain numerous elongated muscle-cells.

This species appears to be nearly related to Lendenfeld's Spongelia laxa from the
South Coast of Australia (66), hut the surface conuli are more or less acute instead of
rounded, the skeleton fibres are apparently much less well-defined, and the sand
cert ex is apparently much thicker.

2 E

210 CEYLON PEARL OYSTER REPORT.

R.N. 7- (outside pearl banks, Gulf of Manaar) : 279 (deep water off Galle and
onwards up West (.'oast of Ceylon).

Psainiuopeinnia, M a kshall.

Spongeliidse in which the skeleton is composed of densely aggregated sand-grains
which are more or less connected together by spongin.

Psaminopemina crassum (Carter), var. clathrata, now

1885, Holopsamma crassa, CARTER (18) : 1889, Psammopemiua crassum, LENDENFELD (66).

The specimens differ from the types as described by Mr. OARTEB in their clathrous
form and in the absence of large conspicuous vents. They are extremely irregular
and triable, and densely charged with coarse sand, over which a distinct pellucid
dermal membrane, tree from sand, is frequently stretched. The surface is very
uneven, sometimes ribbed and sometimes conulose ; the texture incompressible but
fragile, and the colour (in spirit) pale brown.

The sand grains are not arranged in distinct fibres, but in ill-defined bands forming
an irregular reticulation. They are connected at the points of contact by a very
small quantity oi' spongin cement.

The flagellate chambers are large (about 0'08 millini. in longer diameter) and sac-
shaped, and open direct into the exhalant canals by wide mouths. The ground
substance between them is clear and transparent and very sparsely developed, except
in the neighbourhood of the larger canals, which are surrounded bv an abundant
gelatinous collenchvma. with numerous stellate or fibrous connective-tissue cells.

The types of the species came from Australia, and Lkxdenfeld also records it from
New Zealand.

R.N. 64 (type of variety. Gulf of Manaar) : 330 (Ceylon seas).

Family: SPONGIID^E.

Euceratosa with a reticulate horny skeleton and with small, more or less spherical
flagellate chambers, commonly provided with special narrow exhalant canaliculi.
The ground-substance between the chambers is compact and densely charged
with fine granules.

Cacospougia. Schmidt.

Spongiidfie with a very wide-meshed skeleton network, and with distinctly lamellated
horny fibres which are usually of a brown colour and of very variable diameter.

Cacospougia scalaris. Schmidt.

1862, Cacospougia scalaris, Schmidt (47): 1870, Cacospougia scalaris, SCHULZE (72);
1889, Stelospougia scalaris, Lexdexfei.b (66 1.

There are several rather small specimens in the collection which agree closely with

SPONGES. 211

the Mediterranean form as described and figured by SCHULZE. The connecting
fibres are often very irregular in arrangement and very variable in diameter. The
histological features are not very well preserved, but the flagellate chambers probably
have elongated exhalant canaliculi, as described by Schulze.

The main fibres may contain many foreign bodies. What can Lendexfeld mean
by saying that the fibres are all of uniform diameter and never contain foreign bodies,
in direct opposition to the observations of Schmidt and Schulze ?

K.N. 18 (Gulf of Manaar) ; 1G2 (three specimens) ; 347 (all Ceylon seas).

Euspongia, Bitoxx.

Spongiidse of compact structure and with a very fine-meshed skeletal network of
slender and fairly uniform horny fibre. Primary fibres, usually containing foreign
bodies, radiate towards the surface and are connected together by a close but
very irregular-meshed network of more slender connecting fibres free from
foreign bodies.
Lendexfeld (66) has already recorded several species of Euspongia from Ceylon,
viz. : —

E. trincomaliensis, which he identifies with one of Hyatt's American varieties

of " Spongia officinalis" ;
E. irregularis, var. pertusa, which is also recorded from North America, Australia,

and the tropical Pacific ; and
E. irregularis, var. dm-a, also recorded from Madagascar and Australia.

There are two forms in Professor Herdmax's collection which appear to be cmite
distinct from all these.

Euspongia officinalis, Auctorum, var. ceylonensis, nov. — Plate XIV., fig. 3 ;
Plate XVI., fig. 5.

There are in the collection several nice specimens of a bath sponge which obviously
belong to a variety of Euspongia officinalis, closely resembling E. officinalis, var.
rotunda, of Hyatt (69) and Lendenfeld (66). The latter is one of the American
varieties, and there is probably sufficient difference in the Ceylon sponge to merit
recognition under a new varietal name.

The sponge is massive and compact, without vestibules Of the two specimens
which I now have before me, the one (in spirit) is slightly elongated vertically and
at the same time slightly compressed laterally and slightly flattened on the top, while
the base is somewhat constricted and charged with pebbles and other foreign matter.
The vents are rather large and conspicuous, scattered on the top and sides of the
sponge, from 2-5 millims. to 5 millims. in diameter, sometimes more or less prominent.
Each vent forms the termination of a long vertical oscular tube, of the same diameter
as itself. The surface is thickly and uniformly covered with minute, low conuli

2 E 2

212 CEYLON PEARL OYSTER REPORT.

(about 0"8 millini. apart, from apex to apex), from the summits of which fine ridges
radiate into the intervening valleys, where they branch and anastomose with one
another, so as to give rise to a delicate reticulation which is scarcely visible to the
naked eye, and in the meshes of which the dermal pores are situate in small groups.

The colour of the surface in spirit is black, paling to grey below and internally, and
the texture compact, but compressible and very elastic. The specimen thus described
measures about 90 millims. in height bv 73 millinis. in greatest breadth. Another
somewhat similar spirit specimen has been photographed for me by Professor Herdman,
and is represented, about two-thirds natural size, in Plate XIV., fig. 3.

Professor Herdman has also sent me part of the macerated skeleton of a much
larger specimen. The piece sent is a segment of what appears to have been a massive,
hemispherical sponge, with very strongly convex upper surface and broad flattened
base. Numerous large vents, up to 8 millims. in diameter, are scattered singly over
the upper surface, each at the end of a long, vertical oscular tube. Between these
vents the surface is rather uneven (but not channelled or grooved) and honeycombed
by close-set narrow vertical canals (inhalant) about 0"5 millim. in diameter, which
reduce the skeleton reticulation to a mesh-work of thin trabecular which terminate at
the surface in small, slightly projecting villi. The texture, after soaking in water, is
very soft and elastic and not very tough, and the colour is pale greyish-yellow. The
specimen from which this piece was taken must have been about 200 millims. in
diameter by 100 millinis. in height.

The primary fibres of the skeleton (Plate XVI., fig. 5) run parallel with one another
(at distances of about 0-8 millim.) towards the surface, where they end singly in the
conuli. They rarely branch, and apparently never anastomose. They are about
0"04 millim. in diameter and composed chiefly of broken sponge spicules, with
comparatively little spongin cementing them together. They are connected with one
another by a polygonal-meshed network of secondary fibres, in which the meshes vary
greatly in size and shape, while the fibres are of fairly uniform diameter and only
rarely contain foreign matter. Average diameter of meshes, say, about 0-17 millim. ;
diameter of the secondary fibres themselves, when fully developed, about 0"02 millim.,
but often less.

Owing to the quantity of broken spicules which they contain, the primary fibres
are distinctly visible to the naked eye in the macerated sponge, appearing as very
fine threads of a paler colour than the rest of the skeleton, and thus constituting what
is perhaps the most obvious distinctive feature of the variety.

The dermal membrane contains numerous broken sponge-spicules scattered
through it.

In internal anatomy this variety agrees minutely with Schulze's classical account
of the bath sponge (72), so that it is unnecessary to describe the canal-system and
histology in this place. One point perhaps deserves mention, and that is the very
strong development of long bands or cords of granular fibrous cells, running through

SPONGES. 213

the sponge in various directions (but mostly more or less at right angles to the
surface), and not by any means confined to the neighbourhood of the larger canals.
These bands are, as already suggested by ScHULZE for closely similar structures in the
Adriatic bath sponge, probably muscular, and their function appears to me to be to
effect the contraction of the sponge as a whole and thus squeeze the water out very
thoroughly when necessary.

This Ceylon bath sponge appears to differ from most of the varieties of Euspongia
officinalis in the greater slenderness of the very pale-coloured secondary or connecting
fibres of the skeleton. Schulze gives the average thickness of these fibres in
Eusptmgia officinalis as 0-03 millim. to 0-035 millim., while in our variety they
attain only a thickness of about 0*02 millim., and are often less.

Owing to this character the sponge acquires a remarkable softness, but at the same
time loses somewhat in durability. This want of durability may impair its value as
an article of commerce, but in view of its softness and elasticity and its great absorbent
power, combined with its good shape and size, I am inclined to think that it would
be worth while to experiment in the way of placing it upon the market if it can be
obtained in sufficient quantity.

R.N. 37 (several specimens in spirit) ; 101 (dry). All from Trincomalee.

Euspongia tenuiramosa, n. sp.

Sponge consisting of irregular, slender branches, usually only about 5 millims. or
6 millims. in diameter and apparently repent. The branches are more or less angular
or nodose, and usually very crooked. The surface is more or less concealed by coarse,
calcareous debris, such as the shells of large Foraminifera, &c. Where free from
foreign matter, it is covered with small, sharp conuli, scattered at very various intervals.
Vents small and few ; pores not observed. Colour (in spirit) purplish-brown or
nearly black. Texture, where free from foreign matter, which occurs internally as
well as at the surface of the sponge, compressible and resilient. The longest branch
is about 60 millims. in length by 5 millims. in diameter, but the diameter varies
much.

The skeleton is an irregular but fairly close, polygonal- or sometimes quadrangular-
meshed network of rather dark-coloured fibre varying much in diameter. Here and
there primary fibres can be recognised running into the surface conuli, and these may
contain a rather slender core of broken sponge-spicules. The primary fibres have a
diameter of about 0"05 millim. The connecting fibres vary from about the same
diameter downwards to about 0-008 millim. The diameter of the meshes varies so
much that it is useless to give measurements.

The material is not very well preserved for histological investigation, but, so far
as 1 have been able to make out, the internal anatomy offers no features of special
interest and agrees closely with that of other species of the genus. The flagellate
chambers are small, about 0'024 millim. in diameter, and nearly spherical, and the

214 CEYLON PEARL OYSTER REPORT.

ground-substance between them is finely granular. I have not been able to make out
the nature of their openings. The usual cylindrical cords of elongated fibre-cells are
present, and there is a collenchymatous ectosome. There appears also to be a thin
cuticle similar to that of various species of Hippospongia.

As regards skeletal characters this species agrees closely with Lendenfeld's very
comprehensive Euspongia irregularis, but the external appearance is so characteristic
that it seems to deserve a distinct specific name.

li.N. 311 ; 331) (Yard Cove, Trincomalee, shallow water).

Hippospongia, Schulze.

Spongiida; of clathrous structure, but otherwise resembling Euspongia, except that
the skeleton fibre may be much coarser and the whole sponge harder.
Except in its harder and more incompressible character, I cannot see that
Lendenfeld's genus Hyatella differs from Schulze's Hippospongia, yet Lendenfeld
himself describes a Hippospongia dura which is, perhaps, as hard as any Hyatella.

Hippospongia intestinalis (Lamarck).

1813, Spongia intestinalis, Lamarck (73); 1877, Spongelia velata, Hyatt (69); 1884,
Hippospongia intestinalis, Ridley (16) ; 1889, Hyatella intestinalis, Lendenfeld (66).

This species is represented in the collection by several specimens of elongated
tubular form, very intestinal in appearance and of a light brown colour, with their
walls perforated here and there at irregular intervals, and the surface slightly
conulose. The tubes may branch and anastomose, but are for the most part well
separated from one another and usually about 10 millims. in diameter, but variable.
The characteristic surface reticulation of slender horny fibre is very well developed,
and the surface appears to be covered by a remarkable continuous but separable
cuticle, which possibly has some connection with the dermal skeleton, but I have not
been able to elucidate its true nature. The main skeleton is very irregular and
composed of mostly stout amber-coloured horny fibre. Primary fibres cored with
foreign bodies are visible here and there.

The flagellate chambers are small (about 0'03 millim. in diameter) and approxi-
mately spherical, and the ground-substance between them is finely granular. The
special exhalant canaliculi, if present, are short. Stout bands of elongated fibres,
presumably muscular and mostly longitudinal in direction, are developed as in other
Spongidse, and there may be a good deal of collenchyma around some of the larger
canals.

The species has been recorded from the Mediterranean, Zanzibar, the Mascarene
Islands, and the Amirante Group (66), and from Porto Rico (29).

R.N. 65 (Gulf of Manaar); 83 (deep water off Galle and onwards up West
Coast) ; 337 (Ceylon seas).

SPONGES. 215

Hippospongia clathrata (Carter) — Plate XIV., %. 2.

1881, Hircinia clathrata, CARTER (5); 1887, Hircinia clathrata, Dexdy (2); 1889, Hircinia
clathrata, Dundy (3); 1889, Hyatella clathrata, Lendenfeld (66).

There is a single specimen of this well-characterised sponge in the collection.
Lendenfeld (66) has added particulars as to the canal-system to our previous
knowledge. The sponge contains no filaments, but the bands or cords of fibrous
tissue in the choanosome, so characteristic of many other Spongidse, are very well
developed. The species was recorded by Carter from the Gulf of Manaar and the
Red Sea, and it also occurs in Australia (Dendy, Lendenfeld) and on the American
coast of the North Atlantic (Lendenfeld).

The figure represents the characteristic appearance of a spirit specimen, no good
illustration of the external form having yet been published.

R.N. 24 (Gulf of Manaar).

Hippospongia anomala, Polejaeff.

1884, Hippospongia anomala, Polejaeff (74) ; 1889, Hippospongia anomala, Lendenfeld (66).

There is one specimen in the collection which, although of smaller size, agrees very
well in nearly all respects with Polejaeff's description and figure of the type
specimen from Torres Straits. Our specimen consists of an erect, sub-cylindrical
cavernous body, expanding gradually below and giving off two short digitiform
processes on one side. The interior, especially in the lower portion, is sub-divided by
trabecule, and the surface is covered over by a thin, parchment-like dermal membrane
pierced by numerous larger and smaller circular apertures, especially abundant
towards the extremity. These apertures lead into the large vestibular spaces in the
interior of the sponge. The end of the main body and those of the two processes
taper suddenly to rather sharp apices, and are covered with fairly numerous small
conuli, elsewhere the surface is nearly smooth and glabrous and provided with a
delicate cuticle, like that of Hippospongia intestinalis. Polejaeff's expression
" shagreen-like" may refer to the same character. The colour (in spirit) is pale grey,
the texture compressible and very resilient. The height of the specimen is about
120 millims. and the maximum diameter at the base about 50 millims.

The skeleton is a pretty close but irregular polygonal-meshed network of fibres of
very uniform diameter, about 0-02 millim. to 0-03 millim. thick, and free from foreign
bodies. Occasionally only one observes much stouter primary fibres, composed
principally of broken sponge spicules, running towards the surface and sometimes
branching. Occasionally also one sees very slender connecting fibres amongst the
ordinary ones ; these are probably young. There is no specially differentiated dermal
skeleton, but the main skeleton reticulation comes close to the surface. A good many
broken spicules are scattered in the dermal membrane.

The canal system is remarkable for the unusual length of the exhalant canaliculi
of the flagellate chambers, reminding one of the similar condition described by

21(5 CEYLON PEARL OYSTER REPORT.

Schulze (72) in Cacospongia scalaris. Polejaeff (74) has already called attention
to variability in the development of the cameral canaliculi in this species, so that we
can hardly consider the presence of long exhalant canaliculi as a character of specific
value.

R.N. 56 (Gulf of Manaar).

Hippospongia dura, Lendenfeld (66).

The single specimen consists of an irregular massive body tapering gradually
upwards (?) or on one side (?) into a sub-cylindrical fistular process, the conical
extremity of which is perforated by numerous small, round apertures leading into the
internal vestibular space. Similar apertures are scattered more sparsely on other
parts of the sponge. Internally the sponge is cavernous, with wide, sub-cylindrical,
vestibular spaces, more or less sub-divided by trabeculse of smaller diameter than the
vestibules. The vestibular spaces are covered in at the surface partly by a somewhat
parchment-like dermal membrane and partly by superficial extensions of the trabecule
themselves. The surface is rather uneven, but not distinctly conulose ; a consider-
able amount of calcareous and other foreign matter is attached to it. The texture,
in spirit, is hard and only slightly compressible, and the colour rather dark brown
throughout, The specimen measures about 110 millims. in greatest length (height?)
and 60 millims. in greatest breadth at right angles to the length.

The skeleton is an unusually close and fairly uniform network of rather stout,
amber-coloured horny fibre. There is no distinction between primary and secondary
fibres, except at wide intervals, where the network becomes somewhat closer and
forms stout columns radiating towards the surface. In these columns many primary
fibres run side by side and nearly parallel with one another, connected at frequent
intervals by short, transverse secondaries to form a very stout, but ill-defined, com-
pound, trellis-like fibre. Elsewhere the stouter fibres frequently run parallel with
the surface of the sponge and are connected together by more slender fibres which
run transverse to the stouter ones. In other places again the network is quite
irregular. The stouter fibres, forming the bulk of the skeleton, are about 0"04 millim.
or 0'05 millim. in diameter; the more slender ones are very variable. Usually the
fibres are quite free from foreign bodies, but broken spicules may be found occasion-
ally in fibres of the trellis-like groups. There is no special dermal skeleton, but the
ordinary reticulation of fibres comes close to the surface, which is covered by a thin
cuticle like that of Hippospongia intestinalis and II. anomaht. A remarkable feature
of the skeleton fibre is the brilliant yellow colour which it assumes in sections stained
with picro-carmine. The superficial cuticle stains in the same way, and sometimes
appears to be continuous with the more superficial skeleton fibres, but this is a point
which requires further investigation.

The flagellate chambers are nearly spherical, only about 0'024 millim. in diameter,
and the ground-substance between them is finely granular. There is a rather thin,

SPONGES. 217

collenchymatous ectosome, containing a good many brown pigment cells. Cylindrical
bands or cords of elongated fibrous cells are developed as usual.

Lendenfeld describes the species, apparently from a dry specimen, from the
American coast of the North Atlantic. Under these circumstances the identification
may seem somewhat hazardous, but the species is so well characterised by its general
form, its colour and texture, and its peculiar skeleton arrangement, that I do not
think there can be much doubt about it. Of course, it is possible that there has
been a mistake about the locality of the type specimen, which is in the British
Museum Collection.

B.N. 57 (Gulf of Manaar).

Phyllospongia, Ehlers.

Spongiidre of thin, lamellar form, often cup-shaped. With a close-meshed skeleton
network of slender horny fibre.

Phyllospongia papyracea (Esper), var. — Plate XIV., fig. G.

1798 1*00, Spongia papyracea, Espee (6); 1*70, Phyllospongia papyracea, Ehlers (58);
1*77, Phyllospongia papyracea, Hyatt (69) ; 1884, Phyllospongia papyracea, Ridley (16) ;
1889, Phyllospongia papyracea, pars, Lendenfeld (66).

This variety is represented in the collection by a fine dry specimen, of which a
photograph is reproduced in Plate XIV., fig. 6. The specimen is frondose, proliferous
and decumbent, and lias apparently been attached to the substratum at many points.
The thickness of the fronds is about 1"25 millims. The consistence (when perfectly
dry) is stiff and rather fragile, the colour light brownish-yellow. The upper surfaces ot
the fronds are marked with feebly developed concentric and radiating ridges, and also
by numerous narrow grooves, frequently arranged in a branching or stellate manner
and probably containing minute exhalant apertures. The lower surface is entirely
free from such grooves. Both surfaces appear minutely reticulate under a lens, and
neither possesses a continuous sand-cortex, though there is a good deal of sand
scattered on the upper surface.

The skeleton is a close network of very pale-coloured horny fibres usually about
0"02 millim. in diameter. The fibres are mostly free from foreign matter, but the
primary lines, radiating to the surface, contain many comparatively large sand-grains.

Except for the presence of the stellate or branching grooves on the upper surface
and the sand-grains in the primary fibres, this species agrees very closely with the
figures and description of the type given by Esper and Ehlers. As the type came
from Southern India (Tranquebar), it is not likely that the Ceylon form is more than
varietally distinct.

The species has been previously recorded from Tranquebar (Esper) ; Cape of Good
Hope (Hyatt) ; and Mozambique (Ridley). Lendenfeld also records it from

2 F

218 CEYLON PEARL OYSTER REPORT.

Australia and New Zealand, but his identifications are not always trustworthy, and
he appears to me to have got a wrong conception of the species. Thus he states that
" in every case the sponge is attached by a short peduncle." He also includes (perhaps
rightly) the cup-shaped Phyllospongia holdsworthi in the species.
R.N. 104a (Gulf of Manaar).

Phyllospongia holdsworthi (Bowerbank).

1873, Spongionella holdsworthii, Bowerbank (8); 1889, Phyllospongia papyracea, pars,
Lendenfeld (66).

There are several exquisitely cup-shaped specimens of this sponge in the collection.
Bowerbank's figures and descriptions do not appear to me to be very typical, and it
seems not impossible that he had also before him, when writing, specimens of
Phyllospongia papyracea. The specimens which I have examined are regularly cup-

Fig. 5. Phyllospongia holdsworthi, half nat. size.

shaped (see text-fig. 5), with an entire margin and a very well developed peduncle
branching out into root-like processes below. The wall of the cup is only about
1*5 millims. thick, stiff and tough and slightly flexible in the perfectly dry state.
Both surfaces are smooth or nearly so, but show feebly developed concentric and
sometimes radiating ridges. The vents are minute, usually circular in outline, and
abundantly scattered over the inner surface only of the cup, which is covered by a
thin sand-cortex not sufficiently developed to conceal the minutely reticulate character
of the dermal skeleton. There is no sand-cortex on the outer surface, which is also
minutely reticulate. Professor Herdman informs me that the colour of the sponge
in life is purplish-brown, and my dry specimens still retain a distinctly jjurple tinge
in places.

The skeleton is a close-meshed but very irregular network of horny fibre, mostly
about 0*02 millim. in diameter and free from sand, but with stouter primary lines
radiating to the surface and containing numerous comparatively large sand-grains,
especially towards the inner surface of the sponge.

SPONGES. 219

Bowekbank's figures certainly represent a form which is intermediate in external
appearance between what I regard as the typical cup-shaped P. holdsivorthi and the
fbliaceons P. papyracea, and it is quite possible that the two are not more than
varietally distinct.

Bowerbank cpiotes from a letter of Mr. Holdsworth the following interesting
particulars : — " Spongionella is only found on the 9-fathom line of the large pearl-
bank. It is attached to pieces of dead coral or stones. When alive it is of a dark
brown ; and when taken out of the water it looks exactly like dirty wet leather, If
you soak a bit of one of the dark specimens* you will see it with as nearly as
possible the original appearance. This sponge is so strictly confined to the locality
above mentioned, that its discovery by the divers is considered the strongest evidence
that the outer part of the bank has been reached."

Professor Herdman adds, as the result of his much more extended examination of
the Gulf of Manaar, that " although very characteristic of the Periya Paar and other
deeper grounds west of the Cheval Paar, still it is not absolutely confined to these,
hut may be found elsewhere, as on the Muttuvaratu Paar."

R..N. 30 ; and other specimens (dry). (Periya Paar, Muttuvaratu Paar, &c, Gulf
of Manaar.)

Hircinia, Nardo.

Spongiidse with a coarse-meshed skeleton network usually containing much foreign
matter. Denser aggregations of the network along the primary lines frequently
form trellis-like compound fibres. Filaments are usually present in the ground-
substance.

Hircinia fusca, Carter — Plate XIV., fig. 1.

1880, Hircinia fusca, Carter (4), not Hircinia fusca, Ridley (16) and Lendenfeld (66).

This is a very remarkable and well-characterised species. It was originally
described by Carter in less than four lines, and the species was styled " provisional."
It is, therefore, little wonder that Ridley and Lendenfeld have erred in identifying
certain slender branching sponges from other localities with the Ceylon species.
Carter's description of the external form should, however, have been sufficient to
pievent any such misconception, for a slender, branched, cylindrical sponge, narrower
at the base, and with conuli only 1 millim. high, can hardly be identical with one
which is described as " massive, digitate, branched lobate, cactiform on the surface."
In addition to these characters, the dark brown colour and the resemblance to Aplysina
fusca, noted by Mr. Carter, leave no doubt in my mind that Professor Herdman's
specimens really belong to the species in question, an opinion which is rendered almost

* Professor Herdman's dry specimens are very pale in colour.

2 F 2

220 CEYLON PEARL OYSTER REPORT.

certainly correct by the fact of their coming from the same locality. Under the
circumstances it seems desirable to give some details with regard to the species.

There are two good specimens in the collection, one in spirit (R.N. 48) and one dry
(R. N. 99). The former (figured) is massive, irregular, attached by a broad spreading
base, from which compressed digitate or flabellate processes rise vertically upwards,
bearing small vents at their apices (vents about 1"5 millims. in diameter). The
surface is cactiform, with usually sharp-pointed but broad conuli, up to about
3 millims. in height, but usually less. The distance between the conuli varies
greatly, but they are usually widely separated from one another by intervals of about
8 millims. The surface between the conuli is smooth or wrinkled, finely granular,
under a lens very minutely reticulate and porous. The colour on the surface (in
spirit) is warm brown, internally it is much paler, yellowish. The texture is com-
pressible and resilient, but extraordinarily tough and leathery, so that it is very
difficult to cut sections. This leathery character is obviously due to the enormous
quantity of " filaments " which the sponge contains. Internally it is somewhat
cavernous, owing to the presence of numerous cylindrical canals running vertically
upwards towards the vents. This specimen measures about 190 millims. in greatest
breadth of base and 56 millims. in greatest height.

The dry specimen is strongly compressed, flabellate, and only very slightly pro-
liferous, with a narrow margin bearing a row of vents. It contains much more sand
than the spirit specimen, but in other respects agrees closely. It measures about
135 millims. in height by 120 millims. in greatest breadth.

The skeleton is composed principally of large sand-grains, with a comjxrratively
small quantity of spongin ; arranged as follows : — ( 1 ) Very stout columns or tracts
of sand-grains run vertically through the sponge and end in the surface conuli.
These columns are compound structures, in which the sand-grains are held together
by numerous short, slender spongin threads running from one to the other, in much
the same way as Lendenfeld (66) has figured for Psammopemma marshalli. They
apparently represent an exaggerated condition of the trellis-like main fibres of certain
other Hircinise. (2) A very irregular network of more slender secondary fibres,
composed of sand-grains held together by spongin threads as in the main columns,
but the large sand-grains often only in single series. (3) In the dermal membrane
there is a thin layer of broken sponge-spicules and large sand-grains, the former lying
somewhat more superficially than the latter. Numerous broken spicules also occur
along with the sand in the deeper parts of the sponge.

The canal-system appears to be that of a typical Hircinia, but, owing to the large
quantity of sand and :< filaments," it is impossible to get satisfactory sections. The
flagellate chambers are not well preserved, but they appear to be about 0'04 millim.
in diameter and approximately spherical, and the ground-substance between them is
finely granular. The soft tissues are very densely charged with filaments. These
have a maximum thickness of about 0'004 millim. between the heads. The heads,

SPONGES. 221

which stain deeply with picro-carmine, are about 0-008 millim. in diameter and
somewhat variable in shape, sometimes nearly spherical and sometimes more or less
pointed at the end.

Ridley's Dysidea fusca, which that author (16) supposed might be identical with
Carter's Hircinia fusca, appears to be quite a different sponge.

U.N. 48, 99 (pearl banks, Gulf of Manaar) ; 271 (small specimen with fewer
filaments. " Deep water off Galle and onwards up West Coast of Ceylon ").

Hircinia tuberosa, n. sp. — Plate XVI., fig. 2.

The sponge consists of a very irregular, somewhat tuber-like body, from which
irregular, linger-like processes are given off in various directions, the whole much
mixed up and partially coated with calcareous debris. In the largest specimen the
central portion of the sponge measures about 60 millims. in diameter, and the two
larger processes each about 44 millims. in length by 18 millims. in diameter. The
surface is very uneven, but sub-glabrous (not reticulate) between the foreign adhesions,
and only slightly conulose, the conuli being low, irregular, and widely separated from
one another. Internally the sponge is cavernous, being permeated by wide, cylindrical,
meandering vestibules, which are covered in at the surface of the sponge by a thin,
parchment-like dermal membrane, pierced here and there by rounded apertures.
These apertures are evidently vents. They vary from about 0'5 millim. to 4 millims.
in diameter, and are frequently arranged in groups. They are found sometimes on
the central portion of the sponge, but more frequently on the more or less fistular,
finger-like processes. Texture extremely coarse and gritty throughout, but tough.
Colour (in spirit) pale yellowish -grey throughout. There is a distinct but thin sand-
cortex in the parchment-like dermal membrane.

The skeleton (Plate XVI., fig. 2) is an extremely irregular network of more or less
trellis-like horny fibre, partly enclosing and partly connecting together the very
numerous sand-grains and other foreign bodies with which the sponge is filled.
Main fibres are recognisable, but not very well defined, and the whole is so irregular
that it is useless to give measurements.

The flagellate chambers are about 0"03 millim. in diameter and approximately
spherical, and the ground-substance between them is finely granular.

Filaments are present in enormous numbers and usually collected together in more
or less dense bundles. They measure about 0'006 millim. in maximum diameter
between the heads, and their heads are approximately spherical and about 0-008 millim.
in diameter.

R.N. 86, 88, 88a (all from deep water off Galle and onwards up West Coast).

Hircinia schulzei, n. sp. — Plate XVI., fig. 3.

Sponge slender, cylindrical, irregularly branched, and attached at many points to
fragments of calcareous debris, amongst which it appears to creep. A few rather

222 CEYLON PEARL OYSTER REPORT.

short and somewhat club-shaped branches, ahout 25 millims. in length and 3 millinis.
in maximum diameter, probably rose vertically upwards from repent stems of about
the same diameter. The surface is pretty uniformly covered with low conuli, between
which narrow longitudinal canals may be seen running beneath a thin membrane,
which is occasionally broken through by a single small vent or by a sieve-like group
of very small vents. The colour (in spirit) is very pale yellow and the texture pretty
stiff, hut compressible and resilient.

The skeleton (Plate XVI., fig. 3) consists of widely distant, longitudinal main
tihres curving outwards towards the surface and connected at irregular intervals by
a round-meshed lattice work of secondaries. The main fibres are pretty regularly
cylindrical and about 0-12 millim. in diameter, not fascicled, and containing a good
many fragments of sponge-spicules. The connecting fibres are entirely free from
foreign matter ; they vary much in diameter, but are usually pretty stout.

The canal-system is of the type usually met with amongst the Spongiida?. The small
flagellate chambers are approximately spherical and up to about 0"04 millim. in
diameter, with short, wide, exhalant canaliculi. The ground-substance between them
is finely granular. A very conspicuous histological feature is the presence of
numerous long, cylindrical cords of fibrous tissue running longitudinally through the
sponge. These cords are composed each of a compact mass of elongated, finely
granular cells, each with a very distinct, darkly staining nucleus. They closely resemble
the similar fibrous bands found in Euspongia, &c, and are probably contractile.
The characteristic Hircinia filaments are abundantly scattered through the soft
tissues. They are, however, very slender, and I have not succeeded in making out
the nature of their terminations.

This appears to be a very distinct and well-characterised species, differing from
Hircinia dendroidex, Schmidt, which is, perhaps, its nearest ally, in its much more
slender branches and much more regular main fibres. I have much pleasure in
dedicating it to the zoologist to whom we chiefly owe our accurate knowledge of the
Spongiidse.

R.N. 277 (deep water off Galle and onwards up West Coast of Ceylon).

Hircinia anomala, n. sp. — Plate XIV., fig. 5 ; Plate XVI., fig. 1.

Sponge massive, irregular, with a tendency to become lobose or digitate. Surface
uniformly covered with small, sharp conuli, about 1 millim. in height and 2 millims.
apart; with a minute reticulation of fine ridges chiefly radiating from the apices of
the conuli. Sometimes the surface reticulation is supjjressed, and it may be present
or absent in different parts of the same specimen. The surface is not sandy, except
sometimes at the apices of the conuli. Vents inconspicuous, the sponge being,
perhaps, sometimes lipostomous. Pores scattered abundantly in the meshes of the
dermal reticulation. Colour (in spirit) varying from pale brown to black on the
surface ; pale brown internally. Consistence firm, but compressible and elastic. The

SPONCxES. 223

largest specimen is about 7f> niillinis. in length by 42 millinis. in greatest breadth ;
another is about 140 niillinis. long, with a maximum diameter of 25 millinis.

The skeleton (Plate XVI., tig. 1) is an extremely irregular network of highly
arenaceous fibres. There is no distinct differentiation into main and secondary fibres,
hut the reticulation is much closer along tracts which run vertically to the surface
and end iu the surface conuli. These denser parts of the reticulation are evidently
homologous with the trellis-like main fibres of other Hircinise. Between them large
tracts may remain entirely devoid of skeleton. The fibres themselves vary a good
deal in thickness and in the amount of foreign matter which they contain. Usually
there is a very large proportion of sand or sponge spicules, and comparatively little
spongin ; occasionally, however, I have seen fibres without foreign inclusions. The
spongin of the fibres is very distinctly lamellated.

The flagellate chambers are approximately spherical, up to about 0'04 millim. in
diameter (but often smaller), and either eurypylous or with short exhalant canals.
The ground-substance between them is finely granular, though perhaps somewhat less
markedly so than in typical Spongiidaa. The larger canals are surrounded by a very
large quantity of gelatinous, vesicular-looking collenchyma, and commonly more or
less sub-divided by septa. Bands of fibrous tissue penetrate the soft tissues as in
other Spongiidee, but I have not found any of the " filaments " so common in the
genus Hireinia. B.N. 13 contains an immense number of unicellular bodies of a pale
yellow colour (staining brown with picro-carmine) ; these are oval or nearly spherical,
and about 0"02 millim. in diameter; each with a small nucleus. Occasionally they
appear to be broken up into fragments. Probably they are unicellular Algse,
comparable to those which I have described in Hexadella. B.N. 82 and 171 contain
numerous groups of smaller cells which are, perhaps, the same Alga in process of
division.

This species exhibits characters intermediate between those of the Spongeliidaa and
those of the Spongiidae. The skeleton, in its highly arenaceous character, agrees with
that of Spongelia, but the small size of the flagellate chambers and the granular
character of the ground-substance prevent us from including it in that genus.
The absence of filaments, on the other hand, militates against our regarding the
species as a typical Hireinia, but the nature of these filaments and their taxonomic
value are still so obscure that I am not inclined to exclude the sponge from the genus
solely on account of their absence.

In external appearance the species somewhat resembles Cacospongia scalaris.

B.N. 13 and 91 (Gulf of Manaar) ; 82 and 171 (deep water off Galle and onwards).

Aplysina, Nardo.

Spongiidse with distinctly pithed horny fibres forming a coarse-meshed skeleton
network. Of very compact texture, with narrow canals and very small flagellate
chambers.

224 CEYLON PEARL OYSTER REPORT.

Aplysina purpurea, Carter.

1880, Aplysina purpurea, Carter (4); 1881, Aplysina purpurea, Carter (65); 1889,
Aplysina purpurea, Dendy (3); 1889, Psammopemma fuliginosum, Lendenfeld,
pars (66).

There is only a single dry specimen of this sponge in the collection, so that I am
not in a position to add anything to the descriptions of the species given hy Carter
and myself, except by referring back to some preparations of spirit material collected
by Mr. Thurston, which show the sponge to be a true Aplysina and not, as
Lendenfeld has supposed, a Psammopemma.

The skeleton is composed of dense local aggregations of very irregular, branching
and anastomosing horny fibres, accumulated along certain tracts to form the so-called
" compound fibres," while large intervening areas remain free from fibre altogether.
The fibres themselves have a very curious structure, consisting of a very thin outer
layer (if any) and a very thick " pith," the latter exhibiting a granular or often
minutely reticulate appearance. They are free from foreign bodies. The inhalant
pores are abundantly scattered over certain parts of the surface. The structure of
the soft tissues is very compact and the flagellate chambers are small and probably
aphodal or diplodal. In short, the canal-system probably agrees closely with that
described and figured by Schulze in Aplysina aerophoba, although the condition of
my material is not good enough to render a detailed comparison possible.

Lendenfeld has, as already indicated, made the curious mistake of confounding
this species with " Psammopemma .fuligi nosum," a totally different sponge. Carter's
specimens of Aplysina purpurea were from the Gulf of Manaar and Trincomalee, but
he subsequently (65) identified an Australian sponge with the same species. Still
later, however, when describing his Pseudoceratina durissima (18), he showed that
the Australian specimen previously identified by him as Aplysina purpurea should
really be considered as a specimen of Pseudoceratina durissima. Lendenfeld,
accordingly (66), in a manner very characteristic of that writer, observes that
" Carter himself has shown that his Aplysina purpurea and his Pseudoceratina
durissima are identical," which, of course, is by no means the case. At the same
time he omits the locality from which the types of Aplysina purpurea were obtained
from the geographical distribution of Psammopemma fuliyi nosum, under which name
(while admitting that it is not the oldest) he. also includes (rightly or wrongly)
Carter's Pseudoceratina durissima.

Mr. Carter has given a characteristic sketch of the external form of the sponge,
together with figures of the skeletal structure (65, Plate IX., fig. 1), which are
sufficient for the identification of the species.

There can be no doubt that Keller's Psammaplysilla arabica from the Red Sea (61)
is very closely related to, if not identical with, this species. The compound skeleton
fibres, consisting entirely (according to Keller) of reticulate " Marksubstanz," the
cacti form surface, and black-violet colour in alcohol, all point to generic if not

SPONGES. 225

specific identity. In the Ceylon specimens, however, there appears to he (usually af
any rate) no sand in the skeleton fibres. Keller makes his species the type not
only of a new genus, but even of a new family. He considers it to be closely related
to Aplysillu, but, unfortunately, his material did not enable him to investigate the
form and arrangement of the flagellate chambers, or he would probably have seen
that the affinity was rather with Aplysina, as the firm, almost leathery texture of
the living sponge and its stony hardness when dry might alone have indicated. I
follow Mr. Carter in retaining the Ceylon species, at any rate (with which Keller
does not appear to have been acquainted), in the latter genus, from which it differs in
no important respect. In any case the name Psammaplysilla appears to have been
very unfortunately chosen, as the sandy character of the fibre is hardly of generic
value, and the relationship with AphjsiUa is not nearly so close as that with Aplysina,
a very distinct genus.

R.N. 95 (Stat. IV., off Karkopani, 6-9 fithoms, Gulf of Maiiaar).

Aplysina herdmani, n. sp. — Plate XVI., fig. 4.

The single specimen consists of a rather thin, irregular, flattened crust, from which
short, slender, cylindrical, digitiform processes rise vertically upwards at wide and
irregular intervals. The ends of these processes are truncated, and each has a single
small vent in the middle.* The surface, both of the basal crust and of the dicitiform
processes, is glabrous, hut beset with numerous minute conuli, from the apices of
which the ends of the primary fibres sometimes project. The colour, in spirit, is dull
purple throughout ; the texture compact and rather fleshy, but compressible and
resilient. The maximum diameter of the basal crust is about 55 millims. and its
thickness about 4 millims. The digitiform processes are about 11 millims. high by
only 2 millims. or 3 millims. in diameter.

The skeleton, in the basal crust, consists of a reticulation of thin-walled, pithed
fibres of a rather dark brown colour, amongst which distinct primary fibres, running
vertically into the surface conuli, are clearly differentiated (Plate XVI., fig. 4). These
primary fibres are about 0"08 millim. in diameter, and the very thick, granular
" pith " contains abundant broken sponge spicules as foreign inclusions. The
primary fibres sometimes branch, and they are connected together by a network of
secondaries which vary greatly in diameter, being sometimes as stout as the
primaries and sometimes very slender, only about 0'016 millim. in diameter. The
secondary fibres are free from foreign matter and the stouter ones commonly run
across between the primaries and thus form rectangular meshes, but the meshes are
usually irregularly polygonal and very variable in diameter.

In the digitiform processes the main fibres — containing broken spicules — run longi-

* One of the processes forks into two close to its extremity, and each of the very short brunches thus
produced bears a small vent on its truncated end.

2 G

226 CEYLON PEAEL OYSTER REPORT.

tudinallv and give off short branches — also containing broken spicules — into the
surface conuli. The secondaries are arranged as in the basal crust.

The flagellate chambers are very small, only about 0"024 millim. in diameter, and
approximately spherical, but the arrangement of the collared cells, on the inhalant
side of the chamber only, gives them a curious crescentic appearance ; moreover, they
frequently appear in sections to be arranged in single curved rows ; surrounding the
narrow exhalant canals at about equal distances, and doubtless communicating with
them by very long and narrow canaliculi, but the condition of the specimen is not
good enough to enable me to make out minute histological details. The ectosome is
chondrenchymatous rather than collenchymatous, but a large quantity of gelatinous
tissue is developed around the larger canals in the choanosome. In the neighbourhood
of the flagellate chambers the choanosome is abundantly granular, and both ectosome
and choanosome contain numerous pigment cells.

It affords me much pleasure to name this well-characterised species after Professor
W. A. Herdman.

K.N. 340 (Ceylon seas).

Class: CALCAPEA.
Porifera with a skeleton composed of calcareous spicules.

The number of calcareous sponges in the collection is remarkably small, only four
species being represented, two of winch, however, are new. I have discussed the
classification of the group in considerable detail in my earlier writings (76, 78, 80) and
adhere to the opinions therein expressed.

Order 1 : HOMOCCELA.
( Calcareous sponges in which the endoderm consists throughout of collared cells.

Leucosolenia, Bowerbank.
With the characters of the order.

Leucosolenia (Clathrina) coriacea (Montagu), var. ceylonensis, nov. — Plate XIII. ,
fig. 8.

[For literature and synonyms vide Haeckel (7)].

This well-known European species is represented in the collection by a slight
variety belonging to the reticulate section of the genus Leucosolenia as defined by
the present writer (76). The sponge forms massive, closely reticulate colonies of
slender ascon-tubes, each colony with a constricted base of attachment ; with fairly
numerous, small but prominent true oscula formed each by the coalescence of several
tubes in a projection from the general surface. The ascon-tubes are only about
0-16 millim. in diameter and there is no pseudoderm. The entire colony attains a
diameter of some 10 millims, or 20 milbms. The colour in alcohol is pale grey.

SPONGES. 227

The spicules are nearly ;ill regular triradiates (Plate XIII., fig. 8, a., b., a); with
slender, not very sharply pointed rays of pretty uniform diameter and measuring
about 0*088 millim. in length by 0-008 millim. in diameter at the base. The apex is
rather abruptly and rather irregularly pointed. There are apparently no quadri-
radiates at all, but two (R.N. 377, 378) of the three specimens in the collection show
a few very slender oxea (Plate XIII., fig. 8, d) projecting from the surface of some
of the tubes. I have not been able to obtain these spicules in an unbroken condition,
but they apparently closely resemble those found in the next species ; whether or not
they should be regarded as constant features of this variety. I am unable to say
with certainty.

11.N. 377, 378, 379 (all from Cheval Paar).

Leucosolenia (Clatnrina) tenuipilosa, n. sp.- -Plate XIII., fig. 9.

Sponge forming massive, reticulate colonies of ascon-tubes, closely resembling the
preceding variety but somewhat coarser. Here and there on the surface of the
colony the tubes converge to unite in small, prominent, true vents. The tubes
themselves are about 0'5 millim. in diameter, and they form a close reticulation
without any pseudoderm. The colour in alcohol is pale grey. The largest specimen
(R.N. 158, which may be regarded as the type of the species) is cake-shaped and
flattened, measuring about 47 niillims. in length, 37 millims. in breadth, and 16 millims.
in thickness.

The skeleton is arranged as usual in the genus, and the spicules are of three
kinds : —

(1.) Regular triradiates (Plate XIII., fig. 9, a), with rather stout, slightly
fusiform rays, bluntly and rather abruptly pointed at the apex, which is
often somewhat irregular. Rays measuring about 0"1 millim. in length by
0-012 millim. in diameter at the thickest part.

(2.) Quadriradiates (Plate XIII., tig. 9, b, c) abundant ; resembling the triradiates,
but with an apical ray projecting at right angles into the gastral cavity.
This ray is somewhat variable in form and size ; typically it is long and
slender, gradually and sharply pointed, and slightly undulated towards the
extremity ; in the type specimen it attains a length of about 0'14 millim.

(3.) Very slender, hair-like oxea (Plate XIIL, fig. 9, d). sparsely hispidating the
surface of the tubes. These may attain a length of more than 0-4 millim.,
with an average diameter of only about 0"002 millim. They taper very
gradually from the proximal extremity, which is somewhat hastately sharp-
pointed and may be as much as 0'004 millim. thick, to the distal, which is
hair-like and apparently nearly always broken off.

This species is evidently closely related to Leucosolenia coriacea, var. ceylonensis,
but differs in the presence of the quadriradiates, and also in the greater stoutness
and the frequently fusiform shape of the rays of the triradiates.

2 g 2

228 CEYLON PEAKL OYSTER REPORT.

R.N. 158 (Stat. LXIV., 5 fathoms, south-east of Modragam, March 17, 1902);
158a (south of Cheval); 380 and 381 (both from Cheval Paar, March 4, 1902).

Order 2 : HETEROCCELA.
Calcareous sponges in which the collared cells are confined to more or less well-defined

flagellate chambers.

Family : GRANTIID^E.

Heteroccela with a distinct and continuous dermal cortex, completely covering over
the chamber-layer and pierced by inhalant pores. There are no sub-dermal
sagittal triradiates, nor conspicuous sub-dermal quadriradiates. The flagellate
chambers vary from elongated and radially arranged to spherical and irregularly
scattered ; while the skeleton of the chamber layer varies from regularly articulate
to irregularly scattered

Leucandra (Haeckel).

Grantiidae in which the flagellate chambers are spherical or sac-shaped, never
arranged radially around the central gastral cavity, with which (or with the
main exhalant canals derived therefrom) they communicate by a more or less
complicated exhalant canal-system. The skeleton of the chamber layer is
composed of irregularly scattered radiate spicules, but it may still present traces
of its derivation from a radially symmetrical type in the presence of a few sub-
gastral sagittal triradiates.

Leucandra donnani, n. sp. — Plate XIII., fig. 10.

The only specimen in the collection consists of a single Leucon person of sac-like
form ; elongated, sub-cylindrical, but slightly compressed ; rather strongly curved ;
tapering gradually from broadly rounded base to narrower apex, where the terminal
osculum is situated. The outer surface is nearly smooth, but slightly granulated in
appearance. The oscuhun has no spicular fringe, but a slightly developed mem-
branous margin. The total length of the specimen is 23 millims. ; the maximum
diameter at the base is about 9'5 millims., and the diameter of the vent is 2 millims.
The thickness of the sponge- wall in the middle is about 2 millims. The colour (in
spirit) is light brown and the texture firm but brittle.

The dermal cortex, of sagittal triradiates, is only about 0'05 millim. thick ; beneath
it lie large, irregular sub-dermal cavities, without, however, any special supporting
skeleton of their own. The gastral cortex, of sagittal quadriradiates, is about as
thick as the dermal cortex and pierced by the numerous apertures of the exhalant
canals. The skeleton of the chamber layer consists for the most part of very large
sagittal triradiates, generally arranged with the basal ray pointing outwards.

SPONGES. 229

The canal-system is typically leuconoid; the flagellate chambers, abundantly
scattered in the ground-substance between the gastral and dermal cortex, being
spherical or sac-shaped, and having a maximum diameter of about O'OS millim. The
epithelial cells lining the larger canals contain numerous brown pigment granules.

Spicules.— (1.) Dermal triradiates (Plate XIIL, fig. 10, c, d) ; sagittal, with slender
rays of not very unequal length, measuring, say, about 0-276 millim. in length by
0-01 G millim. in thickness at the base, and tapering pretty gradually from base to
apex, which is sharp-pointed. The rays are all straight, and the angle between the
two paired rays is only slightly greater than the other tw7o angles.

(2.) Triradiates of the chamber-layer (Plate XIIL, fig. 10, a, h) ; stout, sagittal,
with the basal ray somewhat shorter than the paired rays. Rays usually straight ;
somewhat fusiform and gradually and very sharply pointed. Angle between paired
rays only slightly greater than the other two angles. The paired rays in a typical
example measured about 07 millim. in length by 0-066 millim. in maximum diameter,
with a basal ray about 0'57 millim. long and of about the same thickness as the
others.

(3.) Gastral quadriradiates (Plate XIIL, fig. 10, e, /) ; strongly sagittal, with the
paired rays extended nearly, or quite, at right angles to the conspicuously shorter
basal ray, and with the still shorter apical ray directed forwards almost in a line with
the basal. The rays are all straight, or nearly so, and only moderately stout, and
taper gradually from the base to the sharp-pointed apex. Length of paired rays
about 0-188 millim., with a diameter at the base of 0"012 millim. ; with basal ray
about 0-072 millim. long and apical ray about 0'04S millim. long; the basal ray of
about the same diameter as the paired rays; the apical ray rather more slender.

(4.) Fusiform oxea (Plate XIIL, fig. 10,//); very slightly curved; gradually and
finely pointed at the inner end. but with the outer end nearly always broken off;
moderatelv stout, measuring, say. about 074 millim. in length by 0-02 millim. in
maximum diameter. Arranged in sparse bundles at right angles to the surface, with
the outer ends projecting but slightly.

This species is perhaps most nearly related to Lendenfeld's Leucandra typica (79)
[= Leuconia typica, var. tuba, of Polejaeff (77)], from the East Coast of Australia
(and Bermudas ?), but differs considerably in details of spiculation. 1 have much
pleasure in naming it after Captain Donnan, the veteran Inspector and explorer of
the Ceylon pearl banks.

E.N. 18G (Donxans Muttuvaratu Paar, Gulf of Manaar).

Fa m i i.y : A MPHOKISCIDiE.

Heteroccela with a distinct and continuous dermal cortex. With conspicuous sub-
dermal quadriradiate spicules with inwardly directed apical rays. Flagellate
chambers varying from elongated and radially arranged to spherical and
irregularly scattered.

230 CEYLON PEARL OYSTER REPORT.

Heteropegma, PoleSjaeff.

Amphoriscidae with elongated flagellate chambers arranged radially around the
central gastral cavity. With a vestigial tuhar skeleton of minute radiates.
With a very thick dermal cortex, composed principally of triradiate spicides.

Heteropegma nodus-gordii, Polejaeff (77).

This remarkable and well-characterised species is represented in the collection by a
single good-sized specimen, which agrees very closely with the types described by
Polejaeff in 1883 from Australia (Cape York) and the Bermudas. The only other
species known is the very closely related //. latitubulata from near Port Phillip
Heads (80), so that the genus is apparently a characteristic Australian one, and the
discovery of H. nodus-gordii (the Northern Australian form) at Ceylon affords
another good example of the close relationship between the Ceylonese and Australian
Sponge-Fauna. The anatomy of this species has been figured both by Polejaeff (77)
and by myself (78).

R.N. 155 (Ceylon seas).

LIST OF THE CEYLON SPONGE-FAUNA

so far as at present known ; showing the classification adopted and the geographical
distribution of those species which have been recorded from localities beyond the
Ceylon ai*ea. The species marked H occur in Professor Herdman's collection ; those
marked D are doubtful* : — •

Class-. NON-CALCAREA. G. Stceba (Samus) simplex (Carter, 4)t, Temate

Order: Myxospongida. (87)-

H 7. Stceba extensa, n. sp.

8. Triptolemus (Samus) parasiticus (Carter, 4).f

9. Nethea (Tisiphonia) nana (Carter, 4).t
Order: Tetraxonida. 10. Sphinctrella 1 (Tisiplumia) annulata (Carter,

Grade; Tetractinellida. 4).t

Sob-order : Homosclerophora. Family : Stellettidse.

Family : Plakinidae. n. Myriastra (Stelletta) crassicula (Carter, 5).t

H 3. DeD-citcpsis ceylonica, n. gen. et sp. H 12. ,, clavosa (Ridley). Off north coast of

H 1. Hexadella indica, n. sp.

D 2. Halisarca (?) ruhitmgens, Carter (5)

Sub-order : Astrophora.

Australia; Philippine Islands; Amboyna (83) ;

Ternate (87); Coast of Cochin China (86).
Family : Pachas trellidse. H 13_ Myriastra tethyopsis (Carter).

H 4. Plakinastrella intermedia, n. sp. 14. Pilochrota ceylonica, Sollas (15).

H 5. ,, schulzei, n. sp. H 15. „ haeckeli, Sollas. Philippine Islands.

* Iu compiling this list no notice has been taken of varietal distinctions. It may be noted that the
sponges described by me in my paper on " The Sponge-Fauna of Madras " (2) all came from the Gulf of
Manaar (vide Thurston, 99). t Fide Sollas (15).

SPONGES.

231

H 10. PUochrota hornelli, n. sp.

II 17. Stelletta herdmani, n. sp.
H 18. „ vestigium, a. sp.

19. Aurora {Stelletta) globostellata (CARTER, 21). t

H 20. Eciont ma carteri, n. sp.
H21. „ laviniensis, n. sp.

Family : Geodiidae.

1122. Geodia perarmata, Bowerbank.

II 23. .. peruncinata, n. sp.

11 24. ,. areolata, Carter.

11 25. ,, raiiiodigitata, Carter.

26. ,, globostelUfera, Carter. Port Darwin,
Australia (RlDLEY, 16).*

27. Erylus carteri, Sollas (15)=SteUetta euastrum
Carter (4).

Sub-order : Sigmatophora.

H28.

Family: Tetillidse.
Tetilh hirsuia, Dexdy.

H54.

55.

H29.
H30.
H31.
H32.
H33.

„ poculifera, n. sp.

„ anomala, n. sp.

,, limicola, n. sp.
( 'raniella elegant, n. sp.
Paratetilla cineriformis, n. gen. et sp.

56.

57.

58.
H59.
H60.

Family : Samidse.

1161.

34.

Samus anonymus, Cray; //</e Carter (4).
Bahia, West Indies, Australia, South Seas,

H62.

Seychelles (15).

63.

Grade : Lithistida.

H64.

35. JDiscodermia papillata, Carter (4).

36. „ aspera, Carter (4).

37. „ Icevidiscus, Carter (4).

38. „ sinuosa, Carter (5).
H 39. „ i margmata, n. sp.

40. Racodiscuia (Diseodermia) spinispiiidifera (Car-
ter, 4).t

41. Racodiscuia ( Discodt rmia) sceptrellifera (Carter,
5).t

42. Corailistes aadeata, Carter (4).

43. „ '-uoisu, Carter (4).

D 44. Goralh ' ' 'i<>>>'< ima, Carter (4).

H 45. Aticulitt - m <■ ntalis, n. sp.

H 46. Taprobane herdmani, n. gen. et sp.
H 47. Petromica massalis, a. sp.

Grade : Monaxonellida.

Sub-order : Astromonaxonellida.

Family : Epipoksidae.

48. Goppatias (Tisiphonia) penetrans (Carter, 4).
H 49. ,, reptans, n. sp.

H 50. Asteropus haeckeli, n. sp.

H51. Gryptotethya aggluti/iums, n. gen. et sp.

Family' : Tethyidae.

H52. Tethya lyncurium, Lin. Semi-cosmopolitan
(especially in the North Atlantic).

H 53. Xenospongia patelliformis, Cray. Torres
Straits.

Family : Spirastrellidse.

Hymedesmin stellivarians, Carter. Azores (48).

,, moorei, Carter (4).

„ spinatostellifera, Carter (4).

„ capitatostellifera, Carter (4).

„ trigonostellata, Carter (4).

,, curvistellifera, n. sp.

Spirastrella vagabunda, Ridley. Torres
Straits ; Ternate (87) ; Aden (92).
Spirastrella tentorioides, n. sp.
Placospongia carinata (Bowerbaxk). Widely
distributed in tropical seas (35).
Placospongia melobesioides, Gray, fide Carter
(4). Widely distributed in tropical seas (35).
Negomho tenuistellata, n. gen. et sp.

Family : Clionidae.

65. Clioua waneni, Carter (5).

66. „ indica, Topsext (37).
H67. ,, margaritifene, n. sp.

68. Tlwosa socialis, Carter (4).

69. Dotona pulchella, Carter (4). Azores (62).

70. Alectona higgini, Carter (4).

Family : Suberitidae.

71. Suberites vestigium, Carter (4).
D 72. „ inconstavs, Dexdy (2).
II 73. „ cruciatus, n. sp.

* Ridley's identification seems a little doubtful (vide Sollas, 15).
t Fide Sollas (15).

CEYLON PEARL OYSTER REPORT.

Family : Chondrosiidse.

74. Chondrilla nucula, Schmidt; fide Carter (5).

Adriatic, Florida, Antilles, Azores (48);

Red Sea (61).
H75. Chondrilla australiensis, Carter. Australia;

Coast of Cochin China (86).
II 76. Chondrosia reniformis, Nardo. Adriatic;

Aden (92); Amboyna (83).

Sub-order : Sigmatomonaxonellida.
Family : Haploscleridse.
Sub-family : Gelliinse.

H77. Gellius fibulatus (Schmidt). Adriatic; Euro-
pean coast of North Atlantic; Azores (62) ;
Australia (16) ; Ternate (87).

II 78. Gellius <tit>ii<ltitus (Bowerbank). British seas
(52) ; Azores (1, 48, 62) ; Iceland (88).

H 79. Gelliodes carnosa, Dendy.

H80. ,, inarustans, n. sp.

H81. ,, petrosioides, n. sp.

H82. Tvxochalina robusta, Ridley. Port Jackson
(Australia).

H 83. Strongylcphora durissima, n. gen. et sp.

Sub-family : Renierinse.

84. Reniera madrepora, Dendy (2). Java (86).

D 85. Reniera albescens (= HaUchondria albescens,
Johnston), fide Carter (4). British seas
(81).

H86. Reniera implexa, Schmidt. Adriatic (50);
Azores (1, 62).

H 87. Reniera pigmentifem, n. sp.

H 88. „ zoologka, n. sp.

H89. Petrosia testudinaria (Lamarck). Queensland
(16) ; Mergui Archipelago (3)

H 90. Petrosia similis, Ridley and Dendy. South
of Cape of Good Hope and between
Kerguelen and Heard Islands (1).

II 91. Petrosia densissima, a. .sp.

H 92. Halichondria panicea, Johnston. Cosmo-
politan (1).

H 93. Trachyopsis halichoridrioides, n. gen. et sp.

Sub- family : Chalininse.

H 94. Pachychalina subcylindrica, n. sp.
II 95. „ delicatula, Dendy.

H 96. „ brevispiculifera, n. sp.

H 97.

H 98.

H 99,
H100
II 101.
H102.
H103.
H104.

H105.
H106.

PachychaUna spinilamella, Dendy.
( 'halina subarmigera, Ridley. Torres Straits
(16); Port Jackson (51); Coast of Cochin
China (86).

Chalina obtusispiculifera, n. sp.
,, clathrata, n. sp.
,, cymmformis (Esper?).
Ceraochalina retiarmata, n. sp.
,, reticutis, n. sp.

„ multiformis, Lendenfeld. Aus-

tralia and New Zealand (51).
( 'eraochalina ceylonica, n. sjj.
Siphonochalina communis (Carter). Port
Jackson (16) ; Kurrachee (16).

Sub-family : Desmacellinse.
H 107. Desmacella tubulaia, n. sp.

Sub-family : Tedaniinae.

108. Tedania digitata, Schmidt, fide Dendy (2).
Cosmopolitan (1).

Sub-family : Heteroxyinse.
H 109. Acanthomfer ceylonensis, n. gen. et sp.

Family : Desmacidonidse.
Sub-family : Esperellinre.

H110. Esperella pdrishii, Ridley. Port Darwin,
Australia (16).

Hill. Esperella plumosa (Carter). Mauritius (33) ;
Mergui Archipelago (20).

H 112. Esperella crassissima, n. sp.

H113. ,, tenuispieulata, n. sp.

D 1 1 4. Esperella tunicata (Schmidt), fide Carter (4).
Adriatic (47) ; Azores (62).

H115. Paresperella serratohamata (Carter). Van-
couver, fide Lambe (85).*

H116. Paresperella bidentata, n. sp.

H 117. Iotrochota pmpurea (Bowerbank). Straits of
Malacca (49) ; North Australia (16) ; Am-
boyna (83); Amirante group (16); ? South
Coast of Australia (10) ; ? West Indies (53).

H 118. Iotrochota baculifera, Ridley. Western Indian
Ocean (16) ; North Australia (16) ; Am-
boyna (83); Celebes (39); Ternate (87);
Coast of Cochin China (86).

Lambe's identification is probably erroneous.

SPONGES.

233

119. Yvesia (?) (lhiVuhniiihiii) aceratospiculum
(Carter, 4).

120. Paramyxilla* (Halichondria) infreguens
(Carter, 5).

Sub-family : Phloeodictyinae.

II 121. Phkeodidyon fistulosum (Bowerbank). North-
ern and Western Australia (1, &c.) ; Amboyna
(83) ; Azores (1 ; 62) ; 1 off Bahia (1) ; Ter-
nate (?) (87).
122. Histoderma (Suberites) fistulatwm (Carter, 4).
West Australia, fide Carter (4).

H 123. Histoderma vesiculalum, n. sp.

Sub-family : Ectyoninse.

H 12-1. Myxiila arenaria, n. sp.

II 125. ,, tenmssima, n. sp.

H126. Clathria frondifera (Bowerbank). Straits of
Malacca and Gaspar Strait (49) ; Queensland
and Torres Strait (16) ; Western Indian
Ocean (Providence, Amirante, Seychelles)
(16) ; Red Sea (91) ; Java Sea (86).

II 127. Clathria indica, Dendy.

II 1 28. „ spiculosa (Dendy).

H129. Raspailia thurstmi, Dendy.

H130. „ fruHcosa, Dendy.

H 131. ,, hornelli, n. sp.

132. Plumohalichondria (HalicJwndria) plumosa
(Montagu); fide Carter (5). British
seas (52, &c); off Bahia (1); Kerguelen (82).

133. Microriona armata, Bowerbank, fide Carter
(4). British seas (52).

134. Microriona atrasa/nguinea, Bowerbank, fide
Carter (4). British seas (52).

135. Microriona ajfinis, Carter (4).

136. Hymeraphia elavata, Bowerbank, fide Car-
ter (4). British Isles (52) ; Azores (62) ;
Amboyna (83) ; Ac. (vide 62).

137. Hymeraphia {Microriona) bulboretorta (Car-
•n.n, 4).

138. Hymeraphia (?) (Microriona) fasrispimilifera
(Carter, 4).

139. Hymeraphia (?) unispicvium (Carter, 4).

11 140. Agelasmauritiana (Carter). Mauritius (54) ;
off Tristan da Cunha (?) (1) ; Ternate (87).

11 141. Agelas ceylonica, n. sp.

11142. Echinodictyum clathratwm, n. sp.

H 143. Aulospongus tubulatus (Bowerbank).

144. „ (?) (Dictyocylindrus) sessilis, Car-

ter (4).

H 145. Acarnus ternatus, Ridley. Bombay (?) (16) ;
Amirante (16) ; Torres Straits (16) ; Tahiti (1) ;
Ternate (87).

H14G. Cyamon quinqueradiaium (Carter).

147. „ (Microriona) qwdriradiaturn (Carter)
(4). West Indies (4).

II 1 18. Plocamia manaarensis (Carter). California,
fide Lambe (85). t

H 149. Bvbaris eruca (Carter).

150. Rhabdoploca (Microriona) curvispicuiifera (Car-
ter, 4). Azores (62).

H151. Rhahderemia indica, n. sp.

Family : Axinellidae.

H 152. Spongosorites topsenti, n. sp.
H 153. „ (?) lamellaia, n. sp.

H 154. „ (?) lapidiformis, n. sp.

H 155. Hymeniacidon petrosioides, n. sp.
H 15G. Thrinacophora agariciformis, n. sp.
H157. „ durissima, n. sp.

H 158. Axinella labyrinthica, Dendy.
H 159. ,, manus, ri. sp.
H160. ,, lenuidigitaia, n. sp.
H161. „ halichondrioides, n. sp.

162. „ (?) (Hymeraphva) erecta, Carter (4).
North and South Atlantic and Southern
Ocean (1) ; Azores (62).
H 163. Phakellia donnani (Bowerbank).

164. „ ridhyi, Dendy (2).
H 165. „ symmetrica, n. sp.
H166. ,, ceylonensis, n. sp.
H167. ,, crassistylifera, n. sp.
II 168. Acanthella carferi, Dendy.
H 169. „ flabellifm-mis, Keller. Red Sea

(61).
H 170. Auletta lyrata (Esper).

* This genus I now propose for the reception of Carter's Halichondria infreguens. It is evidently
nearly related to Myxiila, and may lie diagnosed as follows: — " Esperellinse whose principal megascleres
are spined oxea, with which are associated smooth tylota. Microscleres tridentate isochelse and
sigmata."

t Lambe's identification is doubtful.

•1 H

234

CEYLON PEARL OYSTER REPORT.

II 171. Auletta elongata, n. sp.

H 172. Leucophloms fcetidus (Dendy). Amboyna

(83) ; Ternate (87) ; (?) China Sea (86).
II 173. Ciocidypta fyZm, Bowerbank. Port Elizabeth

(22) ; Southern and Eastern Coasts of

Australia (3, 10).

11 174. Gollocaiypta digitata, n. gen. et sp.

Order: Euceratosa.
Family: Aplysillidee.

11 175. Darwim lla simplex, TOPSENT. Mediterranean,
Azores.

H 176. Megalopastas nigra (Dendy).
H 177. „ pulvittus, n. sp.

Family : Spongeliidoe.

II 178. Spongelia fragilis, Montagu. Cosmopolitan.

H179. „ elastica, SCHULZE. Cosmopolitan.

II 180. ,, incrustata, n. sp.

D 181. „ (?) conica (Bowerbank, 8). Wes-
tern Indian Ocean (Glorioso Island) (16).

H 182. Psammopemma crassum (Carter). Australia
and New Zealand (66).

Family : Spongiidae.

183. Cacospongia cavernosa, Schmidt {fide Len-

DENFELD, 66). Almost cosmopolitan (66).
II 184. Cw'ti.-i/iiuii/iti Sri/In /•;.<,; Schmidt. Mediterranean.
H185. Euspongia officinalis, Auctorum. Almost

cosmopolitan (66).
186. Euspongia irregularis, LENDENFELD (Jide

LENDENFELD, 66). Madagascar; East Coast

of Australia : Torres Straits ; Oceania ;

Bahamas (66).
H 187. Euspongia tenuiramosa, n. sp.

188. „ trincomalensis, Lendenfeld (fide

Lendenfeld, 66). Nassau; Havannah;

Pernambuco (66).
II 189. Hippospongia inteslinalis (Lamarck). Medi-
terranean; Zanzibar; Mascarene and Ami-

rante Islands ; Porto Rico.
H 190. Hippospongia clathrata (Carter). Red Sea ;

Australia ; American Coast of North Atlantic.
II 191. Hippospongia anomala, Polejaeff. Torres

Straits.
H 192. Hippospongia dura, Lendenfeld. American

Coast of North Atlantic.

193. Stelospongia ondaatjeana, Lendenfeld (66).
H 194. Phyllospongiapapyracea(T^FER). Tranquebar;

Cape of Good Hope ; Mozambique ; (?) West

Coast of Australia; (?) New Zealand; (?)

Chatham Islands (66).
H 195. Phyllospongia holdsworthi (Bowerbank).
II 196. Hircinia fusca, Carter.

197. ., vallata, Dendy (2). Port Phillip

Heads, Australia (66).
H 198. Hircinia tvberosa, n. sp.
H 199. ,, schulzei, n. sp.
H 200. ,, anomala, n. sp.
D201. „ arundiiuicea, CARTER (4).
H202. Aplysina purpurea, Carter.

203. „ fusca* Carter (4). Seychelles (16);
South-west Coast of Australia (65).

204. Aplysina spengelii, Lendenfeld. Jamaica
(66).'

II 205. Aplysina herdmani, n. sp.

Class: CALCAREA.

Order: Homocoela.

H 206. Leucosolenia coriacea (Montagu). North-east

Atlantic (7).
H 207. Leucosolenia lenuipilosa, n. sp.

208. Leucosolenia (Ascallis) darwinii (Haeckel, 7).

Red Sea, Java (7).

Order : Heteroccela.
Family : Sycettidse.

209. Sycetta sagittifera, Haeckel (7).

210. Sycon raphanus, Schmidt, fide Haeckel (7).
Mediterranean, Red Sea (7) ; South Coast of
Australia (7, 80).

Family : Grantiidse.

211. Leucandra (Leucetta) primigenia (Haeckel, 7).
Cosmopolitan.

212. Leucandra ( Leucortis) pulvinar (Haeckel, 7).f
West Coast of Australia, Red Sea (7).

II21o. Leucandra donnani, n. sp.

Family : Amphoriscidse.
II 214. Heteropegma nodus-gordii, Polejaeff. North
Australia, Bermudas (77).
215. Leucilla (Leucandra) cucumis (Haeckel, 7).
South Coast of Australia (7).

* According to Lendenfeld (66), this species is identical with Hyatt's Dendrospongia crassa. If so, the
geographical range must be extended to Nassau.

t CARTER (5) adopts Haeckel's varietal name indira for the Ceylon form.

SPONGES. 235

It appears from the above list that no less than 215 species of sponges (including
about seven doubtful ones) have now been recorded from Ceylon waters, which evidently
form an extremely rich centre of sponge distribution. Of these 215 species, 146
occur in Professor Herdman's collection, of which 77 (or 527 per cent.) are here
described as new. 1 consider it very satisfactory to have been able to identify as
many as 69 previously known species in the collection, especially as I always prefer
giving a new specific name to making a doubtful identification ; mistakes of the latter
kind being often far more difficult to coi-rect than those of the former. Seventy-five
out of the total number of species have been recorded from localities beyond the
Ceylon area.

As regards the general character of the Ceylon Sponge-Fauna, the most striking
feature, next to its richness, is its close relationship with the Sponge-Fauna of
Australia and the adjacent islands. In the Report on the " Challenger" Monaxonida
we defined an area of distribution (No. IV. on the chart) including Australia and the
islands north of Australia as far as the Philippines (inclusive). This we called the
Indo- Australian area. No less than 47 out of the 75 species whose range is known to
extend beyond the Ceylon region are common to the latter and our Indo- Australian
region — the majority of them occurring actually along the Australian Coast.

Of these 47 species, the following are not known to occur except in the Indo-
Australian area (in the "Challenger" sense) and westwards as far as Southern India:—
Stceba simplex, Myriastra clavosa, Pilochrota haeckeli, Geodia globostellifera,
Xenospongia patelliformis, Chondrilla australiensis, Toxochalina robusta, Reniera
madrepora, Petrosia testudinaria, Chalina sub-armigera, Siphonochalina communis,
Esperella parishii, (?) Iotrochota purpurea, Histoderma fistulatum, Leucophlceus
fetid us, Hippospongia anomala, Hircinia vallata, LeuciUa cucumis. The following
extend (so far as known) westwards not further than the East Coast of Africa
or eastwards not further than New Zealand : — SpirastreUa vagabunda (Aden),
Ceraochalina multiformis (New Zealand), Iotrochota baculifera (Western Indian
Ocean), Clathria frondifera (Western Indian Ocean, Red Sea), Acarnus ternatus
(Western Indian Ocean, Tahiti), Ciocalypta tyJeri (Port Elizabeth), Psammopemma
crassum (New Zealand), Phyllospongia papijracea (Mozambique, Australia ?, New
Zealand ?), Aplysina fusca (Seychelles), Leucosolenia darwini (Pted Sea), Leucandra
pulvincvr (lied Sea). Another characteristic Indo -Australian species is Agelas
mauritiana, hitherto recorded only from Mauritius, Ceylon, Ternate (under the
synonym A. cavernosa, Thiele), and doubtfully from Tristan da Cunha.

If, on the other hand, we compare the Ceylon Sponge-Fauna with that of the Pted
Sea, as elaborated chiefly by Keller (61), we notice a considerable difference.
Kellek records a total of 88 species from the Red Sea, to which Topsent (91, 92)
has added 13. Of these, only 14 are known to occur in Ceylon, viz. : — Spongelia
fragilis, Euspongia officinalis, Cacospongia cavernosa, Hippospongia claihrata,
Clathria frondifera, Acanthella flabelliformis, SpirastreUa vagabunda, Placospongia

2 H 2

236 CEYLON PEARL OYSTER REPORT.

melobesioides, Chondrilla nucula, Chondrosia reniformis, Leucosolenia darwini,
Sycon raphanus, Leucandra primiyenia, Leucandra pulvinar, of which all but five
(( 'lathria Jrondifera, Acanthella jiabelliformis, Spirastrella vagabunda, Leucosolenia
darwini, and Leucandra pulvinar) are so widely distributed that their occurrence is
of little significance. A few other Red Sea species, however, are very closely related
to Ceylon forms, and it is not improbable that two or three which have been described
by Keller under new names may be identical with Ceylon species, as will appear
from the descriptive part of this report. Amongst the Australian Sponge-Fauna also
a number of species occur which are very nearly related to Ceylon species, and which
have not been taken account of in the above comparison.

A considerable number of sponges have also been described from South and East
African waters outside the Red Sea. The " Challenger " made collections in the
neighbourhood of the Cape of Good Hope, and Kikkpatrick (17) has lately described
45 species collected by Dr. Gilchrist. Keller (61) also gives a considerable list of
species from the East African area, in which, however, while excluding South Africa,
he includes the islands of the Western Indian Ocean visited by the " Alert." In
spite of these researches only four Ceylon sjjecies appear to have been met with
along the African Coast itself (outside of the Red Sea), viz. ■.—Phyllospongia
papyracea (Mozambique), Ciocalypta tyleri (Port Elizabeth), Tedania digitata
(Mozambique), and Spmigelia fragilis (Zanzibar), the last two being cosmopolitan.
On the other hand, a number of Ceylon species extend, as we have seen, to the
islands of the Western Indian Ocean, viz. : — Hippospongia intestinalis, Euspongia
irregularis, Cacospongia cavernosa, Spongelia (?) conica, Aplysina fusca, Tedania
digitata, lotrochota baculifera, Lotrochota purpurea, Clathria frondifera, Acarnus
ternatus, Agelas mauritiana, Chondrilla nucula, Samus anonymus, Leucandra
primigenia. These species have very possibly spread south-westwards from Ceylon
along the chain of small islands formed by the Maldives, Seychelles, Chagos,
Amirantes, &c.

Even if we make allowance for errors and omissions in identification and doubtful
species, and also for the important fact that many more species are known from
Australasian waters than from the Red Sea and East African Coast, it appears
probable from the above data that the Sponge-Fauna of Ceylon is more closely
related to that of the eastern side of the Indian Ocean than it is to that of the
western side, and the Ceylon region may be safely included zoo-geographically in our
Indo-Australian area. The similarity of the Sponge-Fauna of Ceylon to that of
Australia is doubtless to be accounted for by the facilities of distribution and suitable
habitats afforded by the broken coast line which extends between the two, for it is
probable that, owing to the brief duration of their free-swimming larval condition,
shallow- water sponges are rarely able to traverse wide areas of deep ocean.

A few other points concerning the geographical relationship of the Ceylon Sponge-
Fauna are, perhaps, worth noticing in this place. The Sponge-Fauna of the Azores

SPONGES. 237

has lately been worked out in great detail by Topsent (48, 62), who records eight
species from that area which are also met with in Ceylon, while two of these, Dotona
pulchella and Rhdbdoploca unispicuMfera, have not yet been met with elsewhere.
The genus Pvtrvun'cK, also, is only known as yet from the Azores and from Ceylon,
the two species being almost identical. Hence, in spite of the great difference in
point of distance, the Sponge-Fauna of the Azores appears to resemble that of Ceylon
nearly as much as does that of the Red Sea !

There also appears to be a certain amount of affinity between the Sponge-Fauna of
Ceylon and that of the American coast of the North Atlantic, especially amongst the
horny sponges (<'■;/., Hippospongia dura), though, perhaps, some of the identifications
in this group are a little doubtful. The Lithistid genus Aciculites, however, has, so
far as I am aware, only been recorded from Ceylon and Havannah, and the two
species in these two remote localities are closely similar.

Very nearly two-thirds of the total number of species are, however, so far as is
known at present, peculiar to the Ceylon area, and, although the number of such
species (at present 140) will certainly be largely decreased by future identifications
and discoveries in other localities, it will, on the other hand, also be augmented by
fresh additions to the list of the Ceylon Sponge-Fauna. My investigations of the
Sponge-Fauna of other parts of the world also lead me to believe that while a
considerable number of species enjoy a very wide geographical range, the majority
have comparatively small areas of distribution. I strongly suspect, however, that a
large number of our so-called "species" will ultimately be found to lie so closely
connected by intermediate forms that it will no longer be possible to separate them
sharply from one another. In the meantime they must be distinguished by separate
names, and, if the different forms are fully and properly described, it should be easy
for future workers to trace out their true relationships.

REFERENCE LIST OF LITERATURE.

(1.) Eidi.ey and Dexdy. — 'Eeport on the Monaxonida collected by H.M.S. "Challenger,"' 1887.

(2.) Dexdy. — "The Sponge-Fauna of Madras. A Eeport on a Collection of Sponges obtained in the

Neighbourhood of Madras by Edgar Thursthx, Esq." 'Ann. and Mag. Nat. Hist.' (5), xx.,

p. 153. 18*7.
(3.) Dexdy. — "Eeport on a Second Collection of Sponges from the Gulf of Manaar." 'Ann. and Mag.

Nat. Hist.' (6), hi., p. 73. 1889.
(4.) Carter. — "Eeport on Specimens dredged up from the Gulf of Manaar and Presented to the

Liverpool Free Museum by Captain W. H. Cawne Warrex." 'Ann. and Mag. Nat. Hist.' (5),

v., p. 437; (5), vi., pp. 35 and 129. 1880.
(5.) Carter. — " Supplementary Eeport on Specimens dredged up from the Gulf of Manaar," &c. ' Ann.

and Mag. Nat. Hist.' (5), vii., p. 3G1. 1881.
(6.) Esper. — 'Fortsetzung der Prlanzeuthiere,' Part 2. 1798-180C.

238 CEYLON PEARL OYSTER REPORT.

(7.) Haeckel. — ' Die Kalkschwamme.' 1872.

(8.) Bowerbank. — " Report on a Collection of Sponges found at Ceylon by E. W. H. Holdsworth, Esq."

'Proc. Zool. Soe. Loud.,' p. 25. 1873.
(9.) Holdsworth. — "Note on the Occurrence of Xenospongia paiellifm-nds, Gray, on the Coast of

Ceylon." 'Proc. Zool. Soc. Loncl.,' p. 32. 1873.
(10.) Dendy. — "Catalogue of Non-calcareous Sponges collected by J. Bracebridge Wilson, Esq., M.A.,

in the Neighbourhood of Port Phillip Heads." ' Proc. Roy. Soc. Victoria,' vol. vii. (N. S.),

p. 232; vol. viii. (N. S.), p. 14: vol. ix. (N. S.), p. 230. 1895-1897.
(11.) Topsent. — " Materiaux pour servir a l'etude de la Faune des Spongiaires de France." ' Mem. Soc.

Zool. de France,' tome ix., p. 113. 1896.
(12.) Minghin. — "Sponges." ' Lankester's Treatise on Zoology,' Part ii. 1900.
(13.) Lendenfeld. — " Die Tetraetinelliden der Adria." ' Denk. Math.-Nat. Classe K. Akad. Wiss.,'

Rand lxi. 1894.
(14.) TOPSENT. — "Etude Monographique des Spongiaires de France." 'Arch. Zool. Exper. et Gen.,'

ser. 3, tome ii., p. 259; ser. 3, tome iii., p. 193; ser. 3, tome viii., p. 1. 1894-1900.
(15.) Sollas.— ' Report on the Tetraetinellida collected by II. M.S. "Challenger."' 1888.
(16.) Ridley. — "Spongiida. Report on the Zoological Collections made in the Indo-Pacific Ocean during

the Voyage of H. M.S. 'Alert.'" 'Brit. Mns.' 1884.
(17.) Kirkpatrick. — " Descriptions of South African Sponges." ' Marine Investigations in South Africa,

Cape of Good Hope Department of Agriculture.' 1902-1903.
(18.) Carter. — "Descriptions of Sponges from the Neighbourhood of Port Phillip Heads, South

Australia." 'Ann. and Mag. Nat. Hist,' (5), xvi., pp. 277, 347 ; (5), xvii., pp. 40, 112, 431, 502 ;

(5), xviii., pp. 34, 126. 1885-1886.
(19.) Carter. — " Supplement to the Descriptions of Mr. J. Bracebridge Wilson's Australian Sponges."

'Ann. and Mag. Nat. Hist.' (5), xviii., pp. 271, 369, 445. 1886.
(20.) Carter. — " Report on the Marine Sponges, chiefly from King Island, in the Mergui Archipelago,

collected for the Trustees of the Indian Museum, Calcutta, by Dr. John Anderson." ' Journ.

Linn. Soc' (Zool.), vol. 21, p. 61, 1887.
(21.) Carter. — "Contributions to our Knowledge of the Spongida. — Paehytragida." 'Ann. and Mag.

Nat. Hist.' (5), xi., p. 344. 1883.
(22.) Bowerbank. — " Contributions to a General History of the Spongiadse," Part IV. ' Proc. Zool. Soc,

Lond.,' p. 3. 1873.
(23.) Schmidt. — 'Grundziige einer Spongien-Fauna des atlantischen Gebietes.' 1870.
(24.) Fristedt. — "Bidrag till Kannedomen om de vid sveriges vestra Kust lefvande Spongia\" ' Kongl.

Svenska Vetenskaps-Akademiens Handlingar,' Band 21, No. 6. 1885.
(25.) Topsent. — " Application de la Taxonomie actuelle a une Collection de Spongiaires," &c. ' Mem.

Soc. Zool. de France,' tome vii., p. 27. 1894.
(26.) Topsent.— "Introduction a l'etude monographique des Monaxonides de France. Classification des

Hadromerina." 'Arch. Zool. Exper. et Gen.,' ser. 3, tome vi., p. 91. 1898.
(27.) Carter. — "Contributions to our Knowledge of the Spongida." 'Ann. and Mag. Nat. Hist.' (5), iii.,

p. 343. 1879.
(28.) Topsent.--- " Sur les Hadromerina de l'Adriatique." ' Bull. Soc. Sei. et Med. de l'Ouest.' ler trimestre,

p. 117. 1898.
(29.) Wilson. — "The Sponges collected in Porto Rico in 1899," &c 'U.S. Fish Commission Bulletin,'

vol. 2, p. 375. 1900.
(30.) Gray. — "Description of a New Genus of Sponge (Xenospongia) from Torres Strait." 'Proc Zool.

Soc, Lond.,' p. 229. 1858.
(31.) Gray. — "Notes on the Arrangement of Sponges, with the Descriptions of some new Genera."

'Proc. Zool. Soc, Lond.,' p. 492. 1867.

SPONGES. 231)

(32.) Carter. — "Notes introductory to the Study and Classification of the Spongida." 'Ann. and Mag.

Nat. Hist,' (4), xvi., pp. 1, 126, 177. 1875.
(33.) CARTER. — "Some Sponges from the West Indies and Acapulco in the Liverpool Free Museum," &c.

•Ann. and Mag. Nat. Hist.' (5), ix., pp. 266, 346. 1882.
(34.) CARTER. — "Description of a Siliceous Sand-Sponge found on the South-East Coast of Arabia {Tethya

dadyloidea)." 'Ann. ami Mag. Nat, Hist.' (4), iii., p. 15. 1869.
(35.) VbSMAEB ami VERNHOUT. — 'The Porifera of the Siboga Expedition. I. The genus Placosjjonr/ia.'

1902.
(36.) Topsext. — "Contribution a l'etude des Clionides." 'Arch. Zool. Exper. et Gen.,' ser. 2, tome v., Us.

Supt. L887.
(37.) TOPSENT. — "Deuxieme Contribution a l'etude des Clionides." 'Arch. Zool. Exper. et Gen.,' ser. 2,

tome ix., p. 555. 1891.
(38.) HANCOCK. — "On the Excavating Powers of certain Sponges belonging to the genus Glioma; with

Descriptions of several new Species, and an Allied Generic Form." ' Ann. and Mag. Nat. Hist.'

(2), iii., p. 321. 1S49.
(39.) Thiele. — "Studien fiber pazifische Spongien." 'Zoologica,' Heft 24 und 24°. 1898 and 1899.
(40.) SCHULZE. — "Die Familie der Chondrosidfe." ' Zeit. f. Wiss. Zool.,' Band xxix 1877.
(41.) Carter. — "On two new Species of Gumminece, with Special and General Observations." 'Ann. and

Mag. Nat. Hist.' (4), xii., p. 17. 1873.
(42.) Lendenfeld. — "A Monograph of the Australian Sponges." ' Proc. Linn. Soc. N.S.W.,' vol. ix.,

pp. 121, 310, 1083; vol. x., pp. 3, 282, 475, 481, 845. 1884-1885.
(43.) Carter. — " Additional Information on the Structure of Tethya dadyloidea, Cart." ' Ann. and Mag.

Nat, Hist.' (4), ix., p. 82. 1872.
(44.) Schmidt.— "Die Spongien des Meerbusen von Mexico." 1879-1880.

(45.) ToPSENT. — "Eponges nouvelles des Azores." ' Mem. Soc. Zool. de France,' tome xi., p. 225. 1898.
(46.) Haxttsch. — "Revision of the Generic Nomenclature and Classification in Bowerbaxk's 'British

Spongiadss.' " ' Trans. Liverpool Biol. Soc.,' vol. viii. 1894. (-
(47.) Schmidt. — "Die Spongien des Adriatischen Meeres." 1862.
(48.) Topsext. — " Contribution a l'etude des Spongiaires de l'Atlantique Nord." ' Resultats des Campagnes

Scientifiques du Prince de Monaco,' fascicule ii. 1S92.
(49.) BOWERBANK. — "Contributions to a General History of the Spongiache." Troc. Zool. Soc. Lond.,'

p. 281. 1875.
(50.) Schmidt. — "Die Spongien der Kiiste von Algier." 1868.
(51.) Lexdexfeld. — " Die Chalineen des Australischen Gebietes." ' Zoolog. Jahrbiich.,' Band 2, p. 723.

1887.
(52.) Bowerbaxk. — " Monograph of British Spongiada\" 'Kay Society.' 1864-1882.
(53.) HiGGlX. — " Description of some Sponges obtained during a Cruise of the Steam-yacht ' Argo ' in the

Carribean and neighbouring Seas." ' Ann. and Mag. Nat. Hist.' (4), xix., p. 291. 1877.
(54.) CARTER. — "Contributions to our Knowledge of the Spongida." 'Ann. and Mag. Nat. Hist.' (5),

xii., p. 308. 1883.
(55.) Norman. — " On the Genus Haliphysema, with Descriptions of several Forms apparently allied to it."

'Ann. and Mag. Nat. Hist.' (5), i., p. 264. 1878.
(56.) Carter. — "Contributions to our Knowledge of the Spongida." 'Ann. and Mag. Nat. Hist.' (5\

iii., p. 284. 1879.
(57.) Carter. — " Descriptions and Figures of Deep-sea Sponges and their Spicules, from the Atlantic

Ocean," &c. ' Ann. and Mag. Nat. Hist.' (4), xviii., pp. 226, 307, 388, 458. 1876.
(58.) EHLERS. — ' Die Esper'schen Spongien,' &c. Erlangen. 1870.
(59.) Topsext. — " Une Reforme dans la Classification des Halichmidririu." ' Mem. Soc. Zool. de France,'

tome vii., p. 5. 1894.

240 CEYLON PEARL OYSTER REPORT.

(60.) RIDLEY. — "On the Genus Ploamia, Schmidt," &c. 'Journ. Linn. Soe. (Zool.),' vol. 15, p. 476.

1881.
(61.) Keller.— "Die Spongienfauna des Rothen Meeres." 'Zeit. f. Wiss. Zool.,' Band 48, p. 311. 1889;

Band 52, p. 294. 1891.
(62.) TOPSENT. — " Spongiaires des Acores." 'Resultats des Campagnes Scientifiques du Prince de

Monaco,' fascicule xxv. 1904.
(63.) DENDY. — " The New System of Chalininse, with some brief Observations upon Zoological Nomen-
clature." 'Ann. and Mag. Nat. Hist.' (5), xx., p. 326. 1887.
(64.) TOPSENT.—" Spongiaires." 'Resultats du Voyage du S.Y. "Belgica" en 1897-1898-1899.' Anvers,

1901.
(65.) Carter. — " Contributions to our Knowledge of the Spongida. Order II.: Ceratina." 'Ann. and

Mag. Nat. Hist.' (5), viii 1881.
(66.) Lendenfeld. — " A Monograph of the Horny Sponges." ' Roy. Soc' 1889.
(67.) MtTLLER. — "Ueber Dwrwinella aurea, einen Schwamm mit sternformigen Hornnadeln." 'Archiv. f.

mikr. Anat.' Band i. 18G5.
(68.) Carter. — "Description, with Illustrations, of a New Species of Aph/sina from the North-West Coast

of Spain." 'Ann. and Mag. Nat. Hist.' (1), x., 1872.
(69.) Hyatt. — " Revision of the North American Poriferaj." ' Mem. Boston Soc. Nat. Hist.,' vol. ii., 1S75

and 1877.
(70.) Schulze.— " Die Familie der Aplysinicbe." ' Zeit. f. Wiss. Zool.,' Band xxx. 1878.
(71.) Schulze. — " Die Gattung Spongelia." ' Zeit. 1. Wiss. Zool.,' Band xxxii. 1879.
(72.) Schulze.— " Die Familie der Spongidaj." ' Zeit. f. Wiss. Zool.,' Band xxxii. 1879.
(73.) Lamarck.—" Sur les Polypiers empates." ' Ann. Mus. Nat. Hist.,' Paris, vol. 20. 1813.
(74.) POLEJAEFF. — 'Report on the Keratosa collected by H.M.S. " Challenger."' 1884.
(75.) Dendy. — " Observations on the West Indian Chalinine Sponges, with Descriptions of New Species."

' Trans. Zool. Soc. Lond.,' vol. xii., part 10. 1890.
(76.) Dendy. — " The Organization a^id Classification of the Calcarea Homoccela, with Descriptions of the

Victorian Species." ' Trans. Royal Soc. Victoria,' vol. iii., Part 1. 1891.
(77.) Polejaeff.— ' Report on the Calcarea collected by H.M.S. " Challenger.'" 1883.
(78.) Dendy. — " Observations on the Structure and Classification of the Calcarea Heterocoela." ' Quart.

Journ. Micro. Sci.,' vol. 35 (N. S.). 1893.
(79.) Lendenfeld. — ' Descriptive Catalogue of the Sponges in the Australian Museum, Sydney.'

London, 1888.
(80.) Dendy. — " Synopsis of the Australian Calcarea Heteroccela," &c. ' Proc. Roy. Soc. Victoria,' vol. 5.

1893.
(81.) Johnston. — 'British Sponges and Lithophytes.' 1812.
(82.) Carter. -" Zoology of Kerguelen Island — Spongiidae." 'Phil. Trans. Roy. Soc.,' vol. 168.

1879.
(83.) TOPSENT.— "Spongiaires de la Baie d'Amboine." 'Revue Suisse de Zoologie et A nnales du Musee

d'Histoire Naturelle de Geneve,' tome iv., fascicule 3. 1897.
(84.) TOPSENT. — "Diagnoses d'Eponges nouvelles de la Mediterranee et plus particulierement de Banyuls."

' Arch, de Zool. Exp. et Gen.,' tome x. 1892.
(85.) Lambe. — " Sponges from the Western Coast of North America." ' Transactions of the Royal Society

of Canada,' Section iv. 1894.
(86.) Lindgren. — " Beitrag zur Kenntniss der Spongien-Fauna des Malayischen Archipels und der

Chinesischen Meere." ' Zoologische Jahrbueher,' Band xi. 1898.
(87.) Thiele. — " Kieselschwamme von Termite, I., II." ' Abhandlungen der Senckenbergischen natur-

forschenden Gesellschaft,' Band xxv. 1900-1903.
(88.) LUNDBECK. — "Porifera. — Part 1." ' Danish Ingolf-Expedition,' vol. vi. 1902.

SPONGES. 241

(89.) Toi'Sknt. — " yuol(|iirs Spongiaires du Banc de Campeche et de la Pointe-a-Pitre." ' Mem. Soc. Zool.

de Franco,' tome 2. 1889.
(90.) THURSTON. — "Ram&varam Island and Fauna of the Gulf of Manaar." 'Madias Government

Museum,' Bulletin No. 3. 1895.
(91.) TijI'sknt.— " Eponges de la Mer Rouge." ' Mem. Soc. Zool. de France,' tome v. 1892.
(92.) Topsent. — "Note sur quelques Eponges du Golfe de Tadjoura." 'Bull. Soc. Zool. de France,'

tome xviii. 1893.

DESCRIPTION OF PLATES.

PLATE I.

Fig. 1. HexadeUa indica ; vertical section of R.N. 46, combined drawing.

111., inhalant chones ; I'LL., choanosomal lamella; Or., sub-cortical crypts; Eel., ectosome ;
F.C., flagellate chambers; L.E.C., larger exhalant canal; Mem., roofing membrane of
inhalant chones; P.E.I '., primary exhalant canals; P.I.C., primary inhalant canals;
Sy.J., symbiotic algse.
,, 2. Hexadella indica ; small portion of vertical section through choanosomal lamella, more highly
magnified (Zeiss F., Oc. 2).
C.T.C., connective-tissue cells; Ep., epithelium lining primary inhalant and exhalant canals;
I.I'd., inhalant canaliculi. (Other lettering as before.)
,, 3. Hexadella indica : tangential section of the ectosome, seen from below, showing the inhalant

chones, &c. x 52. (Lettering as before.)
,, 4. Plakinastrella intermedia : R.N. 224. x 2. — o., vent.
,, 5. Ecionema carteri; R.N. 175. x 2. — o., vents; j>..s., pore-sieves.
,, C. TeiUla poculifera ; R.N. 230. x 3. Upper part in section.

67., cloaca; L.E.I'., larger exhalant canals.
„ 7. TetUla limicola ; R.N. 70. Nat. size.

p., vents ; r.t., root tuft of spicules matted together with mud.
„ 8. Taprobane herdmani; R.N. 40. Slightly reduced ; from a photograph.

PLATE II.

Fig. 1. Dercitopsis ceylomka; R.N. 139. Spicules. All x 230.
a.-h., calthrops and triods ; i.-o., oxea.
„ 2. Plakinastrella intermedia ; R.N. 224. Spicules.

a.-d., dichotriames, x 52 ; e.-n., oxea, x 52 ; o.-p., microxea, x 230 ; </.-"., oxyasters, x 230.
„ 3. Plakinastrella sch »/ zei ; R.N. 149. Spicules.

a.-d., triames, x52; c-q., oxen, x 52 ; r.-t., microxea, x230; u.-.c, oxyasters, x 230.
„ 4. Pilochrota haeckeli'; R.N. 127. Spicules. All x 52.

a., orthotrisene ; /»., monstrous form of orthotriwne; c, anatriajne ; ('., e., monstrous forms of
anatriajne.
,, 5. Pilochrota hornelli; R.N. 176. Spicules.

a., /'., orthotria'.nes, x 52 ; <•., '/., cladomes of orthotria-nes seen from above, x 52 ; <'., anatriajne,
x 52 ; /., oxeote, x 52 ; g.-i., chiasters, x 230.
,, 6. Stelletta herdmani ; R.N. 137. Spicules.

a.-'., trhrites, x 52 ; '/., oxeote, x 52 ; e.-h., chiasters, x 230 ; i.-k., oxyasters, x 230.
„ 7. Stelletta vestigium ; R.N. 200a. Spicules. All x 230.

*'.-»/., reduced triaenes ; e., oxeote; /., </., spherasters; //.. /.. oxyasters.

2 I

242 CEYLON PEAEL OYSTER REPOET.

PLATE III.

Fig. 1. Ecionema carteri ; R.N. 175. Spicules.

a., b., pkgiotrirenes, x 52 ; c, plagiotriiene (protrisene T), x 52 ; <?., cladome of c, x 230 ;
d., anatriasiie, x 52 ; d'., cladonie of d., x 230 ; e., oxeote, x 52 ; /., microstrongyla, x 230 ;
if., chiasters, x 230.
„ 2. Ecionema laviniensis ; R.N. 265. Spicules.

»., dichotri»ne, x 52 ; b., cladome of dichotriaene, seen from above., x 52 ; c., anatriaene, x 52 ;
c'., cladome of c, x230; d., oxeote, x 52 ; e., chiasters, x230; /., oxyasters (?), x230;
(/., microstrongyla, x 230 ; //., microxeotc (?), x 230.
,, 3. Gcodia penmcmata ; R.N. 223. Arrangement of the skeleton as seen in vertical section, x 52.

a., a'., dichotrisenes ; b., somal anatriaene; c, reduced cortical anatriaenes ; d., oxea ; c, sterr-
asters in cortical layer; e'., young sterrasters in choanosome ; /., large spherasters.
„ 3a. Geodia perwicinata ; R.N. 223. Reduced anatriaene, x 230.
„ 3l5. Gcodia peruncinata ; R.N. 223. Microscleres. All x 230.

a., small spheraster (chiaster ?) ; /*., large spheraster ; ''., outline of sterraster, to show relative
size.
,, 4. TctiUa pondifcra ; R.N. 230. Spicules.

".,/'., plagiotria?nes, x52; c, anatriaene, x 52 ; r'., cladome of c, x 230 ; d., protrisene, x52;
d'., cladome of d., x 230 ; c, oxeote, x 52 ; /., '/•, microxea, x 230 ; h., sigmata, x 230.
,, 5. Tetilla anomala ; R.N. 153. Spicules.

a., part of protriaene, x 230 ; b., anatriaene, x 52 ; //., cladome of b., x230; c, oxeote, x52;
d., sigmata, x 230.
„ 6. Tetilla limicola ; R.N. 70. Spicules from body. All x 230.

a., anatriaene, cladome and part of shaft; b., protriaene, cladome and part of shaft; <., oxeote;
d., sigmata.
,, 7. Paratetilla cineriformis ; R.N. 214. Spicules.

a.-d., modified triaenes, x 230. (The central canal of these spicules is indicated by the dark
shading.) '•., protriaene, x 52 ; e'., cladome of c, x 230 ;/., anatriaene, x 52 ;/'., cladome
of /., x 230 ; '/., oxeote, x 52 ; ;/', i/", ends of ;/., x 230, showing their irregularity in
shape and the central canal ; h., sigmata, x 230.

PLATE IV.

Fig. 1. Craniella elegans ; R.N. 193. Vertical section, x 28.

a., cinatriaenes ; b., protri;enes ; CO., cortical oxea; d.m., dermal membrane; emb., embryo;
f.c, fibrous layer of cortex; i.c.c., intracortical (sub-dermal) cavities; r.b., radiating
spictilebundles of main skeleton.
,, 2. Tapivbane herd/ward ; R.N. 40. Desmas. All x 230.
a., b., young monocrepid desmas; c, adult desma.
„ 3. Aciculites orientalis ; R.N, 150. Spicules. All x 230.

(»., /»., c, young monocrepid desmas; d., adult desma; e., /., strongyla.
„ 4. Discodermia emarginata; R.N. 234. Spicules. All x 230.

a.-d., stages in the development of the tetracrepid desma; c, end of branch of an adult
desma; /., very young discotrisene, side view, s, shaft; ;/., adult discotriasne, surface
view s, shaft ; //., microrhabds.
., 5. Petromica massalis ; R.N. 257. Spicules. All x 52.

"., monocrepid desma; b.-d., monaxonid megascleres, showing variation in form.

SPONGES. 243

PLATE V.
Pig. |. Stvba extensa ; R.N. 167. Spicules.

a.-d., dichotrisenes (sk. = shaft), x230; e., e., microxea, x 530.

2. CoppaUas repta/ns ; R.N. 242. Spicules.

«.-<•., oxea, x 52 ; '/., chiasters, x 5.30.

3. Asteropus haeehti ; K.N. 219. Spicules.

«., oxeote, x 52 ; 6., oxyaster, x 530 ; c, d., sanidasters, x 530.

4. Oryplotefhya agglutinans ; R.N. G2. Half the specimen, after division in the median plane.

Natural size.
ch., choanosome; ect., ectosome; f.h., foreign bodies adhering to the surface; /./., inner fibrous
layer of ectosome; f.p., finger-shaped process of ectosome; .<■., process of ectosome
flattened at the end and containing longitudinal canals.
„ 5. Oryptotethya aggluimans ; R.N. 62. Spicules.

a., I)., oxea, x 52; c, chiasters from ectosome, x 230.
G. Hymedesmia curvistellifem ; R.N. 320. Spicules.
a., tylostyle, x 230; b., r., d., asters, x 530.
,, 7. Spirastrella temtorioides ; R.N. 239. Spicules.

a., b., tylostyli, x 230 ; c.-g., spirasters, x 5.30.
„ 8. Negombo tenuistellata ; R.N. 362. Spicules.
ii-.-c, styli, x230; d., sanidasters, x 530.
9. Cliona margaritiferm ; RN. 261. Spicules.

n., tylostyle, x230; b.-ih, spirasters, x530; <:, ;/., intermediate microscleres, x530; /(., /.,
spined microxea, x 530.
,, 10. Suberites cruciatus ; R.N. 315. Spicules.

it., ]>., tylostyles, x 230; c.-g., heads of tylostyles, showing variations, x 530.

t

PLATE VI.
Xeiwaporujia pateUiforum, R.N. 375.

Fig. 1. Entire specimen (young), upper surface, slightly diagrammatic, x 2.

«., projection caused by foreign body; in./., marginal fringe of spicules; mjp.g., marginal pore-
grooves; c, vents; r.p.g., radiating pore -grooves, feebly developed.
., 2. Radial vertical section, showing the ectosome (cortex), the sand-free portion of the choanosome,
and part only of the sandy layer of the choanosome ; slightly diagrammatic.
cart., cortex; il.b., dermal brushes of styli; ex.c, exhalant canal; f.b., bands of fibrous tissue
running inwards from the cortex to the sandy layer; //(./., marginal fringe of styli;
m.p.g., marginal pore-grooves, showing inhalant pores and canals ; o., vent. ; s.g., sand
grains.
,. 3. Small portion of similar section, including the entire thickness of the cortex and a small portion
of the choanosome, x 230.
fur/., cortex; /'./'., band of fibrous tissue running inwards from the cortex; Ji.c, flagellate
chambers ; scl., scleroblasts, containing small asters.
„ 4. Megascleres.

a., b., styli, x 52 ; b'., //"., basal and apical portions of //, x 230.
5. Larger microscleres, x 230.

a.- 1-., various forms of aster; </., a much-branched aster from the interior of the sponge.
„ 6. Smaller microscleres, x 530.

a. in., various forms of aster.

2 i 2

244 CEYLON PEARL OYSTER REPORT.

PLATE VII. (All the figures from photographs.)

Fig. 1. Siphonochalina communis, var. tenuispiculata ; R.N. 117, reduced to nearly ^.

,, 2. Ceraochalina multiformis, var. mtmaarensis ; R.N. 98, x about i.
„ 3. ,, ceylonica ; R.N. 5, x about i.

„ 4. Pachychalim spinilamella ; R.N. 94, x about i.

„ 5. Gelliodes carnosa ; R.N. 69, x about -J.

„ C. Colhcalypta digitata ; R.N. 74a, nearly natural size.

PLATE VIII. (All the figures from photographs.)

Fig. 1. Plocamia mcmaarensis ; R.N. 107, x about |.

„ 2. Clathria spiadosa, var. tessellata ; R.N. 92, x about f.

„ 3. PhaJcellia ceylonensis ; R.N. 34, x Jj;.

,, 4. Acarnus ternatus ; R.N. 105, x about J.

,, 5. Baspailia frutieosa, var. tenuiramosa; E.N. 100, x about i.

,, G. Acanthella carteri ; R.N. 36, x about §.

PLATE IX.

Fig. 1. Strongylophora duiissima ; R.N. 244. Part of dermal skeleton, x 230.

/»., microxea; s., strongyla.
,, 2. Toowchalina robusta, var. ridleyi ; R.N. 109. Part of main skeleton in vertical section, x 230.

o., oxea ; t., toxa.
„ 3. Gelliodes petrosioides ; R.N. 146. Spicules, x 230. — a., b., c, megascleres ; d., sigmata.
,, 4. Desmacella tubulata ; R.N. 324. Spicules, x 230.

a., b., styli ; c, sigmata ; /!., trichodragmata j e., isolated microxeote.
,, 5. Acantlumfer ceylonensis ; R.N. 247. Spicules, x 230.

«., smooth oxeote ; /'., stylus ; c, spined oxeote ; d., trichodragmata.
,, G. Gelliodes incrustans ; R.N. 112. Part of dermal skeleton, x 230. — o., oxea ; s., sigmata.
,, 7. Gellius cmgulatus, var. ccmaliculata ; R.N. 140. Spicules, x 230.

a., oxeote ; /'., sigmata ; r., toxiform.
„ 8. Eeniera zoologica ; R.N. 262. Oxea, x 360.
,, 9. Petrosia densissima ; R.N. 138a. Oxea, x 3G0.
„ 10. Reniera jrigmentifera ; R.N. 290. Spicules, x 360. — «., oxeote; />., strongylote ; c, d., stylote.

PLATE X.

Fig. 1. Pachychalina subeylindrica ; R.N. 292. Part of dermal skeleton, x 52.
„ 2. „ R.N. 292. Oxea, x 360.

,, 3. Chalina clathrata ; R.N. 102. Part of skeleton in tangential section near the surface, x 360.
,, 4. ( 'eraodwiina retiarmata ; R.N. 342. Oxea, x 360.

,, 5. Chalina subarmigera; R.N. 116, about half natural size (dry specimen, from photograph).
„ 6. Ceraochalina ceylonica ; R.N. 50. Oxea, x 360.
,, 7. Pachychalina brevispiculifera ; R.N. 110. Oxea, x 360.
„ 8. Ceraochalina retieutis ; R.N. 58. Part of dermal skeleton, x 360.

a., well developed oxea, lying for the most part outside the horny fibre; v.s., vestigial oxea,
lying in the horny fibre.
!). Cludina obtusispmdifera ; R.N. 370. Part of surface skeleton, x 360,
10. Trachyopsis halicJwndrioides : R.N. 147. Oxea, x 230,

SPONGES. 245

PLATE XL

Fig. 1. Paresperdla bidentata ; R.N. 263a. Spicules, x530.

a., tylostyle with bidentate apex; &., 6., serrated sigmata; c, anisoehela, side view; >/.,
anisochela, front view.
„ 2. Paresperella serratohamata ; R.N. 220c. Spicules, x 530.

n., tylostyle with mucronate apex; /»., h., serrated sigmata.
,, 3. Paresperella, sp. Gigantic serrated sigma, x 230.
„ 1. Echinodictynm clathmtum ; R.N. 325. Spicules, x 230.

ii., large stylus; &., slender stylus; c, </., oxea ; >•.,/., spined tylostyli.
„ 5. Rfynlla tenuissima ; R.N. 234a. Spicules, x 530.

a., polytylote ; 6., spined tylostyle; ''., isoehela, front view; i/., isochela, side view.
,, 0. Esperella crassissima ; R.N, 240. Spicules., x 230.

a., tylostyle; /'., anisochela, side view; c, anisochela, front view; <l., sigmata; <>., fcricho-
dragmata.
,, 7. Raspailia h&rnelli ; R.N. 59. Spicules, x 230. — a., large stylus; &., c, small surface styli; d., e.,

/., ;/., oxea ; h., Ii., strongyla ; /.-., /., m., spined subtylostyli.
„ 8. Histoderma vesicidatwm ; R.N. 212a. Part of dermal skeleton, x 230.

"., »., o., oxea; •->'., s., strongyla.
,, 9. His/, nlmiiii i-rsii iihitiim .- R.N.. 212a. Isochelse, x 460. — a., side view; h., front view.

PLATE XII.

Fig. 1. Spongosorites topsenti ; R.N. 152. Spicules, x 52.

a.-k., various forms of large megascleres ; /., two small oxea; in., intermediate form.
,, 2. Spongosorites (1) lamellata ; R.N. 236.

a. b. c, styli, x 52 ; d., two small oxea, x 52 ; c, small oxeote, x 230.
,, 3. SpoTU/osorites (1) lapidiformis ; R.N. 145. Spicules, x 79.
,, 4. Hymeniacidon petrosioides ; R.N. 151. Spicules, x 230.
,, 5. Tkrinaeophora dwrissima ; R.N. 355. Spicules, x 230.

a., stylote; b., oxeote; ''., </., trichodragmata.
„ 6. Tlniiitiri'jiliiiiit agariciformis ; R.N. 160a. Spicules, x 230.

a., !>., <\, styli; '/., c, oxea; /'., trichodragmata.
„ 7. Axinella haiichandrioides ; R.N. 75. Spicules, x 230. — a., oxeote; 6., c, stylote.
,, 8. Axinella manus ; R.N. 53. Spicules, x 230.
,, 9. Agelas ceyhnica ; R.N. 312. Spicules, x 230.
., 10. Hlnihih',1 nun indica; R.N. 341.

a., b., i'., rhabdostyli, x 230; </., minute spined styli, x 530; e., contort sigmata, x 530.

PLATE XIII.

Fig. 1. OoUocalypta digUata ; R.N. 74. Oxeote spicule from body, x 230.
„ 2. „ R.N. 74. Part of transverse section of a digitiform process, x 52.

ii.r., axial skeleton of spicules and spongin; d.p., position of dermal pores; ect., collenchy-
matous ectosomal tissue; I.e., longitudinal canals; sept., septum between two longitudinal
canals.
„ 3. Phakellia symmetrica ; R.N. 159. Spicules, x 230.— a, b, short styli; c, oxeote; d, long stylus.
„ 4. Axinella tenuidigitata ; R.N. 202A. Styli, x 79.

„ 5. Phakellia ceylonensis ; R.N. 34. Spicules, x 230. — a, b, styli ; c, d, e, oxea.
,, 6. Phakellia crassistylifera ; R.N. 256. Three spicules, x 230.

'ig.

1.

JJ

2.

»1

3.

»J

4.

T>

5.

)»

C.

7.

24G CEYLON PEARL OYSTER REPORT.

Fig. 7. Auhtta elmgata; R.N. 73. Spicules, x 79.— a, b, styli; c, oxeote; <7, e,/, strongyla.
8. Lewosolenia cmiacea, var. ceylonensis; R.N. 37 7. Spicules, x 360.

a, A, i ■, triradiates ; d, portion of oxeote.
it. Leucosolenia temipilosa ; R.N. 158. Spicules, x 360. — <i, triradiate ; b, quadriradiate, facial view ;
c, quadriradiate, side view, showing apical ray ; d, part of oxeote.
, 10. Leucanclra donnani ; R.N. 186. Spicules, x 79.

a, b, parenchymal triradiates; c, d, dermal triradiates; e,f, gastral quadriradiates ; g, oxeote.

PLATE XIV. (All the figures from photographs.)

Hircinia fusca ; part of R.N. 48, x f.

Hippospongia clathrata ; R.N. 24, x }i.

Euspongia officinalis, var. ceylonensis ; R.N. 37a, x about |.

Spongelia elastica, var. crassa; R.N. 35, x about f.

Hircinia anomala ; R.N. 1 3, x about S.

Phyllospongia papyracea, var. ; R.N. 104a, x \i.

Megalopastas nigra ; R.N. 71, x £.

PLATE XV.

Fig. 1. Iiiuviiirlhi simplex ; R.N. 302. — End of primary fibre, enclosed in conulus, x 230.

„ 2. Darwinella simplex ; R.N. 302. — Portion of primary fibre, in which the pith (p.) is partially replaced
by fungal (?) filaments (/.), x 230.

„ 3. Megalopastas pidvilhs; R.N. 191.— Portion of skeleton as seen in vertical section, showing the old
apices of primary fibres (ap.) enclosed in younger portions of the fibres, x 79.

„ 4. Diagrammatic sketch of the general anatomy of the genus Megalopastas, based upon vertical
sections of a specimen of M. elegans (= Dendrilla elegans, Lendenfeld) from Port Phillip
Heads, Australia, x 52. — r., surface conuli ; ex., embryo capsules ; cm., embryo; ex.e., exhalant
canals; f.c, flagellate chambers; i.e., inhalant canal; p.f., primary skeleton fibres; p.s., pore-
sieves; s.c., subdermal cavities; s.f., secondary skeleton fibres.

,, 5. Megalopastas nigra; R.N. 161a.— Part of skeleton, as seen in vertical section, including the dermal
skeleton (d.-g.) as well as the main skeleton, x 52.

„ 6. Megalopastas nigra; R.N. 161 A. — Dermal skeleton, x 52.

,, 7. Megalopastas nigra ; R.N. 161a.

End of primary fibre, showing mode of growth and origin of secondary fibres, x 230.

„ 8. Megalopastas nigra ; R.N. 161a.

Origin of new primary fibre from a secondary fibre of the dermal skeleton, vertical section,
x 230. — c, surface conulus; d.s., dermal skeleton; p.f., new primary fibre.

PLATE XVI.

Fig. 1. Hircinia anomala; R.N. 82. — Part of skeleton, as seen in transverse section, x 52.
„ 2. Hircinia tuberosa; R.N. 88a.— Part of skeleton, as seen in vertical section through the surface,

including the sand-cortex (s.c), x 52.
,, 3. Hircinia schulzei; R.N. 277. — Part of skeleton, as seen in vertical longitudinal section, x 52.
„ 4. Aplysina herdmani ; R.N. 340. — Part of skeleton, as seen in vertical section of body, x 79.
„ 5. Euspongia officinalis, var. ceylonensis; R.N. 101. — Part of skeleton, as seen in vertical section, x 79.

CEYLON PEARL OYSTER REPORT

SPONGES PLATE 1.

■">'•>«*•: '

Ch.L

.■.*."•••■••. ; -■•..•••.

i, -•. : :■•■■ ;;:;-/..A •■■•■.•.-• •■•"••■ •■'.= .« •

• \„\ ■ PEC ....

> iVew

FlG.3.

enay del .Figs 1-7.

E Wson.Llth. Cambridge

Figs. 1—3, Hexadella indica, n.sp. ; Fig. 4, Plakinastrella intermedia, n.sp. ;
Fig. 5, Ecionema carteri, n.sp.; Fig. 6, Tetilla pocdlifera, n.sp.; Fig. 7, Tetilla limicoi.a, n.sp.;

Fig. 8, Taprobane herdmani, n.sp.

LON l'KARI. OYSTER KKPONT.

SPONGES PLATE TI.

Fig. 3.

1, Dercitopsis ceylokica, u.sp. ; Fn I intermedia, u.sp. ; Fig. 3. Plakinastri

hulzei, n. sp. ; Fig. 4. Pilochrota haeckeli, Sollas; Fig. .r'. Pilochrota horselli, u.sp.;

HERDMANJ i.l.LETTA VESTIGIUM, 11. Sp.

(KYI, ON' PEARL OYSTER REPORT

SPONGES PL VTE III

!. EciONEMA CARTERI, n.sp ; FlG. 2, EclONEMA LAVINIENSIS, n.sp. ; FlG. 3, 3 \, SB.GeODIA PERTJNCINATA, n.sp. :

. 4, Tetilla poculifera, n.sp.; Fig. 5, Tetiixa anomai.a, n.sp.; Fig. 6, Tetilla umicola, n.sp.;

Fig. 7, Paratetilla cineriformis, n.sp

:ylon pearl oyster report

SPONGES PLATE IV.

FIG. I.

.,-.^V

'

ice

%

■:

•■ .

y

- : ,

I

A .jen&y del.

E Wlson.Lith Cambridge

Fig. 1, Craniella elegans, n.sp. ; - Fig. 2, Taprobane herdmani, n.sp. ; Fig. 3, Aciculites
oriejjtalis, n.sp. ; Fig. 4, Discoderiiia emarginata, n.sp. ; Fig. 5, Petromica massalis, n.sp.

< EYLON PEARL OYSTER KKI'OKT

SPONGES I'LATE V

FIG. 2

FIG I.

*

; . \~"

Fig. 4.

Fig. 5.

CO

o

FIG. 10

fig 6

Fig. 1, Skeba Extensa, n.sp. ; Fig. i. Coppatias reptans, n.sp.; Fig. 3, Asteropus haeckei.i, n.sp. ;

Figs. 4, 5, Cryptotethya agglutinans, n.sp.; Fig. 6, Hyhedesmia curvistellifera, n.sp.; Fig. 7, Spirastrella

TENTORIOIDES, 11..-]..; FlG. 8, XEGOMBO TENUISTELLATA, Il.sp.: FlG. 9, CLIONA MARGARITIFER.E, ll.sp.;

Fig. 10, Soberites cruciatus, n.sp.

CEYLON PKAKI, OYSTER KKPOKT

SPONGES PLATE VI

Tj-

£

a

S3
O
-
m
O
S3
H

CEYLON PKAHI. OYSTER k'KPoKT

SPONGES PLATE VII.

V

V *

V

* ^ ■

Afll?

t'CS

*%;-

- •

Fig. 1, Siphoxochalina communis, var. tenuispiculata, n. ; Fig. 2, Ceraochalina multiformis,

var. manaarensis, Dendy; Fig. 3, Ceraochalina ceylonica, n.sp.; Fig. 4, Pachychalina spinilamella,
Dendy; Fig. 5, Gelliodes carnosa, Dendy; Fig. 6, Collocaltpta digitata, d

CEYLON PEARL OYSTER REPORT.

SPONGES PLATE VIU.

J

i*

<0«$rt

!6*f

mm

Fig. 1, Plocamia manaarensis (Carter); Fig. 2, Clathkia spiculosa, var. tessellata, 11.; Fig. 3, Phakelt.ia
ceyloxensis, n.sp. ; Fig. 4, Acarndb TEiiXATCs, Ridley; Fig. 5, Raspaima FRUTICOSA, var. tenuiramosa, n. :

Fig. 6, Acanthella carteri, Dendy.

CEYLON PEARL OYSTER REPORT

SPONGES PLATE IX

N

1 a.

r

■j

10.

n

'•y del

- ambridge

Fig. 3, Gelliodes

Fig. 1, Stroxgvlophora durissima, n.sp.; Fig. 2, Toxochalina robusta, var. ridleyi, n.:
petrosioides, n.sp. ; Fig. 4, Desmacella tubulata, n.sp. ; Fig. 5, Acaxthoxifer ceyloxensis, n.sp. ;

Fig. 6, Gelliodes incrustans, n.sp. ; Fig. 7, Gellius axgilatcs, var. canaliculate, n. ; Fig 8. Renieba
zoologica, n.sp.; Fig. 9. Petrosia densissima, n.sp.; Fig. 10, Renieba pigmentifera n.sp

(F.YLON PEARL OYSTER REPORT

SPONGES PLATE X.

.excFig.5

Figs. 1, 2. Pachychai ixa subcylindrica, Q.sp.; Fn;. 3, Chalina clathrata, n.sp.; Fig. 4, Ceraochalina

RETIARMATA, n.Sp. J FlG. 5, CHALINA 3DBABMIGERA (Ridley); Flu. (i, CERAOCHALINA CEYLOXICA, n.sp. J

Fig. 7, Pachtchalina brevispiculifera, n.sp.; Fig. 8, Ceraochalina ketkttis, n.sp.; Fig. 9, Chalina

OBTUbISPK'i LIFERA. n.sp.: Flu. 10, TrACHYOPSIS HALICHOXDRIOIDES, n.sp.

CEYLON PEARL OYSTER REPORT

SPONGES. PLATE XL

0

FIG.

FIG5

0

Fig ?

I

r

FlG.6.

r\

\J

Fig 7.

\J

FlG-4.

FIG. 9.

FIG 8.

Fig. 1, Paresperella bidhxtata, n.ap.; Fig. i, Parespekei.i.a serratohamata (Cart.)i Fig. 3, Parespkrella. sp.
Fig. 4, EcEHNODicmroM ci.athh.vtlm, n.sp.; Fig. 5, Myxilla tenuisbima, n.sp.j Fig. 6, Esperella
crassissima, nsp.; Fig. 7, Raspailia hornelli, n.sp.; Figs. 8, 9, Histoderma vesiculatdm, n.sp.

CEYLON PKAKI. OYSTER KKI'ORT

SPONGES PLATE XII.

b. 0.

Fig. 2.

Fig.1.

Fig. 5.

Fig. 6.

Fig. 10.

Fig. 1, Spongosorites topsenti, n.ap. ; Fig. -i, Spongosorites (?) lamellata, n.sp. ; Fig. 3, Spongosorites (?)

lapidifohmis, n.sp.; Fig. 4, Hymeniacidon petrosioides, n.sp. ; Fig. 5, Thrinacophora durissima, n.sp. ;

Fig. 0, Thrinacophora agariciformis, n.sp.; Fig. 7, Axinella halicuondrioides, n.sp.; Fig. 8, Axinella

maxus, n.sp.; Fig. '■>, Agelas cejlonica, n.sp.; Fig. 10, Rhabdereiha indica, n.sp.

CEYLON PEARL OYSTER REPORT

SPONGES PLATE XIII

Figs. I, 2, I ollocalypta DioiTATA, n.sp.; Fig. 3, Phakellia symmetrica, n.sp.: Fig. 4, Axinella

tenuidigitata, n.sp. ; Fig. 5, Phakellia ceylonensis, n.sp. ; Fig. 6, Phakellia crassistylifera, n.sp. ;

Fig, 7. Auletta elongata, n.sp, ; Fig. 8, Leocosolenia coriacea, var. ceylonensis, n. ; Fig. 9, Leccosolenia

TENDIPIL08A, n.sp.; FlG, 10, LeDCANDRA DONNANI, n.sp.

(KYI, ON PEARL OYSTER REPORT

-V -

SPONGES PLATE XIV.

■

v

«>'i*

■"w

■J

*& ^

:ste&*^

Fki. I, Hirctnia fusca, Carter; Kirs. 2, Hippospongia clathrata (Carter); Fig. 3, Euspongia OFFICINALIS,

var. eeylonensis, n. : Fig. 4. Sponoelia elastica, var. crassa, n.; Fig. 5, Hircinia anomala, n.ap.;

Fir,. 6, Phyllosfongia papyhacea. var. j hir,. 7. Megalopastas mora (Dendy).

CEYLON PKANI. OYSTER UK PORT

SPONGES PI-ATE XV

em

4 . : Cambridge

Figs. 1, 2, Darwi.nella simplex, Tops.; Fig. 3, Megalopastas pulvillus, n.sp.j

Fig. 4, Megalopastas elegans (Lend.); Figs. 5—8, Megalopastas nigra (Dendy).

EYLttN' PKANI. OYSTER REPORT

SPONGES PLATE XVI

1, tflRCINIA A.NOMAI.A, U.8p. j FlG. 2, HlRCI.NIA TLBEROSA, D.sp. ; FlG. 3, HlRCINIA 8CHULZEI,

4, Aplvsixa herdmaxi, n.sp. ; Fig. 5, EusPONGIA officinalis, var. oeylonensis, n.

asp. ;

[CEYLON PEARL OYSTER FISHERIES— 1905— SUPPLEMENTARY REPORTS, No, XIX.]

REPORT

ON SOME

ALCYONIID.E

COLLECTED BY

Professor HERDMAN, at CEYLON, in L'JOU.

BY
EDITH M. PRATT, D.Sc.

[With THREE PLATES.]

Tins collection of Ceylon Alcyoniidte* is remarkably similar to one made by
Mr. Stanley Gardiner some three or four years ago in the Maldive Islands
(Pratt, 1903). t The resemblance in many cases is so pronounced as to indicate a
very intimate relationship between the Alcyoniid fauna of these islands. The
genera Sarcophytum, Lobophytum, Sclerophytum, and Aleyonium are common to both
localities. These genera are represented in the present collection by 17 species, and
of these 9 occur also in the Maldive collection, 5 are new, and 3 have only been
recorded hitherto from the Pacific.
The Ceylon collection includes : —

I. Five species of Sarcophytum : of these, three are new ; one, & ehrenbergi, has
been recorded from the Maldives and Red Sea, and one, S. plicatum, has been
recorded only from the Pacific (Ternate and North Celebes).

II. Two species of Lobophytum ; one, L. paucifiorum, appears to be widely
distributed in the tropical and southern sub-tropical waters of the Pacific and
Indian Oceans, having been recorded from Tonga, Moluccas, Funafuti, New Britain

* See also the Report on the other Alcyonaria by Professor J. Arthur Thomson, which follows in this
volume. — AY. A. H.

t See List of Literature on p. 265.

:MH

CEYLON FEARL OYSTER REPORT.

and New Zealand in the Pacific, from the Red Sea, and from the Maldive Islands and
Gulf of Mauaar in the Indian Ocean. The second species, L. hedleyi, has only been
previously recorded from the Pacific (Funafuti). Its record from the Gulf of Manaar
indicates a wide distribution of this hitherto little-known species.

III. Eight species of Sclerophytum : of these three, S. marcnzellem, S. polydactylum,
and S. densum, are comparatively well-known species. S. marenzelleri, however, has
only been before recorded from the Pacific (New Britain and New Hebrides) ;
S. polydactylum from the Pacific (British New Guinea, China Seas), the Red Sea,
Indian Ocean (Maldives), and now the Gulf of Manaar. S. densum is the most
common species of the genus, and has a wide geographical range — it is now known
from the Pacific (Funafuti, China Seas, British New Guinea), the Indian Ocean
(Maldive Islands), and now the Gulf of Manaar.

Four species : S. gardineri, S. palmatum, S. querciforme, and S. durum, have been
recorded only from the Maldives. One species is new.

IV. Two species of Alcyonium, one of which is new, and the other, A. pachyclados,
is apparently generally distributed in the warm southern and tropical waters of the
Pacific and Indian Oceans.

Several specimens of species common to Ceylon and the Maldive Islands vary
considerably in size, external form, colour, consistency, size of spicules, &c, so that it
was sometimes impossible to determine the identity of a species from a study of
superficial characters. When examining a specimen of this family in the preserved
condition, it is necessary to note that (1) the size of a specimen may vary according
to age or environment ; (2) external form is, to a certain extent, dependent on the
state of expansion or otherwise of the colony at the time of fixing, and also upon the
nature of the fixing re-agent ; (3) the colouring matter in Alcyonaria, apart from that
of the spicules, is usually acted upon, and in many cases dissolved, by alcohol ; (4) the
consistency of a colony, and to a certain extent the size of the spicules, may be
affected by the use of an acid preservative. When these more or less artificial causes
of difference are eliminated, the identification of most species, when based upon
general anatomical principles, can usually be determined with considerable ease and
accuracy.

List of Species occurring in Professor Herdman's Collection.

Sarcophylvm bicohr, n. sp.

,, oligolrema, n. sp.

„ contorlum, n. sp.

„ plicatu/m, Schenk.

elirenbergi, Marenzeller.
Lobophytwm hedleyi, Whitelegge.

, pauciflorum, Eiirenberg.
Sclerophytum Jierdmani, n. sp.

„ marenzelleri, Wright and St.

,sw< rophytum polydactylum, Dana.

„ gardineri, Pratt.

„ palmatum, Pratt.

„ densum, Whitelegge.

„ querciforme, Pratt.

durum, Pratt.
Alcyonium ceylonicum, n. sp.

„ pachyclados, Klunzinger.

ALCYONIID^E. 249

Sarcophytum bicolor, n. sp. — Plate I., figs. I and 2.

Four complete and several fragmentary specimens were taken from the Gulf of
Manaar and from the coral-reef at Galle. The largest specimen is incomplete, but is
250 millims. high and 190 millims. broad, and is provided with huge marginal lobes.

One complete specimen is irregularly mushroom-shaped, with four lobes, and is
attached basally to hard coral. It is 120 millims. high and the capitulum measures
150 millims. by 85 millims. across its surface. The stalk is 65 millims. high and has
a transverse diameter of 96 millims. by 60 millims. The capitulum has a thick
marginal seam.

The second complete specimen is very much smaller. It is oidy 50 millims. high,
and the capitulum measures 47 millims. by 29 millims. across its surface (Plate I.,
fig. 1).

The stalk varies in length in individual specimens ; its surface is usually smooth,
but here and there are occasional longitudinal wrinkles. In many cases the
autozooids may be seen through the transparent body-wall.

The colour (in spirit) varies from a dark slaty grey to a pale brownish-grey. In all
specimens the autozooids and siphonozooids are of a beautiful creamy white colour.

Some of the specimens are moderately hard and tough, others are soft and fleshy,
while a few are quite brittle. The texture of the species seems to vary with the age
of the colony, locality, and state of preservation.

The small complete specimen differs from the larger specimen in the feeble
development of the marginal seam, which is very thick in the older colony.

The Autozooids are extremely numerous and are usually uniformly distributed over
the upper surface of the capitulum. In the largest colony they are most numerous on
the upper surface of the marginal seam, where they average 12 to the centimetre,
and are few in number or entirely absent on the extreme edge of the seam, where
siphonozooids are very abundant. None of the specimens are completely expanded,
but anthocodia of partially expanded z'ooids are fully 2 millims. in length and
0"5 millim. in diameter.

The tentacles are all contracted, and, although very small, are similar to those
of other species. The stomodseum, mesenteries and mesenterial filaments do not
apparently differ in any important respect from those of other species, but the
specimens are not sufficiently well preserved to permit of histological examination.

The Siphonozooids (in spirit), of a beautiful cream colour, are outlined by a network
of brown pigment, which also defines the outlines of the canals.

The average superficial diameter of a siphonozooid is 0'3 millim. The stomodamm
is well marked and has an average length of 0-28 millim. The mesenteries are similar
to those occurring in other species.

The Canal System is well marked, but exhibits no specially interesting features,
with the exceptiou of the presence in the canals of an orange-brown coloured pigment

2 K

250 CEYLON PEARL OYSTER REPORT.

which may owe its existence to the disintegration of algae cells, so abundant in many-
tropical corals (see below).

The Spicules are long and slender, and in form are characteristic of the genus
(fig. 2). They are smaller in the actively growing superficial tissues than elsewhere.
When an autozooid is in a state of contraction, the external aperture is effectually
closed by a small conical cap of minute tentacular spicules.

This species exhibits considerable variation in form and consistency, but is charac-
terised externally by the distinctive creamy white autozooids and siphonozooids and
the greyness of the remaining portions of the colony.

Sarcophytum oligotrema, n. sp. — Plate I., figs. 3 to 5.

A complete cup-shaped colony attached basally to stony coral was taken in the
Gulf of Manaar (fig. 3).

The specimen is almost stalkless and is 42 millims. high. The capitular cup,
irregular in outline, is 32 millims. in depth and is G2 millims. by 50 millims. in
breadth. The base of the cup is 10 millims. in thickness, the sides are about
5 millims. in diameter near the base, but thin out towards the edge, so that they are
wedge-shaped in vertical section. Zooids are only present on the inner surface of the
cup. The colour in spirit is of a dirty brownish-grey, with autozooids of a more
pronounced brownish hue. The outer surface of the cup is moderately smooth, with
few wrinkles. The specimen is moderately hard and tough.

Autozooids are very small, and are relatively very numerous on the sides and edges
of the cup, but are somewhat scanty in the middle. As the colony is in a state of
extreme contraction, it is impossible to determine the actual size of the autozooids ;
the young autozooids near the edge of the cup are smaller than elsewhere. The
average diameter of a mature autozooid at the surface is about 0'3 millim.

The tentacles, stomodseum, mesenteries and mesenterial filaments do not differ in
any essential respect from those of other species. Reproductive organs were not
observed.

Siphonozooids are remarkably few in number (fig. 4). Usually only one or two
of these zooids occur between two autozooids, but in the middle of the capitulum
the number is slightly increased. The average superficial diameter of a siphonozooid
is 0-35 millim., the mouth is sunk in a slight depression, and the stomodamm has an
average length of 0'25 millim.

The canal system presents no specially interesting features. Transverse ciliated
vessels connecting the ccelentera of neighbouring zooids (Pratt, 1903, p. 507, fig. 3)
are very numerous.

The spicules, in the form of irregular warted spindles, are extremely small, hut are
characteristic of the genus. They are from 0*09 millim. to 0'3 millim. in length and
from 0'0'J millim. to 0'1 millim. in diameter (fig. 5).

This species has a superficial resemblance to S. roseum, from which however it

ALCYONIIDiE. 251

differs chiefly in colour, form of spicules, in its abundance of autozooids and paucity
of siphonozooids.

Sarcophyturn contortum, n. sp. — Plate I., figs. G and 7.

Two complete specimens and one fragment were taken from the reef at Galle.

The largest specimen is 47 millims. high. The capitulum has a curiously lobed
appearance, somewhat resembling a Lobophytum. This is due to the growth inwards
of 8 more or less digitate marginal lobes (fig. 6, e.m.l.). All the lobes arise from the
capitular margin and not from the upper surface of the capitulum as in Lobophytum.
The convoluted character is not so strongly marked in the two other specimens of the
species. The capitulum measures 47 millims. by 35 millims. across its surface. The
lobes are somewhat flattened and rise slightly above the level of the true capitular
margin. They are from 18 millims. to 30 millims. long and are from 9 millims.
to 15 millims. hroad at the base. The edges of the lobes curl upwards and slightly
inwards (fig. 6, c.m.L). The stalk is 40 millims. high, tapers towards the base and is
drawn into furrows by the inward growth of the marginal lobes. Its surface is
minutely granular and is marked by few longitudinal wrinkles.

The colony is in a state of extreme contraction, which may, to a certain extent,
account for the complication of the marginal folds. All the specimens (in spirit) are
of a pale fawn colour. They are hard to the touch, minutely granular, compact,
and are readily cut with a knife.

The Autozooids are very small, and as they are withdrawn some distance below the
surface, it is almost impossihle to perceive their external apertures without the aid
of a lens. The average surface diameter of an autozooid cavity is 0"25 millim.
Autozooids are apparently more numerous in the margin of the capitulum and
capitular lobes than elsewhere.

The tentacles, stomodaeum and mesenteries with mesoglceal thickenings are
similar to those of other species. Ova in several stages of development are present
on many of the mesenteries, development proceeding from below upwards, as is usual
in these forms. In one instance two series, one a little distance below the other,
of developing ova were observed on one mesentary.

The Slphonozotiida cannot be seen without the aid of a lens. They are fairly
numerous, two or three may usually be counted between two autozooids. The
average surface diameter of a siphonozooid is 0"3 millim. The stomodaeum has an
average length of 0'18 millim.

The canal system does not exhibit any specially noteworthy features.

The Spicules are of the usual Sarcophyturn type, and are extremely numerous.
They are from 0'3 millim. to 0-4 millim. in length, and from O'l millim. to 0'15 millim.
in breadth. Occasionallv a club with a pointed handle may be seen (fig. 7b).

This species is characterised by the curiously convoluted marginal lobes, the small
size of the zooids, the hardness and compact character of the colony.

2 K 2

252 CEYLON PEARL OYSTER REPORT.

Sarcophytum plicatum, Schenk (189G).

Two moderately large colonies of a pinkish-brown colour were taken from
Station XXII. , 13 fathoms, in Dutch Bay, near Trincomalee.

This species has previously only been recorded from Ternate, hut in the zoological
laboratories of the Victoria University are two specimens, one taken by Professor
Hickson in North Celebes and the other taken by Mr. J. Stanley Gardiner at
Funafuti (Ellice Islands). The latter has been identified by Miss Hiles. This species
therefore appears to have a very wide distribution within the tropical belt.

Sarcophytum ehrenbergi, Marenzeller (1886).

One large and complete specimen was taken on the reef at Galle.

The colony is 50 millims. high, and the capitulum measures 245 millims. by
215 millims. across its surface. The capitular margin is thrown into deep folds.

The specimen is of a dark grey colour, but towards the middle of the capitulum it
assumes a lighter shade.

The species has previously been recorded from the Red Sea and from the Maldive
Islands (see Pratt, 1903).

Lobophytuin hedleyi, Whitelegge (1897).

A single complete specimen and a fragment were taken from the reef at Galle ;
these agree with specimens from Funafuti in all essential respects, and only differ
slightly in external form. The complete specimen is short and squat, and is
75 millims. high. The capitulum measures 136 millims. by 115 millims. across its
surface. It has a fairly broad marginal seam, and is deeply concave in the middle ; the
lobes are arranged in fairly regular radiating lines from the middle of the capitulum,
and vary in height from 36 millims. to 53 millims. The colony is hard and smooth,
the stalk is here and there marked by wrinkles. The specimens in spirit are of a
dirty greenish-grey colour with brownish patches. The autozooids are contracted,
but in sections are seen to be of a paler shade of grey than the prevailing tint.
They are relatively few in number, but are more abundant on the summits of the
lobes, where they are from 1 millim. to 2 millims. apart, than in the furrows between
the lobes, where they are separated from each other by wide intervals.

The average diameter of a contracted crown of tentacles is about 1 millim. Owing
to their extreme contraction it is impossible to distinguish their pinnules. The
stomodseum appears to be longer than in the species L. paudfiorum or L. crassum,
but the walls are folded in such a manner that it is impossible to measure the
actual length. The mesenteries and mesenterial filaments do not apparently differ
from those of other species. Ova, exhibiting several stages of development, are
present on the mesenteries of the autozooids. They are from 0"08 millim. to
0'2 millim. in diameter, and each is enveloped in a follicle which, in the largest ova,
is 0"03 millim. in thickness.

ALCYONIID^. 25

The Siphonozooids are arranged with great regularity in the spaces between the
autozooids, and are consequently most numerous where autozooids are fewest.
Whitblegge describes the siphonozooids as being numerous and distinct, and states
that there may be as many as 12 siphonozooids between two autozooids. On the
side of one of the lobes 43 siphonozooids were counted in a straight line between
two autozooids, but the distance between the autozooids is not so great on the
summits of the lobes, where they are seldom more than 3 siphonozooids apart.

The siphonozooids are extremely well marked in this species, their average surface
diameter of 0"27 millim. is slightly less than in the species L. pauciflorum and
L. crassum, but their length is greater than in those species. The average length ol
the stomodamm is 0'23 millim.

The canal system does not differ in any essential respect from that of other
species ; transverse ciliated vessels are very numerous, especially near the surface.

Spicules. — The spicules of this species have been described and figured by White-
legge (1897, p. 217). Those of the cortex are tiny spindles or clubs from 042 millim.
to 0'15 millim. in length, and from 0"02 millim. to 0'05 millim. in breadth. The
more deeply seated spicules are tuberculate spindles, and short, stout, sub-cylindrical
spicules. These are from 0"15 millim. to 0'3 millim. in length, and from 0-03 millim.
to 0"19 millim. in breadth.

Lobophytuin pauciflorum, Ehrenberg (1834).

One fairly large fragment and two young mushroom-shaped colonies were taken
from the Gulf of Manaar.

One of the latter was 10 millims. high, with a capitulum 12 millims. by 10 millims.
in diameter, and the other was 12 millims. high, with a capitulum 12 millims. by
10 "5 millims. in diameter.

This species (see Marenzeller, 1886, and Pratt, 1903) appears to be very widely
distributed and has been taken from the Eed Sea, New Zealand, Andaman Islands,
Tonga, Moluccas, Funafuti, New Britain and the Maldive Islands.

Sclerophytum herdmani, n. sp. — Plate II., figs. 8 and 9.

One complete and several fragmentary specimens were taken from the reef at
Galle, and two specimens from the Pearl Banks off Aripu.

The complete specimen from Galle has a much lobed capitulum, the lobes are similar
in size, and are regularly arranged on the capitular surface (fig. 8). The colony is
85 millims. high and the capitulum measures 135 millims. by 140 millims. across its
surface. The stalk is 70 millims. high and 120 millims. broad. The capitulum has
no marginal seam, but passes almost imperceptibly into the stalk. On the wrinkled
stalk, some little distance below the capitulum, are three small tubercles which bear
zooids on their upper surface (fig. 8, tu). This appears to be the only recorded instance
of the occurrence of zooid apertures on the stalk. The colour in spirit is pale drab,

254 CEYLON PEARL OYSTER REPORT.

the summits of the branches have a yellowish tinge. When living, the zooids were
of a pale brown colour, which, however, has been dissolved by the spirit preservative.

A specimen from Aripn is smaller. It is 50 millims. high and measures 45 millims.
by 35 millims. across the surface of the capitnlum, which is seamed at the margin.
In spirit the stalk is light brown and the capitulum yellowish-drab. When living,
the capitulum was a pale yellowish-green colour.

The fragmentary specimens are of a uniform drab colour. All the specimens are
moderately hard, yield slightly to the touch, and are somewhat brittle.

The species is monomorphic. The autozooids appear to be rudimentary. They are
extremely small, numerous, and are arranged with great regularity on the capitular
surface. The average surface diameter of a contracted autozooid is 0-04 millim.
Frequently as many as 20 zooids may be counted in a linear centimetre. It is
unfortunate that the preservation of the specimens will not permit of a more complete
investigation of the anatomy of this interesting species.

The ventral mesenterial filaments are extremely small, but the dorsal filaments are
as well developed as in other species. The superficial and internal canal systems do
not differ in any essential i-espect from those of other species. The endodermal tissues
and canals are crowded with zoochlorellse. The minuteness of the ventral mesenterial
filaments and the smallness of the autozooids is doubtless correlated with the extreme
abundance of zoochlorellae, and is discussed in a paper now in the press (Pratt,
1905).

The Spicules (fig. 9) are similar in character to those of other species. Tuberculate
spicules usually in the form of spindles, but sometimes forked, are comparatively
numerous, and average 3 millims. in length and 0"5 millim. in breadth. Clubs some-
times with spiny handles, and spindles set with minute spines, are closely packed near
the surface, so as to form an outer crust. The average size of the clubs is 0"09 millim.
by 0-04 millim., of the spindles 0'25 millim. by 0-03 millim. A few spiny crosses
also occur.

Sclerophytum marenzelleri, Wright and Studer (1889).

A single specimen of this species was taken in the Gulf of Manaar.

The colony forms an extremely hard cake-like mass, with lobes of varying size
arising vertically from the upper surface of the capitulum. The basal attachment
consists of hard coral encrusted with Polyzoa and shells. The colony is 93 millims.
high ; the capitulum has a transverse diameter of 337 millims. by 257 millims ; the
stalk is 55 millims. in height and is irregular in outline. The capitulum has a well-
marked marginal seam and bears numerous conical lobes — many being fully 20 millims.
high. A few of the lobes are branched.

The specimen in spirit is of a pale pinkish-brown colour, which originally was
doubtless of a more intense hue, as the spirit is of a golden-brown colour. The
capitulum is tough in consistency and slightly yielding to pressure ; that of the stalk

ALCYONIID.K. 255

hard, granular and more brittle. Towards the base the stalk is extremely hard
owing to the great abundance of spicules in this region.

The spicules are relatively small compared with other species. They are adequately
described by WEIGHT and STUDER (1889, p. 251).

The species is monomorphic. The autozooids are all contracted, but appear to be
extremely small. They are much more numerous in the lobes than in the furrows
between, where they are often from 2 millims. to 4 millims. apart. They are also
Very scantily distributed on the capitular margin. I )wing to the extreme contraction
of the autozooids, the pinnate character of the tentacles cannot be distinguished.
The stomodseum is comparatively long and has a well-marked siphonoglyph. As in
other species, the dorsal mesenteries are of considerable size. There are of course
no siphonozooids.

This species has been previously recorded from Api Island, New Hebrides
(60 fathoms to 70 fathoms) and Cape Gazelle, New Britain.

Sclerophytum polydactyluin (Dana).

One complete specimen and several fragments were obtained from the Gulf
of Manaar.

The complete colony is 35 millims. high, the capitulum measures 37 millims.
by 22 millims. across its surface. It is smaller, slightly harder and more brittle than
the Maldive specimen, but does not appear to differ from it in any essential respect.
This species (see Pratt, 1903) has been recorded from the Red Sea, China Strait,
British New Guinea, the Maldive Islands and now the Gulf of Manaar.

Sclerophytum gardineri, Pratt (1903).

Three complete specimens were obtained from the Gulf of Manaar.

Two of the specimens (in spirit) are of a beautiful cream colour, while the third is
of a pale drab colour. These specimens agree with the type specimen from the
Maldive Islands in all essential respects. Two are slightly larger, while the third is
of similar size.

This species lias been hitherto recorded only from the Maldive Islands.

Sclerophytum palmatum, Pratt (1903).

Two complete and several fragmentary specimens were taken from the reef at Galle
and from the Gulf of Manaar.

The complete specimens .ire small, and from their manner of growth do not appear
to he so robust as the typo specimen from the Maldive Islands. The fragmentary
specimens are from shorter and stouter colonies, and have broader and larger lobes.
The zooids are very similar in all the specimens, the tentacles having the double row
of pinnules characteristic of the species (Pratt, 1903, p. 526). None of the Ceylon

256 CEYLON PEARL OYSTER REPORT.

specimens in spirit show the vivid green colour of the type specimen when living, but
several of the specimens have a greenish tinge.*

This species has hitherto only been recorded from the Maldive Islands.

Sclerophytum densuin, Whitelegge (1897).

One very young mushroom-shaped colony (Pratt, 1903, tig. 18) was taken from
Gulf of Manaar ; two specimens, brown in colour, were taken from the reef at Galle ;
one reddish-brown specimen from a coral bank in the Gulf of Manaar, and a light
brown and a fawn specimen were taken from the reef at Galle.

The light brown specimen is complete, and is comparatively young, having not yet
outgrown its early mushroom shape. It is 22 millims. high, and the capitulum
measures 46 millims. by 27 millims. across its surface.

The capitulum is concave, with a central depression of 12 millims. Small lobes
from 2 to 4 millims. high arise from, the middle of the cup, while larger ones from
5 to 6 millims. high occur near the edge.

The fawn-coloured specimen was taken on the reef at Galle. It is 70 millims. high,
and the capitulum measures 240 millims. by 200 millims. across its surface. The
colony forms a large, hard, cake-like mass, with a more or less vertical short and thick
stalk, having almost the same breadth as the capitulum, which has no marginal seam.
Lobes are very numerous near the edge of the capitulum, but are few and small in
the middle. Zooids are numerous on the lobes, but are scantily distributed over the
centre of the capitulum.

This species appears to be the most common as well as the most widely distributed
species of the genus. It is known from Funafuti, China Seas, British New Guinea,
Maldive Islands, and now from the Gulf of Manaar and Galle.

Sclerophytum querciforme, Pratt (1903).
Two specimens similar in form, size, texture, and apparently in all essential respects,
to the specimens from the Maldive Islands, were taken from the Gulf of Manaar.
This species has hitherto only been recorded from the Maldive Islands.

Sclerophytum durum, Pratt (1903).

Four specimens were taken. Three from the Reef at Galle and one from the pearl
banks in the Gulf of Manaar.

Three specimens, one complete and two fragmentary, are of a beautiful orange
colour (in spirit) — of a deeper shade towards the middle and paling towards the
margin. The fourth, a cup-shaped colony, is of a deep cream colour.

As already pointed out (Pratt, 1903, p. 528), specimens of this species from the
limited area of the Maldive Islands exhibit great diversity in form and colour.
Further modifications in these respects occur in the Ceylon specimens. The cup-
shaped colony is similar in form, size, and texture to the cup-shaped Maldive

* Some of these colonies from Ceylon were certainly green when alive.— W. A. H.

ALCYONTTDiE. 257

specimens. The remaining three specimens, which appear to be very much older, are
extremely hard, irregular, mushroom-shaped masses, with convex capitula thinning
out towards the edges, set with short frequently laterally compressed lobes, usually
arranged in radiating ridges.

© DO

The lolied specimen from the Maldive Islands appears to be intermediate in form
and age between the cup-shaped specimens and the hard, irregular, mushroom-shaped
specimens from Ceylon.

The complete specimen is 47 millims. high and the capitulum measures 92 milium
by 80 millims. across its surface. The stalk is short, broad, laterally compressed and
constricted in the middle. It is 25 millims. high and has a diameter at the base of
80 millims by 40 millims. A few of the lobes are branched. Their average height
near the middle is 10 millims. All the specimens give off a peculiar pungent odour
even in spirit.

Only autozooids are present in this species. They are fairly numerous on the
margin of the colony and on the lobes, but are few in number in the furrows between
the lobes. All the specimens are hard and granular, owing to the enormous spicules
which are very thickly set and extremely abundant.

This species has hitherto only been recorded from the Maldive Islands.

Alcyoniuin ceylonicum, n. sp. — Plate II., figs. 10 to 12.

A single complete colony was taken from the reef at Galle.

The specimen has an irregularly ridged capitulum and is creamy white in colour.
The colony is 60 millims. high and the capitulum measures 75 millims. by 55 millims.
across its surface.

It is tough and fleshy in consistency, slightly softer and more fleshy in the middle
than near the surface, where the spicules are much more abundant.

The Zooids are extremely small and are very numerous (13 or 14 to the centimetre),
they are uniformly distributed over the entire upper surface of the capitulum.
Measurements across the partially expanded crown of tentacles average 0'6 millim.
None of the zooids are completely expanded. The apertures of contracted zooids are
so minute as to be almost imperceptible without the aid of a lens. As the tentacles
are only partially expanded, it is difficult to distinguish their pinnate character.
They are extremely short, the longest measuring only 0"16 millim. in length. The
stomodasum averages 0'5 millim. in length and has a moderately large siphonoglyph.

The mesenteries are small, and the musculature only feebly developed. As in
many other tropical forms, the dorsal mesenteries have extremely long and well
developed filaments, but the ventral mesenteries have no filaments (Plate II., fig. 11,
r.ni.). As in several other instances (Pratt, 1904, in the press), this reduction of the
digestive surface is accompanied by an abundance of zoochlorelke in the superficial
endodermal tissues.

The Spicules are similar in form (fig. 12) to those of A. pachyclados, but are

2 L

258 CEYLON PEARL OYSTER REPORT.

slightly larger than in that species and are much less numerous. They have the
form of tuberculate dumb-bells. They are fairly abundant in the superficial tissues, but
are scantily distributed in the more deeply seated portions of the colony.

The average size of the spicules is (VI 4 millim. by 0-l millim.

As in other species of Alcyonium, there is only one system of canals. This, the
internal system, is extremely well developed in the interior of the colony, but becomes
less well marked towards the surface of the capitulum. This system consists chiefly
of short transverse vessels which connect the coalentera of the zooids (Pratt, 1903).
Owing to the great number of zooids, which in this species penetrate deeply into the
interior, the longitudinal canals are comparatively few in number.

This species is characterised by its soft fleshy consistency, its numerous but very
minute zooids, the entire absence of ventral mesenterial filaments, accompanied by an
abundance of zoochlorelke in superficial tissues, and the small spicules.

It approaches A. pachycladoa in the absence of ventral mesenterial filaments and
in the form of the spicules. It differs however from that species in the texture of the
colony, the minuteness of the zooids and in its less numerous but larger spicules.

Alcyonimn pachyclados, Kltjnzinger (1877).

Three complete specimens of this widely distributed species were taken from the
Gulf of Manaar.

They are all similar in external form, consistency and colour, but specimens in the
Victoria University of Manchester from different localities exhibit considerable variation
in these respects. The Ceylon colonies form low, more or less lobate, sessile, encrusting
masses, from 10 millims. to 20 millims. high and from 28 millims by 18 millims. and
42 millims. by 23 millims. broad. The capitulum has a more or less convex surface,
on which are studded, with varying frequency, the short and broad, rounded or
conical lobes ; the largest lobe is 7 millims. high and has a basal diameter of
8 millims. by 7 millims. These colonies are very similar in external form and
character to specimens from the Maldive Islands (Pratt, 1903), China Strait and
New Britain, but differ considerably in colour and manner of growth from the
arborescent specimens which have been attributed to this species from the Cape of
Good Hope (Hickson, 1900). The Ceylon specimens are harder and more brittle
than the forms I have examined from other localities, this being no doubt attributable
to the great abundance of spicules.

The Ceylon specimens (in spirit) are of a creamy-white colour with autozooids of a
darker shade, and are similar in this respect to the specimens from China Strait.
The Maldive specimen (in spirit) is a pale brownish-grey colour, of a darker shade
between the lobes. The specimen from New Britain is a uniform grey colour, while
some of the Cape specimens are white, some yellow and others red in colour.

In distribution this species is known from the Red Sea, Maldive Islands, China
Strait, New Britain, Cape of Good Hope and now the Gulf of Manaar.

ALCYONIID^E. 25!)

ANATOMICAL NOTES.*
Nematocysts (tigs. 13 and 14).

Moseley was unable to find nematocysts in specimens of Sarcophytum, and there-
fore states (1881, p. L19) that they do not occur in this genus. In his description
of the occurrence of nematocysts in several members of the Alcyonaria, Ash worth
(1899, p. L93) records them for Sarcophytum pauciflorum. This species, however,
is now placed in the genus Lobophytum, for which the record of the presence of
nematocysts is new. Hitherto they have not heen observed in any species of
Sarcophytum.

On examining several preparations of a single species stained with iron hrazilin,
some specimens seemed to contain no nematocysts, many contained few nematocysts,
and in only a very few instances were they at all numerous. The apparent absence
or scarcity of nematocysts in individual memhers of a species may be attributed to
the use of a fixing preservative which permeates the tissues so slowly that the
stinging cells are discharged before the colony is fixed. If a rapidly fixing preserva-
tive such as an 8 per cent, aqueous solution of hot formalin he employed, the tissues
apparently suffer no contraction and the tentacles are found to contain innumerable
batteries of nematocysts.

Nematocysts in Alcyonium digitatum were first described by Htckson (1895).

I have been successful in observing nematocysts in all the species of Alcyonium,
Sarcophytum, Lobophytum, and Sclerophytum which I have examined. They are,
however, apparently more numerous in the British species than in the tropical
representatives of the family. They always occur in batteries, and are never
uniformly distributed in the ectoderm. The most common type is the well-known
form occurring in Alcyonium. Modifications of this with regard to size, and length
and thickness of thread are not infrequent, and may be found in Sarcophytum roseum
(fig. 13«) and S. glaucum (fig. 13c), and in Sclerophytum gardineri-. The accompanying-
table shows that they vary considerably in size in specific members of the family, and
are apparently more uniform in this respect in Alcyonium and Lobophytum than in
Sarcophytum, where they are from 6/jl to '1'Ijj. long, and from 2/x to 14/a broad, and
in Sclerophytum, in which they are from 5/x to 12/jl long, and from 2(i to 4ya broad.

Nematocysts attain their greatest size in Sarcophytum glaucum (fig. 13c), in which
species they are comparatively few in number. The thread is extremely short and
broad, being usually fully 2/x in breadth. In sections, the internal portion exposed
at the cut surface stains more deeply than the external portion of the thread.
Within the cell the thread lies loosely and irregularly coiled, there being usually only
from 3 to 4 turns of the spiral. The smallest nematocysts occur in Sclerophytum

* I have given an account of the physiology and comparative anatomy of the digestive organs of the
Alcyonaria in a paper which is in the Press at the time of writing (Pratt, 1905).

2L2

260

CEYLON PEAEL OYSTER REPORT.

durum. In this species they are extremely difficult to see, and are only about
5/u, long and 2/z, broad.

Throughout the family the thread is apparently unbarbed ; nematocysts of Alcyo-
nium with discharged threads are figured by Hickson (1895). The nucleus of the
cnidoblast is usually long and curved (fig. 13), and stains very readily. Nematocysts,
similar to those occurring in batteries in the tentacles, are frequently to be found in
the ectoderm of the mouth disc, in the stomodseum, imbedded in the peripheral
portions of the mesenterial filaments, in the endoderm of the canals, and in some cases
in the mesoglcea some little distance below the surface. They are extremely rare in
the ectoderm covering the general surface of the colony between the zooids.

Genus. Species.

Size of nematocysts.

Author.

Akyonium digitatum ....

7^/i hy 2fi to 3ft

Hickson

„ pachyclados

8/x to 9/% by 2/i to 3/<.

Pratt

Lobophytum paudflorum .

6/x by 2/i

Ashworth

„ validum . .

6/x to 7/x by 2/i to 2'5/j.

Pratt

,, arasswm .

5-9/i to 6-9/t by 2/x

I?

Sarcophytum roseum .

8 -9/i to 9/x by 4/t

j>

„ latum . .

6/x by 2/t to 3/x

)>

,, glaucum . .

16/i to 22/x by 10/i to 14/x
Thread 2/i broad.

i'

,, ehreribergi .

6/x by 2/x

15

Sclerophytum tuberauksum

7/x to 8/x by 2/x

1)

„ densum .

7/x by 3/x

1)

„ capitals . .

8/x to 9/x by 3/x to 3-5/1.

>>

,, hiiiuin .

8/x to 9/x by 3/x to 4/t

>>

,, gardineri

12/x by 4 '5/i

>>

„ durum . .

5/x by 2/x

)»

,, palmatum .

6/x by 2/x

n

Heteroxenia elizabeffue. .

9/x by 2'5/x

Ashworth

Heliopora casrulea .

9/x by 2/x to 3/x

Moseley

Xenia hicksoni ....

8/i by 2/x, to 3/x

Ashworth

Clavularia prolifera .

10/x to 15/t by 2/x to 3/x

1

V. Koch

Tentacles.

"When the tentacles of Alcyonium are expanded, their ectoderm is extremely thin,

and is composed almost entirely of batteries of cnidoblasts, columnar interstitial cells,

scleroblasts, and a few mucous cells. I have observed no granular gland cells such as

occur on the mouth disc. In the ectoderm of the body wall, nematocysts are

comparatively few in number, and are rarely seen in the ectoderm covering the

colony.

ZOOCHLORELI^E.

Zoochlorellse are more or less fully described in a publication now in the Press
(Pratt, 1905). In a specimen of Sclerophytum densum some of the zoochlorellae
are seen to have lost their cellulose cell walls (fig. Lb). In such a case the
nucleus, chromatophore, and protoplasm of a cell exhibit unmistakable signs of

ALCYONIID^S. 261

division. This probably represents au early stage of sporogenesis. Very minute
zoochlorelho (fig. 16, c.) are also seen in the canals, but I have been unable to determine
the intermediate stages of spore formation.

Yellow Cells.

The more or less rounded cells of a yellow colour which occur in the ectodermal as
well as endodermal tissues are to be found in many tropical forms, but are more
numerous in the specimens of Sarcophytv/m (Pratt, 1903, fig. 4) than in those of
Lobophytum and Sclerophyt.um. They appear to be very abundant in certain
Madreporaria, and have been described and figured by Duerden (1903, p. 439).

As small portions of yellow granular matter, similar in all respects to the contents
of the yellow cells, may be frequently observed in the zoochlorellse, it is suggested
that the yellow cells represent an early stage of sporogenesis. The presence of yellow
granular matter may be due to the breaking up of the chromatophores, but this point
can only be satisfactorilv settled by a systematic study of the lifediistory of the
zoochlorellas in living corals.

Generative Organs.

Comparatively few specimens in the collection contain generative organs, and in such
instances they are entirely of a female character. Hickson (1895, pp. 376-380) has
already described the oogenesis, and to a certain extent the spermatogenesis, of
Alcyonium, and Ashworth (1898, pp. 207-209) has described the spermatogenesis of
Xenia. In the British Alcyonium digitatum the development of the sexual cells
occupies a very considerable period, as the ova begin to appear about April or May,
and are usually not sexually ripe until December or even later. I have observed the
discharge of ripe ova at Port Erin during the early part of April. In the British
species the period of the sexually mature condition is extremely short. Apparently
the colonies inhabiting a particular district discharge all their ripe ova and spermatozoa
almost simultaneously. In tropical members of the family, however, such is not
the case.

In his description of the spermatogenesis of Xenia, Ashworth (1898, p. 207)
states that the spermatozoa closely resemble those of Alcyonium in development and
structure, but spermatozoa occur in a single specimen in all stages of development, so
that the discharge of ripe spermatozoa extends over a considerable period ; this he
attributes to the fact that as the coral lives on reefs in the shallow waters of tropical
seas, it is not subjected to any great variation in temperature and food supply.

The development of the ovum in Sarcophytum, Lobophytum, and Sclerophytum
closely resembles that of Alcyonium, and is typically Alcyonarian in character
(tigs. 17 and 18). Frecpiently (in specimens of Sarcophytum latum, Sclerophytum
durum, and Sclerophytum gardineri), however, ova in several stages of development
may be observed on a single mesentery, the development proceeding from above

2(52 CEYLON PEARL OYSTER REPORT.

downwards, so that the youngest ova usually occur nearest the lower end of the
stomodseum. In a specimen of Sareophytum latum, ova on a single mesentery vary
in size from 0'03 millim. to 0-5 millim. in diameter. Two series of developing ova, one
a little distance below the other, were observed on a single mesentery in Sareophytum
contortum.

On comparing the oogenesis of Sareophytum, Lobophytum, and Sclerophytum with
Ashworth's description of the spermatogenesis of Xenia, it would appear that the
sexually mature condition, in these tropical genera, extends over a considerably longer
period than in the case of corals inhabiting temperate waters. The complete absence
of male colonies, and of generative organs in many specimens, in the collection, is very
noteworthy. I can offer no explanation, for many of the colonies are of considerable
size, and the zooids appear to be mature in every other respect.

In the Alcyonaria generative organs usually occur on the ventral and lateral mesen-
teries only (fig. 17). In a specimen of Sclerophytum gardineri, however, ova occur also
on the dorsal mesenteries (fig. 18). They are not connected with the ectoderm of the
mesentery, but are borne and supported by lateral outgrowths of the mesenterial
endoderm. Ashworth (1898, p. 207) records the presence of spermaria on the
dorsal mesenteries of Xenia, but these are apparently the only records of their
occurrence on these mesenteries in the Alcyonaria. As both forms in which they occur
present modifications in other respects, this must be regarded also as a secondary
feature, and of no special morphological interest.

Siphonozooids.

I have slight additions to make to the description of the siphonozooids of Sareophytum,
Lobophytum and Sclerophytum already given in a previous publication (Pratt, 1903).

In Sareophytum and Lobophytum the siphonozooids are small in comparison with
the size of the autozooids, but in Sclerophytum they are either very minute (fig. 20)
or absent. In all well marked cases of dimorphism the stomodseum is large, usually
ciliated throughout, and has a well marked siphonoglyph furnished with long flagella
(fig. 19). The dorsal mesenteries are long and have large, grooved and ciliated
filaments. Even when of considerable size siphonozooids are usually only slightly
contractile, and have only small ventral mesenteries. The specialisation of these
zooids to fulfil a circulatory function has been accompanied by a reduction and
atrophy of those organs which are concerned in the capture and digestion of food.
Throughout the Alcyonaria siphonozooids apparently have no tentacles and their
ventral mesenteries no filaments. The increase of the flagellate surface of the
stomodueum has been at the expense of the digestive surface. The increased current
of sea water within the zooids would carry away the food particles so rapidly, that
they would be unable to receive the secretion of the gland cells, which, however,
would be rendered so dilute as to be quite useless for digestive purposes. These

ALCYONIIP.R 263

zooids therefore have hecome incapable of capturing and digesting food. The increased
circulation brought about by their specialisation would relieve the autozooids to a
certain extent of that function, hence the feebly marked siphonoglyph in the autozooids
of Sarcophytum and Lobophytum.

The stomod;eal ectoderm of the siphonozooids in Sarcophytum and Lobophytum is
composed of the same elements as in the autozooids, but owing to the physiological
division of labour they occur in very different proportions.

In the stomodseum of the siphonozooid there is a pronounced increase in the
number of flagellate cells, so that the sipbonoglvph is proportionately larger than in
the autozooid — this is accompanied by a very pronounced reduction in the number of
granular gland cells, which are extremely rare in these two genera, but are entirely
absent in Sclerophytum. The stomodseum of the siphonozooid in Sarcophytum and
Lobophytum consists chiefly of flagellate cells, which constitute the well marked
sipbonoglvph and the ciliated columnar epithelial cells which line the remaining
portion of the stomodseum. Among the latter are interspersed a few interstitial cells,
nematocysts, scleroblasts, and a very few gland cells.

Siphonozooids and autozooids frequently communicate with each other by means of
short ciliated canals, which differ from the vessels of the canal system in the presence
of long columnar endoderm cells provided with fairly long cilia (fig. 22).

In Sclerophytum modification of the stomodaeal ectoderm has proceeded still
further. In this genus the siphonozooids are very much reduced. The stomodseum
has neither flagellate cells nor gland cells, but consists in the species S. densum and
S. hi rt u in (fig. 20) of a short simple tube opening to the exterior and lined apparently
by undifferentiated ectoderm cells. In the species S. palmatum and S. capitate there is
no aperture to the exterior ; in S. durum and S. polydactylum, an aggregation of a very
few ectoderm cells is the degenerate representative of the stomodseum in other forms.
The stomodseum is entirely unrepresented in S. gardineri and S. querciforme.
Mesenteries are only present in very few cases and are extremely small and very
feebly developed (fig. 2(1 m).

Development by Budding.

The siphonozooid buds differ from those which are destined to become autozooids
in that their development is arrested at a very early stage, which appears to be
constant for the species, and also to a certain extent in development, although the
very early stages of both forms of zooids are identical.

In Sclerophytum the siphonozooids are extremely small and retain throughout life
their intimate connection with the transverse vessels of the canal system from which
they originated. In the development of the siphonozooids of S. densum a caecum is
formed by an outgrowth from an endodermal canal near the surface of the colony. This
is accompanied by a multiplication and aggregation of the ectoderm cells at the surface,
immediately opposite the caecum (fig. 20). These cells arrange themselves so as to

2(54 CEYLON PEARL OYSTER REPORT.

form a tube, usually one cell in thickness, which is at first blocked by a plug of
mesogbeal-like substance (pi.). Rudiments of mesenteries frequently appear (m.),
and by the removal of the plug an aperture is formed to the exterior. If a bud is
destined to become an autozooid, mesenteries appear at a very early stage and
usually attain a considerable size before the completion of the stomodseum or
formation of the tentacles. My study of the development of the buds is yet incom-
plete, but there can be no doubt of the ectodermal origin of the stomodaeum and the
endodermal origin of the mesenteries. I have not yet worked out the origin of the
ventral mesenterial filaments.

Canal Systems.

The canal systems of the Alcyonaria have been described by Hickson (1895) for
Alcyonium, Moseley (1881) for Sarcophytum, Ashworth (1898) for Xenia, and
Pratt (1903) for Lobophytum and Sclerophytum.

The canals in Alcyonium are very similar to each other in form and structure, but
very irregular in their course ; there appear to be no large vessels which maintain a
course parallel to the periphery or to the longitudinal direction of the zooids, so that
it is impossible to divide the canals into superficial and internal systems as in
Sclerophytum. The vessels of the canal system are most numerous near the surface — in
the deeper portions of the colony they are very sparingly distributed and are confined
to the basal portions of the secondary polyps and the adjacent portions of their
neighbours (Hickson, 1895, p. 362, Plate 37, fig. 8).

In Sclerophytum the superficial and internal canal systems are extremely well
marked throughout the colony ; near the base the internal system is very complex,
owing to the branching and anastomosing of numerous vessels. In the presence of
two systems of canals, and in their manner of distribution, Sclerophytum resembles
Xenia.

The canal system of Sarcophytum is very similar to that of Lobophytum, and
differs from Sclerophytum and Xenia in the absence of a superficial system, and
in the fact that the principal longitudinal vessels are direct prolongations of the
ccelentera of the siphonozooids.

New canals frequently arise as outgrowths from pre-existing vessels in the follow-
ing manner : — One or more endoderm cells wander from a canal into the mesoglcea
and divide and multiply to form a strand of cells ; the individual cells of the strand
multiply and form a fairly solid cord of endoderm, which usually remains attached to
the canal from which it grew. Eventually a longitudinal splitting occurs and the
lumen of the canal is formed (fig. 23).

This investigation of the Ceylon Alcyoniidas has been carried out in the Zoological
Laboratories of the Victoria University of Manchester. I am greatly indebted to
Professor Hickson for much valuable advice and kind assistance in my work.

ALCYONIID/E. 265

LIST OF LITERATURE.

1834. Ehrenbero, C. G. — 'Die Corallthicre rles Rothen Meeres.'

1877. Klunzinger, C. B. — 'Die Korallthiere des Eothen Meeres,' erster Theil, "Die Alcyonarien und

Malacodermen." Berlin.
1881. MOSELEY, H. N. — ' " Challenger " Reports,' Zoology, vol. ii., Corals, " Heliopora and Sarcophyton."
1S86. Marenzeu.er, C. von. — "Ueber die Sarcophytum benannten Alcyoniiden." ' Zool. Jahrb.,' I.

Jena. 1886.
1889. Wright, C. P., and Studer, T. H. — ' "Challenger" Reports,' Zoology, vol. xxxi., " Alcyonaria."
1889. Famintzix, A. — "Beitriige zur Symbiose von Algen und Thieren." 'Mem. de l'Acad. Imp. des

Sci. St. Petersburg,' Ser. vii.
1895. Hickson, S. J. — "The Anatomy of Alcyowiwn digitatum." 'Quart. Joum. Micro. Sci.,' vol. xxxvii.,

part iv. 1895.
189C. Schexk, A. — " Clavulariiden, Xeniiden und Alcyoniida1 von Ternate." ' Abhandl. Senckenb.

Ges.' Frankfurt-a.-M.

1897. Whitet.egge, T. — "Alcyonaria of Funafuti." ' Mem. Austral. Mus.,' part iii. 1897.

1898. Ashworth, J. H. — "The Stomodamm and Mesenterial Filaments of Xenia." 'Roy. Soc. Proc.,'

vol. lxiii. 1898.

1899. May, Walther.- — "Beitriige zur Systematik und Chorologie der Alcyonaceen." ' Jena. Zeitsch.

Naturw.,' Band xxxiii., N. F. xxvi.

1900. Hickson, S. J., and Hii.es, Isa. L. — " Stolonifera and Alcyonacea of New Britain." 'Wii.i.ey's

Zool. Results,' part iv. Camb. Univ. Press.

1 900. Hickson, S. J. — " Alcyonaria and Hydrocorallinre of Cape of Good Hope." ' Marine Investigations
in South Africa.' Cape Town. 1900.

1903. Pratt, Edith M. — 'The Alcyonaria of the Maldives,' part ii. " Sarcophytum, Ldbophytvm, Sclero-
phytum and Ahi/miium." 'Fauna and Geog. of Maklive and Laccadive Archipelagoes'
(J. Stanley Gardiner), vol. ii., part i.

[/n the Press.] 1905. Pratt, Edith M. — "The Digestive Organs of the Alcyonaria and their Relation-
ship to the Mesogkeal Cell Plexus." ' Quart. Journ. Mic. Science.'

2 M

266

CEYLON PEARL OYSTER REPORT.

EXPLANATION OF PLATES.

List of Reference Letters.

amb.e., amoeboid cell in mesogloea.

nji.c, aperture of canal.

on., contracted autozooid.

lt.cn., battery of nematocysts.

c.iii.l., capitular marginal lobe.

r.uii., expanded autozooid.

r.c, ccelenteric cavity.

cap., capitulum.

cap.mar., capitular margin.

<•/., clubs.

col.end., columnar endoderm.

c.v., ciliated vessel.

tl.in.f., dorsal mesenterial filament.

ect., ectoderm.

end., endoderm.

end.b.w., endoderm body-wall.

end.can., endodermal canals.

g.v., germinal vesicle.

in.m.sp., inter-mesenterial space.

l.m.f., lateral mesenterial filament.

/. lobe.

m., mesentery.

in. up., mouth aperture.

mar.lob., marginal lobe of capitulum.

mg., mesogloea.

mg.h.tv., mesogloea body-wall.

nu/.c, mesogloeal cells.

uti/.il., dense layer of mesogloea surrounding

siphonozooids.
nuc, nucleus,
of., ovum.
s., stomodaeum.
si., siphonozooid.
sp., spicule.

sp.h., hole left by spicule after decalcification.
st., stalk.
tmt., tentacle.

tr.s.c, transverse superficial canal.
tu., tubercle-like zooid on the stalk.
cm., ventral mesentery.
v.m.f., ventral mesenterial filament.
■i/.aii., young autozooid.
■i/l\, yolk of ovum.
zo., :.'., zoochlorell.'e.

ALCYONIID.E, 267

PLATE I.

Fig. 1. Sarcophylum bicolor, n. sp. — Drawing of the smaller specimen, which has no capitular marginal
seam. Autozooids and siphonozooids are of a beautiful creamy white colour, the remaining portion of the
colony being dark grey, x 2.

Fig. 2. Sarcophylum bicolor, n. sp. — Different forms of spicules, x 60.

Fig. 3. Sarcophylum oligotrema, n. sp. — Drawing to show cup-like form of colony, x \i. The autozooids
on die margin are smaller and relatively more numerous than in the middle of the capitulum. Siphono-
zooids arc very few in number in this species.

Fig. 1. Sarcophytum oligotrema, n. sp. — Surface view of a decalcified fragment of the capitulum, about
an inch from the margin, to show the relative abundance of autozooids and of siphonozooids. x 27.

Fig. 5. Sararphytum oligotrema, n. sp. — Three warted spicules, x 60.

Fig. 6. Sarcophylum contortum, n. sp. — Drawing to show the complicated folding inwards of the capitular
marginal lobes (e.m.l.). All these lobes arise as marginal ingrowths and not, as in Lobophylum and Sclero-
phylum, from the general surface of the capitulum. Autozooids and siphonozooids in this specimen can
only be seen with the aid of a lens and are, therefore, not indicated in the drawing, x l.1..

Fig. 7. Sarcophylum contortum, n. sp. — Three warted spicules, x 60.

PLATE II.

Fig. 8. Sclerophytum herdmani, n. sp. — From a photograph, showing the zooids on the stalk. About
A nat. size.

Fig. 9. Sclerophytum herdmani, n, sp.- — Drawings of the different forms of spicules found in this species.
a.h.c x7. d.c.f. x60.

Fig. 10. Alcyonium ceylonicum, a. sp. — From a photograph. About A nat. size.

Fig. 11. Alcyoniwm ceylonicum. — A contracted autozooid showing 2 large dorsal mesenterial filaments
and a ventral mesentery which has no filament, x 60.

Fig. 12. Alcyonium ceylonicum. — Three spicules characteristic of the species, x 140.

Fig. 13. Nematocysts of a. Sarcophylum roseum. — Actual size 8-5/i by 4 '5/*.
b. Sclerophytum densum. — Actual size 6/x by 4-5/x.
r. Sarcophylum glaucum. — Actual size 20//. by 12//.
(/. Lobophylum pauciflorum. — Actual size 7/t by 2'5/x.

The nematocysts of Sarcophytwn glaucum are enormous. The thread is loosely coiled within the cell.
The internal portion of the thread stains more deeply than the external.

Fig. 14. Alcyonium digitalum (British). — Drawing of an extended living tentacle, showing the batteries
of nematocysts and the single bi lateral row of pinnules (b.cn.). Cam. Inc. x 26.

Fig. 15. Lobophylum pauciflorum. — An autozooid tentacle, showing the asymmetrical arrangement of
pinnules, x 60. Cam. luc. The tentacles of the tropical species are usually much smaller than those of
the British species (fig. 14).

Fig. 16. Sclerophytum densum. — Zoochlorellae : a and b show early stages of sporogenesis. The
cellulose cell walls have disappeared, and the protoplasm and nuclei of the cells show unmistakable signs
of division. In b the division is more complete than in a. c represents two very young zoochlorelke I
have been unable to observe the intermediate stages between b and c. «xl300, &xl600, ex 1200.
Cam. luc.

2 M 2

26 S CEYLON PEAEL OYSTER REPORT.

PLATE III.

Fie. 17. Sclerophytwm gardineri. — Transverse section through a ventral mesentery bearing a typical
Alcyonarian ovum, x 930. Cam. luc.

Fig. 18. Sclerophytwm gardineri. — Transverse section showing the unusual occurrence of two young ova
on a dorsal mesentery, x 930. Cam. Inc.

Fig. 19. Lobophytwm pauciflorwn. — Transverse section through a siphonozooid just below the surface.
All the columnar cells lining the stomodseum are ciliated, but those of the siphonoglyph are more closely
packed and have very long flagella. The endoderm is greatly distorted owing to the presence of
zoochlorellae. x 80.

Fig. 20. Sclerophytwm damim. — Longitudinal section showing stages in development of siphonozooids,
Sii, Si2, Si3, by budding. Sii is the youngest and is formed as a bud from the endodermal canal leading
from Si3, and by an aggregation of ectodermal cells to form the stomodseum, which in the section consists
at this stage of only two cells. A plug of mesoglcea at this stage closes the mouth aperture. Si2 is an
older bud. The development of the stomodseum has advanced, but the mouth aperture is still closed.
6"i3 is a siphonozooid which has attained its normal development in this species. The mouth aperture is
present and a very rudimentary mesentery (in.) is seen on the right, x 600.

Fig. 21. Sclerophytwm densum. — Longitudinal section through a developing autozooid. The mesenteries
are fairly well developed before the mouth aperture and tentacles are formed. Only one mesentery is
indicated in the drawing, x 360.

Fig. 22. Lobophytwm pauciflorwn. — Transverse section showing the ciliated communication between two
siphonozooids. The endoderm cells differ from those lining the cculeiiteron in that they are long, columnar,
and are provided with fairly long cilia, x 600.

Fig. 23. Sclerophytmn densum. — Drawing showing the formation of an endodermal canal by the splitting
of a solid cord of cells, x 700.

CEYLON PEARL OYSTER REPORT

..Cap mar

ALCYONIICLE PLATE 1

Si

■ r <

Fig 3

tent
.Au.

Si.

') ■

yltt

Fig.4.

Fig. 7.

Fig. 6

Pigs .1, 3&6 .1
E.M.PMI

laon . Cambridge

Figs. 1, 2, Saroophytum bicolor, n.sp.; Figs 3 5, Sarcofhytom oligotrema, n.sp. ;

Figs. 6, 7, Saroophytum contortum, n.sp.

CEYLON PEAU1. OYSTER REPORT

ALCYON HD.E PL ATE 1 1 .

FIG. 8

Fig. 12.

" &

Fig. 13.

b

'

. ■ -

a

< £1

FIG. II.

Fig. 15.

b OTb-

6. en.

FlG. 14.

-vm.

. d. m
. drn.f

- C.C

Fig. 10

■

13, Nematocys Figs. 14, 15, Ten . 16. Zoochlor

CEYLON PEARL OYSTER REPORT

ALCYONIIILE PLATE HI.

dl mes.

- ect

.. .

i

■

FIG. 18.

Tnm, sjj.

FIG. 19.

•-•.i. del

i ATOMY.

[CEYLON TEARL OYSTER FISHERIES— 1905— SUPPLEMENTARY REPORTS, No. XX.

REPORT

ON THE

ALOYONARIA

COLLECTED BY

Professor HERDMAN, at CEYLON, in 1902.

BV

Pkofessok J. ARTHUR THOMSON, M.A.. University of Aberdeen.

AND

W. O. HENDERSON, M.A., B.So., Uaknec;ie Scholak* Univeksity of Auekdeen.

[With SIX PLATES.]

The l'ich collection of Ceylonese Alcyonarianst here reported on was made in 1902
from the Pearl Oyster Banks in the Gulf of Manaar, by dredging within the
100-fathom line off Trincomalee and off Galle, and in adjacent littoral areas. The
localities are more precisely referred to in Professor Herdman's " Narrative " in Part I.
of the general Report (1903).

As only a few Alcyonarians from Ceylon have been previously recorded, e.g., by
Thurston \ (1890) and Ridley § (1883), Professor Herdman found an almost fresh
field, which his faunistic genius recognised as extraordinarily rich. We have not been
able to exhaust the material which his skilful collecting has garnered. His collection

* It is more than a formal duty to make grateful acknowledgment of the assistance of the Carnegie
Trust : Mr. HENDERSON did part of his share in this Report as a Carnegie Seholar, and I received from
the Trust £30 towards the expenses of preparations and drawings. — J. A. T.

t The "fleshy" Aleyoniida1, e.g., Alcyoniwm, Lobophytum, Sarcophytum, Sclerophytum, are reported on
separately hy Miss Edith M. Pratt, D.Sc. (see this volume, p. 217).

I Thurston's list is as follows: — Echinogorgia pseudosasappo, E. sasappo, E. term, E./wfaraced, Plexaura
flabellum, Juncella jv/iuxa, Gorgonia (Leptogorgia) miniacea, Gorgonella umbella, Svherogorgia subavsa, Pteroeides
javanicum, PL esperi, Virguhtria jwncea and Liiuariasp.

§ Ridley's list is as follows: — Alcyonium polydaclylvm, ./. submurdle, Sarcopkytwm pauciflorum
Spongodes sp., Rhizoxenia sp., MenaceUa reticularis, Echinogorgia pseudosasappo, Plemv/ra fiabellwm, Jvmcell
j a nn a, Svherogorgia verrmdata, S. svherosa, and (?) Coralliwm while.

270 CEYLON PEARL OYSTER REPORT.

is "littoral" in the wide sense, but it may be noted that the explorations of the
" Investigator" show that the deep water off Ceylon is also very rich in Alcyonarians.
In reporting on Professor Herdman's collection, we have had to deal with at least
75 species, representing at least 40 genera. About 32 of the species seem to us to be
new, and the number might have been considerably increased if we had exercised less
restraint. Our restraint in multiplying species has been mainly due to the ignorance
that obtains in regard to the internal structure of many Alcyonarian types — an
ignorance that cannot be removed without specimens specially preserved for histo-
logical purposes. Wonderfully diagnostic as spicules often are, they have been used
too confidently and liberally in the establishment not only of species, but of genera,
and a secure natural classification of Alcyonarians is still far to seek. Moreover, in
some of the forms which we have studied, e.g., Spongodes, there is great specific
variability both as regards spiculation and branching. As an astounding illustration
of diversity of spiculation in a single species, we may refer to our description of the
form which we have called Echinogorgia multispinosa, n. sp.

Before proceeding to the systematic report, we may call attention to a few
facts of general interest : —

(a) Very striking is the frequent illustration of " convergence," i.e., the occurrence
of superficially similar forms which are not in reality nearly related. Thus, on one
occasion, three similar bright red forms were collected in one haul and loosely bound
together with thread. An examination of this small parcel disclosed three species
belonging to separate genera — Keroeides gracilis, Verrucella rubra, n. sp., and Telesto
rubra. Equally marked is the resemblance between Echinomuricea indo-malaccensis,
Ridley, and Echinogorgia pseudosasappo, Kolliker, to which Ridley calls attention
(1883, p. 235), and many other examples might be given.

(b) We have compared the Ceylon collection with two others entrusted to us, one
made in the Indian Ocean by Major A. Alcock, F.R.S., the other made at Zanzibar by
Mr. Cyril Crossland, M.A., and we find that the three collections have extremely
little in common. This is mainly because Alcock's collection was from deep water,
and Crossland's chiefly from close in-shore, but it may also indicate that many
Alcyonarians have a restricted local distribution. There is also a great difference
between the collection here discussed and those made off the Maldives by
Mr. Gardiner, and off New Britain, New Guinea, &c, by Dr. Willey. It may be
of interest to note that three Alcyonarians brought from Patani (Siam) by Mr. Nelson
Annandale and Mr. H. C. Robinson are unrepresented in Herdman's collection
(though included in this report), viz., Telesto trichostemma, Astromuricea rdmosa, n. sp.,
and Juncella trilineata, n. sp.

(<■) Some of the distributional facts, on the other hand, point in a different
direction, showing that species occur in Ceylonese waters which enjoy a wide range
elsewhere. Thus the exceedingly beautiful Primnoa ellisii is known from the

ALC YON ARIA. 271

Mediterranean and from off the Andaman Islands, and Gorgonia capensis is known
from the Cape and further south. In many respects the Ceylonese Alcyonarian fauna
is, as regards genera, remarkably cosmopolitan. Few naturalists, however, have
contrived to collect so exhaustively as Professor Herdman did in his Ceylon
expedition.

(<() There is in some species great variability in colour, in mode of branching, in
dimensions of spicules, and in the proportionate numbers of different forms of spicules.
In illustration we may refer to Spongodes pvlchra, n. sp., Spongodes bicolor, Wright
and Studer, Cliironephthya rariabilis, Hickson, Bebryce hicksoni, n. sp., Verrucella
Jlexuosa, Ridley. In many cases we have given details in regard to these variations,
partly as a contribution to the data of variability in general, and partly to facilitate a
future revision of species. It can hardly be doubted that many of the Alcyonarian
species, e.g., of Xenia, Spongodes, Scirpearella, Juncella, and Verrucella, are in a
state of evolutionary flux, and yet it is at present difficult to reduce the number of
species. As we have had numerous specimens of many of the apparently most
variable forms, we have not felt justified in regarding these as instances of merely
individual variation.

(p) As a simple illustration of the way in which even these Alcyonarians may
interact with the life of such animals as pearl oysters, we may refer to Clavularia
margaritiferm, n. sp., which spreads on the pearl oyster shell, and to the occurrence of
young pearl oysters attached to Alcyonarians, e.g., to the branches of Hetero-
gorgia sp. Not a few specimens, e.g., Nephthya lobulifera, were found attached to
dead pearl oyster shells.

( /') We have had the pleasure of confirming in Gorgonia capensis Professor
S. J. Hickson's interesting discovery of viviparity.

It may be convenient to place here a classified list of the species reported on :—

ALCYONAPJA (excluding most of the Alcyoniid.e).

Order I. : Stolonifera. *Nephfhya ceylonensis, n. sp.

Family: Clavulariidse. *Eunephthya purpurea, n. sp

*Paraspongodes striata, n. sp.
*Capnella manaarensis, n. sp.

*C'larntaria margaritiferce, n. sp.

Family: Xeniidae. * Spongodes' pulchra, n. sp,

Xt iiiu ternatana, Schenck. „ tricolor, Wright and Studer.

„ umbellata, Sav. „ „ „ H Var. ceylon-

ORDER II. : AlOYONACEA. CnS!S' "-

Family : Alcyoniidre.

* „ aurantiaca, n. sp.

*Bellonella indica. n. sp. i^v.™,

' v „ rosea, Kukenthal.

Family : Nephthyidav n armata, Holm., var. ceylonensis, n.

Nephthya chabrolii, Arn., var. ceylonensis, n. ,, dendrophyta, Wright and Studer.

„ lobulifera, Holm. „ splendent, Kukenthal.

* The forms with an asterisk are reported as new.

272

CEYLON PEARL OYSTER REPORT.

Family : Siphonogorgiidae.
*Paramphfhya pralli, n. sp.
Chironepbthya varidbilii, HlCKSON.
Siphonogorgut puslulosa, Wright .and Studer.
„ miniacea, Kukenthal.

„ Mlikeri, Wright ami Studer.

Order III. : Pseudaxonia.
Family : Briareida?.
Solenocaulon torhtomm, Gray.

Family : Sclerogorgidse.
Keroeides gracilis, Whiteeegge.
Suberogorgia verriculala, Esper.

Order IV. : Axifera.
Family: Primnoidse.
Primnoa ellisii, v. Koch.

Family : Muriceidse.

* Acanthogorgia media, n. sp.

„ muricata, Verrill, var. ceylonensis, n.

„ ceylonensis, n. sp.

*Astromuricea ramosa, n. sp.
Ei'li inomwricea indc-malaccensis, Ridley.

„ ceylonensis, n. sp.

Echinogorgia pseudosasappo, KOlliker.

* „ multispinosa, n. sp.
* Heterogorgia verrUli, n. sp.
*Bebryce hicksoni, n. sp.

*

*Lophogorgia irregularis, n. sp.
Leptogorgia auslraliensis, Ridley, var. flavotincta.

„ perflava.
„ (?) sp.
*Stenogorgia ceylonensis, n. sp.
Gorgonia capensis, Hickson.
Rhipidogorgia sp.

Family : Gorgonellida?.

Scirpearella awantiaca, n. sp.
divisa, n. sp.
sp. a.
sp. (1.
sp. y.
Sciipearia sp.

Juncella gemmaeea, Valenciennes.
„ juncea, Pallas.
„ fragilis, Ridley.

var. rubra, n.
* „ iriline'ata, n. sp.
*Vemtcella rubra, n. sp.

„ flexuosa, Klunz., var. aurantiaca, n,
,, „ var. gallensis, n.

*Acamptogorgia atra, n. sp.

,, spinosa, Hii.es

„ ,, var. ceylonensis.

*Acis indica, n. sp.

* ,, alba, n. sp.
„ wiemtalis, Ridley.

* „ cei/lonenfiis, n. sp.
*Muricella ramosa, n. sp.

* „ ceylonensis, n. sp.
„ nitida, Verrill.
,, complanata, Wright and Studer

Family : Plexauridre.
Plexaura praslonga, var. typica (Ridley).
„ elongata, n.
„ amiipathes, var. flexuosa.

Family : Gorgonidse.
Lophogorgia lutkeni, Wright and Studer.
rubrolincta, n. sp.

Order V. : Stelechotokea.

Section I. : Asiphonacea.

Family : Telestid®.

Ti'lesto rubra, Hickson.

„ (Carijoa) trichoslemma, Wright and Studer.

Section II. : Pennatulacea.

Family : Umbellulida;.
Umbellula sp.

Family : Virgularidfe.

Firgularia multiflora, Kner.
„ loveni, Kolliker.
* „ tuberculata, n. sp.
sp.

Family : Pennatulidce.

*Halisceplrum periyense, n. sp.

„ gustavianum, Herklots.

Pieroeides lacazei, var. spmosum, KOlliker.

Family : Veretillidse.

Cavermlaria obesa, Valenciennes.
*Stylobekmnoides herdmani, n. gen. et sp.

* The forms with an asterisk are reported as new.

ALCY0NAR1A. 273

DESCRIPTION OF THE SPECIES.

Order I. : STOLONIFERA.
Family: CORNULARIID^E.
Clavularia niargaritiferae, n. sp. — Plate III., fig. 8.

The stolon consists of a number of thin and narrow branching threads, forming a
network on the surface of a pearl-oyster shell. The strands of the stolon vary in
breadth from 0"4 millim. to 0-6 millim. They do not form any membranous expansion.
The polyps arise vertically from the branches of the stolon at intervals of 2 millims.
to 2 -5 millims. ; there does not seem to be any interpolation of new polyps among
the older.

The polyps measure 3 millims. in height, with a distal diameter of* 0-9 millim., and
a proximal of 0-5 millim. to 0'65 millim. They are thus broader than the stolon
branches, cf. Clavularia reptans, Hickson (1895). They have, in all cases, their
crowns retracted ; there is no indication of longitudinal grooves or lines.

The walls of the polyps and of the stolons show thickly crowded spicules of minute
size (0-06 millim. to 0-07 millim. X 0-(J3 millim. to 0-04 millim.) interlocked by their
wart-like projections, cf. Clavularia Jlava, Hickson (1895). Most of the spicules are
substantial rods bearing blunt wart-like projections often in a whorl of four towards
each end. The short and simple tentacles, bearing a few blunt pinnules (not
sufficiently fixed to allow of precise examination) are supported by numerous
longitudinally disposed elongated and curved rods, which may measure 0-l millim. in
length. They are smooth and only occasionally bear projections. All the spicules
are colourless, and the whole colony is white in its preserved condition.

Locality : — Pearl banks, Gulf of Manaar.

Family: XENIID^.
Xenia ternatana, Schenck.

A single specimen attached to a sandy worm-tube, with polyps in three groups,
agrees on the whole with Xenia ternatana, Schenck, notably in having only two
rows of pinnules on each side of the tentacles, and about 1 8 slender pinnules in each
row, as also in the measurements of the polyp-body and of the tentacles.

Xenia umbellata, Sav.

Attached to the substratum of another specimen, which we unfortunately omitted
to note in the business of assortment, there was a species of Xenia which appears to
be referable to A', umbellata. The locality was low water at Pamban. We assume
that the presence of only two species of Xenia in the collection simply means that
Professor Herdman's work was outside the ATmm-zone. There must be many species
of Xenia in Ceylon.

2 N

274 CEYLON PEARL OYSTER REPORT.

Order II. : ALCYONAOEA.*

Family: ALCYONIID^E.

The genera Sarcophytum, Sclerophytum, Lobophytum, and Aleyonium are discussed
in Miss Pbatt's Report, this volume, p. 247. There has been left to us to describe
what seems to be a new species of the little-known genus Bellonella.

Bellonella indica, n. sp. — Plate VI., fig. 5.

A small cylindrical specimen of a crimson-vermilion colour with yellow calyces and
white polyps. It is attached to a fragment of rock and stands 24 millims. in height,
with a basal diameter of 10 millims. and an apical diameter of 6 millims. The lower
half is a sterile trunk, the upper half bears crowded polyps, whose white colour
contrasts well with the red of the general ccenenchyma and the rich yellow of the
calyces. Many of the calyces measure 1 millim. in diameter, and the interval between
them is often the same. Smaller forms occur among the larger, but there is no
evidence of dimorphism of zooids. The margin of the calyx is neatly 8-lobed, and
here and there the white polyps are expanded. The tentacles, which seem to run
somewhat markedly to a triangular point, bear on each side about 18 finger-like
pinnules. A longitudinal cut shows the large longitudinal canals traversing the
bright red ccenenchyma, and the eight longitudinal strands in each canal stand out
sharply as bright white lines.

The spicules of the ccenenchyma are of an orange-red to yellowish colour, mostly
like knobbed capstans, or double clubs with large warty heads, or double wheels with
a very slight constriction between them. They form a granular pavement over the
surface and densely fill the ccenenchyma. In the sterile trunk there seem to be no
double clubs in the strict sense ; the form is more like a knobbed dumb-bell with an
exceedingly slight and short constriction in the middle. In all cases the warts are
few, large, and blunt. The spicules are thus unlike the fusiform echinate forms
reported as characteristic of Bellonella, but the genus has not been well defined, and
it may be noted that two isolated polyps showed a few colourless or faintly yellowish
fusiform spicules with a few thorns. It is not certain, however, that these belonged
to the polyp ; they may have been artificial inclusions.

The following measurements of the typical spicules were taken, length and breadth,
in millimetres : —

008 X 0-06 (0-04 at middle); 0-06 X 0*06 (0'04 at middle); 0*06 X 0"045 ;
0-06 X 0-04; 0-05 X 0'04 (0-03 at the middle); 0'045 X 0"0375 (0-025 at the
middle).

The specimen is very different from Bellonella {Cereopsis) boeagei, Sav. Kent, and
B. variabilis, Studer, but neither of these agrees conspicuously well with Gray's

* Exclusive of the Alcyomidae described by Miss Pratt (this vol., p. 247).

ALCYONARIA.

275

original definition. Gray's description oi'Bellonella granulatum (' Proc. Zool. Soc.,'
1862, p. 35) is unfortunately short and vague. He speaks of the "angular tips" of
the calyces ; he figures longitudinal furrows on the stem, and these features are
here absent. It seems better, therefore, to start afresh with a new species.
Locality : — Deep water south of Galle.

Family : NEPHTHYID^E.

Sub-family : SP0NG0DI1SLE.

Nephthya chabrolii, Aud., var. ceylonensis, n.

To this species, known to include several varieties (Kukenthal, 1903, p. 157), we
refer a specimen which differs from ordinary forms of N. chabrolii in being dull
greyish-white, and in having a " Stiitzbundel" spicule projecting beyond the polyp for
0*5 milliin. The following comparison with Kukenthal's description of N. chabrolii
may serve to illustrate dimensional variability.

Polyp.

Polyp spicules.

Stem spicules.

Length.

Breadth.

Length.

Breadth.

Length.

Breadth.

Kukenthal's measurements .
Herdman's specimen . . .

millims.

0-5 to 0-7
0-6 „ 0-7

millims.

0-5 to 0-7
0-6 „ 065

millims.

0-08 to 0-45
0-15 „ 0-35

millims.

0-01 5 to 0-06
0-02 „ 006

millims.

11 to 1-9
0-65,, 2

millims.

0-12 to 0-26
0-16 „ 0-26

For description, see Holm (1895).

Locality: — Pearl banks, Gulf of Manaar, March, 1902, and Donnan's Paar.

Nephthya lobulifera, Holm.

From an elongated base attached to an oyster shell there rise 7 main branches
apparently flaccid in character, bearing numerous lobes of unequal size and variable
arrangement. On these the polyps occur in groups often united by a few individuals
scattered in the intervening spaces. The polyps are light yellow and stand out
prominently against the general white colour of the colony. They measure 0"7 millim.
to l-3 millims. in length. The " Stiitzbundel " spicules are strong, slightly curved,
warty spindles up to 1"9 millims. in length; the three longest project beyond the
polyp. The polyp spicules measure 0"06 millim. to 0'25 millim., and the stem spicules
1*1 millims. to T95 millims. in length. For description, see Holm (1895).

Locality : — lj miles off Old Dutch Modragam Paar, Gulf of Manaar.

Nephthya ceylonensis, n. sp. — Plate I., fig. 4.

A small rigid colony fixed to a stone rises to a height of 38 millims., giving off

2 N 2

27ft CEYLON PEARL OYSTER REPORT.

three lobes crowded with polyps. The general colour of the lower part of the stem is
oreyish-white, but in the region of the polyps the colour is light violet, and the
polyps themselves are canary-yellow. The lower part of the stem has a stiff leathery
character and a granular appearance ; the upper part is entirely covered with large
spicules which are for the most part disposed longitudinally.

The polyps occur in small almost contiguous groups or singly. Each is an
elongated cylinder, standing at right angles to the " Stutzbundel,'' measuring
1*2 millims. in length by 0*8 millim. in breadth.

The spicules of the lower part of the stem consist of short thick spindles
(11 millims. to 1*4 millims. in length by 01 millim. to 0-2 millim. in breadth),
sometimes bifurcate or trifurcate at one end, and also of triradiate and quadriradiate
forms. All exhibit numerous blunt warts.

In the upper part the spicules are spindle-shaped, covered with numerous spines,
and tinged with violet. Some of them are forked at one end, and a few give off a
short branch about the middle of their length. They vary in length from 1*1 millims.
to 2-4 millims., and in breadth from 0-l millim. to 018 millim. Those of the
"Stutzbundel" measure T55 millims. to 1*8 millims. There is only a slight
projection beyond the polyp.

The polyp spicules are very small, deeply tinged with yellow, measuring 01 millim.
by 0'08 millim.

Eunephthya purpurea, n. sp. — Plate I., fig. 3 ; Plate V., fig. 5.

This species is represented by several specimens which present a striking appearance
owing to the contrast between the yellow polyps and the purple-red of the stem and
branches. The stem is large and flaccid, and gives off numerous small branches and
twigs. The superficial coenenchyma is rough, and has a granular appearance due to
the arrangement of the numerous small spicules. The twigs, however, have a
different appearance, for there the spicules are longer and are mainly disposed
transversely. Two good specimens measured in centimetres, — 11 X 7 and 12 X 3.

The spicules of the lower part of the stem are very short, irregularly branched rods
with prominent spines. They measure 01 millim. in length by 0-08 millim. in width.
The spicules of the twigs and the tip of the stem are slender spindles, varying in
length from 0'23 millim. to 0"25 millim. and in breadth from 0-02 millim. to
0-023 millim.

The polyps are scattered over the stem and branches, occurring either singly or in
small groups. They are yellow in colour, and they contrast well with the red colour
of the twigs and branches. The length of the polyps is 0-9 millim. to 1 millim. and
the width varies from 0"5 millim. to 0*6 millim. The polyp spicules are very small
yellow spindles with prominent warts and spines. Their measurements are 014 millim.
to 0"3 millim. in length and 0"015 millim. to 0-03 millim. in breadth. Below the
anthocodial part of the polyp there is a collar formed of several rings of spicules, red

ALCYONARIA. 277

in colour, which measure from 0-5 millim. to 0"7 millim. in length by 0"04 millim. to
0'06 millim. in breadth.

Localities : — -Pearl banks off Aripu, Gulf of Manaar ; and deep water off Galle.

Paraspongodes striata, n. sp. — Plate II., figs. 2 and 7.

A specimen without " Stutzbimdel," with no spicules in the canal walls, and
polyps disposed in well-defined bundles must be referred to the genus Paraspongodes
(see Kukenthal, 1896).

The colony studied stands 67 millims. high by 49 millims. in maximum breadth ; the
general colour is whitish. A short substantial stem gives off three main branches
which divide and re-divide, so that the appearance of an irregular corymb of umbels
results. The average length of a terminal branch is 4 millims. The polyps occur in
bundles of 6 to 7, and these bundles are grouped to form an umbel. A much finer
specimen, which was left at Liverpool, measured 22 centims in height, 11 centims.
before the first branching, and 4 centims. across at the base.

The superficial coenenchyma of the branches is deeply striated transversely, we
may almost say ringed, and is tough in character ; on the main stem the surface is
irregularly wrinkled and rough in texture.

The polyps vary in length from 1*1 millims. to 2 "3 millims., and in breadth from
0*65 millim. to 0-9 millim. They are supported by spindle-shaped spicules uniformly
disposed in contiguous longitudinal rows, and measuring 0'18 millim. to 0"6 millim.
in length by 0'04 millim. to 0-06 millim. in breadth. The tentacles measure
0"6 millim. in length, and two or three pairs of plump pinnules are visible.

In the coenenchyma the spicules are minute but very numerous. They occur as
rough spindles or rods irregularly branched, and also as what may be called irregular
stars. They vary in length from 0"09 millim. to 0'16 millim., and in breadth from
0-03 millim. to 0"06 millim.

Locality : — Deep water outside pearl banks, Gulf of Manaar.

Another larger specimen, 169 millims. in length by 69 millims. in maximum
breadth, from the same locality seemed at first different.

It had a relatively much longer stalk with basal rhizoid branches ; it showed
greater transparency of texture, less abundant superficial spicules, less marked
transverse wrinkling, and a larger number (7 to 11) of polyps in each bundle.
Closer examination showed essential agreement as to spicules, polyps, tentacles, &c.
The measurements for this specimen show considerable variation, as the following
table indicates : —

278

CEYLON PEARL OYSTER REPORT.

Polyps.

Polyp spicules.

Length.

Breadth.

Length.

Breadth.

millims.
1-2
1-3
0-9

1-0

millims.

0-8

0-85

0-7

0-8

millims.
0-6

0-63
0-7

ro

0-9

0-25

0-18

millims.

0-05

0-05

0-05

0-10

0-08

0-016

0-03

Capnella manaarensis, n. sp. — Plate II., fig. 4 ; Plate V., fig. 14.

A specimen without " Stutzbundel," with numerous spicules in the canal walls, and
with the polyps arranged not in bundles but in conical lobes or catkins, we refer to
the genus Capnella, as reconstructed by Kukenthal (1902).

The colony is fairly rigid and stands 79 millims. high by 75 millims. in maximum
breadth. The general colour is a greyish-white or a dirty white. A short stem
gives off several main branches, which in their ultimate divisions give rise to polyp-
bearing lobes or catkins.

The superficial ccenenchyma of the branches and the main stem is rough in texture
and has a stringy appearance owing to the arrangement of the spicules.

The polyps have an average length of 0'95 millim. by an average breadth ot
0-65 millim. They are supported by spindle-shaped spicules which are arranged
longitudinally, and vary in length from 0-29 millim. to 0'55 millim. and in breadth
from 0'04 millim. to 0'08 millim.

In the ccenenchyma the spicules are spiny spindles. They are usually straight or
curved, but several show a bifurcation at one end, or give off a short branch near
their middle point. They show considerable variation in size, measuring from
0'5 millim. to 1*35 millims. in length and 0"10 millim. to 0"16 millim. in breadth.

Locality : — Pearl banks off Aripu, Gulf of Manaar.

Spongodes pulchra, n. sp.* — Plate I., fig. 5 ; Plate V., fig. 10.

This species is represented by many specimens, 3 of which give the following
measurements : —

* The third instalment of Professor W. Kukenthal's " Versuch einer Revision der Alcyonarien "
(' Zool. Jahrb.,' xxxi., 1905, pp. 503-726, 7 pis., and 61 figs.) appeared while this report was being passed for
press, and it could not be utilised. He splits the old genus Spongodes, with which one is loth to part, into
the two new genera Dendronephthya and Slereonephthya, with 90 species to the former and 8 species to the
latter. Everyone will welcome Kukenthal's important contribution to the study of a difficult and
polymorphic genus, but what is gained by trying to do away with ' the old title Spongodes, which is
practically equivalent to Dendronephthya, just as Spongodia is to Stereonephthya ? Kukenthal has studied

ALCYONAKIA. 279

(A.) (B.) (C).

Total height 66 millims. 66 millims. 36 millims.

Length of trunk. ... 14 „ 13 ,, 9 „

„ head .... 52 „ 53 „ 27 „

Maximum width of same . 39 ,, 43 „ 21 ,,

The trunk or bare part of the stem is granular in appearance and leathery in
texture. In (B.) and (C.) it is greatly wrinkled, owing to the large canals and the
weak partition walls between.

The branching is very profuse, the stem giving off a large number of primary
branches which break up into secondaries, and these in their turn into the twigs
bearing the polyps. The twigs or tertiary branches bear the polyps in corymb-
fashion, so that they all reach the same level, and owing to the large number of the
polyps the stem and branches are completely hidden.

The polyps are arranged in groups of 4 to 6, and they in their turn are aggregated
into larger bundles. They are red, with markedly white tentacles. Although some
of the spicules close to the polyps become a little stronger, none of them project, so
that the colony has not the slightest appearance of being spiny.

The surface of the stem is rough to the touch and has a granular appearance. In
the primary and secondary branches, however, the spicules appear more prominently
and give the surface an irregular stringy appearance, and finally they take an almost
longitudinal arrangement in the twigs.

In (A.) the stem, primary and secondary branches are almost white, while in (B.)
and (C.) they are yellow. In (A.) the twigs are white, but become pinkish below the
polyps, while in (B.) and (C.) they are yellowish, but also become pink below the
polyps. Thus there is considerable colour-variation.

The spicules of the trunk are very short and furnished with long spines. They are
very varied in shape — short rod-like bodies, 3 -rayed stars, and irregular crosses. In
all, however, the spines are well developed. They measure, length by breadth, in
millimetres, as follows :— 0-15 X O'l ; 02 x G"05 ; 0"20 x G"l ; (V5 X 0'15. All these
measurements include the spines, which often measure 0*04 X 0-02 ; 0'03 X 0*02 ;
and 0-045 X 0-01.

The spicules of the stem and branches are long slender spindles, usually curved or
bent and covered by fairly prominent spines. They measure, length by breadth, in

213 specimens in all, but in many cases he only studied one (32) or two (21). He makes 30 new species,
some of which may correspond to those here described as new, though we have not been able as yet to
satisfy ourselves as to any identity. We venture to predict, however, that further study of this prolific
genus, whether it be called Spongodes or DendronepMhya, with say ten specimens of each species, will
increase the number of quite distinct species, and will lessen the number of closely adjacent species, reducing
them to varietal forms. Even on a single colony there is often great diversity, and the impression of
great lability is increased when we compare diffei-ent specimens of the same species collected on the same
date from the same locality. — J. A. T.

280 CEYLON PEARL OYSTER REPORT.

millimetres, as follows :— P05 X 0-56; 0"55 X 0-06; 073 X 0-04; and 0"47 X 0*05.
The longest reach a length of 17 millims., 1"6 millims., and 1'5 millims.

The specimens differ markedly from all the previously described species of this large
and difficult genus.

Locality: — Station LXVL, off Mutwal Island, 10 to 35 fathoms; Cheval Paar,
Gulf of Manaar.

Spongodes bicolor, Wright and Studer.

This species is represented by a fairly large complete colony, which consists of a
short trunk and a long head, formed by a large number of branches given off on all
sides. The base is formed of numerous rhizoid-like offshoots, which are attached to
grains of sand, to sponge spicules, and to particles of shells.

The stem is soft and flaccid, and gives off flabby branches on all sides. The majority
of the lower branches are flattened, the flattening taking place not from above
downwards but from side to side.

The stem and branches have thin semi-transparent walls with numerous fine
spicules scattered all over, some of them projecting so as to give the surface a
pubescent appearance. The measurements of the polyps and of the spicules agree
with those given by Wright and Studer (188 9).

Locality : — Deep water off Galle.

Spongodes bicolor, Wright and Studer, var. ceylonensis, n.

The trunk of this specimen is very short, and does not seem to be distinctly marked
off from the stem except in the spiculation.

Small branches are found coming off to within a few millimetres of the lower end of
the trunk. The spicules are more distinct in the stem and give a net-like appearance
to its surface, while the surface of the trunk is granular.

Two of the lower branches are flattened from above downwards and nearly surround
the stem, but just below the two spaces left between them there are two ordinary
branches.

The polyps are red, with white tentacles, and a " Stutzbiindel " spicule projects
behind each polyp.

The colour of the trunk, stem, and branches is a yellowish-white, that of the polyps
and twigs being red ; the lower part of the twig may be yellowish.

This form differs from Spongodes bicolor in several respects, but yet comes very
near it.

Spongodes bicolor, Wright and Studer, var. dubia, n. — Plate V., fig. 13.

Another variety is represented by a specimen measuring 71 millims. in height and
61 millims. in maximum breadth. It may be a portion of a larger colony or an entire
colony in which the base is wanting.

ALCYONAEIA. 281

The stem is flaccid and gives off branches almost to the base, the lower branches
are flattened and give off ordinary branches from their upper surfaces.

The polyps are arranged in umbels of 10 to 13 polyps. The polyps are white, but
have a red appearance, owing to the spicules ; the tentacles have white or colourless
spicules.

The spicules of the stem are long, slender, colourless spindles covered with spines,
and varying in length from 1 "55 millims. to 0'6 millim., and in width from 0"12 millim.
to 0-08 millim. The spicules of the terminal twigs and polyps are slender spindles,
yellowish-red in colour and thickly beset with spines. They vary in length from
(VI millim. to 1'65 millims., and in breadth from 0'02 millim. to 0"09 millim.

As this form agrees in many respects with Spongodes bicolor, we have ranked it
provisionally as a variety.

Spongodes aurantiaca, n. sp. — Plate I., fig. 9 ; Plate V., fig. 6.

In this species the trunk is long, stiff, leathery in character, and granular in
appearance, but the individual spicules become evident at the upper end.

The stem is divided into two main branches, from which the primary and secondary
branches arise. The four lowest branches are flattened from above downwards, and so
form a complete collar to the trunk. They are also recurved, so that they hide the
upper part of the trunk. Their edges are very much divided, and they carry the
polyps singly on the edges. From their upper surface spring ordinary branches.
The other branches stand at right angles to the stem, but the upper ones are directed
obliquely upwards.

The polyps are borne singly at the ends of the twigs, and also in bundles of 6 to 7,
all reaching the same level. They are white in colour and backed by strong
" Stutzbtlndel " spicules.

The colony presents a striking appearance, due to the marked contrast between the
orange-coloured stem and branches and the white polyps. The trunk is orange-
coloured at the upper end, but it gradually becomes a whitish-grey as the lower end
is reached.

The " Stiitzbiindel " spicules vary in length from 27 millims. to 3-3 millims., while
the stem and branch spicules vary from (V5 millim. to 2"1 millims. in length.

Spongodes rosea, Kukenthal — Plate V., fig. 1.

What we regard as a variety of this species is represented by a stiff colony, with a
bare stalk very much shrunken, owing to the large canals and the thin partitions
between them. Height of colony, 70 millims. ; length of stalk, 22 millims. ; diameter
of same, 1 1 millims.

The lower branches, 6 in number, are flattened from above downwards, breaking up
at their edges into single polyps and groups of polyps, and giving off cylindrical
branches from their upper surfaces. Immediately above these the stem gives off a

2 o

'282 CEYLON PEARL OYSTER REPORT.

large branch which repeats the structure of the stem. Over the whole surface of stem
and branches smaller branches are given off which also branch.

The colour of the stem is greyish-white below, white above, while the smaller
branches are orange-coloured. The polyp-cups are a dark purple-red, the tentacles

white.

The surface texture of the specimen presents two quite diffei'ent aspects. On the
bare stalk the surface has a granular appearance caused by the large number of very
small closely packed spicules which grow larger as the upper part of the stalk is
reached. On the branch -bearing part, and also on the branches, a stringy appearance
is produced by the larger spicules.

The polyps occur in bundles of 4 to 8. The projecting spicules are usually red
throughout, but sometimes the lower part is yellow. The spicules are as follows :—
(a.) Of polyps and terminal twigs. — Yellow, or red, or partly yellow, partly red,
covered by numerous small spines; Stutzbundel, 2"6 millims. X 0-17 milliin. ;
polyp spicules, 0-03 millim. X 0'04 millim. ; others, 1'5 millims. X O'l millim.,
17 millims. X 0'12 millim.
(b.) Of the stem, upper part, — Similar to those of (a.), but the majority are
colourless, or very faintly tinged with yellow. Spines more prominent
and appear rougher.
(c) Of the bare lower part of stem. — Shorter, thicker, and with more prominent
spines than (a.) and (b.). Spines more numerous and of greater
diameter than the above. The spicules of (c.) measure 076 millim. to
0-69 millim. X 0-16 millim., 0'42 millim. X 0-18 millim., 0-26 millim.
X 0-10 millim. Many of the spicules in (c.) are in the form of crosses.
In several an X -shaped marking is seen as if the 4 arms of the cross were
dovetailed into one another.
Another specimen similar in appearance to the above has its branching restricted to
one plane.

In its texture and flattened branches, &c, it agrees with the above. The tentacles

are white, but owing to their state of retraction they do not show so much. The base

of attachment shows the rhizoid outgrowths frequent in Spongodes and allied genera.

While these specimens do not rigidly correspond to the description given of

Spongodes rosea, the resemblance is too close to warrant separation.

Localities : — Deep water outside pearl banks, Gulf of Manaar ; and Trincomalee.

Spongodes armata, Holm, var. ceylonensis, n. — Plate I., fig. 6.
This foliate divaricate colony consists of a cylindrical barren trunk, attached at its
base by means of rhizoid outgrowths to sand and fragments of Mussel shells, and a much
branched upper portion which forms an irregularly oval-shaped head. Everywhere
spicules project, thus giving a spiny appearance to the head portion. The little-
violet coloured polyp heads contrast well with the orange coloured twigs.

AUYOXARIA. 283

Specimen A. Specimen B.

Length of trunk (bare part of stem) . 21 millims. 32 millims.

Diameter of same 10 ,, 13 „

Length of head 52 ,, 55 ,,

Maximum width of same 32 ,, 41 ,,

The trunk (or bare part of the stem) has a leathery rigid character and a rough
granular appearance. It is slightly thicker at the top, and at its lower end gives off
thin rhizoid-like outgrowths. In specimen B these are also given off for a short
distance up the trunk.

The stem gives off larger and smaller branches, and divides at the tip into 2 or 3
branches. It appears longitudinally streaked, owing to the disposition of the large
spicules. The lower branches are flattened (6 in one, 4 in the other). The others
rise at right angles to the stem, the upper ones being directed obliquely upwards.
Some of the lower ordinary branches are slightly flattened. The 4 to 6 flat branches
are recurved and directed downwards.

Locality : — Deep water off Galle.

Spongodes dendrophyta, Wright and Studer.

This species is represented by several specimens, one of which is evidently a young
form. The base of attachment consists of rhizoid-like outgrowths. The lowest
branches are flattened from above downwards, and bear the polyps on the edge of the
flattened portion. The appearance of the stem in the older specimens is quite
granular, but this gives place in the younger form to a slightly stringy appearance,
due to the spicules being larger. In the older specimens the polyps are more
markedly arranged in bundles, while they have a tendency to stand singly in the
younger. Several of the lower branches are colourless, and there is a great
variability in the colour of the spicules.

The geographical range of this species is worthy of note, as its previous record is
from the Philippines.

Locality : — Gulf of Manaar.

Spongodes splendens, Kukenthal — Plate I., fig. 2.

This species is represented by numerous stiff and rigid colonies. The following

measurements are taken from two of them : —

A. B.

Length of trunk 29 '5 millims. 17 millims.

Diameter of same 13*5 „ 12 ,,

Length of head 69 ,, 30 ,,

Maximum width of same .... 61 ,, 40 ,,

The stem in its lower portion is tough and rigid in character and granular in

2 o 2

284 CEYLON PEARL OYSTER REPORT.

appearance, which in the upper portion gives place to a streaked appearance due to
the larger spicules being disposed for the most part longitudinally.

The lowest branches, two in number, are flattened from above downwards, and are
bent downwards so as to hide the upper portion of the trunk. The other branches
are more or less cylindrical and rise at right angles to the stem, the uppermost being
directed obliquely upwards. The secondary branches give rise to the twigs which
bear umbels of 6 to 8 divergent stipitate polyps.

The lower portion of the stem has a yellowish -white colour, the upper portion of
the stem and the branches are yellow, while part of the twigs and the polyp spicules
are brick red to dull red in colour. Several of the spicules consist of a colourless
sheath surrounding a central core of a deep yellow to orange-yellow which extends for
about half the length of the spicule.

The spicules vary very much in the specimens. The " Stutzbt'tndel " spicules vary
in length from 2 "9 millims. in some to 2-2 millims. in others, and even to 1*5 millims.
The stem spicules vary from 0"75 millim. to 2 '3 millims. in length, but in one
specimen they reach 4*5 millims. in length. The polyp spicules also show considerable
variation, but it does not cover such a range as in the others.

Some other specimens correspond more closely than the above do to previous
descriptions of this species. It appears to us that many of the species of Spongodes
are exceedingly variable, and we have therefore refrained from accentuating minor
differences.

Locality : — -Cheval Paar ; pearl banks, Gulf of Manaar.

Sub-family : SIPHONOGORGIN^E.

Paranephthya pratti, n. sp. — Plate II., fig. 6 ; Plate V., fig. 18.

A peculiar specimen with only the basal attachment missing was included in the
collection. We are indebted to Miss Pratt for a figure and the following notes.

The colony, which is apparently young, is small and measures as follows : —

Height 21 millims. ; stalk 15 millims.

Breadth of stalk, 13 millims. to 11 millims., of capitulum, 24 millims. to 20 millims.

The capitulum is lobate, with deep grooves between the digitate lobes.

The lobes (7 in number) are digitate and are comparatively large. The largest is
13 millims. in height by 7 millims. to 5 millims. in diameter. The smallest is
3 millims. in height by 5 millims. to 4 millims. in breadth. Some of the lobes are
beginning to branch dichotomously.

The capitula are dotted with minute tubercles, and when observed under the
mici'oscope have a curious polyzoon-like appearance. These tubercles indicate the
presence of the zooids which are small, degenerate in character, and are so overgrown
with minute spicules as to be only slightly contractile. They have apparently only
rudiments of tentacles. The mesenteries are crowded with brown cells which are

ALCYONARIA. 285

probably zoochlorellse. In many cases there appears to be no aperture to the
exterior.

The colony is extremely hard, the surface being rough and granular to the touch.
The colour of the spirit specimen is medium brownish-grey, which becomes deeper
towards the base of the stalk.

There is no doubt that this interesting form is near Paranephthya capitulifera,
Wright and Studer ; but it is a quite distinct species.

Chironephthya variabilis, Hickson.

This species is represented by a magnificent specimen, measuring 21 centims. by
20 centims. The base measures 3 "5 centims. by 4 centims. There is a very marked
" weeping willow " appearance, as almost all the twigs are pointing downwards. This
is probably the finest specimen of its kind that has been hitherto obtained. There
are also numerous fragments, and one or two specimens which may be complete young-
forms.

The colour is very variable in the different specimens and even in the different
parts of the same specimen. In several fragments the general ccenenchyma is white
with red spots here and there on the surface ; in others it is yellowish-white with red
spots. When the general coenenchyma is white in colour, a pinkish tinge becomes
gradually more marked in the branches until it finally predominates, thus giving the
tips of the branches and twigs a deep pinkish-red colour. In the specimens with a
yellowish-white general coenenchyma the pinkish tinge increases as before, but in this
case the tips of the branches have an orange colour. In other cases the tips of the
branches may be pinkish-red with the projecting edges of the calyces an orange-
yellow.

There is also great variation in the colour of the polyp spicules ; in some the spicules
of the crown and points are red, in others they are yellow, while in others the crown
spicules are red and the point spicules orange-yellow. Considerable variation is also
found in the measurements of the crown and the point spicules ; so great is the range
of variation, in fact, that they do not appear to be of any value as a specific
distinction. The crown spicules vary in one specimen from 0"24 millim. to 0'5 millim.
in length, and the point spicules from 0*3 millim. to 0"55 millim. in length, while in
the same specimen the breadth of the crown spicules varies from 0-026 millim. to
0-06 millim.

Locality : — Deep water off Galle ; Pearl banks, Gulf of Manaar.

Siphonogorgia pustulosa, Wright and Studer.
Locality : — Ceylon seas.

Siphonogorgia miniacea, Kukenthal.

Locality : — Ceylon seas.

286 CEYLON PEAEL OYSTER REPORT.

■
Siphonogorgia kollikeri, Wright and Studer — Plate I., fig. 7.

Locality : — Deep water off Galle.

In the examination of the specimens in the collection belonging to the genera
Chironephthya and Siphonogorgia the following distinction suggested by Hickson
(" Fauna of Maldives, 1900") has been adopted : "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 anthocodise rarely completely retracted in preserved
specimens and with four principal spicules arranged en chevron in the points of the
anthocodise ; and that the name Siphonogorgia be retained for species or facies of more
massive Gorgonia-Y\k.Q form of growth, with anthocodise capable of complete retraction
within the general ccenenchym and with spicules irregularly placed or arranged in a
fan-like manner in the points of the anthocodise." While adopting this suggestion
we have been impressed by the great similarity of the two genera. There are many
reasons for thinking that they ought to be merged in one.

Order III. : PSEUD AXONIA.

Family: BRIAREIDjE.
Solenocaulon tortuosum, Gray.

The collection includes magnificent specimens of a brilliant red Solenocaulon,
probably a variety of S. tortuosum. We have studied a representative piece about
8 inches in length.

The substantial axis, measuring 5 millims. in diameter at the base of the specimen,
is composed of closely packed colourless spinose needles, all of the same type, varying
from 0'4 millim. to 1 millim. in length by 0"016 millim. to 0'018 millim. in breadth,
and thus much longer than those measured by Hickson. When boiled with caustic
potash the axis remains coherent, but crumbles readily. In a section 20 nutritive
canals are seen between the ccenenchyma, and the axis across the base of the specimen.

Towards the top of the stem the axis gradually decreases in size and finally
disappears. The rest of the stem forms a continuous tube with large lateral openings,
the margins of which are prolonged into the slender polyp-bearing branches. These
are mainly directed away from the side bearing the axis ; they often measure
17 millims. in length by 2 millims. in breadth.

The ccenenchyma of the tubular stem contains (a.) colourless needles from 0'4 millim.
to 07 millim. in length by 0"018 millim. in width, with few and short spines;
(b.) shorter and broader rods (0"28 millim. to 0"4 millim. in length by 0'06 millim. to
0"07 millim. in width), both coloured and colourless, thickly beset with rough wart-
like projections ; and (c.) coloured somewhat irregular ovals (0"1 millim. to 0"12 millim.
by 0-06 millim. to 0"07 millim.), covered with blunt tubercles. The coloured spicules
are pale yellowish-red ; and an orange variety occurs among the other brilliant red
specimens.

ALCYONARIA. L>87

The polyps are white, and are for the most part restricted to the twigs. A few
occur on the stem on the side away from the axis. The polyp-spicules are arranged
in a crown with slightly projecting triangular points. Each point consists of about
7 spicules, the longest median one being supported by three on each side. Below
these running transversely are the rows of crown spicules. All the spicules are
colourless warty spindles, varying in length from 0-2 millim. to 0'6 millim., and in
breadth from 0'02 millim. to 0*08 millim. The verruca? are truncated cones, with a
height of I millim. to 2 millims., and a basal diameter of 2 millims. A small crab
was found in the lumen of the tubular stem.

Localities : — Station LIX., 6 to 9 fathoms ; deep water outside banks, Gulf ot
Manaar ; Periya Paar, 9 fathoms ; deep water off Galle.

Keroeides gracilis, Whitelegge.

This species is represented by a complete colony and several fragments. The
complete specimen stands 85 millims. in height, with a maximum width of 56 millims.
The verruca? alternate on the sides of the stem and branches. They are low and
conical in shape, 1 millim. in height and in basal diameter. The ccenenchyma is thin,
smooth, and filled with large closely packed compound-tuberculated spicules.

The irregularity of the spicules is fairly well marked and agrees with Whitelegge's
description. By transmitted light they are of a bright red brick colour.

The spicules are on the whole slightly larger than those of Whitelegge's specimen.
The measurements obtained were: — (1) For the spindles, 1'4 millims. to 2 '2 millims.
in length by 0-2 millim. to 0'5 millim. in width; (2) Irregularly shaped spicules of
verruca? up to 1"6 millims. in length by 0'5 millim. in width ; (3) Spicules of the axis
from 0-3 millim. to 0"6 millim. in length.

With the exception of the slightly larger measurements of the spicules, the
specimen approximates very closely to that described by Whitelegge (1897), and
also to that described by Miss Hiles (1899).

Locality : — Deep water off Galle, Ceylon.

Previously recorded from : — Milne Bay, British New Guinea, depth 20 fathoms, and
Funafuti, Ellice Islands.

Family : SCLEKOGOKGHLE.
Suberogorgia verriculata, Esper.

Several incomplete specimens, mostly about 7 ceutims. by 5 centims., forming
networks in one plane. One colony with its base complete has a height of 9 centims.
In a portion measuring 5 "5 centims. by 3 '5 centims., there were 22 complete meshes
and 11 incomplete meshes. The meshes differ widely in size, but 7 millims. by
5 millims. is a size which occurs very frequently. The average thickness of branch
is 1*5 millims. The colour is brown or drab, and the general texture of the
ccenenchyma is granular.

The verruca? are thickly crowded and occur all round the branches. They project

288 CEYLON PEARL OYSTER REPORT.

very slightly and have a diameter of 1 millim. They are often separated about
1 millim. from one another. The verruca is a truncated cone, in some cases like a
widely open crater, in other cases with eight spiculated lobes projecting inwards like
small sepals in an almost open bud. A low-power view shows that there are eight
opercular coverings formed from spindle-shaped spicules.

Careful dissection of the included polyp, under low power, shows a cylindrical body
with eight vertical external muscle-strands and with a low conical apex. Around the
base of this apex there seems to be a ring of horizontally disposed spindles in two
or three rows, and from this there arise the eight triangular converging opercular
coverings, with spicules more or less at right angles to those of the basal ring. But
the specimens were extremely brittle, and the gentlest touch of the dissecting
needle almost invariably shivered the microscopic architecture.

The extremely calcareous sclerogorgic axis is densely packed with more or less
spindle-like spicules embedded in the matrix of horny fibres. It varies from 1 millim.
to 2 millims. in diameter.

The ccenenchyma spicules are mostly minute spindles with numerous regularly
arranged blunt tubercles; a common measurement is 0-l millim. to 0"125 millim. in
length by about 0'025 millim. in maximum thickness. There are also minute warty
double clubs or capstans, with a very short middle zone and broad disc-like ends ;
they measure about 0-025 millim. by 0*02 millim. Some of the spindles have very
few tubercles.

We have given these details to supplement the previous descriptions of Suberogorgia
verriculata, to which, we believe, these specimens must be referred.

Locality : — Deep water off Galle.

Another specimen is 281 millims. in height and 134 millims. in maximum breadth,
but this does not represent the full size of the specimen, as a part has been
broken off the main stem. The branching is confined to one plane, and there is a
very complete anastomosis of the branches, thus giving rise to a fine reticulate fan-
shaped colony. The chief branches rise at an acute angle and are directed upwards,
and assume a position roughly parallel to the main stem. The tips of the branches
are clavate.

The polyps come off from all sides of the stem and branches and appear to be
spirally arranged in some parts. They are completely retractile, and in the specimen
they are all withdrawn within the small verrucse.

The spicules are small spindles with prominent rough warts, larger spindles and
minute double wheels or capstans. The measurements length by breadth in milli-
metres are as follows : —

Capstans, 01)37 X 0-02 ; 0'035 X 0"02 ; 0"03 X 0"02.

Slender spindles with few warts, 0-09 X 0-018 ; 0"15 X 0"02 ; 0-10 X 0"03.

Other spindles, 0"10 X 0"03 ; 0"12 X 0"037 ; 0-16 X 0045.

The axis measures 6 '4 millims. in diameter at the base.

ALCYONAKIA. 289

We were at first inclined bo record this specimen as a new species of Suberogorgia,
but further study lias convinced us that it is only a variety of S. verriculata.

[Note. — Family : CORALLIID^E. Ridley reported with some hesitation the alleged occurrence of a
specimen of GoraUium rubrum on the Ceylon shore. It is interesting, therefore, to record that Professor
HERDMAN found in deep water (34 fathoms) oft' Guile a specimen which he named Corallivm (?). It was
scarlet in colour, with yellow polyps, and measured about an inch in length. The sketch in his note-book
suggests a young specimen of the noble coral, but the unsatisfactory point is that the specimen has not
been found in the collection.]

Order IV. : AXIFERA.
Family : ITJMNOID/E.
Primnoa ellisii, von Koch.

This beautiful species is represented by two specimens, the larger of which
measures 08 millims. in height by 39 millims. in maximum width. The specimens
agree with the description given by von Koch (1890), and the measurements obtained
are in absolute accord with those given by him.

The geographical distribution of the species is interesting, ranging as it does from
the Mediterranean to the Andaman Isles. The exact localities from which it has
been recorded are : The Gulf of Naples and the coast of Sardinia in the Mediter-
ranean ; deep water off Galle, Ceylon ; and the Andaman Sea, 275 to 45 fathoms
(Alcock's " Investigator " Collection).

Acanthogorgia media, n. sp.— Plate II., fig. 8 ; Plate III., fig. 7.

The specimens are similar to A. ridleyi in mode of branching and in the arrange-
ment of the polyps, but they are like A. muricata, Hiles, in the disposition of the
spicules on the calyces. But they do not seem referable to either species, or to any
other whose description is known to us. In the figure on Plate III. the gracefulness
of the colony has been inadequately represented.

The dimensions of the polyps are as follows : —

Heiirht 1*3 millims. 1"1 millims. 0*8 millim.

Diameter at base . . 0"95 millim. 0'95 millim. 07 ,,

„ apex . . 0-8 „ 0"8 „ 013 „

The polyps and the ccenenchyma of the stem and its branches have many projecting
spicules. Round the apex of the polyp a number of spicules project. The polyps
are arranged in some parts opposite one another, in other parts spirally. The
arrangement is complicated by the interpolation of smaller polyps between the older
ones. The branching is irregular and not confined to one plane. On the whole,
the branching is alternate. The distances between branches varies from 8 millims. to
2 millims., and even to 1 millim. The axis measures U'35 millim. in its thickest part,

2 P

290 CEYLON PEARL OYSTER REPORT. .

is brownish -yellow in older, yellow in younger portions, and is chambered. The
spicules are long slender spindles, with few warts, and tri- and quadri-radiate stars.
Among the spicules a considerable number are seen with an abrupt bend and one arm
long in comparison to the other (golf-club-like). In these there is a poor development
of the warts, the long shaft being free from the warty protuberances for a considerable
portion of its length, while the short head has a good number of rough warts.

The spicules measure, length by breadth, in millimetres : —

Projecting opercular spicules, 0*4 X 0-05 ; 0-45 X 0"04 ; 0'4 X 0'045.

Ordinary spindles, 0"5 X 0'036 ; 0"3 X 0"04 ; 0-6 X 0'045.

Quadri-radiate forms, 0-25 X 0'15 ; 0"25 X 0'02; arms about 0'1 millim. in length.

Tri-radiate forms, 0'3 X 0'25 ; the arms being about 0-12 millim. in length. Some
apparently sex-radiate and quinque-radiate forms occur.

Locality : — Trincomalee.

Acanthogorgia muricata, Verrill, var. indica, n. — Plate IV., fig. 11.

As a variety under this species we rank a beautiful complete colony, which has a
height of 78 millims. and a maximum width of 105 millims.

The basal attachment is a flat spreading portion from which the main stem rises.

The axis is dark brown in colour, fading to a light yellow in the younger part. It
is covered with close set furrows at the base, and at that point has a diameter of
3 millims., but it becomes extremely thin at the ends of the branches. The branching
is very profuse and is confined to one plane, but the type of branching gives the main
stem a slightly zig-zag arrangement.

The calyces are densely crowded all round the stem and branches. They measure
from 2 "5 millims. to 3 minims, in heiffht and have a basal diameter of 0"9 millim.
and a width of 1"1 millims. to 1'2 millims. at the crown. At the free end of the
calyces there are 8 projecting spicules. The ccenenchyma on the stem and branches
is very thin.

The specimen comes near Acanthogorgia spinosa, but differs from it in having only
8 single projecting spicules, while A. spinosa has 8 groups of 2 or 3. The spicules
also differ slightly. It agrees closely with A. muricata except for slight differences
in the polyps and spicules. The species A. muricata illustrates wide distribution ; it
has been previously recorded from Barbados and Funafuti. See Hiles (1899), p. 48.

Locality : — Trincomalee.

Acanthogorgia ceylonensis, n. sp. — Plate IV., fig. G ; Plate V., fig. 12.

This species is represented by one specimen which measures 55 millims. in maximum
height and 30 millims. in maximum width. The branching is both alternate and
opposite and is confined to one plane. The polyps are arranged in threes or in twos
or in loose spirals. The verrucse vary in height from 0'G millim. to 07 millim. and
are placed at intervals of about 1 millim. The spicules are arranged on the verrucas

ALCYONARIA. 291

in 8 rows en chevron. The figure (Plate IV., fig. 6) is unfortunate in exaggerating
the thickness of the stem and branches.

The spicules are of two types : (1) spindles which vary in length from 0"5 millim.
to 0*6 millim. and have an average diameter of 0-04 millim. ; and (2) quadri-radiate forms
which measure length by breadth in millimetres as follows: 0-08 X 0-06, 0'13 X 0*12,
0-14 X 0-12. We have been unable to refer this specimen to any of the described
species.

Locality : — Trincomalee.

Astromuricea ramosa, n. sp. — Plate I., fig. 8 ; Plate V., fig. 17.

A beautiful deep crimson colony, 8 centims. in height by 4 centims. in maximum
breadth, but lacking its basal portion. Most of the numerous branches spread out
laterally, but a few project forwards and a few backwards. Most of the branches
arise at an acute angle and rapidly attain a vertical direction, giving off short
secondary branches, almost always to the outside. The almost uniform thickness is
2 minima, ; there is a slight swelling at the tips.

The surface of the coenenchyma is rough with stellate spicules ; the slightly
prominent verrucce (about 1 millim. in diameter) occur all round the axis ; from their
summits the yellowish polyps are projecting (about 0'5 millim. in diameter). The
horny axis is bright yellow, with apparent longitudinal striation ; it has a basal
diameter of 1 millim.

The crimson spicules are mostly irregular, warty stars, with 4, 5, 6, or more rays.
Some of the measurements taken are 0-2 millim. by 0-15 millim. ; 0-2 millim. by
0-2 millim. ; 0"3 millim. by (K3 millim. ; 0"3 millim. by 0'2 millim.

There are some irregular triangles, numerous discs, too irregular to be called
stellate, and some small double wheels.

A few straight or curved spindles of diverse dimensions occur, 0"3 millim. by
0-03 millim. ; 0-4 millim. by 0'05 millim. ; 0'25 millim. by (VI millim. ; 0-2 millim. by
0-025 millim.

The species is nearly related to Astromuricea iheophilasi, Germanos, but is quite
distinct. There is superficial resemblance to Eehinogorgia pseudosamppo.

Locality : — Off Patani (Annandale and Robinson).

Echinomuricea indo-malaccensis, Bjdley.

This species is represented by a single specimen, from the upper part of which the
coenenchyma has been rubbed off'. The specimen measures 57 millims. in total
height.

The basal attachment consists of a flat, spreading portion, from which the stem
arises. The axis is dark in colour in the older parts, but becomes lighter in the
younger branches. Branches arise from all sides of the larger branches, but the

2 P 2

292 CEYLON PEARL OYSTER REPORT.

primary branches arise from the stem in one plane. The arrangement of the branches
gives the colony a bushy appearance.

The specimen agrees closely with the description of JEchinomuricea indomalaccensis
given by Ridley (1884). There are several differences, however, thus (l) the verrucae
in this specimen are fully twice as long as in Ridley's, they measure 0'8 millim.
instead of 0-3 millim. in height ; (2) the verrucae spicules with ramifying basal portion
vary in this case from 07 millim. to 0'9 millim. in length instead of a maximum
length of 0"65, as given by Ridley. For discussion of four known species, see
Hedlund (1890), pp. 14 and 15.

Locality : — Pearl banks. Gulf of Manaar.

Previously recorded from Port Curtis, 5 fathoms to 1 1 fathoms ; Port Molle.
12 fathoms to 20 fathoms ; Warrior Reef, Torres Straits, Australia.

Echinomuricea ceylonensis, n. sp. — Plate VI., fig. G.

Colony much branched in one plane, 1 1 centims. in height by the same in breadth
at its broadest part. The colour is reddish-chocolate to crimson. The polyps may
occur all round the branches, but in one specimen they are absent from the concave
side of the plane of ramification. There is no regularity in the manner in which the
branches arise from the stem. The verruca?, which project almost at right angles,
are about 1 millim. in height. The whole surface of the verrucse and of the
branches is rough with the sharp projecting points of crimson spicules, which are
sometimes imbricating. Within the verruca cup there is a distinct special cone of
tapering spicules forming an opercular covering. The polyps are entirely retracted.
The horny axis is almost black at the base, about 1 millim. in breadth ; it tapers
to a breadth of 0"35 millim. near the tips of the twigs, where it shows a horny
yellow colour. The ccenenchyma is relatively thin. The following types of spicule
occur : —

(a) Numerous roughly triradiate forms, with a usually tuberculate main spine and
with irregular foliaceous and ramifying expansions connecting the two other rays ;
0'3 millim. to 0"5 millim. in length by 0"275 millim. to 0"3 millim. in breadth at base
are common measurements ; (b) irregular approximately stellate forms ; and (c) curved
spindles and boomerangs bearing a few tubercles and spines and sometimes forked or
irregularly trifid at one end.

Locality : — West of Periya Paar.

Echinogorgia pseudosassapo, Kolliker.

A single specimen of a crimson-red colour. It rises from a spreading base, about
15 millims. by 5 millims., and consists of two branches, about 50 millims. and
20 millims. in height. The longer branch gives off a branch at its base and another
a third of the way up ; the shorter branch gives off a branch near its end. The
diameter varies from 2 millims. to 3 millims. It is probably a young specimen.

The surface is very rough owing to the sharp projecting points of the large spicules.

ALCYONARIA. 293

There are practically no verrucae. The polyps appear yellowish, and the mouth is
seen in most cases as a very precise oval aperture.
The red spicules are very variable : —

(1.) Fusiform, pointed at both ends, with distant tubercles; common lengths are
0'2 millim., 0*3 millim., 0*4 milium

(2.) Larger fusiform, with tubercles and a few teeth, up to 07 millim.

(3.) Large and variable " Blattkeulen," often roughly triangular, with a shaft often
divided into tubercled branches and a foliar expansion with 2 to 5 teeth ; some
measure 0*6 millim. in length.

(4.) Irregular stellate forms, with transitions to irregular discs ; 0-2 millim. by
0"3 millim. is a common size.

(5.) Triradiate to hexradiate forms, with transitions to the stellate type.

The spiculation is nut quite in agreement with the description and figures given by
Wmght and Studer, but the differences do not seem important.

Echinogorgia inultispinosa, n. sp. — Plate VI., fig. 1.

An imperfect colony, branching irregularly in one plane without anastomosis,
11 centims. in height by 6 centims. in breadth. The horny axis measures 3 '5 millims.
in diameter at the base. Some of the twigs bear galls. The surface of the
ccenenchyma is finely granular and of a cream- white colour. Verrucas occur mostly
on the sides of the branches, and are almost quite absent from the convex surface of
the plane of ramification. In some regions they are crowded, almost touching one
another, in other parts they are separated by intervals varying from 2 millims. to
4 millims. A few spines jDroject slightly round the margin of the verruca opening,
and there is an opercular covering with small spicules arranged in eight rays. Under
low power the surface is seen to be uniformly covered with what seem to be
tubereulate spindles irregularly arranged. When these spindles are examined an
extraordinary diversity of form is disclosed.

(l.) Large tubereulate or spiny spindles, straight or curved, 04 millim. in length
by 0'1 millim. in maximum breadth ; some show a prominent spine on one side.

(2.) Small tubereulate spindles, 0"2 millim. by 0"05 millim.

(3.) Small almost smooth spindles, 0-2 millim. in length by 0-05 millim. ; 0'3 millim.
by 0-05 millim. ; 0"275 millim. by 0-04 millim.

(4.) Slender boomerangs, 0"35 millim. by 0-05 millim.

(5.) Irregular tubereulate and denticulate clubs, 0-4 millim. by 0-2 millim. ;
0"45 millim. by 0"15 millim.

(G.) Foliate clubs, 0-4 millim. in length by 0-2 millim. at broadest part; 0"3 millim.
by 0*2 millim. ; 0-275 millim. by 0'2 millim.

(7.) Irregularly-shaped warty scales with denticulate margins, e.g., 0-3 millim. by

294 CEYLON PEARL OYSTER REPORT.

0"2 millim. ; 0-25 millim. by 0*2 millim. ; 0'2 millim. by 0*2 millim. ; 0"25 millini. by
0"15 millim.

(8.) Almost stellate scales, 0'25 millim. by 0-25 millim. ; 0*15 millim. by 0'15 millim.

(9.) Warty triradiate forms, 0'25 millim. by 0'2 millim. ; 0-2 millim. by 0*13 millim. ;
0*2 millim. by 0"2 millim.

(10.) Tetraradiate forms, 0'23 millim. by 0*1 millim. ; 0-25 millim. by 0'15 millim.

Tbe variety of spicides is greater than we have seen in any other Alcyonarian.

Heterogorgia verrilli, n. sp. — Plate VI., fig. 2.

A handsome branched colony, greyish-black in colour, 20 centims. in height by
8 centims. in maximum breadth. The short common stem, from which two main
branches arise, is 5 millims. in diameter. The branches are generally alternate, and
there is a tendency to swelling at the tips of the twigs. The surface of the
ccenenchyma appears granular and bears many sponge spicules. The verrucse are
usually, but not constantly, disposed on the sides of the branches ; they are conical
warts under 1 millim. in height. In some cases the upper part of the anthocodia
protrudes from the cone, and is seen to bear spindle-shaped spicules. In the retracted
state a whorl of spindles is seen projecting vertically outwards within the margin of
the crater-like verruca. The characteristic spicules of the ccenenchyma are more or
less regular, bluntly tuberculate crosses of varying dimensions, e.g., O'l millim. by
O'l millim. ; 0"2 millim. by 0-2 millim. There are also tuberculate double clubs and
rough discs of various sizes, like crosses with the arms scarcely discernible.

The specimen, apparently a Muriceid, is very difficult to locate ; we have referred
it to Verrill's genus Heterogorgia (' Am. Journ. Sci.,' xlv. (1868), p. 413).

Bebryce hicksoni, n. sp. — Plate III., fig. 1 ; Plate VI., fig. 9.

An imperfect specimen, irregularly branched in one plane, about G"5 centims.
in height by 6"5 centims. in breadth. There are four branches, measuring about
2 millims. to 2 "5 millims. in diameter, which is thicker than the main stem at its
base (l-5 millims.). There is distinct flattening of the branches and twigs. The cortex
has a smoothly granular punctate appearance and a greyish-white colour. The
verrucse, which are disposed mainly on the sides of the branches and twigs, usually
with intervals of 2 millims. to 2*5 millims. separating an adjacent pair on the same
side, are truncate rounded cones, crater-like when the polyps are dislodged. In many
cases the anthocodise of the polyp project at the apex of the cone, and are supported
hy triangular strands of small spicules longitudinally and transversely disposed.
Under low power the superficial spicules appear like rounded grains of sand or
like fish-roe, each grain appearing as if it had a clear nucleus surrounded by
a whitish rim. Higher magnification shows that there are multi-lobate warty
spheroidal discs, some of which show a short central toss projecting inwards ; the

ALC YON ARIA. 2 9 5

diameter of the disc varies considerably, O'l millini., 0'15 millini., 0"2 millim.,
and 0"25 millim.

In some cases the boss is borne on a distinct stalk, and this type measures
0-25 millim. in height bv 0-2 millim. in the diameter of the disc Some small spicules
are capstan-like, double clubs in fact, and various transitional forms show a gradual
reduction of onedialf of the capstan till only a boss is left, or not even that.

In the polyps there are curved tuberculate spindles, 0'25 millim. in length.

The spicules are so characteristic that we have no hesitation in referring this
specimen to the genus Bebryce, and it cannot be identified with B. mollis, von Koch
(1887), or B. studeri, Whitelegge (1897), or B. philippii, Studer. Hickson has
reported Bebryce, sp. (?) from the Maldives ; B. mollis is Mediterranean ; B. studeri was
obtained at Funafuti ; B. philippii from the Arafura Sea — a remarkable distribution.

This seems to be a somewhat variable species, and the five specimens obtained
differ considerably in general appearance, in the degree of prominence assumed by the
verruca?, in their distance apart, and in the shapes and sizes of the spicules. In all
cases, however, the characteristic Bebryce spicules are unmistakable.

In a fragment of a greyish colour, 23 millims. in length, with two short branches,
the verruca? are usually about 1 millim. apart, along one line, and seem to occur
equally all round the branches. Their diameter is 1 millim., and that of the light
brown non-calcareous axis is the same.

We made a study of two other specimens. The more complete of the two
represents a young colony, the other is a fragment of a large colony. In both the
branching is confined to one plane, the branches and twigs being given off at an
angle which varies little from a right angle. The branches grow out for a short
distance in this direction and then bend upwards.

The tips of the twigs and the branches are clavate, but in one or two cases they
are flattened. The two specimens present a greenish appearance, due to the presence
of a siliceous sponge on the branches. The sponge has covered each branch with a
thick coating, through which the tips of the verruca? are seen. The spicules of the
sponge are monaxonial and are oxytylotes.

The general ccenenchyma is thin and granular in appearance, and is composed of
small spherical or irregularly oval bodies covered with numerous rough warts. There
are also some more definite spindles, but they are very few in number.

The axis is horny and black in the lower parts, becoming lighter in colour in the
twigs and younger portions of the branches. The axis is marked by irregular ridges
and has a core of lighter coloured material. This core diminishes in size with the
age of the portion of the axis. It is crossed by a number of bands, thus appearing
to be chambered, and a longitudinal section shows a number of curved lamella?
crossing the core at about equal distances.

The verruca? occur in the younger portions on all sides of the twigs and branches,
but in the older portions they are confined more or less to the lateral faces. They

•296 CEYLON PEARL OYSTER REPORT.

are small and seem to be conical in shape. After carefully cutting away the sponge,
we found that the verrucas were conical and truncated.

There is no differentiation between the spicules of the general ccenenchyma and
those forming the verruca? ; in both cases they are mostly oval or spherical discs
covered with rough warts and sometimes with an internal boss. There are also some
rough warty spindles. The spicules measure, length by breadth in millimetres, as
follows : —

(1.) Globular discs, OT X OT ; 014 X 0-14 ; 0T4 X 0-08 ; 0T1 X 0-105.
(2.) Spindles, 0-26 X 0'038 ; 0T2 X 0'025 ; 0'16 X 0"03.

The colour of the colony when devoid of the sponge is a greyish-white.

Acamptogorgia atra, n. sp.

This species is represented by a complete colony, measuring 211 millims. in height
and 98 millims. in maximum width.

The axis is horny, almost black at the base, but light brown in the younger
branches. The stem is almost of the same diameter throughout its whole length.
It measures 4 millims. in diameter at the base and 3 '9 millims. at the tip of the main
branch. The branching is confined to one plane. The branches arise at a right
angle, but soon turn upwards and run roughly parallel with the main stem.

The polyps are restricted to the lateral edges of the stem and branches, but here
and there they appear either on the back or on the front. They stand out almost
perpendicularly to the stem and branches at intervals of about 2 millims. Each
branch bears close to its tip 3 or 4 polyps, none of which can be said to occupy the
end of the branch.

The colony is black in colour, though it presents a greyish appearance owing to its
being closely coated with a sponge.

Locality : — Ceylon seas.

Acamptogorgia spinosa, Hiles.

This species is represented by several colonies, two of which give the following
measurements : —

(A.) (B.)

Height of colony 76 millims. 74 millims.

Maximum width of same 39 ,, 42 ,,

A finer, but much weathered specimen afterwards found in the collection measured
15 centims. by 15 centims. ; it was branched in one plane and showed much anastomosis.

The polyps on the branches vary from 0-6 millim. to 07 millim. in length and from
0-5 millim. to 0'G millim. in diameter. In a few of the polyps a height of 0"9 millim.
may be reached. The spicules in the present specimens are slightly larger than in

ALCYONARIA. 297

Acamptogorgia spinosa, but even then the relation of length to breadth is almost the
same. The specimens agree with Acamptogorgia spinosa in axis, colour, and thick-
ness, and in the chambered central core. For description, see Hiles (1899).
Locality : — Deep water off Galle.

Acamptogorgia spinosa, var. ceylonensis, n. — Plate I., tig. 1.

This variety is represented by one complete colony and a colony in which the base
is wanting. The complete colony has the following measurements : — Maximum
height 114 millims. and maximum breadth 176 millims., while the incomplete one
measures 154 millims. in maximum height and 142 millims. in maximum width. The
colony is fan-shaped, caused by the profuse branching which is strictly confined to
one plane. Anastomosis of the branches is, however, common.

The stem is dark in colour near the base, but gradually becomes lighter in the
younger parts of the stem and branches. The axis in the stem and principal
branches is oval in shape, the longest diameter being perpendicular to the plane of
branching, but it gradually becomes more cylindrical in shape in the younger
branches and in the younger part of the stem.

The ccenenchyma is thin and allows the axis to shine through ; it presents a very
rough appearance under the lens.

The verrucas are arranged on three sides of the axis, they have a series of spines
projecting round the mouth, and have a maximum height of 0'8 millim. and a
maximum diameter of 0*6 millim.

A few expanded polyps were found at the end of a branch, but almost all were
withdrawn within the verrucre and showed a tentacular operculum, conical in shape,
formed by the spindle-shaped spicules which lie on the aboral surface of the
tentacles.

The spicules are (1) triradiate, with large irregular processes in the angles which
often fuse, thus leaving plate-like portions with perforations; (2) a modified form of
the triradiate with a long arm, which gradually thickens and then gives off two
smaller arms from the thicker end ; and (3) spindles and club-shaped half-spindles,
curved or straight, either with very rough projections or fairly smooth.

Locality : — Deep water off Galle.

Acis indica, n. sp. — Plate II., fig. 3 ; Plate V., fig. 7.

The colony is large and fan-shaped, rising to a height of 149 millims. and having a
maximum width of 167 millims. From a conical base, which has a fiat spreading
margin and is attached to a mass of worm-tubes, the short main stem arises. At a
distance of 14 millims. from its origin, where it has a diameter of 3 '5 millims., it
divides into two principal branches. The branching is for the most part confined to
one plane, but several of the smaller branches and twigs arise at right angles to the

2 Q

■298 CEYLON PEARL OYSTER REPORT.

principal plane of branching. The branching is very profuse and at several points
shows anastomosis of the branches, but this is by no means common. The branches
are cylindrical, but there are traces of slight flattening in the plane of branching.
The twigs arise usually at right angles to the branches, and their tips as well as those
of the branches are slightly clavate.

The polyps are small and are scattered over the whole surface of the stem and
branches. In no place can it be said that they are confined to three surfaces, nor can
any attempt at lateral arrangement be seen. The verrucae are very small and the
polyps can be completely retracted within them. The edges of the verrucaa show a
variable number of spines which project above the slightly conical operculum formed
by the tentacular spicules when the polyps are withdrawn.

The superficial cuenenchyma of the stem and the branches presents a striking
appearance, due to the arrangement of the large flat whitish spicules and to their
being outlined against the darker ground-colour of the stem and branches.

The spicules of the general ccenenchyma are flat and multituberculate, varying very
much in size and shape. The tubercles are low and rough and very numerous.
Many of the larger spicules extend the whole distance between two adjacent polyps,
and sometimes even exceed this length. They fall into three groups, fairly distinct
in shape : (<() large modified fusiform spicules, which taper more or less towards the
ends and measure from 0*9 millim. to 3 millims. in length by 0'25 millim. to
0"45 millim. in breadth ; (1>) squamous or scale-like spicules, often with slightly lobed
margins, which measure from 0-8 millim. to I'l millims. in length by 0"4 millim. to
0-6 millim. in breadth ; and (c) large modified squamous spicules, consisting of a
flattened tuberculate basal portion and of a projecting part which forms the projecting
spine of the verrucae. They measure, in length by breadth in millimetres, as
follows :— 07 X 0-5 ; 0-6 X 0*4 ; 0"5 X 0'3.

In the polyps there are slender spindle-shaped and club-shaped spicules. They are
often slightly curved and either taper to both ends or are blunt and rounded at one
end and pointed at the other. Many of these exhibit fairly prominent spines towards
the thicker end. They vary considerably in size, being from 0"3 millim. to 0-5 millim.
in length and from 0'02 millim. to 0-06 millim. in breadth. They are found chiefly in
the tentacles, where they form an operculum to the retracted polyp ; but an
incomplete and irregular crown or collar is formed by them at the base of the
tentacles.

In colour the spicules vary from white to semitransparent, while the whole colony
has a whitish-brown appearance.

This species differs from Acis pustulata in not having violet-coloured opercular
spicules and in the branches not being compressed in the plane of branching. It also
differs from Acis orientalis in having the polyps on all sides of the stem and branches
and in the branching not being confined to one plane.

Locality : — Deep water off Galle.

ALCYONAEIA. 299

Acis alba, n. sp.— Plate III, fig. 9 ; Plate V., fig. 4.

This species is represented by a complete colony which gives the following measure-
ments : — Height of colony, 127 millims. ; maximum width of same, 115 millims. ;
diameter of axis at base, 2"5 millims.

The basal attachment consists of a flat plate-like expansion from which the main
stem rises. The stem reaches a height of 26 millims. and then divides into two
principal branches, but branching is represented between this point and the base by a
stumjj. The branches are given off at an angle which closely approaches a right
angle ; they soon turn upwards, however, and run roughly parallel to the main
branches. The tips of the branches and the twigs are sometimes slightly clavate.
The colony presents a fan-shaped appearance, due to the branching being strictly
confined to one plane.

The axis is horny, greyish-black in colour in the older parts, and amber-yellow in
the twigs and younger portions of the colony. It is cylindrical and marked by a
number of irregular ridges in the lower part.

The polyps occur on all the faces of the stem and branches, and are arranged so
that they give the appearance of three or four irregular lines. The verrucae are
small, slightly conical in shape, with a blunt tip and fairly wide base. They are
covered by spicules similar to those of the ccenenchyma. In several cases a verruca
seems to occupy the tip of a branch or twig.

The general ccenenchyma is thin and granular in appearance, the surface appearing
scaly owing to the arrangement of the large flat spicules.

The spicules appear to be in two layers, the outer consisting of thick irregularly
shaped flattish spicules, closely covered with rough warts, while the inner layer
consists of spindles often with blunt ends and more irregularly formed spicules. The
measurements, length by breadth in millimetres, are as follows : —

(1.) The outer layer, 0"48 X 0"23 ; 0"5 X 04 ; 0-43 X 0'29 ; 0"45 X 0"2 ; 0"2 X O'l.

(2.) The inner layer, 0-25X0-07; 0-18X0-05; 0-22x0-08; 0'14 X 0"04 ; 0"12 X 0-04.

The appearance of two layers may be due to the fact that the flattish spicules are
thicker and so stand up above the level of the other spicules. If this be so, then the
spicules are in one layer only.

Towards the base of the tentacle on the aboral surface two slender converging
spicules are present and at their base one or more transversely placed, thus giving a
triangular arrangement of spicules at the base of each tentacle.

Galls are present on several of the branches, mostly caused by cirripedes.

The colour of the colony is greyish-white.

Localities : — Trincomalee ; deep water off Galle.

Acis ceylonensis, n. sp. — Plate VI., fig. 3.

A small specimen, 4 centims. in height, with five branches, three of which are
broken. The stem measures 2 millims. in diameter at the base and is scarcely

2 Q 2

300 CEYLON PEARL OYSTER REPORT.

reduced in thickness higher up. Verrucse occur all round and have the form of
truncated cones, about 0'5 millim. in height and about 1 millim. in diameter at the
base. The intervals between them vary greatly, from 2 millims. to almost nil. The
surface is covered with pebble-like spicules predominantly oval in shape and neatly
fitted together like a mosaic of rounded fragments, which show almost no trace of
overlapping or imbrication. Smaller oval scales occur outside the verructe and
spindles project very slightly from within.

The large ccenenchyma spicules are mostly warty irregular ovals, but some are
almost rectangular and others are like flint axe-heads.

They give the following measurements in length and breadth in millimetres : —
0-55 X 0-35; 0'5 X 0*32; 0*4 X 0\3.

Various kinds of spindles occur : ('«) large thick forms, very warty, measuring
0"3 millim. by 0'15 millim., 0'45 millim. by 0*175 millim. ; (/>) long thin curved or
straight forms, with smaller tubercles or almost smooth, measuring 0'5 millim. by
0'03 millim., 045 millim. by 0"03 millim. ; and (c) minute warty spindles, 0*06 millim.
by 0-02 millim.; 0"08 millim. by 0"02 millim.

Besides these forms there are others which approach the club-type, and are
probably modifications of the large spindles.

Locality : — Deep water off Galle.

Acis ceylonensis, n. sp., var. imbricata, n.

Another specimen, measuring 4 centims. in height, with only hints of branches,
showed a general superficial resemblance to Acis ceylonensis, n. sp., except that the
large ccenenchyma spicules were much less regular in form and were much imbricated.
An examination of the spicules showed the following forms : —

(1.) Large, modified fusiform, elaborately tuberculate, of very various dimensions,
eg., 0-5 millim. by 0*25 millim.; 0"65 millim. by 0"4 millim.; 1 millim. by 0"5 millim. ;
without the monticular prominences characteristic of Acis orientalis.

(2.) Large rounded or almost oblong tuberculate scales, of various dimensions, e.g.,
0*35 millim. by 0"4 millim. ; 0"5 millim. by 0'45 millim.

(3.) Tuberculate clubs, with the ends often much swollen, often 0'45 millim. in
length by 0"2 millim. in maximum width ; 0-35 millim. in length by 0-2 millim. in
maximum width.

(4.) Tuberculate spindles, about 0"35 millim. to 0-4 millim. in length by 0-l millim.
in breadth, some 0'25 millim. in length by 0-05 millim. in breadth.

This specimen obviously differs markedly in the details of its spiculation from that
which we have named Acis ceylonensis, but it may be sufficient to erect a variety
within the species. Until more data accumulate it seems quite impossible to tell how
far marked differences in spiculation can be relied on as specific.

Locality : — Trincomalee.

ALCYONARIA. 301

Acis orientalis, Ridley.

This species is represented by three complete specimens, of which the largest
measures 136 millims. in height by 86 millims. in maximum breadth. The specimens
agree with the description given by Ridley [1882 (II), pp. 126-128].

The terminal branches reach a length of 45 millims. to 50 millims., and the tips of
the branches and twigs are clavate, measuring 3-5 millims. to 4 millims. in diameter,
and are occupied by three polyps.

The axis measures 3 millims. at the base, but in branches which have a diameter of
2-6 millims. it measures only 0'5 millim. or 0-45 millim. in diameter.

The spicular measurements agree very closely with those given by Ridley (1882),
the differences being so slight as not to be of any importance. We may perhaps,
having found other Ceylonese species of Acts, quote Ridley's note : — " The occurrence
of this otherwise West-Indian genus in the Indian Ocean, and in its western portion
in particular, has a peculiar significance for the student of geology, as showing that
a communication probably existed between these two areas at a period later than
that at which the genus was differentiated from the main stem of the family."

Locality : — Deep water off Galle.

Previously recorded from Mauritius, 80 fathoms.

Muricella ramosa, n. sp. — Plate III., figs. 2 and 3.

The colony is branched in one plane, and forms a beautiful network with abundant
anastomoses. The branches and twigs are given off, sometimes opposite, sometimes
alternate, at an angle which varies little from a right angle. One of the specimens,
though not complete, measures 170 millims. in height and 273 millims. in maximum
width.

The axis is black in colour in the lower part, but becomes brownish-yellow in the
tips of the branches and in the twigs. It is rather irregular in shape, being some-
what oval shaped, the longer diameter being at one place in the plane of branching,
at another place at right angles thereto. The general ccenenchyma is thin and
practically composed of warty spindles which may be either straight, curved, or
slightly S-shaped.

The verrucse are small and vary greatly in appearance. A. number of spicules
horizontally arranged form the lower part of the verruca, and a number of longi-
tudinally disposed spicules form the upper part. The appearance of the upper set of
spicules varies according to the stage of retraction of the anthocodial part of the
polyp. The spicules composing the verruca? do not differ from those of the general
ccenenchyma except in size. They are smaller, but are as closely covered with warts
as the other spicules.

The polyps are arranged on the stem and branches in what appear to be short
spirals, but on some of the twigs they assume a more or less bilateral disposition.

The spicules are thick spindles, either straight or curved, or slightly S-shaped,

302 CEYLON PEART, OYSTER REPORT.

covered with numerous rough warts. Their measurements, length hv breadth in
millimetres, are as follows: — 1 "6 X 0"2; 0'8 X 0-14: 1"3 X 0-2; 0'9 x (P16 ;
0-25 X 0-06 ; (VI X 0"02 ; 2*1 X 0'26.

In the complete colony the outer spicules often stretch the u-holc distance between
the polyps. In certain conditions of the polyps there appear to be 8 projecting
spicules on each calyx.

Localities : — Deep water outside pearl banks, Gulf of Manaar ; and deep water oh1
Galle and onwards up West Coast of Ceylon.

Muricella ceylonensis, n. sp. — Plate VI., fig. 4.

The colony is branched in one plane, with a strong tendency to anastomosis even
between twigs of different main branches. It measures 12 centims. in height by
8 centims. in breadth. Most of the twigs arise at right angles from the main branches
and almost all the short ones remain perpendicular. Many arise in opposite pairs,
others are alternate, and in other regions the twigs may be restricted for an inch or
more to one side of the branch. A common interval between the origin of two
adjacent branches on the same side is 4 millims. The verrucse stand out prominently
at right angles to a height of about 0-5 millim., and the interval between two on the
same side is on the average 1 millim. The surface of the colony is covered with huge
white spicules readily visible to the naked eye. Spicules slightly curved at the base
run up the sides of the verruca, and their tips project at the mouth. The black
horny axis measures 2 millims. in diameter at the base, it tapers to a diameter of
0-l millim. at the tips of the twigs, and in these delicate parts it shows a brownish
tint. The general colour of the colony is grey. In another specimen measuring
8 centims. by 6 centims. the colour was whiter owing to the more densely packed
large spicules; there was no hint of anastomosis, and the perpendicular mode of
branching was not so pronounced.

The spicules, which are longitudinally disposed, are straight or curved, densely
tuberculate spindles, the largest of which seem here and there to be disposed in
definite strands or lines. They vary in size from 3 "5 millims. to 0'35 millim.

Compared with M. ramosa this species is much less thickly branched ; it shows
much less anastomosis, it bears polyps at wider intervals, and has larger spicules.

Muricella nitida, Vekrill.

This species is represented by a colony with a maximum height of 105 millims. and
a maximum width of 80 millims. The base consists of a broad flattened portion from
which the fan-shaped colony rises. The axis is dark brown, almost black in colour in
the lower parts, but it becomes yellowish in the upper portions of the branches and in
the twigs. It has a basal diameter of 2 millims.

The branches, which are confined to one plane, come off at an angle which
approaches a right angle and then bend upwards, so that they appear to be parallel

ALCVONAKIA. 393

to one another. The younger branches and twins often present a pinnate appearance,
which is sometimes disguised by a larger development of one of the branches. The
stem and branches are flattened in the plane of branching, and the tips of the twigs
form blunt broad expansions due to the two divergent terminal polyps.

The polyps are more or less contiued to the lateral edges of the branches; they are
all withdrawn, lint their tentacular portions are yellowish in colour.

The spicules vary in length from 0"9 millim. to 1"65 millims., and in breadth from
Oi millim. to 0'2 millim.

The colour of the colony is a deep coral-red ; and the spicules of the ccenenchyma
are visible to the naked eye.

The specimen presents close resemblances both to Muricella perramosa and
.1/. nitida, but because of the greater resemblance in the arrangement of the polyps
and the more marked coral-red colour, it was thought best to place it under
M. riit Ida.

Locality : — Ceylon seas.

Muricella coinplanata, Weight and Studer.

This species is represented by a fragment of a colony, consisting of the lower part
of the stem with basal attachment, and a branch bearing secondary branches or twigs.
The basal attachment consists of a slightly conical spreading portion, from the
summit of which the stem arises. The stem measures 2'2 millims. near the base, and
at a distance of 11 millims. from the base the first branch, which is represented by a
fragment, was given off. The secondary branches or twigs are given off at right
angles, and may reach a length of 25 millims. to 37 millims.

The polyps are arranged on the lateral edges of the stem and branches, and are
directed some to the front and some to the back. On the whole they are arranged
alternately, but this is not quite constant. The polyps measure 0"95 millim. in
height by 074 millim. in width, and are contained within verrucse which measure
0"35 millim. in height by 0"8 millim. to 1 millim. in width at the base.

The stem is of a violet tint with yellow polyps and verructe, but this tint gives
place gradually on the twigs to an orange-yellow. The tips of the twigs end in two
divergent polyps, and this expansion measures 3 millims. in width.

The spicules of the ccenenchyma vary in length from 07 millim. to 1'3 millims., and
in breadth from 0"1 millim. to 0*16 millim. They are straight spicules covered with
warts.

This specimen, though not agreeing with Muricella complanata in every detail,
may be included in this species, which is variable in the colour and the size of its
spicules.

We studied a second specimen of a more pronounced violet colour and with much
larger spicules. It was a complete colony which is attached to a piece of coral, and
measures 68 millims. in height and 53 millims. in maximum width. The basal

304 CEYLON PEARL OYSTER REPORT.

attachment consists of a flattened portion spreading over the surface of the coral.
The stem measures 2 millims. at the base, and at a height of L7 millims. the
first branch is given off. The branching is confined to one plane, the primary branches
being given off almost at a right angle. They in turn give off secondary branches or
twigs at an angle which approximates even more closely to a right angle.

The verrucae are placed on the lateral edges of the stem and branches, and are
arranged more or less alternately. They are conical and abruptly truncated, with
usually a few of the spicules projecting at the opening. The spicules in the calyces
are much smaller than those of the general coenenchyma.

The colour of the whole colony is a beautiful violet tint, with the tentacles of the
polyps yellowish in colour.

The spicules of the coenenchyma are spindle-shaped, either straight or curved and
sometimes broad and flat in comparison to their length, or bluntly rounded at the two
ends. They can be divided into two fairly distinct types : —

(1.) The larger spicules with rough warts, which vary from 07 millim. to 2'7 millims.
in length and from 0'075 millim. to 0'32 millim. in width ; and

(2.) The smaller spicules with warts more pointed than rough, which vary in length
from 0"11 millim. to 0"55 millim., and in width from 0"03 millim. to 0"05 millim.

Another very beautiful form with large violet and colourless spicules seems at first
sight very different from the foregoing species and variety. This is mainly due,
however, to the fact that there is a prominent opercular covering of small but very
conspicuous rose-red spicules sometimes disposed in eight rays.

Localities : — West of Periya Paar, Gulf of Manaar ; deep water oft' Galle and
onwards up West Coast of Ceylon .

Family : PLEXAURID.E.

Plexaura prselonga, var. typica (Ridley).

Plexauroides praelonga (Wright and Studeii).

This species is represented by one specimen, which reaches a height of 425 millims.
The branching is confined to one plane and forms an irregular dichotomy. The main
stem is 4 millims. in diameter at its base, tapering to 3 millims. at its termination,
thus agreeing with the measurements given by Wright and Studer.

The verruca; are numerous and irregularly scattered and the polyps are completely
retractile. The ccenenchyma is comparatively thin and very friable. The spicular
measurements are the same as those given by Wright and Studer. The colour is a
deep brownish-red in spirits.

Locality : — Outside pearl banks, Gulf of Manaar.

Previously recorded from Port Curtis, 5 fathoms to 1 1 fathoms ; Port Denison,
4 fathoms; and Station 186, Cape York, 8 fathoms. (See Wright and Studer
(1889), '"Challenger" Peports,' vol. xxxi. : Ridley (1884), '"Alert" Expedition,'
p. 339, Plate xx.wi., tig. F, and Plate xxxviii., fig. <j, (/'.)

ALCYONARIA. ,305

Plexaura praelonga, RiDLBY, var. elongata, n.

Several incomplete specimens, dull crimson in colour, with two or three elongated
cylindrical branches, 3 millims. to 4 millims. in diameter and up to 30 centims. in
length. The axis, blackish-brown, 1 millim. to 2 millims. in diameter, is entirely
horny. The surface of the colony is covered with the pore-like apertures of the
polyps ; around each aperture the spicules are disposed in a ring 0-3 millim. to
0-5 millim. in diameter.

The most characteristic spicules are very variable discs, " Blattkeulen," 0'4 millim.
to 0-5 millim. in length, with a rounded foliaceous expansion (0"3 millim. in breadth)
at one end and an irregular ramification at the other. The surface of the disc bears
a few large blunt tubercles. When the thin foliaceous expansions are broken oft1,
the rest of the spicule may be described as irregularly stellate. Actually stellate
forms also occur, and 3- to 6-rayed forms are represented. A few tuberculate spindles
occur here and there.

As regards types of spicules, this form resembles Plexaura prcelonga, but the
details of the " Blattkeulen," the rough cortex and the crimson colour, and some
other features, mark it as quite distinct from Ridley's varieties typiea and cinerea.

Plexaura antipathes, Klunzinger, var. flexuosa, n.

To a variety of this species we refer two specimens, the larger of which measures
120 millims. in height and the same in width. The colonies are bush-shaped and
very profusely branched, with the branching not confined to one plane. The branches
are sometimes thickened during a part of their length, while all are swollen at the tip
into a knob or club-shaped portion, which measures from 2 millims. to 3 millims. in
diameter.

The polyps are scattered over the whole surface of the stem and branches. No
projecting calyces are present, but the position of the polyp is marked by an oval or
circular opening, round which the coenenchyma is slightly raised.

The axis is thick at the base, measuring 3 '5 millims. in diameter, but it becomes
slender and thread-like at the tips of the branches. The colour is a dark brown, very
little difference being seen throughout the whole ramifications of stem and branches.
The base consists of a broad, flat, spreading portion, from which in one specimen
several stems rise in addition to the main stem.

The spicules are short, many branched, rod-like needles or small spherical bodies,
which bear many protuberances. They measure, length by breadth in millimetres, as
follows:— 012 X 0-08; 0i2 X 0-10 ; 0"14 X 0'08 ; 0"15 X 0"10 ; 0"20 X 0'08.

The colour of the colony is a dirty greyish-white, with a light brownish tinge in
many places.

Though differing slightly from Plexaura antipathes, yet the differences are not
sufficient to warrant the formation of a new species. See Hiles (1899), pp. 51-2;
Whitelegge (1897), pp. 317-8.

Locality : Gulf of Manaar.

2 R

306 CEYLON PEARL OYSTER REPORT.

Lophogorgia lutkeni, Wright and Studer. — Plate III., fig. 6.
This species is represented by many pieces of colonies ; the largest piece measures
33 centims. in height by 38 centims. in breadth, and has three main branches. It
measures 18 minims, across at the base, narrows to 12 millims., and expands again to
28 millims. where the branches arise. The axis is dark brown in colour with a
lighter coloured central core. It is very calcareous as well as horny and measures
fully 2 millims. at the lower end of the branch which we studied. It has two
shallow grooves which correspond to the two grooves on the stem and the branches.
The branching is usually confined to one plane and is fairly profuse. In the
largest specimen one of the branches comes off from the main stem almost at
right angles to the plane ot branching, but it bends over almost directly and
comes to lie in that plane. In this specimen, also, several of the secondary
branches rise in a similar manner to the above. In the other specimens, however,
the branching is strictly confined to one plane. The stem and the branches are
flattened in the plane of branching and are marked on the flat sides by a narrow
winding groove or furrow which is continued along the flattened surfaces of the
secondary branches. The tips of the branches and twigs are almost cylindrical. In
the twigs the nutrient canals occur all round, but in the larger branches and in the
stem they correspond to the grooves on the flattened surfaces.

The polyp-bearing surface is divided into two lateral bands by means of the two
grooves. On both of the bands the polyps occur in irregular rows. Verruca? can
scarcely be said to be present, as the general ccenenchyma seems only to be slightly
raised at the point where the polyp issues. On the polyp being retracted, the edge of
the pore-like opening curves in and forms a star-shaped figure in which the rays are
slightly variable. The slight thickening at the point of issue of the polyp seems more
marked in some of the specimens than in others.

The polyps are either completely or partially withdrawn into the general
ccenenchyma ; the expanded part in the case of those partially withdrawn is white
in colour. On the body of the polyp there are eight bands of spicules running up
towards the base of the retracted tentacles. In each band the spicules, which are
numerous, are arranged en chevron.

The spicules are spindle-shaped, but slightly flat in appearance, and sometimes
curved and with fairly large prominent projections. They measure (in situ) from
0'18 millim. to 0"2 millim. in length. They are colourless.

The general ccenenchyma of the stem and the branches is thick, measuring about
] millim. in depth, and friable in texture. In colour it is yellowish, with darker
brownish patches scattered over the whole surface. It is composed of short spindle-
shaped spicules with rough wart-like projections arranged in whorls, the whorls
varying in number with the size of the spicules.

They measure, length by breadth in millimetres, as follows: — 0"1 X 004 ;
0-L2 X 0-04; 0'14 X 0"05 ; 0"14 X 0"02 ; G'13 X 0035; 0'14 X 006.

ALCYONARIA. 307

The specimens agree with the description of Lophogorgia lutkeni in most respects,
e.g., the branching, the flattening, the absence of verruca1, the disposition of the
polyps and nutrient canals. On the other hand, there are some peculiarities :
(1) there is a distinct groove and not a wavy line; (2) the spicules of the general
ccenenchyma are smaller in the Ceylon specimens ; and (3) the pore-like openings left
by the retracted polyps form star-like figures, not slits.

These differences do not seem important enough to warrant us in separating it from
L. lutkeni.

To this species two other fragments belong. The larger measures 168 milium in
height and has a diameter of 5 millims. at the lower end of the main stalk. These
two specimens differ from those described above in having no projections of the
ccenenchyma at the point of issue of the polyps. The openings left by the polyps
when withdrawn are more like slits than in the former specimens. The specimens
are also more brick-red in colour. In every other respect they agree with the
specimens above described. As the two sets of specimens come from the same locality
we are furnished with a good example of individual variation.

Locality : — Cheval Paar, Gulf of Manaar.

Lophogorgia rubrotincta, n. sp. — Plate IV., fig. 12.

The base of the colony is expanded for 25 millims. on a shell ; the branching is
approximately in one plane ; the measurements are about 100 millims. in height by
a maximum of 30 millims. in breadth ; the longest branches are about 30 millims. in
length ; the primary branches arise irregularly on the two sides of the main stem and
some give off secondary branches. As regards spicules and polyps, it seems to be a
Lophogorgia. Most of the coenenchyma spicules are rough spindles ; 0"1 millim. by
0"04 millim. is a common measurement, but many are much smaller. On the median
line of each flattened surface of the stem and its branches there is a line of red spicules
making a characteristic feature in contrast to the general orange-yellow. If it be not
worthy of recognition as a new species, it is a very distinctive variety.

Lophogorgia irregularis, n. sp. — Plate VI., fig. 7.

An irregularly branching specimen of a dirty white colour, about 3 centims. in
maximum breadth by 1"5 centims. in height. The extraordinarily quaint mode of
branching may be described as somewhat antler-like, but the tips of the branches are
mostly bent downwards. The branches expand, constrict, and re-expand in a peculiar
way, and there is considerable flattening in the general plane of branching. The
polyps occur all over, completely retracted within slightly prominent rounded verrucse,
which are closest together and best defined towards the ends of the branches. Many
of the verrucas measure about 1 millim. in diameter, but the aperture of the polyp
on the apex of the gently rounded elevation is usually very much less (about
0"25 millim.).

2 R 2

308 CEYLON PEARL OYSTER REPORT.

The axis is black in colour and seems to lie non-calcareous. It measures 2 millims.
in breadth at the base, soon expands to 4 millims., and then narrows again. It is
markedly flattened. Towards the ends of the branches, which measure about 3 millims.
in greatest breadth, the axis is thread-like, brown in colour, and very flexible. The
ccenenchyma is thick, its texture recalls that of Iciligorgia ; several sections through
a branch showed three large longitudinal canals symmetrically disposed close round
the axis.

The spicules are broad spindles, 0-2 millim. by (PI millim. ; narrow spindles; clubs;
and double clubs.

Locality: — Pearl banks, Gulf of Manaar.

Leptogorgia australiensis, Ridley, var. flavotincta. — Plate IV., fig. 10.

This species is represented by numerous complete specimens, which differ considerably
in size, the largest being 154 millims. in height and 130 millims. in maximum breadth.
In all the specimens the ccenenchyma is rubbed off near the base, thus exposing the
axis to view. The axis rises from a flat spreading basal portion, it is black in colour
and slightly oval in shape, measuring 2-2 millims. in diameter at the base, but
becoming lighter in colour and thread-like in form in the tips of the branches and in
the pinnae. It is horny in texture and very tough. The branching is confined to one
plane, and each of the branches has a pinnate appearance. On the main stem and
branches there is a distinct groove running up both the surfaces, thus dividing
the polyp-bearing part into two lateral bands. In some parts it seems as if the
groove had disappeared and left a ridge in its place. Some of the specimens
show galls.

The verrucas are arranged on the two lateral faces of the branches and the pinnae
in two alternating rows. They are scarcely elevated above the level of the general
ccenenchyma, and their openings are slit-like in shape, the slit running parallel to the
long axis of the branch.

In every respect the specimens agree with the description given by Ridley (1884).
The spicules in the present specimens seem on the whole to be smaller than those
measured by Ridley either for the species itself or for any of the varieties, varying,
as they do, from 0'09 millim. to 0'16 millim. in length, and from 0-04 millim. to
0-05 millim. in breadth. Few of the spicules, however, attain the latter dimension
in breadth.

The colour of the specimen, not agreeing exactly with any of the colours given
for the species, may be best described as deep crimson. The yellow verrucas stand
out conspicuously in the younger branches, and also in two young colonies (Station I. )
30 millims. in height.

Localities :— Cheval Paar, Gulf of Manaar ; Station I., oft'Negombo, 12 to 20 fathoms;
deep water off" Galle and onwards to Colombo.

Previously recorded from Torres Straits (see Ridley, 1884, p. 342, Plate 30).

ALCYONARIA. 309

Leptogorgia australiensis, Ridley, var. perflava — Plate TV., fig. 10.

This variety is represented by a complete colony which has a maximum height of
80 inillinis. and a maximum width of 88 millims. The type of branching agrees very
closely with that described, even to the pinnate arrangement of some of the pinnae
themselves. The groove or furrow is more tortuous than in Ridley's description,
and in several places it shows the transition between a groove and a ridge. The size
and the arrangement of the verruca), and their manner of opening by longitudinal
slits, are all in close agreement with Ridley's descriptions. The axis just above the
flat spreading basal portion has a diameter of 1*35 millims.

The spicules are spindle-shaped, with whorls of warts, distinct free spaces being
visible between the whorls near the middle portion of the spicule. Their measure-
ments, length by breadth in millimetres, are as follows: — CC12 X 01U ; 0'15 X 0*05 ;
0-11 X 0-04; 0-16 X 0-04; 0'17 X 0'03 ; 0'08 X 0'04.

In this specimen the spicules show a greater range in size than is found in the
specimen described by Ridley.

Neither in this variety nor in the variety described above can any trace be found
of the double-headed spicules which Verrill has described. They are almost
certainly, as Ridley has suggested, pieces of broken fusiform spicules.

A point of some interest in regard to the varieties described above is, that the
spicules of the crimson variety are wholly crimson and the spicules of the yellow
variety are wholly yellow. In no case can any trace of both red and yellow spicules
be found in one specimen. Ridley, however, found that red spicules occurred in the
cortex of the yellow varieties, and that in the cortex of the main branches of the red
form there were almost, if not quite, as many yellow as red spicules. We have
examined our specimens very carefully, and it may be that we have here the initiation
of a more definite splitting up of the species into distinct colour-varieties.

Leptogorgia (?) sp.

We have been unable to come to a definite decision in regard to an interesting
form which we rank provisionally with the genus Lepto(/orgia. It is represented by
an incomplete specimen showing two branches arising from a common basal portion.
The longest branch is 84 centims. in length, 2 millims. to 2'5 millims. in breadth, and
gives off two short branches at an acute angle. The branches are flattened, the
edges show a crenate row of gentle elevations and depressions indicating the position
of the verrucse and the minute intervals between them. The colour is greyish with
a tint of vandyck brown ; the texture of the ccenenchyma appears smooth to the
naked eye, finely granular under the hand lens. The axis measures 0'85 millim.
across at its base, and a surprising feature is its markedly calcareous composition.

The spicules are : — (a) Small warty double spindles, measuring in length by
maximum breadth in millimetres, 0*08 X 0-025; O'lxO-025; 0-15X0-03; 0-16x0-04;
and (b) minute warty double clubs, 0"03 X 0'02 ; 0"05 X 0-03 ; 0-075 X 0"03 : 0"08 X 0"03.

310 CEYLON PEARL OYSTER REPORT.

Stenogorgia ceylonensis, a. sp. — Plate II., figs. 1 and lA; Plate V., fig. 9.

This new species is represented by two complete colonies, which measure as
follows : —

(A.) (B.)

Height of colony 125 milium 73 millims.

Maximum width of same ... 117 ,, 76 „

The colonies are tan-shaped and, on the whole, branched in one plane. The basal
attachment consists of a broad, flat, plate-like portion which spreads over the surface
of the object to which it is attached. From the basal portion the short main stem
arises and soon divides into a number of principal branches. These give off secondary
branches, which in their turn give rise to twigs of similar structure. The branches
are, on the whole, given off in one plane, but several of the smaller branches and
twigs come off in a different plane. The tips of the branches and twigs consist of
flat expansions formed by two divergent terminal polyps ; the expansion in many
cases measures as much as 4 millims. from side to side.

The axis is horny, rigid in the lower part, but flexible and light brown in colour in
the younger branches and twigs.

The verruca? are very abundant on the branches and twigs, a little less numerous
on the principal branches. They are cylindrical in shape, and measure, length by
breadth in millimetres, as follows: — 1"9 X 1 "2 ; L5 X l'l. They occur on all sides
of the branches and twigs, and may be said to occur in short spirals which are closely
crowded on the branch or twig. When the polyps are withdrawn the verruca?
present all the phases between a mere pore with a lobed margin and a minute 8-rayed
star, in proportion to the extent of the retraction.

The ccenenchyma is thin and is easily rubbed off. It is mainly composed of short
rough spindles closely covered with warts and spines. They are light yellow to deep
yellow in colour, and measure, length by breadth in millimetres, as follows : —
0-4 X 0-08; 0-5 X 040 ; 0'41 X 0"079 ; 0*5 X 0*08.

The colour of the colony is a chocolate-brown.

Locality : — Trincomalee.

Gorgonia capensis, Hickson.

Several complete specimens and many fragments belong to this species. Two of
the complete specimens measure 428 millims. and 439 millims. respectively in length.
The basal attachment is in the form of a flat, plate-like expansion, from which the
thread-like axis arises.

The axis is dark brown to black in colour, very tough and flexible, never measuring
more than 0-5 milliin. in diameter. In the larger complete specimen it has a basal
diameter of only 0"15 milliin. The branching is not profuse, the one specimen having
two branches, the other only one, though in some of the broken fragments several
branches are present.

A LC YON ART A. 311

The ccenenchyma is moderately thick, both on the main stem and on the branches,
the terminations of which are knob-like. The colour is practically white, not yellow
as in IIickson's specimens.

Verrucas are absent, the position of the polyps being marked by oval-shaped
openings.

The spicules agree in measuring not more than 0-2 millim.

From these specimens it has been possible to corroborate Hickson's discovery of
viviparity in this species, but the disc-like embryos only measured 0'4 millim. by
0"39 millim. in size. Here we may be allowed to note our discovery of embryos in
Juneoptilwm sp., from Alcock's Indian Ocean collection.

Localities :— Station I., off Negombo, 12 fathoms to 20 fathoms; deep water off
Galle and onwards to Colombo, hauls off Kaltnra and off Mount Lavinia.

Previously recorded from 10 miles south of Cape St. Blaize (HlCKSON, 1902).

Rhipidogorgia (?) sp.

A large fan-shaped reticulate colony was found, but so much weathered that secure
identification was impossible.

Family : GOltGONELLIDyE.
Scirpearella aurantiaca, n. sp. — Plate IV., fig. 7 ; Plate V., fig. 15.

The collection included several portions of what seems a new species of this genus.
We have studied a fork-shaped fragment which has a total length of 366 millims.
The lower part of the colony is wanting, but the axis at the point where it divides
measures 2 millims. in diameter. It is cylindrical in shape, very calcareous and
smooth in the one branch, but marked by two or three slight winding grooves in the
lower part of the other. The general colour of the branches is yellowish-white. A
larger specimen, 470 millims. in length, bore a small comatulid on one of its branches.

The verrucse occur on all sides of the branches. They are conical in shape,
truncated at the tip, and measuring 2 millims. in maximum height and l-5 millims.
in basal diameter. In colour they resemble the stem in the lower part, but the tip is
orange-yellow, thus standing out against the general colour of the brandies. The
edges of the oral end curve inwards, and all stages from an opening with an 8-lobed
margin to a simple pore-like opening and finally to a completely closed tip may be
seen.

The polyps are all completely withdrawn into the ccenenchyma. The general
coenenchyma is yellowish-white in colour, granular in texture, and only of medium
thickness. It is practically composed of spindles and double clubs.

The spicules are small in size, and measure, length by breadth in millimetres, as
follows : —

(1.) Spindles, 0'06 X 0'02 ; 0-08 X 0-02; 0'085 X 0-03.

(2.) Double clubs, 0055 X 0-03; 0"07 X 0"04 ; 0-06 X 0-04.

312 CEYLON PEARL OYSTER REPORT.

This specimen does not agree with any described species of Scirpearella, but comes
nearest to Scirpearella rubra, with which, however, we cannot identify it.
Locality : — Deep water outside pearl banks, Gulf of Manaar.

Scirpearella divisa, n. sp. — Plate VI., fig. 8.

A fragment of a reddish-orange colony with four 1 (ranches, 7 centims. in height, about
2 millims. in diameter. The verrucas are very low and gently rounded ; towards the
end of the highest branch, where they are closely crowded and very distinct, the
arrangement appears to be in four rows with a suggestion of a spiral ; in the older
parts the verrucse are very inconspicuous, not close together, and somewhat irregularly
disposed. There are two enormous cirriped galls, the broadest almost 1 centim.
across. The ccenenchyma is finely granular, almost smooth to the naked eye. The
axis is very calcareous, light yellow in colour, with ten shallow grooves on the part
examined. It measured there about 1 '4 millims. in diameter out of the total branch
diameter of 2 millims. Thus the thickness of the ccenenchyma is very moderate.

The spicules are chiefly minute, warty, double clubs, measuring 0'04 millim to
0'06 millim. in length, but usually 0'05 millim. ; most are yellowish, a minority
colourless. The bare middle part of the double club is very short. There are also
yellowish and uncoloured warty spindles, measuring O'l millim. in length by 0*02 millim.
in maximum breadth.

Scirpearella sp. a.

One specimen with a length of 307 millims. The stem is divided at a distance ot
39 millims. from the flat base, and is flexible. Prominent verrucas occur all round
the stem in rows. The axis is deeply grooved in the lower part, but the grooving is
hardly noticeable higher up.

The fragment approaches Scirpearella moniliforme, but differs in colour and in
predominance of warty clubs over spindles, and in not having shallow verrucas.

Scirpearella sp. /3.

Another fragment, measuring 205 millims. in length. Verrucas scarcely prominent,
polyps white, a longitudinal white line is seen on one side. The axis is horny and
very calcareous, faintly and irregularly marked by grooves.

The spicules, length by breadth in millimetres, are as follows : —

Stellate forms, 0-06 X 0"05 ; 0"07 X 0*06.

Double clubs, 01)6 X 0-03.

Spindles, 0"06 X 0"02 ; 0'08 X 0-025.

This does not agree with any of the formerly described species.

Scirpearella sp. y.

Another fragment which measures 258 millims. in length is included in the
collection.

ALCYONARIA. 313

The stem has a groove on one side, and on the side directly opposite there is a
hare strip. The polyps are in three rows on each side of the groove. The verrucse
are very small.

The following measurements were taken : —

Diameter of stem at lower end, 3 "2 millims.
„ axis ,, ,, 2 ,,

,, stem at upper ,,2 ,,

The spicules measure, length hy breadth in millimetres, as follows : —

Spindles, 0"06 X 0'03 ; 0-07 X 0-03.

Double clubs, 0^05 X 0D4 ; 0-06 X 0-04.

Double spindles, 0"06 X 0'04 ; 0"05 X 0-035.

Scirpearia, sp. (?)— Plate IV., fig. 1 ; Plate V., fig. 16.

This specimen is a fragment without the base. It measures 410 millims. in length
and divides at a distance of 38 millims. from its basal end into two long whip-like
branches.

The axis is deeply grooved. There is a median bare space on each side. The polyps
are darker in colour than the stem.

The spicules measure, length by breadth in millimetres, as follows : —

Double clubs, 0*06 X 0"04 ; 0-063 X 0'04. Darker coloured, 0"08 X 0-025 ;
0'07 X 0-02. In its spicules it is quite distinct from any Juncella.

It may seem of little service to suggest problematical species based on a study of
fragments, but as we have given some description of each, our procedure is probably
preferable to that of some other students of Alcyonacea, who have given names nude
of any description.

Our impression is that the elongated forms of Sdrpearella, Juncella, and the like,
so monotonous in general appearance, so perplexingly diverse when one gets beneath
the surface, are subject to great variability. The same remark applies to Spongodes,
Acts, Muricella, Verrucella, and many other Alcyonarians.

Juncella g-emmacea (Valenciennes) — Plate IV., figs. 4 and 5.

There are several specimens of this species in the collection. The finest one was
3 feet 7 inches in length, with a yellow basal expansion 3 centims. across, and with a
stem measuring 9 millims. in diameter at the base. The stem increased to 1 2 millims.
in diameter, and 10 millims. was a fair average. The lateral lines were spirally
twisted. Another specimen measures 193 millims. in length and tapers from 7 millims.
in diameter at the base to 4 -5 millims. at the tija ; another measures 292 millims. in
length with a basal diameter of 4 millims. The thickest piece measured 9 millims. in
basal diameter, the axis alone amounting to 5 millims., and had an apical diameter of
3 millims. The side-line on either side of the stem is distinct. The axis is

longitudinally grooved.

o JO

2 s

314 CEYLON PEARL OYSTER REPORT.

The spicules agree with those described and figured in the ' " Challenger " Report,
and are (a) unsymmetrical double clubs, and (b) double stars. The measurements,
length by breadth in millimetres, are of (a), 0-12 X 0"04 ; Oil X 0-04 ; 0*10 X 0'04 ;
and of (/>), O'lO X 0"06. A few single clubs and rough needles were seen.

The colour of the specimens in spirit is bright red with tints of orange when
closely examined ; the polyps appear white.

Locality : — Gulf of Manaar ; Vankali Paar; Muttuvaratu.

Previously recorded from Red Sea ; Torres Straits ; Queensland and Mermaid's
Straits, N.W. Australia.

Juncella juncea, Pallas.
The collection includes numerous fragments of various species of Juncella, more
numerous than we could examine ; some of these belong to Juncella juncea which
Ridley reported from Ceylon (1883).

Juncella fragilis, Ridley.

This species is represented by several fragments, in which the axis is marked by
longitudinal grooves or striae. In some of the fragments the verrucas measure nearly
2 millims. in height and the axis has a diameter of 1 millim., while in the other
fragments the verrucse are much smaller and the axis measures 2 millims. in diameter.
The spicular measurements in the one set of fragments are almost identical with
those previously given for Juncella fragilis, but in the other set they measure as
follows, length by breadth in millimetres : —

Unsymmetrical clubs, 0'08 X 0-03 ; 0-0G5 X 0-03.

Double clubs, 0-08 X 0-04 ; 0-075 X 0'03.

Juncella fragilis (Ridley), var. rubra,, n.

This species is also represented by one complete specimen on which this new variety
has been founded. The specimen shows no trace of branching, and measures
508 millims. in length. It is long and flexible, and tapers very gradually throughout
its entire length to the pointed apex. The cortex is thick and red coloured, with
numerous reddish-brown verrucas. There is no trace of a lateral line or groove
throughout the entire length of the stem, but the cortex has been rubbed off the
lower part.

The verruca? are numerous and closely appressed, measuring about 1 millim. in
height. The axis is grooved and very flexible, measuring 3 millims. in diameter at
the base and becoming hair-like near the tip where the stem measures 2 millims. in
diameter. The base consists of a broad flat plate which has a diameter of 14 millims.

The spicules are double clubs and double stars, which measure, length by breadth
in millimetres, as follows : —

Double clubs, 0"08 X 01)3 ; 0*087 X 0'04

Stars, 005 X 0-03 ; 01)0 X 0U3.

ALC YON ARIA. 315

The specimen approximates closely to the description of Juncella fragilis given
by Eidley (1884). The only marked differences are in the colour and in the diameter
of the axis at the hase. The larger diameter can be easily accounted for by the
large size of the specimen, 508 millims. But it seems useful to record a new colour
variety.

Locality: — Gulf of Manaar.

Previously recorded from Port Denison, Queensland ; 4 fathoms.

Juncella trilineata, n. sp.

Height of specimen 238 millims. It is sparingly branched in one plane, the first
branching occurring at a height of 67 millims. There is a distance of 10-5 millims.
between this and the next branching. The colony is exceedingly flexible and very
graceful. The base is wanting, but the stem at its lower end measures 3*5 millims. in
diameter.

Polyps arise in three different bands, leaving three narrow bare strips, each of which
has in its centre a slight rib or keel. Under each bare strip lies a large longitudinal
canal. The axis shows longitudinal grooves.

The polyps, which measure from l'l millims. to 1'5 millims. in height, are arranged
in transverse rows of 3 or 4, but many smaller polyps occur which break this
regularity.

The spicules resemble those of Juncella gemmacea in several of their forms.
Unsymmetrical clubs, length by breadth, 0"08 millim. by 0'03 millim. ; 0"065 millim.
by 0 03 millim. ; 0-07 millim. by 0"04 millim. ; 0-07 millim. by 0-03 millim.

Double stellate forms, 0-08 millim. by 0-04 millim. ; diameter of spindle, 0*02 millim.
„ 0-09 „ „ 0-035 „ „ „ 0-02 „

Locality : — Patani, Siarn.*

Verrucella rubra, n. sp. — Plate IV., fig. 13 ; Plate V., fig. 8.

A fragment of a colony, which has a total height of 98 millims. and a maximum
width of 69 millims., appears to require the establishment of a new species. The
base and the lower part of the colony is wanting, and many of the branches seem to
be without the upper portion.

The axis, which measures 2-25 millims. in diameter at the lower end, is greyish-
brown in colour, with a white central core, very calcareous, rigid, and brittle.

The branching of the colony is very profuse and is confined to one plane. The
branches and twigs are given off at a right angle or at a close approximation thereto.
There does not appear to be any fixed arrangement in the branching.

* This specimen, along with two others (Asirmnmicea ramosa, n. sp., and Telesto Irichosiemma), was given
to me on a visit to Professor Hekdman's laboratory by Mr. Nelson Annandale, who asked that it
might be incorporated in this Report. — J. A. T.

2 S 2

316 CEYLON PEARL OYSTER REPORT.

The verruca? are small, low, conical, wart-like bodies. They are confined on the
whole to the lateral faces, but on the older branches, and on some parts of the stem
and of the twigs, they seem to be scattered over three surfaces.

The general coenenchyma is thin and easily rubbed off (in several places the axis is
laid bare) ; it is mainly composed of double clubs, spindles, and double spindles. The
spicules are covered with large rough warts, which are arranged in whorls. The
spicules measure, length by breadth in millimetres, as follows : —

Double clubs, 0-10 X 0-047 ; 0'09 X 0"05 ; 0"09 X 0-06.

Spindles, 0*10 X 0-035 ; 0'08 X 0-03 ; 0-09 X 0-035.

Double spindles, 0"07 X 0"04 ; 0-08 X 0'04 ; 0'06 X 0-035.

The colour of the specimen is a bright crimson.

Locality : — Gulf of Manaar.

Verrucella flexuosa, Kltjnzinger, var. aurantiaca, n. — Plate TIL, fig. 4 ; Plate IV.,
fig. 8.

This form is represented by several complete specimens and many fragments. One
complete colony has a height of 210 millims. and a maximum width of 180 millims. ;
another is 74 millims. by 80 millims. The colonies rise from a flat portion which
spreads over the surface of a stone. In the largest specimen branching begins at the
base, in a smaller the stalk rises to a height of 30 millims. before it gives off the
first branch. In the lower portion of the stalk, which has a diameter of 2 millims.,
there are no polyps present. The verrucas begin to appear at a height of 14 millims.
from the base, but they are not abundant in the older basal region. There are
numerous galls.

The branching is profuse and confined to one plane. The branches are given off
at various angles — acute for some of the main branches, and almost always a right
angle for the minor branches or twigs.

The tips of the branches and the twigs are sometimes slightly flattened in the
plane of branching. The branches and the twigs come off very irregularly.

The verrucas on the younger portions of the branches and on the twigs are some-
what more numerous towards the lateral edges, but on the stem and the older
portions they occur all round. The verrucas are conical in shape and truncated,
measuring about 0-75 millim. in height and having a basal diameter of 1 millim.

The general coenenchyma is granular in appearance, not of great thickness, and
composed of spicules in the form of spindles and double clubs. They measure, length
by breadth, in millimetres, as follows : —

Spindles, 0-08 X 0'03 ; 0"07 X 0-02; 0-10 X 0"04 ; 0"08 X 0-02.

Double clubs, 0-0G5 X 0"04 ; 0-08 X 0"04 ; 0-08 X 0-045.

The specimens described here approach very closely to Verrucella Jlexuosa,
Kltjnzinger. but as this species appears to be very variable, it may be useful to class
them as an orange variety.

ALC YON ARIA. 317

In another specimen the brandling was not at all in one plane; in another there
was abundant anastomosis.

Locality : — -West of Periya Paar.

Verrucella flexuosa, var. gallensis, n. — Plate III., fig. 9 ; Plate V., fig. 11.

This variety is represented by a slightly damaged colony and several large and small
pieces of colonies. The complete specimen has a height of 47 millims. and a maximum
diameter of 53 millims. The basal attachment consists of a flat plate-like spreading
portion from which the main stem rises. The stem is rigid and brittle, very
calcareous in nature, and rises to a height of 13 millims. before it gives off the first
branch. The branches are given off at right angles and they in their turn give off
branches in a similar manner. The tips of the twigs and branches are occupied by
two divergent polyps which gives them a slightly flattened appearance. The branches
are given off in an irregularly alternate manner from the main stem and from the
larger branches. The branching is confined to one plane. The colour varied from
rose-red to vermilion.

The verrucse are blunt and conical shaped, measuring 07 millim. in height and
having a basal diameter of 0'8 millim. to 1 millim. They are confined to the lateral
surfaces of the stem and branches, but are directed a little more to one surface than
to the other. They are arranged in such a way that the verruca? on the one lateral
face alternate with the verruca? on the other face.

The general coenenchyma is very thin and is composed of spindles, double clubs and
double spindles.

This variety is represented by a large number of fragments, mostly small. The
branching in some of the fragments is confined to one plane, but in others the
branches are given off on all sides. The axis is cylindrical, white in colour and very
calcareous. In the several fragments it shows considerable differences in diameter.

The verruca? are small, conical shaped, with a wide base. They measure 0'5 millim.
in height and have a basal diameter of fully 1 millim. They occur all round the main
stem or branch and also on the other branches, but on portions of the older branches
and on the younger branches they are more or less confined to the lateral surfaces.

The general coenenchyma is thin and is chiefly composed of spindles and double
clubs. They are covered by whorls of rough warts. In the double clubs there is a
very short bare shaft in the middle. Their measurements, length by breadth in
millimetres, are as follows : —

Spindles, 0-10 X 0-04; 0*08 X 0"03 ; 0-10 X 0"03 ; 0"12 X 0"03.

Double clubs, 0-08 X 0-00 ; (VIC) X 0"04 ; 0'07 X 0'04 ; 0'085 X 0"04.

Localities : — Deep water off Galle ; Gulf of Manaar.

Verrucella sp. (?).
A small fragment, from which several branches are given off in one plane, measures
63 millims. in height.

318 CEYLON PEARL OYSTER REPORT.

The axis is very calcareous and brittle.

The polyps on the main stem or branch are more or less restricted to the lateral
edges, but on the younger branches they occur on the four surfaces in such a way that
they all alternate and appear as if arranged in a spiral manner. They are contained in
verruca; which are small, slightly conical and truncated at the top. The polyps
themselves are all withdrawn into the verrucas, but they present yellowish-white
apices at the opening of the verrucas. Towards the base of the tentacles on the
aboral surface two or three pairs of spicules are arranged en chevron, and at their
bases a few are arranged transversely so as to form a collar. These spicules are short
warty spindles, either straight or slightly curved.

The cnenenchyma is thin and is composed of two layers of spicules, the outer layer
being coloured and giving the characteristic colour to the colony, while the inner layer
is colourless.

In the coloured layer the spicules show a considerable range in shades, varying
from a light to a deep amber tint.

The spicules are spindle-shaped, thick in comparison to length, or slender, and
covered with rough warts. The coloured, or outer layer, contains the bigger spicules.
Their measurements, length by breadth in millimetres, are as follows : —

(a.) Coloured, 0-3 X 0-075; 0'4 X 0'08 ; 0'22 X 0"05 ; O'll X 0"05 ; (HO X 0-03;
0-08 X 0-02.

(/>.) Colourless, 0-14 X 0'04; 0-175 X 0-04; 010 X 0'035 ; 0"28 X 0'025.

Order V.— STELECHOTOKEA.
Section I.— ASIPHONACEA.

Family: TELESTID^E.
Telesto rubra, Hickson.

This species is represented by two fragments, one of which (A) measures only
30 millims. in height.

Specimen A agrees closely with the description given by Hickson. Thus, in the
basal portion of the youngest calyces, the spicules are fused to form tubes, the axis
has eight longitudinal ribs, and prolonged boiling in caustic potash does not
disintegrate it.

Specimen B seemed at first sight to have a black axis, but this was found to be
due to debris which had collected in the hollow tube.

The whole surface of specimen (A) is covered by an encrusting sponge, which has
completely destroyed the ccenenchyma. The spicules of the sponge are monaxonial
and triaxonial.

Locality : — Trincomalee.

ALCYONARIA. 319

Telesto (Carijoa) trichostemma, Wright and Studer.

This species is represented by one specimen, which agrees with the description
given by Wright and Studer,

The spicules are of two distinctly marked types : —

(1.) Long slender spicules with few lateral processes, often bifurcated at the end ;

(2.) Strong, relatively broader spicules, with prominent lateral processes which
interlock and form a felted covering.

They measure, length by breadth in millimetres, as follows : — ■

(1.) 0-3 X 0-02; 0'28 X 0-019; 015 X 0-012; 0"26 X 0-014.

(2.) 0-2 X 0-03 ; 0'25 X 0"02 ; 0'23 X 0'03.

The distance between the bifurcations at the ends of the spicules of type (1) often
measured 0-02 millini.

Locality : Patani (Siam). This was one of Mr. Nelson Annandale's three speci-
mens already referred to.

Section II. : PENNATULACEA.

Family : UMBELLULIDJL
Umbellula, sp.

The collection included* a small specimen, doubtless a young colony, of a species of
Umbellula. As the specimen is a poor one, we have not studied it in any detail.
We may refer to our forthcoming report on Professor Alcock's " Investigator "
collection of deep-sea Alcyonacea from the Indian Ocean, which includes numerous

Umbellulids of large size.

Family : VIEGULAIUID.E.

Virgularia multifiora, Kner.
The specimen representing this species is fragmentary. It consists of a piece of the
rachis, which seems to come from the lower portion of the colony. It measures
51 millims. in length and has 20 pairs of pinnules. The pinnules are crescent-shaped,
3 millims. in breadth and 2 millims. in height, with 11 polyps-on each. The axis is
white, almost perfectly cylindrical, with a pitted appearance on the surface and with
a diameter of 1 2 millims. The specimen agrees very well with the description given
by Kolliker. The following comparison will show the close similarity between the
specimens :- Kolliker's. Herdman's.

Breadth of pinnule .... 3'3 millims. to 4 millims. 3 millims.

Height „ „ .... 2 millims 2

Number of polyps in a pinnule . 11 to 15 11.

Diameter of axis 1 "29 millims T2 millims.

Locality : — Triucomalee.

* From Station LX., Gulf of Manaar, 20 to 30 fathoms. See Professor Herdman's " Narrative," this
Report, Part I. (1903), p. 75.

320

CEYLON PEARL OYSTER REPORT.

Virgularia loveni, Kolliker.

This species is represented by three fragments, the largest of which measures
106 millims. in length, and appears to consist of nearly the whole rachis, while the
other two fragments represent the upper portions of two specimens.

The axis shows slight differences in form ; in one it is cylindrical, in another
quadrangular, with two of the sides rounded, while in the third it returns to the
cylindrical shape, hut with one surface slightly flattened. It is covered by transverse
ridges or by tubercles, which gives it a pitted appearance. On the prorachidial face of
the rachis there is a well-defined narrow groove. The diameter of the axis is
1*5 millims. ; the rachis measures 106 millims. in length by 5 '3 millims. in width.

The pinnules are folded so as to form a double S-shaped figure. They are almost
touching one another near the tip, but lower down they stand considerably apart
(3 '5 millims.). They vary from 3 "5 millims. to 4 '8 millims. in breadth, and from
2 millims. to 2-5 millims. in height. The polyps are well defined, arranged in one
row on the pinnule, and vary in number from 19 to 24.

The specimens agree with the description of Virgularia loveni given by Kolliker
(1872), although there are a few slight differences.

Locality : — Trincomalee.

Virgularia, sp. — Plate IV., fig. 2.
The collection includes two fragments which approach Virgularia loveni. As they
may be young forms, in regard to which little is known, and as they are only fragments,
we have been content to give a brief description. The measurements are stated in
the following table : —

Rachis.

Pinnules.

Number

of

polyps on

pinnule.

Number

of

rows.

Axis
diameter.

Stalk
diameter.

Length. Width.

Breadth.

Height.

miHims. millims.

millims.

millims.

millims.

millims.

(A.) . . .

53 4-5to5-5

4-8 2-5

23

1

1-5

2

(B.) . . .

56

4-5 „ 5-5

4-7

2-4

24

1

1-5

(white in
colour).

2

The pinnules are folded so as to form a double S-shaped figure. The polyps are
well defined. There is a distinct groove on the prorachidial surface which measures
0'4 millim. from edge to edge.

At the tip the pinnules are almost touching, but lower down the distance between
them is 3 '4 millims.

The axis is almost rectangular in one specimen, with two shorter sides slightly

ALCY0NAK1A.

321

rounded. In the other it is almost circular, but one surface is slightly flattened. In
both cases it is covered by tubercles, which gives it a characteristic appearance. The
tip of the axis is blunt.

A third specimen gave the following measurements : —

Total
length.

Stalk.

Ruchis.

Number

of
polyps.

Number
of

rows.

Zooids.

Length.

Width.

Length.

Width.

millims.

57

millims.

22-7

milliins.
1-5

millims.
34-3

millims.

At base

2,

half-way up

3,

near top

5.

26

1

Occurring

over the

whole surface

of rachis.

Brownish

colour.

Specimen is club-shaped and tapers gradually from base till it reaches maximum
width at a point near the tip of rachis.

Abundant spicules both in the stalk and in the rachis.
Calyx ridged.

A distinct groove runs the whole length of the rachis.
Locality : — Trincomalee.

Virgularia tuberculata, n. sp. — Plate II., tigs. 5 and 9.

This species is founded to include six fragments which it was found impossible to
place under any of the known species. The fragments are very incomplete and are
not in good preservation. The two largest specimens are 110 millims. and 8 1 millims.
in length.

The axis is cylindrical and presents a characteristic pitted appearance, due to the
tuberculated structure of the surfaces. The tubercles appear at some parts to be
irregularly arranged, at other parts to be arranged in whorls. The colour of the axis
is a deep coral red shading into a yellowish -red. The diameter is 0"6 millim.

One of the fragments represents the lower part of the rachis, while other two
represent the upper extremity. In the lower part the pinnules are small and appear
as ridges, which run across the lateral surfaces of the rachis. The polyps are borne
in one row on the edge of the pinnule and number six on each pinnule. Higher up
the pinnules become more mature, larger, and separated by a longer interval.
They present a quite different appearance, becoming slightly crescent-shaped. The
polyps are somewhat barrel-shaped and are quite distinct, except for a small portion
at their bases, which are fused together to form the pinnule.

The coenenchyma is very thin, but on the pararachidial surface it is produced into two

2 T

322 CEYLON PEARL OYSTER REPORT.

ridges, on which the siphonozooids are found. Their arrangement cannot he exactly
ascertained, hut they seem to form an interrupted line on each of the ridges, and also
occur either singly or in small groups helow each pinnule.

The polyps on the separated parts vary from barrel- to flask-shape in the retracted
condition, and measure in this state from 0-5 millim. by 0'4 millim. to 0"65 millim. by
0'4 millim. In their expanded condition they are more cylindrical in shape and
measure 0"9 millim. by 0*4 millim. They still present a slightly swollen appearance
about the middle of their length.

The six fragments represent parts of the rachides of at least three different
specimens. In no case is there any part of the stalk left, so that it is impossible to
give any idea of its shape and size.

Locality : — Trincomalee.

Family: PENNATULID/F,

Halisceptrum periyense, n. sp.

The collection included a couple of specimens of what appears to be a new species
of Halisceptrum. Both specimens were complete and the natural colour is recorded
by Professor Herdman as pink and white.

The lower portion of the stalk is curved and stands at right angles to the axis of
the rachis, with its lower extremity marked by a ball-shaped swelling. The stalk
passes almost imperceptibly into the lower portion of the rachis, which forms a long
spindle-shaped swelling marked by two narrow V-shaped lines of pores, one on each
lateral surface. The upper part of the rachis bears the pinnules, which show a
marked difference in appearance. On the upper portion they are fairly large and
show well-marked polyps on a wavy outline, while on the lower part they are
reduced to straight ridges across the lateral faces of the rachis. The proportion of
the mature to the immature pinnules is rather striking, there being in one specimen
14 mature to 53 immature, and in another 10 mature to 36 immature.

The pinnules in the immature state have the polyps arranged in one row, but soon
the pinnule begins to get folded and the polyps begin to alternate, thus giving the
appearance of multiple rows of polyps. The distance between the origins of the
pinnules is 2*5 millims. in the case of the mature, but the immature lie almost
touching one another. The pinnules are translucent, and they show a stringy
appearance when held between the eye and the light.

The prorachidial surface of the rachis is free from pinnules and is marked by a
distinct narrow groove. This groove fades away and disappears entirely between
the immature polyps.

We add a table of measurements of the two specimens. There is in each a single
row of polyps on the pinnules, but they alternate in arrangement so as to appear like
two or more rows.

ALCYONARIA.

323

Stalk.

Rachis.

Pinnules.

Number of

polyps on

each lower

pinnule.

Distance
between
origins of
pinnules.

Length. Width.

Length.

Width.

Breadth. Height.

(A)

(B)

millims.
41

36

millims.

3
2-2

millims.

78

49

millims.

6-5

6

millims. ' millims.

3-3 31

3 2-8

29 to 31
28 „ 30

millims.

2-5
25

Locality : — Periya Paar, Gulf of Maiiaar, 9f fathoms.

Halisceptrum gustavianum (Herklots).

This species is represented by one complete specimen. Of the 23 millims. forming
the stalk the lower 11 millims. consist of nothing but the axis which tapers to a point.
At the upper end of the 1 1 millims. the axis has a diameter of 1 "3 millims. and is
cylindrical in shape.

In the rachis the stalk becomes grooved and approximates to a quadrangular shape.
A groove is distinctly seen in the prorachidial face of the rachis, but this fades away
and disappears in the lower part of the rachis.

The pinnules show a large number of polyps which appear to be arranged in a
number of rows. In the immature pinnules the polyps are in a single row, but they
soon begin to alternate, and in the mature pinnules they seem to be in six rows. The
number of mature pinnules is very small compared with that of the immature, the
numbers being 16 pairs to 40 pairs.

Pinnules: Breadth, 2-5 millims. ; height, 2'5 millims.

Distance between origins of pinnules, 4 millims.

Length of whole colony, 86 millims.

Locality : Station LIL, Gulf of Manaar, 3 to 6 fathoms.

Pteroeides lacazei, vai. spinosum, Kolliker.

This species is represented by a fragment only of the upper portion of the rachis,
measuring about 36 millims. in length, with 15 pinnules on one side, 16 on the other.

The principal measurements are as follows : — Width of rachis, 48 millims. ; breadth
of pinnule, 20 millims. ; height of pinnule, 12 millims. ; number of main rays, 12.

There is a single row of zooids running; down the centre of the meta-rachidial
surface.

The wart-like termination of the rachis is almost completely hidden between the
two uppermost pinnules. The rachis seems to have been swollen towards the tip and
the fragment seems to be a part of the swelling.

The zooid plate is basal. The pinnules are marked by a number of V-shaped black
markings on their margins. The polyps are situated both on the upper and on the
lower surface of the pinnule, but they are more abundant on the upper surface.

2 T 2

324 CEYLON PEARL OYSTER REPORT.

This specimen resembles Pteroeides lacazei, var. spinosum, in having one row of
zooids, in the termination of the axis, in the number and composition of the main rays
and in the texture of the pinnules.

Locality :— Station XXL, Back Bay, Trincomalee, 8 to 12 fathoms ; Station LXIIL,
West of Periya Paar, 40 fathoms.

Family : VERETILLID^.

Cavernularia obesa, Valenciennes.

This species is represented by a number of specimens obtained from shallow water
in the northern part of the Gulf of Manaar. The largest specimen is evidently only
the upper portion of a large colony. The part measures 61 millims. in height, and
has a diameter which varies from 10 millims. near the tip to 17 millims. near the
lower end. In shape it is irregularly elliptical.

The autozooids are scattered over the surface, and the parts not occupied by the
autozooids are thickly beset with siphonozooids. The position of the autozooids,
which are all retracted, is marked by bluish- to brownish-black spots.

A cross-section in the lower part shows four large central canals which run parallel
to the long axis of the specimen and are apparently continuous. Radiating from the
central part thus formed numerous canals can be seen, some of which run down to
the central canals, others end blindly, and others join with one another. Some of
the larger canals seem to open into one or other of the four central canals.

The spicules are perfectly smooth rods, with blunt ends, some cut abruptly across,
others slightly tapered and rounded.

Another specimen, which is quite complete, probably belongs to the same species.
It has a total length of 53 millims. and a rachidial diameter of 7 millims. near the
tip. The rachis is slightly elliptical in cross-section, but becomes almost cylindrical
near the tip. The whole surface not occupied by the autozooids is covered by
numerous minute wart-like bodies which represent the upper parts of the siphono-
zooids.

The autozooids occur scattered over the whole surface, and all are almost com-
pletely withdrawn. The incurved tentacles are just visible on the surface as
brownish- to bluish-black circles. There is a marked difference between the
autozooids in size, but this is probably due to a difference in the state of retraction.

The stalk is marked off from the rachis by an abrupt diminution in size, and it,
with the lower part of the rachis, is marked by a number of grooves which run
parallel to the length of the specimen.

Localities: — Station LIL, near Vankali Pceef, Gulf of Manaar, 3 to 6 fathoms;
and (Jheval Paar.

ALCYONARIA. 325

Stylobelemnoides, 11. gen.

Veretillid with polyps over the whole surface ; spicules long, rod-like spindles ;
axis cylindrical ; calyces with 8 double rows of spicules.

Stylobeleninoides herdmani, n. sp. — Plate III., fig. 5 ; Plate IV., fig. 3.

The specimen is a broken club, 32-5 niilliins. in length, 2-8 milium in breadth at
thickest upper part, 2 millims. at lower end. The axis is straw coloured, scmare,
with somewhat prominent angles, 0'85 millim. in thickness. The whole surface of
the club is covered with small rod-like spicules, which form tubular calyces with
8-lobed apertures. The polyps are arranged all round the club, in two intersecting
(but not exact) spirals. They form nine irregular rings. The siphonozooids fill up
the interstices, and no bare streak is left.

The radial disposition of the polyps indicates a Veretillid ; that they cover the
whole surface indicates a Veretillean. As the limy bodies are long rod-like spindles,
the specimen must be near Stylobelemnon.

But it differs from Stylobelemnon in many ways : e.g., the axis is round not square ;
the calyces show 8 double rows of spicules ; and the spicules are much larger,
2-25 millims. in length,

Under a magnification of 32 diameters the surface is seen to be closely covered with
small straight spicules, among which the siphonozooids can be seen.

The polyp calyces are built up of similar spicules very closely packed together, and
lying with their long axes parallel to the length of the calyx, thus forming a stiff
protective case.

The spicules are present both in the cutis and in the deeper tissue.

Locality : — Trincomalee.

LITERATURE REFERRED TO.

1900. Ashworth, J. H.— " The Xeniidse." ' Willey's Zoological Results,' Part iv.

1896. Germanos, N. K. — " Gorgonaceen von Ternate." 'Abhandl. Senekenberg. Nat. Ges.,' Band xxiii.

1 890. Hedi.und, T. — "Einige Murieeiden der Gattungen Aeanthogorgia, Paramuricea und Echinomnricea."

'Bihang Svensk. Vet.-Akad. Handl.,' xvi., afd. iv., No. 6, pp. 1-18, 3 pis.
1895. Hickson, S. J. — "A Review of the Genera of the Aleyonaria Stolonifera." 'Trans. Zool. Soc.

London,' vol. xiii.
1900. Hickson, S. J.- — " The Aleyonaria and Hydrocorallinaa of the Cape of Good Hope." ' Department

of Agriculture : Marine Investigations in South Africa,' vol. i., pp. 67-96.
1900. Hickson, S. J.—" The genus Telesto," &c. 'Fauna of the Maldives,' vol. ii., Part I., p. 480.
1900. Hickson, S. J., and Hiles, Isa L. — "Stolonifera and Alcyonacea from New Britain." ' Willey's

Zoological Results,' Part IV.
1899. Hii.es, Isa L. — "Report on the Gorgonacean Corals collected by Mr. J. Stanley Gardiner at

Funafuti." ' Proc. Zool. Soc. London,' 1899, pp. 46-54, 4 pis.

32(5 CEYLON PEARL OYSTER REPORT.

1899. Hiles, Isa L. — "The Gorgonacea collected by Dr. Willey." ' Willey's Zoological Results,'
II., pp. 195-206. 2 plates.'

1895. Holm, Otto. — "Beitrage zur Kenntniss der Alcyoniden-Gattung Spongodes, Less." 'Zool.

Jahrbiicher,' Band viii.
1877. Klunzinger, C. B. — 'Die Korallthiere des Rothen Meeres.' Erster Theil. " Die Alcyonarien und

Malacodermen."
1887. Koch, G. von. — "Die Gorgoniden des Golfes von Neapel." 'Fauna und Flora des Golfes von

Neapel,' Band xv.
1890. Koch, G. von. — " Die Alcyonacea des Golfes von Neapel." ' Mitth. Zool. Stat. Neapel,' Band ix.,

Heft 4.
1865. Kolliker, A. von. — "Icones Histiologies," Leipzig.
1872. KOlliker, A. VON. — " Anatoraische systematische Beschreibung der Alcyonarien. — I. Abth. Die

Pennatuliden." ' Abhandl. Senckenberg. Nat. Ges.,' Band vii. und viii.
1877. Koren and Danielssen. — 'Fauna littoralis Norvegise,' Heft. iii. Bergen.
1883. Koren and Danielssen. — ' Nye Alcyoniden.' Bergen.

1896. Kukenthal, W. — " Alcyonaceen von Ternate." 'Abhandl. Senckenberg. Nat. Ges.,' Band xxiii.
1902. Kukenthal, W.— "Versuch einer Revision der Alcyonarien. — I. Die Familie der Xeniiden."

' Zool. Jahrbiicher,' xv., pp. 635-662.

1902. Kukenthal, W. — " Diagnosen neuer Alcyonarien aus der Ausbeute der Deutschen Tief-See-

Expedition." ' Zool. Anzeig.,' xxv., pp. 229-303.

1903. Kukenthal, W. — "Versuch einer Revision der Alcyonarien. — II. Die Familie der Nephthyiden."

' Zool. Jahrbiicher,' xix., pp. 99-172.

1899. May, W. — "Beitrage zur Systematik und Chorologie der Alcyonaceen." 'Jenaische Zeitschrift

f. Naturwiss.," xxxiii.
1862. Mobius, K. — "Neue Gorgoniden des Natur-Histor. Museums zu Hamburg." 'Nova Acta Akad.
Leopoldina,' Band xxix.

1900. Putter, H. — "Alcyonaceen des Breslauer Museums." 'Zool. Jahrbiicher,' Band xiii., Heft 5.

1882. Ridley, S. O. — "Contributions to our Knowledge of the Alcyonaria." 'Ann. Mag. Nat. Hist.,

IX., Part I., p. 184 ; X., Part II., pp. 125-133, 1 pi.

1883. Ridley, S. O. — "The Coral Fauna of Ceylon, with Descriptions of New Species." 'Ann. Mag.

Nat. Hist.,' vol. xi., pp. 250-262.

1884. Ridley, S. O.—' Eeport on Zoological Collections made in H.M.S. " Alert." '

1887. Ridley, S. O. — " Report on the Alcyonid and Gorgonid Alcyonaria of the Mergui Archipelago.
' Journ. Linn. Soc. London.,' vol. xxi.

1896. Schenck, A.—" Clavulariiden, Xeniiden und Alcyoniiden von Ternate." 'Abhandl. Senckenberg.

Nat. Ges.,' Band xxiii.
1887. Studer, A. — " Versuch eines Systems der Alcyonaria." ' Archiv Naturges.,' liii.

1889. Studer, A. — " Supplementary Report on the Alcyonaria." ' "Challenger " Reports,' vol. xxxii.

1890. Thurston, E. — " Notes on the Pearl, &c, Fisheries of the Gulf of Manaar." ' Government Central

Museum, Madras.'
1863-1869. Verrill, A. E.— "List of the Polyps and Corals." 'Bull. Mus. Comp. Zool. Harvard,'

vol. i.
1883. Verrill, A. E. — "Report on the Anthozoa and on some Additional Species dredged by the

'Blake '(1877-1879), and by the U.S. Fish Commission Steamer 'Fish Hawk' (1880-1882)."

'Bull. Mus. Comp. Zool. Harvard,' vol. xi., No. 1.

1897. Whitelegge, T. — "Alcyonaria of Funafuti." 'Mem. Australian Museum,' iii., pt. 5, pp. 307-20,

3 pis.
1889. Wright, E. P. and Studer, A. — " The Alcyonaria." ' "Challenger" Reports,' vol. xxxi.

ALCYONARIA. • 327

EXPLANATION OF PLATES.

PLATE I.

Fig. 1 . Expanded polyps of Acamptogorgia spinosa, var. eeylomensis, n. x 25.
Spongodes splendms. x 20.
Eunephthya purpurea, n. sp. x 20.
Nephthya ceylonensis, n. sp. x 20.
Spongodes pulchra, n. sp. x 20.
Spongodes armafa, var. ceylonensis, n. x 20.
Siphonogorgia kollikeri. x 4.
Aslromuricea ramosa, n. sp. x 12.
Spongodes aurantixica* n. sp. x 20.

PLATE II.

Fig. 1. Stenogwgia ceylonensis, n. sp. x 12.

,, 1a. Branching of Slenogorgiu ceylonensis, n. sp. Nat. size.

„ 2. Paraspongodes striata, n. sp. x 1£.

„ 3. Acis indica, n. sp. x 15.

„ 4. Capnella manaarensis, n. sp. x 15.

„ 5. Virgularia tubcrculala, n. sp. x 20.

„ 6. ParanepMhya pratti, n. sp. x 25.

,, 7. Paraspongodes striata, n. sp. x 15.

„ 8. Acanihogorgia media, a. sp. x 20.

,, 9. Virgvdaria tuberadata, n. sp. x 12.

PLATE III.

Bebryce hicksoni, n. sp. x 12.

Blanching of Muricella ramosa, n. sp. Nat. size.

Mnricella ramosa, n. sp. x 8.

Ven-ucella flexuosa, var. aurantiaca. x 2 J.

SI yhibelemnoMes herdmani, n. gen. et sp. x 5.

Lophogorgia lidkeni. x 1J.

Acanihogorgia media, n. sp. x 20.
„ 8. C'lamilaria margaritiferm, n. sp., spreading on an oyster shell, x 2.
,, 9. ^/«.s alba, n. sp. x 12.

PLATE IV.

Fig. 1. Scirpcaria sp. Nat. size.
„ 2. Virgxdaria sp. xl|.

„ 3. Sfylobelemnoides herdmani, n. gen. et sp. x 2.
„ 4, 5. Juncella gemmacea. Nat. size.

„ 6. Acanthogorgia ceylonensis, n. sp. x 3. The stem and branches have
been made much too substantial.

* This has been by mistake printed S. flabellifera on the plate.

Fig.

1.

»»

2.

J»

3.

1)

4.

»>

5.

»»

6.

i)

7.

328

CEYLON. PEARL OYSTER REPORT.

Fig. 7. Scirpea/rella aurantiaca, n. sp. Nat. size.

„ 8. Verrucella flemiosa, var. aurcmtiaca, n. Nat. size

„ 9. Femnrlhi flexuosa, var. gallensis, n. Nat. size

,, 10. Leptogorgia (wstraliensis. x2.

., 11. Acanthogorgia muricata, var. indica, n. Nat. size

„ 12. Lophogorgia rubrotincta, n. sp. x 2.

„ 13. VernifdJa rubra, n. sp. Nat. size.

PLATE V

Fig. 1. Spongodes rosea.

2. Mwrkella ramosa, n. sp.

3. Acanthogorgia muricata, var. indica, n.

4. ^4as crifta, n. sp.

5. Euncpht hi/a purpurea, n. sp.

6. Spongodes aurantiaca, n. sp.

7. ^4as indica, n. sp.

8. Verrucella rubra, n. sp.

9. Stenogorgia ceylonensis, n. sp.

10. Spongodes pukhra, n. sp. (From stem and branches only.)

11. Verrucella flexuosa, var. gallensis, n.

12. Acanthogorgia ceylonensis, n. sp.

13. Spongodes bicolor, var. <Mia, n.

1 4. Capndla manaarensis, n. sp.

15. Smpearella aurantiaca, n. sp.

16. Scirpearia sp.

1 7. Astromuricea ramosa, n. sp.

18. Paranephthtja pratti, n. sp.

PLATE VI.

Fig. 1. Echinagorgia multispiwsa, n. sp. x 15.
,, 2. Heterogorgia verrilli, n. sp. § nat. size.
„ 3. ^ezs ceylonensis, n. sp. x 12.
,, 4. Murkella ceylonensis, n. sp. § nat. size.
„ 5. Bellonella indica, n. sp. x 2L
„ 6. Echinomuricea ceylonensis, n. sp. § nat. size.

Lophogorgia irregularis, n. sp. Nat. size.

Schpearella divisa, n. sp. Nat. size.

8.
9.

Bebryee hicksoni, n. sp. s nat. size.

CEYLON PEARL OYSTER REPORT

ALCYOXARIA. PLATE I

\ ^

,

^"^<<^C'

*\

■ I

8

Fig. 1, Acamptogorgia snxosA, var. ceylonensis, n. ; Fig. 2, Spongodes splendens; Fig. 3, Eunephthya

purpurea, n.sp.; Fig. 4, Nephthya ceylonensis, n.sp. ; Fia. 5, Spongodes pulchra, n.sp. ;

Fig. 6, Spongodes armata, var. ceylonensis, n. ; Fig. 7, Siphonogorgia kollikeri ;

Fig. 8, Astromuricea ramosa, n.sp. ; Fig. 9, Spongodes flabellifera, n.sp.

CEYLON PEARL OYSTER REPORT

ALCYONARIA. PLATE II.

cOn,Cambridge-

Figs. 1, 1a, Stenogorgia ceylonexsis, n.sp.; Fig. 2, Parasposgodes striata, n.sp.; Fig. 3, Acis indica, n.sp. ;

Fig. 4, Capnella maxaarexsis, n.sp.; Fig. 5, Virgclaria tcberculata, n.sp.; Fig. 6, Paranephthya

prattii, n.sp. ; Fig. 7, Parasposgodes striata, n.sp. ; Fig. 8, Acanthogorgia media, n sp. ;

Fig. 9, Virgclaria tdberculata, n.sp.

CEYLON' PEARL OYSTER REPORT

ALC YON ARIA. PLATE III

del.

Fig. 1, Bebrtce hicksoxi, n.sp. ; Figs, l', 3, Muricei.la bamosa, n.sp.

vat. aurantiaca ; Fig. 5, .Sttlobelemxoides herdjiaxi, n.g. et sp. ;

Fig. 4, Verrucella fiexitosa,
Fig. 6, Lophogorgia i.itkexi;

Fig. 7, Acaxthogorgia media, n.sp. ; Fig. 8, Clavui.aria margaritifer.e, n.sp. ; Fig. 9, Acis alba, n.-p.

CEYLON 1'EAHl. OYSTER REPORT

ALC YON ARIA. PLATE IV

i, Scirpearia, sp.; Fig. 3, Stylobelemnoides berdmani, n.g.,et.sp.;

Figs. 4, 5, Junceixa gemmacea; Fig. 6, Acahthogorgia ceylonensib, n.sp.; Fig. 7, Scirpearella

ACRANTIACA, n.sp. ; FlG. 8, VERRUCELLA FLEXUOSA, var. aurantiaca, n. ; Fig. !), Verrucella flexdosa,

var. gallensis, d. ; Fig. 10, Leptogorgia acstraliensis ; Fig. 11, Acanthogorgia mdricata, var. indica, n. ;

Fig. 12, Lophogorgia m brotincta, n.sp.; Fig. 13, Verrucella rubra, n.sp.

CEYLON PEAHL OYSTER REPORT

ALCYJQNARIA . PLATE V

2

£ Wilson , C&mt ■

SPICULE S
( For list of Species see ion of Plate;; )

CEYLON PEARL OYSTER REPORT.

i.

ALCYONARIA. PLATE VI.

FlG. 1. ECHINOGORGIA MULTISPIXOSA, tl.sp. : FlG. 2, HETEROGORGIA sp. : FlG. ■'!. AdS CEYLONENSIS, n.sji. ; FlG. 4, HURICELLA

i eyloxexsis, n.sp.; Fig. 5, Bellonella indica, n.sp.; Fig. 6, Ei hinomi ru ea ceyloxexsis, n.sp;

Fig. 7. Lophogorgia irregularis, n.sp. . Fig. 8, S< irpearella divisa, n.sp.;

F i < ; . '.'. Beeryce hicksoxi, n.sp.

[CEYLON PEARL OYSTER FISHERIES— 1905— SUPPLEMENTARY REPORTS, No. XXL]

REPORT

ON THE

OPISTHOBRANOHIATE MOLLUSCA

COLLECTED BY

Professor HERDMAN, at CEYLON, in 1902.

BY
GEORGE P. FARRAN, B.A.,

ASSISTANT NATURALIST (FISHERIES BRANCH) DEPARTMENT OF AGRICULTURE AND
TECHNICAL INSTRUCTION FOR IRELAND.

[With SIX PLATES.]

There are about 50 species of Opisthobranchs in tbe collection, and of these 16 are
now described as new to science.

The Report is divided into two sections, I. Nudibranchfata and II. Tecti-
branohiata, with an Appendix containing (1) Onchidium and (2) Marsenia.

I. NUDIBRANCHIATA.

The collection of Nudibranchs made by Professor Herdman in Ceylon and handed
over to me for identification amounts to 30 species. Of these 9 must be described as
new, 12 are fairly well known tropical species, 4 are now identified with species
previously described by Abraham or by Quoy and Gaimard, but of which little was
known, and 5 are too small to identify satisfactorily.

The list of species is as follows : —

Hervia ceylonica, n. sp. Doto sp. 1

Galvi'na prodwta, n. sp. Melibe fimbriate, A. and H.

Eolis sp. 1 Scyllcea pelagica, L.

,, sp. t Plewrophyllidia formosa, Kel.

2 u

330 CEYLON PEARL OYSTER REPORT.

IAnguella cinerea, n. sp. Casella cincla, Bgh.

Eexabranchus marginattts (Q. and G.). ( ' ratosoma cornigerum, Gray.
Iim-wlwis hibifrra (Am;.). „ ornatum, Bgh.

Platydcris inframaculata (Aer.). Dons sp. ?

„ speciosa (Abi;.). „ sp. ?

„ herdmani, n. sp. Doriopsis awrea, Q. and G.

,, (?) spinulosa, n. sp. Phyllidia varicosa, Leuck.
Halgerda punctata, a. sp. „ nobilis, Bgh.

Thordisa (?) caudata, n. sp. jEgires villoms, a. sp.

Chromodoris reticulatus, Pse. Trevelyana tricolor, A. and H.

,, tenuilinearis, n. sp. Kalinga ornata, A. and H.

It is, perhaps, premature to institute comparisons between the Nudibranch fauna
of Ceylon and those of other localities, as so little is known comparatively of the
distribution of the group in the warmer seas, and as the identification of spirit
specimens from written descriptions is attended with great uncertainty.

The capture of four species of Eolids and a Doio on the hanging oyster cages is ot
interest, as it seems to show that absence or rarity of these families in tropical seas
is, perhaps, due more to the difficulty of finding them than to their actual scarcity,
the larger and more showy Dor ids attracting the attention of collectors to the
exclusion of more insignificant forms.

There has been but little work done in the past on Ceylonese or Indian
Nudibranchs. Dr. Kelaart, in 1859 (3) (4), described 33 species of Doris from
specimens collected in Ceylon, but his descriptions are based mainly on the colour and
texture of the living animal. Alder and Hancock (5), in examining a collection of
Nudibranchs from Madras, re-discovered six of Kelaart's species with the aid of his
original drawings, and added 32 more, besides recording three others previously
described.

Of more recent works should be mentioned a paper by Collingwood on Nudi-
branchs from the Eastern seas (8) and a list of the Dorididse in the British Museum
Collections bv Abraham (7), who there describes the external characteristics of a
number of spirit specimens. Reference should also be made to the work of
Sir Charles Eliot, who has published papers on the Nudibranchs and Tectibranchs
of Samoa, the Maldives and Laccadives, and East Africa (11) (12) (13).

Most, however, of the adequate descriptions of Oriental Nudibranchs are due to
the long-continued work of Dr. Rudolph Bergh, culminating in the magnificent
series of papers in Semper's ' Reisen im Archipel der Philippines'

In working out the collection no attempt has been made to investigate the anatomy
of the specimens beyond what seemed to be necessary for their identification.

In conclusion, I should like to express my thanks to Professor Herdman for his
kindness in allowing me to work out this collection and use his notes, and for the
facilities which he atforded me, and also to Mr. E. A. Smith for kindly allowing me
to examine the Nudibranchs preserved in the British Museum Collections.

OPISTHORKANCHIATE MOLLUSCA 331

NUDIBRANCHIATA CLADOHEPATICA
Family: MOHDITDM.

Sun- f A M l l y : CKATENIIN M .

Hervia ceylonica, n. sp. — -Plate I., figs. 1 to 5.

Amongst the Nudibranchs obtained from oysters hung over the ship's side, in the
Gulf of Manaar, were 12 specimens of a small Eolid, length 8 millims., which seems
to belong to the genus Hernia. This genus is closely allied to Cratena, and is dis-
tinguishable externally by the large tentacles, smooth rhinophores, grouped papillae,
and produced anterior angles of the foot ; and internally by the single series of arched
teeth strongly denticulate on either margin.

The specimens here described have the body pale and semi-translucent, the papillae
with fawn-coloured hepatic contents, probably faded, long and slender, with distinct
white cnidogenous sacs. They are arranged in 5-6 irregular transverse rows, the
first row widely separated from the rest. The tentacles are long, very thick at the
base, and tapering evenly to a point. The rhinophores are short, slender, and smooth,
the eyes being visible at their base. The anterior angles of foot are moderately
produced, the tail slender and moderately long (Plate I., fig. 1).

The jaws are of the usual shape in the Cratenidae, with rather short cutting edge,
irregularly denticulate. The radula has a single row of arched teeth, with an acute
apex and 5 strong lateral denticulations on either side (Plate I., fig. 2).

The genital organs did not show any sign of armature*.

Bergh refers three species to this genus, two from the Mediterranean, and one,
H. rosea; Bgh., from Amboyna ; and in addition to these Alder and Hancock (5)
have described another species from Ceylon, under the name of Eolis militaris, which
appears to be undoubtedly a Hervia, though Bergh (9), probably by an oversight,
refers it to Coryphella. It is distinguishable from the present species by its much
larger size and the possession of 7 lateral denticulations on the radula teeth.

Kelaart has given names to 8 species of Eolis from Ceylon, but his descriptions
are so vague as to be practically worthless.

«
Sub-family: TERCtIPEDIN^E.

Galvina producta, n. sp. — Plate I., figs. 6 to 9.
In company with the last-mentioned species was one specimen whose characters,
with one exception, agree with those of the genus Galvina.

Length, 9-5 millims., papillae inflated, tapering, constricted below the tips, which
are somewhat opaque, hepatic contents vesicular-looking, rhinophores moderately long,
tentacles short, anterior angles of foot produced.

Colour, faded, a light transparent brown without markings ; hepatic contents

2 u 2

332 CEYLON PEARL OYSTER REPORT.

whitish-yellow ; jaws and teeth as in the genus Galvina. The median teeth of the
radula robust, with short point, and 5, sometimes 6, lateral denticles. This species
differs from the other members of the genus in having the anterior angles of the foot
produced, but otherwise it is a typical Galvina. All hitherto recorded Galvinse are
either Atlantic or Mediterranean, with the exception of G. crigua, which Bergh (15)
has recorded from the Pacific. This adds the Gulf of Manaar.

Eolis, spp. ?

There are two other small Eolids, representing two different species, in the
collection, both from the oyster cages hung over the ship's side in the Gulf of Manaar,
which are too small to identify. One of these, length 4 millims., is, perhaps, a Hervia.
It has the anterior angles of the foot produced, the tentacles long and slender, the
rhinophores small ; the papillse are crowded, long and slender, with acute white
cnidogenous tips and dark greenish-black contents. The body colour is a transparent
reddish-yellow.

The other sjiecimen, length 4 millims., has the papillse in 5 distant transverse rows ;
the larger papilla? are ringed in a manner that recalls the genus Zatteria, described
by Sir C. Eliot (11, p. 62) from Zanzibar specimens. The rhinophores are moderately
large and rugose, the tentacles small and the anterior angles of the foot produced.

Family: DOTONID^E.
Doto, sp. ?

One specimen of a Doto, length 2 '5 millims., too small to identify, was obtained
from the above-mentioned oyster cages in the Gulf of Manaar. Externally it bears
a close resemblance to D. pinnatifida, Mont., the form, coloration, and black spots on
the papillar tubercles being in agreement with that species.

Bergh (9, p. 795) has recorded a species of Doto (D. indica) from Mauritius, to
which it is possible that the specimen should be referred.

Family : TETHYMELIBID^.

Melibe fimbriata, Ald. and Hanc.

There are three specimens of this species from separate localities on the West
Coast of Ceylon. One from off Negombo, one from Periya Paar, (J fathoms, and one
from Chilaw Paar, 11 fathoms. As usual, they are in a very imperfect state, having
lost several of their dorsal papillae.

Professor Herdman notes in his diary the colour of the living animal as follows :
" Pale amber brown, in front, with brownish papillse and white sjjots — clear,
gelatinous." This species, as was first pointed out by Alder and Hancock (5), and
subsequently confirmed by Sir Charles Eliot (11), differs from the rest of the genus
in having no trace of labial or buccal armature.

M. fimbriata is found throughout the Indian Ocean.

OPIRTHOBRANCHIATE MOLLUSCA. 33.3

Family: SCYLL^EID^.
Scyllaea pelagica, Linn.

Amongst the nudibranchs found on the oyster cages which were kept hung over
the ship's side from February 15 to March 15, 1903, in the Gulf of Manaar, were
two specimens of Scyllcea which agree fairly closely, both in general appearance and
also in form of radula and jaws, with S. pelagica. Externally they also rather closely
resemble S. marmorata, Ald. and Hanc, from Madras, agreeing with it in the
rounded dorsal wings and slight development of the caudal ridge, but it seems to me
that there are hardly sufficient grounds for regarding the latter as a distinct species.

The colouring of the Ceylon specimens, the largest of which measures 3'0 centims.,
is mainly due to large irregular areas of light-brown stellate chromatophores, amongst
which are vacant pale blotches. There are scattered dark spots formed by aggregations
of stellate chromatophores which seem to correspond to sub-dermal patches of pigment
and also a few small ocellated spots, the centre formed of contracted brown
chromatophores, possibly of a different colour during life, and the surrounding paler
area of very finely-branched chromatophores. These ocelli occur on the sides and in
a row down the centre of the back.

The body bears a few small colourless tubercles. The buccal mass and, more
noticeably, the oesophagus are pigmented with brown stellate chromatophores.

Family : PLEUEOPHYLL1DIID.E.

Pleurophyllidia formosa, Kel. — Plate I., figs. 14 to 16 ; Plate II., fig. 1.

One specimen from northern part of the Gulf of Manaar.

Length of preserved specimen 5 2 centims., width 4 centims., width of foot
2 "3 centims. Colour an opaque greyish-brown, paler on the sides of the foot, with
remains of black pigment between the dorsal ridges and on the dorsal surfaces of the
tentacle shield.

The form of the animal corresponds closely to the figure given by Bergh (10,
plate i., fig. 1). The tentacle shield is broad, with thin projecting upper margin.
The back is marked with about 36 slightly raised longitudinal lines, larger and
smaller ones irregularly alternating.

The gill lamellae are about 5 millims. in length, and lie in a crowded transverse
band, which runs at about right angles to the direction of the anterior lateral lamellae.
The lateral lamellae are numerous, about 50, and run transversely from the body
almost to the margin of the mantle, the direction of the anterior ones being somewhat
diagonal.

The jaws (Plate I., fig. 14) are powerful, of a dark horn colour, and of the usual
form. The radula (Plate I., figs. 15 and 16) consists of 70 rows of 140-1-140 teeth.
The median tooth is very broad (0-3 millim. x 0"12 millim.), with four lateral
denticles and a strong, rather blunt median point with toothed margins, the basal

334 CEYLON PEARL OYSTER REPORT.

pair of denticulations being very large. The 1st lateral tooth is strong, with a broad
somewhat rectangular base and an irregularly serrulate or smooth outer margin. In
the 2nd tooth the margins are smooth and the point entire. The 3rd tooth has a
small sub-apical denticle on the outer margin. This denticle is found on all the teeth
outwards till the outermost one is reached ; it increases in size outwards, so that the
lateral teeth are for the most pai*t bifid or, in a few cases, trifid. The outermost
tooth is simple. This specimen differs slightly as to its radula from that examined
by Bergh, in which the apical denticulation begins on the 5th tooth and is absent
from the 3-6 outermost. It agrees, however, in other respects. The other Ceylon
species, P. ceylanica, has, like P. formosa, the lateral teeth bidentate, but differs in
having a much smaller number of lateral teeth and in having the outermost teeth
finely serrulate.

Lingiiella cinerea, n. sp. — Plate I., figs. 10 to 13.

A single specimen of this species was dredged 10 miles north of Cheval Paar.

Length about 1'5 centims., width 1*4 centims., width of foot 0*9 centim.

Mantle moderately smooth, of a greenish-grey colour, which is denser on the head
and front of cloak. The margin of the cloak studded with distant tubercles marked
with a ring of grey, becoming larger towards the centre of the back and merging into
irregular raised areas separated by deeply pigmented grey furrows.

Lateral margin of underside of mantle with a band of opaque yellow cnidopores,
2 to 3 deep. Lateral lamella? about 32, slightly oblique.

Radula of 35 rows, formula 46-1-46. The central tooth (Plate I., fig. 10) has a
median spine with 3 lateral serrulations, and 3 lateral denticulations. The 1st
lateral tooth with about 9 denticulations on the outer edge. The teeth increase in
length outwards, the number of denticulations decreasing (Plate I., fig. 12), and the
two outermost teeth being smooth (Plate I., fig. 13).

The jaw plates (Plate I., fig. 11) of the usual form, strongly curved (in the figure
they are flattened out), of a dark horn colour.

This species seems outwardly to come rather near to L. sarasinica, Bgh., from
Ceylon, but differs in not having the tubercles densely crowded on the mantle edge.
The radula also is very different both in form and number of teeth.

NUDIBRANCHIATA HOLOHEPATICA.

DO RIDIAD^ CRYPTOBRANCHIAT.E.)

Sub-family : HEXABRANCHID^E.

Hexabranchus marginatus, Q. and G.

There are in the collection two specimens of Hexabranchus, both from Periya Paar,
which I think it is probable belong to this species. One specimen is very small,

Ol'ISTHOBRANCHIATE MOLLUSCA. 335

length only I'l centims.. the other is somewhat larger, length 3 5 centims., but both
fall far short of the size usually attained by members of the genus, and are probably
immature.

Professor Herdman's notes, referring to t lie larger specimen, are " scarlet and
white, swimming species," and "red blotched with white, especially at sides of
mantle." The spirit specimen is somewhat depressed, of a dirty semi-translucent
white colour. The mantle wide and thin and frilled at the margins, total width
2"4 centims., free margin 0-55 centim. The free edge of the mantle in front of the
head is 2-5 millims. wide and has a smooth entire margin.

Rhinophores far apart, of the usual shape, with about 28 lamella?. Tentacles
large; flattened, with crenated margins. Branchias in 6 groups surrounding the short
tubular anus.

The jaws are powerful, with about 8 transverse ridges.

The radula has about 32 rows of teeth, the formula being 46-0-46. The innermost
2-3 teeth on each side are imperfect.

The identification of species of Hexabranchus from preserved specimens is, as both
Bergh and Eliot (12) have remarked, a matter of difficulty, but in this instance
there seems little doubt that the specimen is Hexabranchus marginatus, as the white
marginal blotches, crenate margin to tentacles, and distinct anal tube, as well as the
agreement of the radula, seem to indicate.

Collingwood's Albania formosa (8, p. 132), from Formosa, seems to be undoubtedly
the same species, as also, perhaps, Alder and Hancock's " Undescribed species of a
new genus apparently related to Goniodoris," from Madras (5, Plate XXXIII.,
fig. 20).

Sub-family : DISCODORIDID^E.
Discodoris labifera (Abr.). — Plate I., figs. 17 and 18.

There are three small specimens of Discodoris, one from the pearl banks oil Aripu
and two from Oheval Paar, which I doubtfully refer to the above species. The
animal is quite unrecognisable from Abraham's original description (7) ; but on
examination of what appeared to be the type specimen in the British Museum, I found
that, making allowance for difference of size, it agreed fairly well in external features
with the Ceylon specimens. The British Museum specimen is undoubtedly a Disco-
doris, as the partially protruded buccal armature is distinctly visible, though there is
no mention of this in Abraham's description. There are also in the British Museum
collection some unnamed specimens from South Africa which apparently belong to the
same species.

The Ceylon specimens measure about 1*6 centims. in length, and the width of
the most symmetrical is 0'95 centim. The colour of the Aripu and one of the Cheval
Paar specimens is a dirty greyish-brown. The other Cheval Paar specimen, though
much distorted, has preserved its colour and texture better than the rest. It is of a

336 CEYLON PEARL OYSTER REPORT.

slaty blue-grey colour above aud below. The mantle has a number of moderate
brownish spots, absent from the centre and extreme margin. The underside of the
mantle and foot are finely dotted with brown. The mantle is covered with fine
tubercles, somewhat pointed, giving it a slightly pilose appearance ; this is not so
marked in the other specimens. The branchiae are 6 in number, compoundly
pinnate. The branchial opening is placed far back and has the form of a transverse
slit. The jaws are of the usual form, made up of blunt rods ; colour, a pale brownish-
straw. Radula with 18 rows, the last two immature, formula 26-0-26, the teeth
increasing in size gradually outwards (Plate I., fig. 18), the outermost four teeth
again rapidly decreasing (Plate I., fig. 17).

Abraham's original specimen of D. labifera came from the Seychelles Islands.

Sub-family : PLATYDORIDID^.

Platydoris inframaculata (Abr.). — Plate II, figs. 2 to 7 ; Plate III., figs. 8, 9.

Amongst the Nudibranchs obtained at Galle is one well-preserved specimen of
Platydoris, which is, I think, referable to the above. The general appearance,
markings, and tuberculation of the mantle of the type specimen from Amboyna
preserved in the British Museum agree fairly well, but it is impossible to decide with
certainty upon the identity of closely allied species of Platydoris without an exami-
nation of the radula and the usually very characteristic armature of both penis and
vagina.

The ground colour of the Ceylon spirit specimen is pale reddish -brown, the mantle
having a few irregular dark rusty markings formed by minute reticulations round the
tubercles ; the lower surface of the mantle bears large circular purplish-black blotches,
rather larger and fewer than in the type specimen. The mantle is much depressed
and covered with minute, close-set. spiculose tubercles. The rhinophore pores are
slightly raised. The branchiae are 6 in number, large, 3-pinnate surrounding
the long, tubular anus, partially retracted into a wide, deep collar (Plate II., fig. 3).
The total length of the animal is 47 centims., width 3 centims., width of foot
1 centim.

The radula consists of 44 rows ; there are 70 lateral teeth on each side and no
median tooth. The teeth are simply hooked, increasing in size outwards (Plate II.,
fig. 4; Plate III., figs. 8 and 9).

The retracted penis (Plate II., figs. 5 and 7) is armed internally with numerous
hook-bearing discs, crowded and rather large below, becoming smaller and more
distant as it passes upwards, and taking the form of small prickles on the glans.
The vaginal wall bears 6 stout, chitinous discs (Plate II., fig. 6) arranged in opposing
pairs. These discs are peltate, the boss being formed of 5-6 rounded nodules,
and the whole surface being finely granulated. They measure 0'65 millini. in
diameter. The size of the largest penial discs is about 0'19 millim. by 0*155 millim.

OPISTHOBKANCHIATE MOLLUSCA. 337

Platydoris speciosa (Abr.). — Plate II., figs. 8 to 12, and Plate III., fig. 10.

The type of this species from Amhoyna is, like the preceding, preserved in the
British Museum, and has heen described by Abraham (7), but nothing is known of
the form of its radula or genital organs. A specimen of Platydoris from the Gulf
of Manaar appears to belong to this species. Its length is 57 centims., width
4-5 centims., width of foot, much contracted, 0*9 centim. The mantle is very wide,
covered above with closely-set, minute tubercles with radial spicules. The ground
colour of the specimen is an opaque muddy brownish -grey. The mantle is marked
above with collections of small, purplish-brown spots, irregularly placed over the
whole surface.

The lower sides of the mantle are marked with large, purple-brown, circular
1 ilotches, with smaller ones on the side of the body. The rhinophore pores have their
margin slightly raised. The branchial pore is 6-lobed, stellate, tightly closed over
the retracted branchiae. The walls of the pore are moderately high. The branchiae
are 6 in number, compoundly pinnate surrounding the long tubular anus, the margin
of which is crenulated.

The radula (Plate II., figs. 8 to 10 ; Plate III., fig. 10) contains 46 rows, with
about 58 teeth on either side of the naked rachis. The teeth are simple, hooked,
decreasing in size slightly towards the centre of the radula. Compared with those
of the previous species, they are somewhat smaller and much more slender.

The penis (Plate II., figs. 11 and 12) is armed as usual with hooked discs, rather
large and closely set below, much smaller surrounding the glans, which bears a few
still smaller spines. The armature of the vagina, owing to an accident during
examination, could not be noted.

This, or the preceding species, may perhaps prove to be identical with the Doris
formosa or Dons elliotti described by Alder and Hancock from Madras, but this
cannot be determined without an examination of specimens in their natural colours.

Platydoris herdmani, n. sp. — Plate II., figs. 13 to 17.

A single specimen of this species was taken at Galle, and three others, without
recorded locality, are in the collection. Length 3 centims., width 1*4 centims., width
of foot, edges approximated, 0"4 centim., height 0'5 centim., length of foot 2'1 centims.

The mantle is large, minutely tuberculate, with its edges thin and very much
frilled. Its centre is marked by a large, irregular, pale reddish-brown blotch, spotted
with darker, and breaking up into smaller spots towards the margins of the mantle.

The extreme margin of the mantle is marked by a series of black sub- dermal spots,
equally visible from above and below. The under side is very faintly dotted with pale
brown, the margins of the foot being marked with darker brown.

The ground colour is a waxy flesh colour. The texture of the body is stiff and
brittle, the spicules are numerous, small, slightly roughened, and arranged in dense,
closely-packed, stellate clusters. The branchial and rhinophoral pores are raised and

2 x

3.38 CEYLON PEARL OYSTER REPORT.

closed, not distinctly stellate. The tentacles are comparatively large and conical.
The foot is very narrow, with its margins approximated in all the specimens.

The radula consists of 44 rows with no median tooth and about 70 lateral teeth on
each side increasing gradually outwards from the centre. The outermost tooth is
rudimentary, without a lamina, the 2nd is fringed, the 3rd broad and flattened
with a slightly denticulate edge, the 4th normal.

The penis is armed with comparatively few small, distant, hooked plates which are
not found on the glans. The vagina bears four large chitinous hooks arranged in
pairs one above the other on opposite sides of the lumen, This species belongs to
quite a different group of the genus from that of the two preceding. It appears to be
distinct from any described species, though approaching somewhat to P. angustipes,
Bgh., from the Antilles.

Platydoris (?) spinulosa, n. sp. — Plate II, figs. 20 to 23 ; Plate III., figs. 1 to 3.

There is one specimen of an apparently undescribed Dorid from 10 miles north-east
of Cheval Paar, which I have provisionally referred to the genus Platydoris, though
it differs from it in some particulars.

Its length is 2 "8 centims., width 1"9 centims., height 0*8 centim. The mantle is
wide with straight entire margin, covered above with moderate unequal tubercles
placed fairly widely apart on the centre of the back and more closely near the margin,
but not crowded anywhere. The largest tubercles lie in an irregular ring, 3 to 4
deep, parallel to and some distance from the edge of the mantle, the centre and
margin being occupied by smaller tubercles. The tubercles are rounded or slightly
conical, spiculose, the ends of the spicules protruding distinctly. There are also
larger spicules arranged in reticulating bundles between the tubercles.

The foot is contracted and rather narrow, the front margin bilaminate and divided.
The tentacles are very small, pointed and recognisable with difficulty.

As to colour, Professor Herdman has made a note in his diary on a specimen from
this locality which I think there is little doubt refers to the species in question.
" Flat red Doris, 1 inch, spotted with dark purple on under-side, wide mantle."
There is little trace of colour remaining, the ground colour is white above and below,
and the larger tubercles surrounding the centre of the mantle have each a faint light
brown ring, the under-side is rather thickly spotted with darker brown, the spots
having a deep nucleus and a fainter halo ; their diameter varies from 0-5 millim. to
1 '5 millims. The rhinophores are white, the branchiae yellowish, shaded with brown
internally at their base, rachides white. The branchia? are 6 in number, triqjinnate,
the branches small, the rachides long and thick, joined at the base and each with
a median groove exteriorly. The branchial pore is raised, circular but not produced
into a tube. The rhinophoral pores are slightly raised.

The radula (Plate III., figs. 1 to 3) differs somewhat from that usually found in
Platydoris on account of the rapid decrease in size inwards of the innermost 14 teeth.

OPISTHOBRANCHIATE MOLLUSCA. 339

The number of lateral teeth on each side is 48, the 6 or 7 outermost being fringed
instead of simply hooked. The penis seems to have been armed as in Platydoris
with hooked plates, but the preparation was unfortunately mislaid before examination.
The vagina (Plate II., figs. 22 and 23) is very strongly armed with two longitudinal
rows of peltate discs, very convex, the boss formed of a single powerful spine ; there
are about 16 discs on each side opposite each other, but not arranged in pairs. They
are largest in the centre of the rows and become much smaller towards each end.

The stomach (Plate II., fig. 21) is of an unusual structure, the intestine leaving it
medio-dorsally instead of anteriorly, as is usual. The radula sac is long and curved
upwards on the left side of the oesophagus (Plate II., fig. 20).

This species differs from the usual type of Platydoris in the long radula sac, form
of the radula, structure of stomach and larger size of dorsal tubercles. It agrees
with it in its general flattened form and genital armature. It is probable that an
examination of further specimens will make it clear that a generic distinction exists.

Sub-family : DIAULULI1SLE.
Halgerda punctata, n. sp. — Plate III., figs. 4 to 7.

There are in the collection (Gulf of Manaar) two specimens of a moderately large
glabrous Dorid with the mantle raised into tubercles, more or less connected by
ridges, for which I propose the above name. I have placed it in the genus Halgerda,
as it agrees in most points with the single known species Halgerda form.osa.
Professor Herdman's notes describe what must be this species as follows : — " Dorid,
W. of Donnan's Paar, 29 faths., March 11th, 1902; jmle whitish-grey, rather
translucent, with yellow papillae on mantle, and purple spots on foot and a few also
on mantle."

The ground colour of both the preserved specimens is white on both sides of the
mantle, and on the foot. The mantle is flecked with a few black spots, which
measure from 0-5 millim. to 1*5 millims. in diameter, appearing to be of tougher
consistency than the rest of the body, and to extend to a slight depth. In one
specimen there are only four such spots, very small, near the posterior end, but in the
other they are more numerous, about twenty, arranged more or less symmetrically.
There is a pinkish tinge on the apices of some of the tubercles. The under-side of
the body of both specimens is marked with black spots similar to those found on the
notaeum, but more numerous, especially on the specimen which is more spotted above.
The size is about the same in both specimens: length about 3 centims., width
2'8 centims., height 1'5 centims. The mantle is glabrous, moderately tough, and
bears numerous large blunt'tubercles arranged in more or less diagonal lines connected
by ridges which form an irregular triangular network. This arrangement is much
more distinct in one specimen (figured on Plate III., fig. 4) than in the other, which
has much fewer tubercles, and in which the ridges are almost obsolete. The branchial

2x2

340 CEYLON PEAEL OYSTEE REPORT.

opening is circular and measures about 0'6 millim. in diameter, its edges are thin
and raised into a collar with smooth even margin. The rhinophoral pores are oval,
placed longitudinally with even edges somewhat raised. The pseudo-peritoneum is
of a blackish -brown colour. The foot is moderately wide : in the preserved specimen
it is much contracted and measures 0-45 centim. across. It is bi-lipped in front, the
anterior lip bifid. The head is small, the tentacles very small and apparently
tubercular.

The branchiae (Plate III., fig. 6) are 4 in number and tri-pinnate, the anterior
pair larger, the posterior pair shorter and divided deeply into two main branches.
The rachides are rather thick, white on the outer side and black interiorly. The
lamellae are black. The anus is rather long and tubular, white with a black crenulate
margin. The rhinophores are stout, with black lamellated tips and thick white base.

The radula sac is unusually long and curved towards the left side. The radula
consists of 50 rows of hooked teeth ; the lateral teeth on each side number about 50,
a median tooth being absent. The 20 innermost teeth (Plate III., fig. 5) are very
small, increasing outwards ; the 3 outermost teeth (Plate III., fig. 7) are small and
thin, but not fringed or denticulate. There is no buccal armature, but the buccal
tube has a stiff chitinous lining in which were embedded several grapnel-shaped
sponge spicules.

The penis and vagina are as far as could be made out unarmed. The stomach and
gut were both crowded with tapering rod-shaped spicules roughened at the apex and
slightly bent at the base.

This species agrees with Bergh's diagnosis of the genus Halgerda in the following
particulars : The smooth leathery mantle raised into tubercular ridges, the bi-lipped
foot, small tentacles, small number of branchiae, absence of jaws, long curved radula
sac, radula without median tooth and with the innermost teeth markedly smaller,
unarmed penis and vagina. The genus Dictijodoris, though placed by Bergh in a
different sub-family, also approaches it in some respects.

Halgerda formosa from Mauritius differs from the species now described in having
more numerous and apparently narrow dorsal ridges, and in having two outermost teeth
of the radula on either side denticulate. It is marked with crimson spots and yellow
lines, and Professor Herdman states that the black spots found on H. punctata are
purple during life.

Thordisa (?) caudata, n. sp. — Plate II., figs. 18 and 19.

I have provisionally placed this species in the genus TJwrdisa, though one of the
most noticeable features of the genus, the villose mantle, is absent.

A single specimen was taken to the south-east of Cheval Paar in February, 1902.
The length of the animal is 2 '9 centims., width 1:1 centims., height l-6 centims. ;
the contracted foot is 0'8 centim. wide. The colour is a creamy white without
markings, the rhinophores grey, the branchiae white.

•OPISTHOBRANCHIATE MOLLUSCA. 341

The body is elevated, evenly rounded ; the mantle well developed posteriorly,
slight!}' longer than the foot, the postero-lateral margins bent inwards towards the
foot; the distance from the branchial opening to the end of the mantle 0-8 centim.
The branchial opening is oval, transversely placed, measuring 5 millims. by 3 millims.,
its edges raised into a low smooth collar. The edges of the rhinophore pores are
slightly raised. The texture of the mantle is firm and spiculose, it is covered with
rather small but not very crowded low truncated or flattened tubercles with the ends
of the spicules showing through their surface. The front margin of the foot is
bilobed, the head is small, the tentacles very small and flattened.

The radula (Plate II., figs. 18 and 19) consists of about 38 rows. The lateral teeth
on each side number 39. The 10 iunermost teeth are rather small, hooked, and have
a minute denticulation on the outer side. The three outermost teeth are small, the
two outermost being fringed and the next simply hooked.

There are both salivary and ptyaline glands present on the buccal mass.

The penis and vagina are unarmed, but it is possible, since the genital organs bear
a great resemblance to those of Jorunna, that the presence of a stylet may have
been overlooked.

Sub-family : CHROMODOEIDIN^E.

Chromodoris reticulatus, Pease.

One specimen of the above (= Chromodoris alderi, Collingwood) was taken at a
depth of 45 fathoms in the Gulf of Manaar, in February, 1902.

The colour when living, as noted by Professor Herdman, was " white speckled
with red-brown ; bright yellow border. Gills dark red and white ; long white tail."

The colour of the preserved specimen is a pale orange-yellow, with fine reddish
markings and reticulations, the latter rather wider towards the margin. Branchiae
simply pinnate, light red, with two deep red lines running up the inside of the rachis.
Rhinophores dark red. Margin of mantle with faint narrow orange line. General
colour effect reddish-orange, darkest anteriorly. Length of preserved specimen
2*8 centims., width 1'8 centims., height 1'8 centims., width of foot 1*1 centims. The
free margin of mantle is narrow, turned up posteriorly, with median posterior notch.
Branchiae 24, arranged in two spirals as in Casella. Pseudo-peritoneum tough, white,
but without nodules.

The radula corresponds very well with the figures given by Bergh for C. reticulatus,
and the labial armature is formed of bifid rods, as noted by him.

I think that there is little doubt but that this species is both the C. reticulatus of
Pease (6, p. 205), and the D. alderi of Collingwood (8).

The figure given by Collingwood is more like it than that which Bergh gives
after Garrett, and shows the notch at the end of the mantle which Bergh does not
allude to. The number of branchiae is given by Bergh as 9 to 15, and by Collingwood
as 12, but this seems to be of no specific importance in the genus.

342 CEYLON PEARL OYSTER REPORT.

I have no drawing of the form assumed by the animal when alive, but the shape
of the preserved specimen approaches so near to Casella that, taken with the manner
of arrangement of the branchiae, it raises a doubt as to whether the distinction
between the genera Chromodoris and Casella can be preserved or rather whether
the line of separation has not been drawn in the wrong place.

Chromodoris tenuilinearis, n. sp. — Plate III., figs. 11 to 15.

There is a single specimen of the above, taken on the Cheval Paar.

The length of the preserved specimen is 2-5 centims., its width OG centim. and
its height 0'8 centim. The mantle is smooth, long and narrow, slightly wider than
the foot and not reaching as far as it posteriorly. The foot has its margins adpressed
and extends behind the body for 0-4 centim.; width of the body 0'5 centim.

The ground colour is a dirty transparent white. The mantle is marked by inter-
secting diagonal broken lines of a pale greyish-green colour (Plate III., fig. 11).
The sides of the foot are marked with traces of similar lines.

The branchiae are 8 in number, long, slender and limp, simply pinnate, issuing
from a collar about 1 millim. high. The rhinophores are long, slender and colourless.

The radula consists of 66 rows, formula 71-0-71. One or two of the innermost
teeth on each side are trifid, the rest bifid. There is a trace of a median tooth on
the rachis. The outermost tooth is a flat plate with median notch, the 2nd tooth is
slightly hooked and distinctly bifid.

The jaws are strong, divided into 3 parts, one median and two lateral. They are
made up of hooked plates regularly arranged in alternating rows.

There was no further examination of the animal made. The markings are of a
rather unusual type in the genus Chromodoris, in which they usually consist of
brightly-coloured lines parallel to the sides of the mantle or longitudinally arranged,
or else of spots.

The armature of the jaws and radula are paralleled in several species, notably in
C. semperi, Bgh. (9, p. 482, plate lii., figs. 1 and 2; plate liii., figs. 13 and 14), in
which they are almost identical.

Casella cincta, Bgh.

One specimen from Muttuvaratu Paar, length 2-3 centims., has its colour fairly
well preserved. It is of a dark claret or plum colour, fairly uniform, with a dark
green margin to the mantle and, less evidently, to the foot.

The branchiae are about 28 in number, simply pinnate, reddish with a dark green
rachis. They are arranged apparently in two spirals, one on each side of the anus.
Anterior to the anus is one gill plume distinctly larger than the rest.

The rhinophores are finely laminate, plum coloured, with a tinge of green at
the tips.

The radula contains about 140 rows, formula about 63-1-63. The teeth are

OPISTHOBUANCHIATE MOLLUSC A. 343

hooked, with from 8 to 10 fine lateral denticulations on the outer side. The innermost
lateral tooth is denticulated on both sides.

The labial armature is formed of bent rods with pointed tips slightly hooked.

Professor Herdman's notes contain a sketch of the animal in a living state which
shows very well the characteristic shape, and very closely corresponds witli the figure
of C cincta given by Bergh (9, plate lxxviii., fig. 9). Professor Herdman describes
the colours in the living animal as follows : — " Pink with green edgings ; edge of foot
and mantle white, then a marked green line, then inside that a yellow band, then a
pink and white dotting covers the rest. Branchiae in two spirally arranged tufts ;
four large processes in front."

Ceratosoma cornigerum, Grey (?).

It is difficult to name with any certainty preserved specimens of Ceratosoma unless
the colour of the living animal has been noted, as the radula and jaws of all the
described species are almost identical and the external form is liable to a considerable
amount of variation. There are in the collection three specimens of Ceratosoma, all
very small, under 2 centims., which I have doubtfully referred to the above species.
They vary considerably in the length of the dorsal horn, but agree in having the line
of separation between the back and sides marked by a not very prominent ridge,
which is not produced into lateral lobes. They are of a uniform cream colour, one
specimen, from south-east of East Cheval, being distantly spotted with opaque white
dots. The other two specimens were taken at Talaivillu Paar and jr mile south-east
of Dutch Modragam Paar, respectively.

[Ceratosoma ornatum, Bgh.]

In Professor Herdman's notes there is a description with accompanying sketch of
a species of Ceratosoma, but the specimen to which it refers has apparently not been
preserved, or has disappeared from the collection. There is no doubt, however, that
the species is C ornatum, which has been described by Bergh from a specimen taken
in Amboyna (9, p. 946).

Professor Herdman's description, which is fuller than that given by Bergh, is as
follows : " Spotted all over with yellow on a pale mauve ground, deeper on dorsum,
getting to white on foot, with deep mauve or purple edgings to the front and lobes.
Horn with mauve edgings on pale ground ; tail with yellow spots on mauve. Head
with yellow spots on white. Front of foot with mauve spots on white. Rhinophores
brown at base, violet at tip, laminated. Anus violet edged. Gills white at base,
yellow in middle, brown at tip. Off Galle, 34 fathoms, February 18th, 1902."

The capture of a small violet-spotted specimen on March 8, 1902, towards the
north of the Gulf of Manaar, is also noted, but the specimen is absent. This may be
an undescribed species, as none of the described species are violet spotted, except as
subsidiary to other coloration.

344 CEYLON PEAEL OYSTER KEPOKT.

Doris, spp. ?

There are several small very much depressed Dorids from the east of the Gulf of
Manaar and the south end of Cheval Paar which I have not been able to identify.
They appear to belong to two closely allied species. One is represented by 8 specimens
from the Gulf of Manaar and 2 from Cheval Paar. The largest specimen measures
1*1 centims. by (P8 centim. The back is covered with uniform, blackish-brown, very
minute marbling or reticulation. The ground colour of the Manaar specimens is a
mahogany-brown, probably owing to the action of spirit ; of the Cheval specimens,
pale brownish-grey. The radula in the largest specimen consists of 45 rows of
18-0-18 teeth; in a specimen measuring 0'8 millim. by 0-5 millim. it is 39 rows of
K3_0-16. The innermost 6 or 7 teeth are bifid at the apex, the rest are simply hooked.
From the margin of the radula the first 4 teeth increase in size, and from the fourth
to the centre again decrease regularly and rapidly.

The foot is moderately broad ; its front margin not distinctly bilobed, but sharply
folded inwards in the middle line, the fold being flattened and fitting into a recess in
the under-surface of the notaeum. The tentacles, which are minute and pointed,
appear to be attached to the sides of this recess. The notaeum is tough and densely
spiculose ; the branchial pore is small and 6-lobed. An examination of the gonads
failed to show any armature.

The other species resembles the above in form and in structure of the front of the
foot, but the markings are more minute and diffuse. It is represented by 2 specimens
from Cheval Paar, the largest measuring 1'2 centims. by 1 centim., most of this size
being taken up by the wide margin of the notaeum.

The radula consists of 42 rows of 24-0-24 teeth, and, except for the inner teeth
being simple instead of bifid, closely resembles the first-mentioned species.

Family: DORIOPSIDJS.

Doriopsis aurea (Q. and G.).

The Doriopsidae are only represented by a single species, D. aurea, of which
one specimen was taken off Karativo on March 10th, 1902.

The colour, when alive, as noted by Professor Herdman, is " orange red with brown
orange spots, those along margin of mantle pale blue." The same coloration, though
much fainter, was noticeable in the spirit specimen, the orange red having become
yellow.

Quoy and Gaimard, in the "Voyage de 1' Astrolabe" (1, p. 265, plate xix., figs. 4
to 7), give a figure and description of Dons aurea. It is described as orange-red with
white spots (blue spots ringed with white in the figure) scattered over the back,
" parseme en dessus" (in the figure they are represented as in three rows, one median
and two lateral), branchiae 5, compoundly pinnate ; back elevated ; smooth rhinophores
with 1 5 lamellae and a distinct longitudinal ridge (in figure) ; length a little over

OPISTHOBRANCHIATE MOLLUSCA. 345

2 inches. It was taken in fairly deep water off Jervis Bay, Tasmania. The
description, though not perfectly in agreement with the Ceylon specimen, yet is
sufficiently near to make the identification most probable.

Doris awrita, Gould (2, p. 299, fig. 394), is similarly coloured, but has 8 branchiae.

The same type of coloration is also found in a Tectihranch, Pleurobranchus punctatus
and it would be interesting to discover whether these similar markings, occurring in
different groups, were associated with a similar habitat, or were merely fortuitous.

Family : PHYLLIDIID^.
Phyllidia varicosa, Lmk.

There is a single specimen of the above, taken on Muttuvaratu Paar. The colour
of the spirit specimen is as follows : — Mantle of a deep bluish-black ; tubercles with
a halo of faint bluish-white ; their tips with a purplish tinge. The tubercles are
roughly of two sizes, large and small ; there is a line of tubercles down the centre of the
mantle, each large tubercle alternating with two small ones, the edges of their halos
just touching. There are 2 ill-defined rows of more distinct large tubercles on either
side, the smaller tubercles being scattered irregularly around them. The margin is
occupied with very small crowded tubercles arranged irregularly 2 to 3 deep.

The underside of both mantle and foot is a dark greyish-black. There is no median
black line on the sole of the foot.

This species is widely distributed throughout the Indian Ocean and has, according
to Bergh, been recorded from Ceylon by Kelaart (3) under the name of Phyllidia
ceylanica.

Professor Herdman's notes contain references to the caj)ture of another Phyllidia,
probably this species, at Periya Paar. It is described as being of a " rich dark green
colour with prominent bright yellow papillae ; edge of mantle yellow all round ; foot
dark green."

Phyllidia nobilis, Bgh. (?).— Plate III., figs. 16 and 17.

There is in the collection a small specimen of Phyllidia or Phyllidiella, taken off
Talaivillu in April, 1902. I have been unable to distinguish satisfactorily between
these two genera, and have been compelled to adopt Sir C. Eliot's view (12) that
the distinction cannot be maintained.

The length of the spirit specimen is 2*5 centims., width 1*25 centims., width of foot
0'7 centim. The tentacles are short, pointed, placed close together at base, divergent,
laterally grooved. The lateral branchial lamellse are about 60 in number, triangular.
The margin of the mantle is white, moderately narrow, bearing very small tubercles.
Inside this is a very narrow black line bearing a few tubercles, entire in front,
slightly broken behind. The centre of the mantle is of a bluish-white colour bearing
white knobbed tubercles and marked by 3 black bands which take the form of
a three-pronged fork directed backwards. The outer branches of the fork are parallel

2 Y

346 CEYLON PEARL OYSTER REPORT.

to the margin and on the outer side bear three short branches projecting at right
angles and nearly reaching to the margin. The innermost prong of the fork is shorter
than the 2 outer and bears 3 white tubercles. A very slight alteration of the white
bands in P. nobilvs, as figured by Bergh, would produce the pattern here described.
It may be, however, that it should be referred to the variety of P. varicosa figured
by Bergh (9, plate xxv., fig. 7). The underside of the Ceylon specimen is of
a pale bluish-grey with 3 to 4 lateral ill-defined black spots.

(DORIDID.E PHANEROBRANCHIATiE.)
Sub-family : POLYCERIN^.

Aegires villosus, n. sp. — Plate III., figs. 18 to 22.

One specimen (Plate III., fig. 18) of the above, length 0-9 centim., was taken to the
north-west of Cheval Paar. The body is elevated, with a medio-dorsal bunch of long
clavate processes or elongated tubercles (Plate III., fig. 19) surrounding the branchial
pore ; two of these processes are slightly branched. There are two irregular lateral
lines of slightly clavate tubercles on each side of the body, and small tubercles are
crowded on the nape of the neck. There are longitudinal patches of purplish -brown
pigment lying between the tubercles on the back and sides. The ends of the tubercles
are shaded with brown. The ground colour is a dirty greyish-white. The surface of
the body and tubercles is densely spiculose, the spicules projecting and giving the
animal under slight magnification a densely pilose appearance.

Phinophore sheaths (Plate III., fig. 20) with 3 outer moderately long tubercles,
and 1 small inner tubercle. The radula shows that the specimen is evidently
immature, it contains only 12 rows with 16 simply hooked teeth (Plate III., fig. 22)
on each side. The jaws are not developed, but the chitinous collar shows slight
lateral thickenings and a well-marked upper jaw (Plate III., fig. 21).

This genus has previously been known only from the Atlantic and Mediterranean.

Trevelyana bicolor, Ald. and Hang

One specimen of what appears to be this species was found on Galle coral reefs in
August, 1902 ; length, 1*2 centims. The colour has been lost and the skin is trans-
parent, showing the internal organs. The liver occupies the posterior two-thirds of
the body and is of a bluish -grey colour. The branchiae are 12 in number arranged in
a circle round the anus on the centre of the back. The area of skin surrounding the
branchiae is more opaque than elsewhere, which seems to suggest that it is of a
different colour during life. There is a branching white arrangement, probably
vascular, anteriorly on the surface of both sides of the body. The rhinophores are
retracted. The size and number of branchiae of this specimen seem to point to its
being T. bicolor. The radula was not examined.

OPISTHOBRANCHIATE MOLLUSCA. 347

Kalinga ornata, Ald. and Hanc. — Plate III., figs. 23, 24.

A small specimen, length 2"5 centims., was taken 10 miles north of Cheval Paar.
Professor Hkrdman has made the following note on its coloration when alive : " Pale
grey, nearly transparent, with red internal organs showing through, and beautiful red
papilke on surface of body. Rhinophores lamellated at tip and coming from fringed
sheaths. Colourless papillae on body. Branchise tri-pinnate."

Bergh (9, pp. 959 to 962) has given a full account of the anatomy of this species with
the exception of the radula, which was missing in his specimen owing to the pharynx
having been everted through the front of the foot. Curiously enough, Professor
Herdman's specimen had suffered an almost similar misfortune, the whole buccal
mass having been everted through the mouth opening, so that the radula lay along
the under-surface of the everted organ. The teeth (Plate III., figs. 23 to 24) are very
numerous and all similar. Their shape is remarkable, each tooth having three long
curved pointed cusps. They are arranged in about 130 rows, the radular formula being
90-0-90. There is no median tooth, but the rachis bears at intervals a few very
small spurious teeth. The innermost tooth on each side differs slightly from the rest
in having the inner spine slightly shortened ; the outermost tooth is a flat plate, the
2nd and 3rd have traces of a single hook, the 4th has two hooks, and the 5th three.
The teeth increase in size gradually from the margin inwards to the centre. The
teeth in the specimen examined were widely separated from each other, perhaps
owing to the stretching of the radula.

II. TECTIBKAJNCHIATA.

There are in the collection 17 species of Tectibranchs, 7 of which are here
described as new to science.

Of those already known, two, Philine aperta and Doridium depictum, are well-
known Atlantic or Mediterranean species; two, Aplysia cornigera and Dolabrifera
maillardi, have been identified on the strength of the resemblances of the shells, as
nothing seems to have been published on the animals themselves ; the rest are fairly
well-known Indian Ocean or Pacific species. The list of species is as follows :—

Philine aperta, Linn. Dolabrifera maillardi, Desh.
Doridium marmoratwn, Smith. „ rnarginata, n. sp.

„ depictum (Ren.), var. Dolabella scapula, Martyx.

Aplysia cornigera, Sow. Notarchus indicus, Schweig.

„ elongata, Pse. „ eeylonicus, n. sp.

,, intermedia, n. sp. Pleurobranchcea brocki, Bgh.

Phyllaplysia albomaculata, n. sp. Pleurobranchus cilrinus, R. and L.

„ pelludda, n. sp. ,, burnetii, n. sp.

Aplysiella mollis, n. sp.

In an Appendix, I place : —

Onchidium verruculatum, Cuv., and Marsenia perspicua (Linn.).

2 y 2

348 CEYLON PEARL OYSTER REPORT.

Family : PHILINIDiE.

Philine aperta, Linn. — Plate IV., fig. 1.

This is one of the most abundant species in the collection. It is represented from
the Gulf of Manaar, Galle, deep water off Galle, Trincomalee, off Kaltura, the
pearl banks off Aripu, and south of Modragam — in all 42 specimens, of which
33 came from the Gulf of Manaar. The specimens vary considerably in appearance
owino- to their different degrees of contraction, but the shells are identical in texture
and almost so in shape. The shells (Plate IV., fig. 1) are of a moderately elongated
oval form with a smooth somewhat opalescent surface, and agree with P. angasi,
from Australia ; but I do not think that the slight difference in form between these
shells and those of British specimens of P. aperta can be regarded as amounting to
a sjDecific distinction. There was no constant difference to be noted between the
gizzard plates and those of the typical P. aperta, though in most cases the ventral
plate was more asymmetrical and the ends of the lateral plates more produced.

The radula was as in P. aperta.

Bergh (12) has recorded P. aperta as occurring not uncommonly in the Gulf of
Siam, and it seems not improbable that it has a cosmopolitan distribution, and is
represented by local races in many places ; as such perhaps may be : — P. vaillanti,
P. erythrcea (though that species is said to be distinguished by the peculiar serration
of its gizzard plates), P. angasi, P. orientalis, P. caurina, P. vitrea and P. coreanica.

Doridium marmoratum, Smith.

One specimen from the Gulf of Manaar should, I think, be referred to D. mar-
moratum. Its length is 2-2 centims., width 1"1 centims. ; length of head shield
1'3 centims., width 1"1 centims. The ground colour is a dark brownish plum with
numerous small pale oval or circular spots, those on the foot being rather larger than
those on the rest of the body. The margins of the foot and the front margin of the
head shield are marked with a bluish-black line. The sole of the foot is slightly
darker than the rest of the body. The lateral margins of the foot are closely
adpressed to the sides of the animal. The head shield appears to be emarginate
posteriorly, owing to its posterior extremity being turned up. The gill protrudes
through the mantle slit on the posterior end of the body, probably owing to an
injury, as the specimen is in rather bad condition, the pharynx having been everted
through a rupture in the front of the head shield and the shell crushed to fragments.
D. marmoratum has only been recorded from Torres Straits (7).

Doridium depictum (Ren.), var.

A small specimen dredged on Periya Paar, in 9 fathoms, had the following colora-
tion, as noted by Professor Herdman, while alive, " brown, mottled with pale green,
and having two yellow lines from the head backwards ; front edge of foot blue, lobe

OPISTHOBRANCHIATE MOLLUSCA. 349

of mantle over gill paler, with edging of violet." In the spirit specimen the markings
are as described, but the blue and green colours have disappeared ; the ground colour
is a pale brown.

The length of this specimen is 1'45 centims., width l'Oo centims., length of head
shield 0"9 centim., width 0"85 centim. The shell was very much broken. The foot
is pale, without markings, its margins are only slightly incurved.

There is nothing in the form of the animal to separate it from D. depictum of the
Mediterranean, and, as that species is known to be extremely variable as regards
coloration, I have recorded the Ceylon specimen as a variety.

Family: AFLYSIID.E.

There appears to be a great amount of uncertainty as to the extent of the genera
Dolabrifera, Aplysiella and Phyllaplysia, each author using the names with a
different meaning. The distinctions have been based mainly on the form of the animal,
which must vary considerably in the same species in preserved specimens. I have
not material sufficient to throw any light on the matter, and have for convenience
merely used the names in the following manner : —

Aplysiella. — Specimens with short median mantle slit, swollen body and

broad foot.
Phyllaplysia. — Specimens with depressed body, median mantle slit and broad

foot ; shell in this and the preceding genus resembles that of Aplysia.
Dolabrifera. — Specimens with moderately depressed body, broad foot, posterior

mantle slit and bat-shaped shell.

The circum-cesophageal nerve ring shows the same type in all, the commissures
being short, the cerebro-pedal being almost as long as the cerebro-visceral. In
the two species of Dolabrifera which occur in the collection there is a resemblance in
the radulas in the fact that the inner cusp of the lateral teeth is always smaller than
the outer.

Aplysia cornigera, Sow. — Plate IV., figs. 2 to 7.

It is impossible, on account of the many vague descriptions, with or without
figures, which have been published, to say with certainty what name should be
applied to what appears to be the common Ceylon species of Aplysia. I have
identified it as A. comigera, Sow, on account of the resemblance to the shell of that
species, of which oidy the shell is known. It is represented in the collection by
seven specimens, two from the Gulf of Manaar, two from the pearl banks off Aripu,
one from the pearl banks off Manaar and two without definite locality.

The coloration of these specimens is a groundwork of fine reticulating brownish-
olive lines, with clear spaces here and there, as in A. punctata. There are also

350 CEYLON PEARL OYSTER REPORT.

small black spots, about 1 millim. in diameter, scattered over the surface of the body.
The amount of pigment on the inner side of the pleuropodia varies ; usually there is
very little, and it is almost absent in some specimens.

The length of a well-preserved specimen was 6 "3 centims. The pleuropodia were
free to within from 5 millims. to 8 millims from the end of the foot. The excurrent
mantle siphon was moderately produced, from 5 millims. to 10 millims. in length.
The tentacle flaps were mostly flattened, but varied according to the amount of
contraction. The mantle foramen was closed in some specimens, but the mark of its
position was apparent ; in others it was about 2 millims. in diameter. The markings
and clear spots showed a sort of radial arrangement round the foramen. The hyaline
gland (Plate IV., fig. 6) consisted of a group of large globular cells lying beneath the
sub-dermal muscle layer and penetrating it at intervals. There was no sign of
a single common duct. The gland in appearance resembled a bunch of grapes
attached laterally to the muscle layer, and would seem to be intermediate in form
between the two types of gland described by Blochmann (6) as occurring in Aplysia,
since the cells appeared to open separately on the surface or else two or three into a
common opening.

The shell (Plate IV., fig. 4) is thin, with a fragile calcareous layer; the inside is
opalescent, of a delicate pale pink tint ; the outside is glossy, pale whitish-brown, with
distant growth ridges.

The radula (Plate IV., figs. 2, 3, 5) consists of about 40 rows of 21-1-21 teeth ot
the usual form ; the 16th and 17th from the centre have rudimentary hooks and the
four outermost teeth consist of flat plates. The labial armature is made up of
blunt rods.

A. pulmonica, Gould, from Samoa, may perhaps prove to be the same as this
species, but seems to be separated by having the pleuropodia fused for a larger
proportion of their length.

Aplysia elongata, Pease — Plate IV., figs. 9 to 12.

A specimen of the above was taken on Jokkenpiddi Paar in April, 1902.

The length of the preserved specimen is 1"2 centims., height 0 9 centim. The
colour is an opaque whitish fawn with brownish markings irregularly arranged,
leaving vacant spots, and in addition a fairly uniform but rather distant spotting of
opaque white pigment. The eyes are distinctly visible, the skin over them being
devoid of pigment. They lie lateral to and a short distance from the base of the
rhinophores. The rhinophores (Plate IV., fig. 10) are tubular and split for about
f of their length ; they resemble greatly a hare's ear in appearance. The tentacles
are much shorter than the rhinophores and are split to the base. The inside of both
rhinophores and tentacles is strongly pigmented. The pleuropodial lobes are short,
meeting behind the mantle siphon ; they are slightly pigmented with light brown on
the inside. The mantle is transparent and colourless except for a black line along

OPISTHOBRANCHIATE MOLLUSCA. 351

the margin of the mantle foramen. The latter is large and oval, measuring
5'5 millims. by 3"5 millims. The shell (Plate IV., figs. 11 and 12) is large, strong,
opaque and deep, with well-marked growth ridges and overhanging adnate callus
hood ; it measures 9 "5 millims. hy 7 millims. Judging by the size of the shell and
the way in which it has been forced out of the mantle cavity, the whole animal is
very much contracted.

The radula (Plate IV., fig. 9) is of an orange-yellow colour and contains 34 rows of
13-1-13 teeth, the 4 outermost being imperfect.

This specimen agrees very closely with A. elongata from the Sandwich Islands,
the very large mantle foramen and large strong shell with well marked callus being
very characteristic.

Besides the above there are in the collection 2 specimens from Chilaw Paar which
are most probably young examples of A. elongata. They resemble the Jokkenpiddi
specimen in form, but differ in being semi-transparent and colourless except for
a narrow black edge to the rhinophores, tentacles, front of foot, mantle foramen and
posterior part of pleuropodia. The shell is membranous with a very delicate white
calcareous layer which has mostly disappeared.

The radula has only 27 rows, the teeth being as in the larger specimen ; the lengths
of the two Chilaw specimens are 1'3 millims. and 1*1 millims. respectively.

Aplysia intermedia, n. sp. — Plate IV., fig. 8.

There is a single well-preserved specimen ol Aplysia from S. Cheval Paar which
does not fit in with any of the published descriptions. It combines the opposite
characters of a large mantle foramen and long free pleuropodia.

Length of the preserved specimen 1'8 centims., width 1*4 centime., height 1 centim.,
length of foot 1*6 centims., distance from fusion of pleuropodia to end of foot
2-4 millims., mantle pore 4 "25 millims. by 3 millims., length of mantle siphon
2 millims. Colour of living animal, green ; colour of spirit specimen, greyish ; general
effect made up of fine recticulating broken black lines of the usual Aplysia type,
following no regular plan; mantle with black specks rather concentrically than radially
arranged. The mantle foramen has not a black edge.

The tentacles are large, produced anteriorly into broad labial flaps. The shell is large
(11 centims. long), thin and membranous ; calcareous layer only present as a few small
patches at the apex of the shell.

The radula (Plate IV., fig. 8) consists of 30 rows of 16-1-16 teeth, the 3 outermost
being without hooks. The teeth resemble those of A. elongata in form.

Phyllaplysia albomaculata, n. sp. — Plate IV., figs. 13 to 16 ; Plate VI., fig. 6.

I have thought it best to create new species for this and the two following forms,
rather than run the risk of identifying them incorrectly with any of the many vaguely
described species which are in existence. The single specimen of P. albomaculata was

352 CEYLON PEARL OYSTER REPORT.

taken south of Adam's Bridge, 30 fathoms, in March, 1902. Professor Herdman notes
concerning it : " Olive hrown spotted with white, foot paler, little spiky papillae on
mantle," accompanying his note with a sketch which shows the living animal in
configuration and proportions almost exactly as in the preserved specimen.

The total length is 2*8 centims., width 1*5 centims., height 0'7 centim.

The form of the animal (Plate VI., fig. 6) is an oval, produced anteriorly to form
the head. It is very much depressed. The foot is conterminous with the body
laterally and posteriorly ; anteriorly it extends beyond the head on either side.

The branchial slit is situated rather nearer the right side, the margins not closing
over the slit. The tentacles and rhinophores are of moderate size, tubular and slit as
usual. The eyes are distinctly visible in front of the rhinophores. The surface of the
back is dotted with little spiky papillae, simple or compound. The shell (Plate IV.,
fig. 15) is calcareous, very fragile, white with a slight brownish tinge, striated
by growth lines and with a small but well-marked embryonic spine (Plate IV.,
fig. 16).

The radula (Plate IV, figs. 13 to 14) has 32 rows of 44-1-44 teeth. The teeth,
as is usual in the genus, have two broad spatulate cusps and a small basal denticulation.
In the 2nd and 3rd innermost teeth the inner cusp is distinctly larger than the outer ;
in the 4th they are about of equal size, and from the 5th outwards the outer cusp is
the largest. The nature of the buccal armature was not noted.

Phyllaplysia pellucida, n. sp.— Plate IV, figs. 17 to 21 ; Plate VI., fig. 12.

There are, in the collection, 2 specimens of a smooth flattened Phyllaplysia from
Chilaw Paar, for which I propose the above name. They are of a gelatinous
consistency, semi-translucent and colourless. The length of the largest specimen
(Plate VI., fig. 12) is 1'7 centims., width 1*5 centims., height 0-7 centim.

The foot is as wide as the body. The head is very much contracted, with short
tentacles and rhinophores. The branchial slit, length 6 millims., lies rather nearer
the right side, the right margin overlapping the left, except posteriorly, where the
slit is open. The surface of the body is smooth. The eyes are not visible in the
preserved specimens.

The radula (Plate IV, figs. 17, 19 to 21) consists of about 30 rows of 30-1-30 teeth.
The teeth are of the usual form, with two main cusps and a smaller accessory one.
From the 1st to the 10th tooth from the centre the innermost cusp is the largest,
and from the 11th to the 25th the outermost. The five outermost teeth are slender,
curved, and simple. The labial armature (Plate IV., fig. 18) is made up of square
rods with denticulate inner margins.

The shell (Plate V., fig. 1) is of the same form as in A'plyaia. It is thin and
membranous, with a very slight calcareous layer which comes oft' at a touch. It
measures 7 "8 millims. by 5 "3 millims. An embryonic spine was not noted, but it may
have crumbled away.

OPISTHOBRANCHIATE MOLLUSCA. 353

Aplysiella mollis, a. sp.— Plate IV., figs. 22 to 26 ; PI. V., figs. 2, 3 ; PI. VI., fig. 1 1.

There arc two specimens from South Cheval Paar.

The length of the larger specimen is 2"4 centims., width T2 centims., width of
foot 1 -.3 centims., height I/O centim., length of branchial slit l-0 centim., front of
head to commencement of branchial slit 1*2 centims.

The colour of the preserved specimens is white, semi-opatpie. They are of soft,
wrinkled, and of a somewhat flabby consistency.

The animal (Plate VI., fig. 11) is moderately swollen, Aplysia-like in form, with
short branchial slit. Over the surface of the back are scattered small spiky papilke,
much fewer in one specimen than in the other. The rhinophores are moderately long,
of the usual form. The tentacles are short, split as usual, and flattened. There is a
slight development of labial flaps. The eyes are distinctly visible.

The shell (Plate V., figs. 2 and 3) is shaped as in Aplysia. It is membranous, with
a very brittle, opaque white calcareous layer covered with small nodules.

The radula (Plate IV, figs. 22 and 24 to 26) consists of 33 rows of 48-1-48 teeth,
the inner cusp being the larger in the 1st to 4th tooth from the centre and the
outer cusp the larger from the 5th tooth outwards. The outermost 5 or 6 teeth are
narrow, curved, and simple. The labial armature (Plate IV, fig. 23) is made up of
rods, bent and bifid at the tips.

Dolabrifera maillardi, Desh. — Plate V., figs. 4 to 10.

One specimen from Muttuvaratu Paar, 8 fathoms, March 1902.

The original species of Deshayes was only described from the shell which came
from Reunion (5). His description and figure agree fairly accurately with the shell
of the Ceylon sjiecimen. The preserved specimen (Plate V., figs. 4 and 5) is smooth,
somewhat plump, moderately elongate, of a dirty white colour. Its length is
l-9 centims., width T3 centims., height 0'9 centim. Professor Herdman's notes
contain a sketch of the living animal and state, " Port-wine colour, with a limited
number of white spots having each a spiky papilla rising from its centre."

The foot is as broad as the body, with its margins slightly frilled. It does not
extend beyond the body posteriorly. The body is elevated posteriorly and slopes
forward somewhat, as in Dolahella. The branchial slit is short and narrow, length
0"6 centim. It commences at about the posterior third of the body and runs
hackwards. The tentacles and rhinophores are tubular and slit externally.

The radula (Plate V., figs. 6, 7, and 10) consists of 35 rows of about 62-1-62 teeth.
The median tooth has a central cusp, and two lateral cusps on each side. The innermost
teeth of the lateral rows have two large hooked cusps, and a small basal cusp on the
outer side. This basal cusp is found on about the 20 innermost teeth, after which it
disappears. The outermost teeth are long, slender, and bifid at the tips.

The shell (Plate V., figs. 8 and 9), length 4"6 millims., has an embryonic spine just
below the apex, the apex itself being formed by a flattened plate or callus.

2 z

354 CEYLON PEARL OYSTER REPORT.

The labial armature is made up of bluut rods.

This species may perhaps prove to be identical with one of the many forms
described by Pease from the South Pacific (3), (4).

Dolabrifera marginata, n. sp. — Plate V., figs. 11 to 15.

One specimen dredged on South Cheval Paar.

The length of the preserved specimen is 1*95 centims., width 1*3 centims., height
075 centim. The form of the body is ovate, somewhat narrower in front, the body
passing into the head without a distinct neck (Plate V., fig. 11). The branchial slit
is situated far back, about 4 millims. from the posterior end of the body. The slit
is 7 millims. in length, open at both ends, the right pleuropodial margin being lapped
over the left in the centre. The genital furrow is well marked. The margin of the
foot extends beyond the body laterally and posteriorly, forming a frill 1*5 millims.
wide. The body is smooth, the skin being somewhat tense. The rhinophores and
tentacles are short, tubular, and slit as usual. The eyes are visible in front of the
rhinophores.

The shell (Plate V., fig. 15) is very thin and membranous, of the same form as in
D. maillardi, but perfectly transparent except for a chalky powder over parts of its
surface which rubs oft' at a touch. The apex of the shell was injured in extraction,
so it could not be ascertained whether an embryonic spine was present or not. It is
probable that this species has not been already described, as the shell is not well
fitted for preservation in a conchological cabinet.

The radula (Plate V., figs. 12, 13, 14) is of the same type as in D. maillardi, but
the lateral teeth are not so slender. It consists of 37 rows of 57-1-57 teeth.

Dolabella scapula, Mart. — Plate V., figs. 16 and 17.

There are three specimens of this, the common species of Dolabella of the Indian
Ocean ; one from South-east Modragam and two from south of Adam's Bridge.

The coloration of the preserved specimens consists in a faded yellowish-white
ground colour mottled over the body and foot with irregular olive-brown blotches.
Professor Herdm an notes of another specimen from S. of Cheval, 6|- fathoms : ' ' Chestnut
brown in general effect, yellowish basis mottled with red and brown, many yellow
tags or spines all over, broad flat foot rather paler." In the preserved specimens the
body was covered sparsely with short fringed papilla?, longest on the margin of the
posterior disc. These papillae were most numerous in the largest specimen, which
measured 4*8 centims. in length, and almost absent from the smallest.

Some of the published descriptions state that D. scapula is uniformly coloured,
but this is probably a matter of individual variation. The shell of the Ceylon
specimens agrees with that of D. scapula, having kindly been compared with
specimens in the British Museum by Messrs. E. A.. Smith and E. R. Sykes.

The radula consists of 44 rows of about 120-1-120 teeth, those near the centre

OPISTIIOURANOHIATE MOLLtTSCA. 355

being simply hooked (Plate V.. fig. 17), while towards the margin the hook becomes a
long scythe-shaped Made (Plate V., fig. 16).

Notarchus indicus, Sohweig.

1 have, en account of the resemblance of the radula to the figures given by Bergh
(11, plate vii.), referred to this species 20 little immature specimens of Notarchus,
from Ceylon. The largest of these measured 1'2 centims. in length. The label
indicating the exact locality where they had been taken has unfortunately been lost.

Notarchus ceylonicus, n. sp. — Plate V., figs. 18 to 23.

There is in the collection one specimen of NotarcJius, from i mile east of Dutch
Modragam Paar, for which I propose the above name.

The form of the body (Plate V., tig. 18) is ovate, very plump, and inflated, except
as regards the head. The foot is moderately wide and pointed behind. There is no
shell. The branchial slit is short, length 6 millims. The length of the animal is
3*3 centims., width 2'1 centims., height 1*6 centims., width of foot 1*1 centims.

The surface of the body and head bears numerous small fringed papillae which
occur more thickly on the head, probably owing to its contraction. The rhinophores
and tentacles are fringed with similar papilla?. Very small labial processes are
present.

The radula (Plate V., figs. 20 to 23) consists of 30 rows of 33-1-33 teeth, the first
8 rows being imperfect. The first 5 teeth from the centre have four lateral denticles
on the outer edge, the following 9 three denticles, the basal denticle being small and
blunt, and becoming hard to recognise at a little distance from the centre. The next
4 teeth have two lateral denticles. From the 18th to the 24th tooth the distal
denticle only persists, and the 9 outermost teeth are smooth and scythe-shaped, thus
differing from those in N. indicus, which are serrulate.

Family : PLEUEOBRANCHID^.

Pleurobranchaea brocki, Bgh. — Plate V., figs. 24 to 28.

Tt is witli some doubt that I refer to this species a well preserved specimen of
Pleurobranchcsa, which was taken 5 miles north of Cheval Paar in March, 1902.

The colour of the preserved specimen is an opaque brownish-yellow, due evidently
to its preservation in spirit. On the mantle and upper surface of the hinder parts
of the foot there are rusty-brown markings in coarse irregular reticulations, occupy-
ing to some extent the furrows on the mantle.

The form of the body (Plate V., figs. 24 and 25) approaches the Mediterranean
species P. mecl-elii. The total length is 3 2 centims., width (of foot) 17 centims.,
height l'l centims., length of mantle 2-6 centims., width 1"5 centims., length of foot
behind bodv 07 centim., length of gill from genital papilla T2 centims., height of
caudal papilla 3 5 millims.

2 z 2

356 CEYLON PEARL OYSTER REPORT.

The mantle is not very large, reaching beyond the body, both laterally and
posteriorly, for about 1 millim. and rather more over the gill. The foot is somewhat
wider than the mantle laterally, and reaches considerably beyond it posteriorly. The
tail is somewhat pointed and bent upwards, terminating in a little conical caudal
papilla. The frontal margin of the mantle is beset with two rows of minutely
tuberculate papillae ; there are 7 of these in the upper row and 6 in the lower.
There are also very small scattered intermediate tubercles along the margin, which
is laterally produced on either side into a tentacular appendage. The gill has
36 lamellse, each of them pinnate ; it reaches to the level of the end of the mantle.

The rhinophores as usual rise from the mantle and are split down the outer margin.

The surface of the mantle is not distinctly tubercular, but is furrowed and
rough, owing perhaps to contraction. The foot, mantle and body walls are very
muscular and tough. The intercrossing of the muscle bands gives the appearance of
a woven material, particularly on the inner surface of the foot after removal of the
viscera.

The genital openings are separate. The vaginal pore has a short tubular opening
at the commencement of the gill rachis. The penis is retracted, its opening directed
forwards and lying about 1 millim. in front of the female opening. The anus lies
dorsal to the membrane which attaches the free end of the branchial plume.

The jaws are made up of hexagonal rods (Plate V., tig. 27), the edges of the inner
marginal rods being very irregularly denticulate. Teeth of radula (Plate V., figs. 26
and 28) in about 29 rows, formula 56-0-56. No sign of a median tooth was seen.
Form, as figured by Bergh for P. brocki (11). Shell absent.

The colour of the Ceylon specimen differs considerably from that described by Bergh,
which was of a reddish -brown with darker reticulations. Vayssiere mentions a
specimen of the same colour, but says that the pigment is superficial, and easily
rubbed off, which may account for its absence on this occasion.

Pleurobranchus (Berthella) citrinus (Rupp and Leuck.). — Plate VI, figs. 7 to 10.

There are 3 specimens of Pleurobranchus apparently belonging to this species
which were taken at Jokkenpiddi Paar, Cheval Paar and to the South-east of East
Cheval respectively. They differ slightly in colour and appearance, the Jokkenpiddi
specimen being of an opaque brownish-yellow and considerably more contracted than
the others, which are white, rather translucent and plump. The South-east of East
Cheval Paar specimen, which is the best preserved of the 3, measures 1 7 centims. by
T05 centims. Its height is 0'5 centim. The mantle is elevated, somewhat inflated,
a little wider than the foot, covering the head, and slightly emarginate in front. Its
colour is a transparent white with opaque white dots. When examined under a lens
of moderate power, the white spots are seen to be subdermal opaque flask-shaped
structures tapering towards the surface and surrounded by distinctly marked clear
hexagonal areas. The surface of the mantle is smooth. The gill is short and

OPISTHOBRANCHIATE MOLLUSCA. 357

compact, length 0'7 centim., without rachidian tubercles. The lamellae number about
2.t. The front of the toot Is divided into two lips by ;i transverse furrow, the lower
lip being the largest. The furrow does not extend laterally. The labial flap is
produced on either side into 2 very short tentacles split laterally at their extremities.

The rhinophores are close together and rather short, tubular and split as usual.
The genital openings are close together at the commencement of the gill, but not
surrounded by a common investment. The penis is retracted.

The shell, length 0"5 centim., lies well forward at about the anterior third of the body ;
the apex is pale, with embryonic spiral well marked ; the expanded portion is of a
horny yellow colour, strong, with well-marked longitudinal and transverse striae.
The teeth are of the usual type in Bcrthella, as defined by Vayssiere (9), knife-
shaped, with strongly denticiilated edges. They bear about 14 denticulations.

The colour of the living animal as noted by Professor Herdman is " brilliant orange,
with white dots in little systems on the back." This differs somewhat from Ruppel's
description of P. citrin us as " pale lemon yellow, with large whitish irregular spots,"
but as they agree otherwise there does not seem sufficient grounds for separating
them. P. angasi, Smith, from Port Jackson, agrees in many ways with this species,
but has only 16 gill lamellae. The figures of P. oblongus, Aud., given by Savigny, are
very like the Ceylon specimens and the figure of the shell of that species shows an
evident attempt to represent an embryonic spiral.

Kelaart (2) has recorded this species as being of common occurrence in Ceylon ;
he gives the colour as orange spotted with white.

Pleurobranchus hornelli, n. sp. — Plate VI., figs. 1 to 5.

There are in the collection some specimens of Pleurobranchus which I have not been
able to identify with any of the described species, and for which I accordingly
propose the above name. These are 7 specimens, 6 of which were taken by
Mr. Hornell from a buoy-rope at Galle in July, 1902, and one on Galle coral-reef
in August, 1902. Their colour, as far as is preserved, is a dirty white ground colour,
with the mantle covered with fairly close reticulations of brownish -purple. The
length of the largest specimen is 1"8 centims., width 1*2 centims., height 10 centim.
The mantle is about equal to the foot in front and behind, and a good deal wider
laterally. It is tuberculate, but the tubercles are indistinct, being crowded together
so as to produce a level surface (Plate VI., fig. 5), the margins of the tubercles being
only noticeable on account of the previously mentioned purple reticulation which
occupies the space between them. In the Galle coral-reef specimen, which seems in
a different state of preservation from the rest, the apex of each tubercle is raised into
a small papilla.

The foot is moderately broad, with the front margin divided into two lips by a
transverse furrow. The upper lip is much thinner and somewhat shorter than the
lower. The furrow is continued laterally down one margin as far as the commence-

358 CEYLON PEARL OYSTER REPORT.

ment of the gill and to an equal distance on the other side. The pedal gland
is well developed in the Galle coral-reef specimen, but hardly noticeable in the others.

The tentacular shield is well developed, but not produced into tentacles laterally.
The anterior portion of the lateral margins is grooved. The rhinophores are mode-
rately long and split as usual. The gill is moderately long, 1 centim. in the largest
specimen, reaching to the end of the body, bipinnate, the rachis with two rows of
tubercles, pinnae 25-26, pinnules about 16.

The genital openings are close together at the base of the gill and appear to be
surrounded by a common investment. The shell (Plate VI., fig. 1) is very thin,
with an opaque white calcareous coating very fragile and easily rubbed off. It seems
to be very faintly striated in the lines of growth. It is situated at about the
anterior third of the mantle and is distinctly visible through it as a white patch ; its
length in a specimen of 1*5 centims. is 2"9 millims. by 1'9 millims. The shell
has an ill-defined membranous edge without calcareous coating. The radula in one
specimen examined consisted of 46 rows of about 70-0-70 teeth, the inner teeth being
simply hooked (Plate VI., fig. 3), the outermost rod-like (Plate VI., fig. 2), slightly
bent. The jaws are rather broader than usual, measuring 1 "12 millims. by 2 '35 millims. ;
they are made up of plates of the usual form (Plate VI., fig. 4), with two lateral
denticles at the ajDex, a large distal and a smaller proximal one.

APPENDIX.
Family : ONCHIDILILE.

Onchidium verruculatum, Cuv. — Plate VI., figs. 13 to 22.

There is one specimen of this widely distributed species, very well preserved, from
the Gulf of Manaar. The length is 3"1 centims., width 3-4 centims., height 2 centims.,
width of foot l-5 centims. The colour of the preserved specimen is grey, with
irregular purplish -brown blotches over the back.

The back is completely covered with simple and compound tubercles of differing
sizes. The simple tubercles are most numerous and vary in size from an extreme
minuteness to about 1'2 millims. in diameter. The compound tubercles measure
about I "5 millims. in diameter and are made up of from 5 to 7 simple tubercles. On
the posterior fourth of the back are about 10 branchial tubercles, which appear to
be made up of from 20 to 40 short papillae crowded together. These apparently
represent frondose processes in the living animal. There are only 6 ocular tubercles
in the Ceylon specimen, all situated on the anterior half of the body. They bear
from 2 to 5 ocelli in a slightly depressed pit on their summit. The general effect of
the tuberculation resembles that seen in Archidoris tuberculata.

The loop of the intestine follows the same course as figured by Plate for the
species, being much less curved than in the closely allied O. tumidum.

OPISTHOBRANCHIATE MOLLUSCA. 35i)

The male gonads (Plate VI., fig. 17) agree fairly closely with Platk's description.
The penis rod is long and slender, the glans short and closely covered with minute
honks ahout <>'<>4 millim. in length.

The length of the "cartilaginous" rod is 16 niilliins., that of the spinulose glans
I "8 niilliins. The dart sac with its ampulla is of the usual form, the ampulla long
and rather narrow and sharply curved at its anterior end, where it overlies the
oesophagus. The length of the ampulla is 10 niilliins., the distance from the ampulla
to the papilla of the dart sac 8 millims. The dart (Plate VI., fig. 13) is straight,
regularly tapered ; length 2 -8 millims. Its free extremity shows a slight curve and
has a solid lateral lip.

The radula (Plate VI., figs. 19 to 22) is of a pale yellow colour very slightly
chitinised. It is made up of 60 rows of 83-1-83 teeth of the usual form. The
central tooth is slightly broader than long ; the lateral teeth, except those close to the
middle line, are produced downwards into a broad adze-shaped hook and upwards
into an irregular thin tongue, which takes part in their attachment to the radula
ribbon.

The contents of the stomach consisted of quartz sand grains and a quantity of
white flocculent matter, amongst which a few Foraminifera were noticed.

Onchidium verruculatum is very widely distributed in the Indian and Pacific
Oceans, occurring from the East Coast of Africa and the Red Sea to Japan and
Australia.

Family : MARSENIAD^E.

Marsenia perspicua (Linn.) — Plate VI., figs. 23 to 26.

There are eleven specimens of Marsenia in the collection, most of them from
Cheval Paar and its neighbourhood and the remainder from South-east Modragam,
Gulf of Manaar. and Galle. Externally they are not to be distinguished from the
common M. perspicua of European waters, with which species also the radula
(Plate VI., figs. 24 to 26) and shell are in close agreement. The largest specimen,
one of those from South-east Modragam, is a female, and measures 2 "5 centims. by
2"2 centims., the shell (Plate VI., fig. 23) measuring 1*6 centims by 1"2 centims.
These measurements are somewhat in excess of those usually attained by European
specimens, but the difference in size and locality does not seem to warrant the use of
a separate specific name.

360 CEYLON PEARL OYSTER REPORT.

LITERATURE REFERRED TO.

I. NUDIBEANCHIATA.

(1.) QUOY and GAIMAKD.—' Voyage de PAstrolabe.' Moll. II., 1832, pp. 248-292,

(2.) Gould.— 'U.S. Exploring Expedition.' Moll, 1852, pp. 293-310.

(3.) Kelaart.— ' Ann. Nat. Hist.,' 3rd ser., vol. iii., 1859, pp. 291-304, 488-496.

(4.) Kelaart.— ' Ann. Nat. Hist.,' 3rd ser., vol. iv., 1859, pp. 267-270.

(5.) Alder and Hancock. — 'Trans. Zool. Soe.,' vol. v., 1864, part 3.

(6.) Pease.— ' American Journ. Conch.,' 1866, vol. ii.

(7.) Abraham.— ' Proc. Zool. Soc.,' 1877.

(8.) Collingwood.—' Trans. Linn. Soc.,' ser. 2, vol. ii., 1881, pp. 123-140.

(9.) Bergh.— " Malacol. Untersuch." (in Semper's ' Reisen im Archipel der Philippinen '), 1870-92.
(10.) Bergh. — ' Verhand. der zool. bot. Gesell. in Wien,' xix., 1869, pp. 224-244.
(11.) Eliot.— ' Proc. Zool. Soc.,' 1901, ii., p. 62.

(12.) Eliot. — 'Fauna Geogr. Maid. Lace. Arch.,' vol. ii., part i., pp. 540-573, 1903.
(13.) Eliot.— ' Proc. Zool. Soc.,' 1903, ii., pp. 354-385.

(14.) Bergh. — "Neue Nacktschnecken der Siidsee," ' Journ. Mus. Godeffroy,' 1875 and 1878.
(15.) Bergh.— ' Malacol. Untersuch.,' 1904.

II. TECTIBEANCHIATA.

(1.) Ruppel and Leuckart. — ' Neue Wirbell. Thiere des Rothen Meeres,' 1828, p. 20.

(2.) Kelaart. — ' Ann. Nat. Hist.,' 3rd ser., vol. iii., 1859, p. 495.

(3.) Pease.—' Proc. Zool. Soc.,' 1860, p. 22 ; 1861, p. 245.

(4.) Pease.—' Amer. Journ. Conch.,' vol. iv., 1868, pp. 76, 77.

(5.) Deshayes.— ' Catal. Moll. Reunion,' 1863, p. 53.

(6.) Blochmann.— ' Mitt. Zool. Stat. Neapel,' v., 1884, pp. 32-34.

(7.) Smith.— ' Zool. Coll. "Alert,"' 1884, p. 87.

(8.) Semper. — ' Reisen im Archipel der Philippinen,' Th. 2, Bd. iii., pp. 262, 263.

(9.) Vayssiere. — ' Ann. des Sciences Naturelles,' 8th ser., vol. viii., 1898, pp. 210-402.
(10.) Vayssiere. — 'Ann. des Sciences Naturelles,' 8th ser., vol. xii., 1900, pp. 1-85.
(11.) Bergh. — " Malacol. Unters." (in Semper's ' Reisen im Archipel der Philippinen '), Bd. vii., Abth. iv.
(12.) Bergh.— ' Danske Selsk. Skr.,' 1902, xii., pp. 153-218.
(13.) Plate.— ' Zool. Jahrb.' (Anat. u. Ontog.), Bd. vii., 1893-4, pp. 93-234.

OPISTIIOBRANCHIATE MOLLUSCA. 361

EXPLANATION OF PLATES.

The figures, with the exception of Plate II., figs. 2, 3, 13, 14, 17, 20, and 21 ;
Plate III., figs. 4, 11, 16, 17, and 20; Plate V., figs. 4, 5, 11, 18, 24, and 25;
Plate VI., figs. 6, 8, II, 12 and 14, were all drawn with the aid of a camera lucida.

PLATE I.

Fig. 1. Herein ceylonica, n. sp., side view of animal, x 9.

2. „ „ 3 teeth from radula, dorso-lateral view, x 470.

3. „ „ tooth from radula, seen from above, x 470.

4. „ ,, jaw plate, x 83.

5. „ „ cutting edge of jaw plate, x 340.

6. Galvma producta, n. sp., median tooth with one of side plates, x 700.

7. ,, ,, 4 teeth from radula, with side plates, lateral view, x 350.

8. ,, „ view of under-side of head.

9. ,, ,, jaw plate, restoration from crushed specimen, x 55.

10. LimjuiHii rinemi, n. sp., median and 2 lateral teeth from radula. x 340.

11. ,, „ jaw plate, x 16.

12. „ „ lateral tooth from radula, 39th from median, x 340.

13. ,, ,, 2 outermost teeth from radula. a = outermost, x 340.

14. I'h iiiiijihi/Ilidia formom, Kel., jaw plate. x8'3.

15. ,, „ outer tooth from radula. x 285.

16. ,, ,, median and 2 lateral teeth from radula. x 285.

17. Discodoris labifera, Abr., 4 outermost teeth from radula. x 330.

18. „ ,, innermost teeth, a = rachis of radula. x 330.

PLATE II.

Fig. 1. Plewqphyllidia formosa, Kel., lateral teeth from radula, 8th to 13th from middle, x 136.

2. Plaiydoris inframaculata (Abr.), under-side, anterior.

3. ,, ,, branchiae and collar.

4. „ „ teeth from radula, middle of lateral row. x 132.

5. „ „ armature of penis, x 29.

6. „ ,, discs from vagina, x 29.

7. „ „ upper part of penis.

8. Platydoris speciosa (Abr.), lateral teeth from radula, 70th to 74th from middle, x 132.

9. „ „ worn teeth from lateral row of radula. x 1 10.

10. „ „ lateral tooth, side view, x 110.

11. ,, ,, lower portion of penis, showing armature, x 35.

12. ,, „ upper portion of penis, with glans. x 35.

13. Platydoris In nhniini, n. sp., view of under-side, anterior, x 2.

14. ,, ,, dorsal view. xl'5.

15. ,, „ outermost teeth of radula, 15th and 16th rows, x 370.

16. ,, ,, penis, upper portion, x 25.

17. „ „ „ and vagina.

18. Thordisa (?) caudata, n. sp., innermost teeth of radula. <i = rachis. x 305.

19. ,, „ 5 outermost teeth of radula. x 305.

3 A

362 CEYLON PEARL OYSTER REPORT.

Fig. 20. Plaiydoris sjnnulosa, n. sp., dorsal view of bulbus pharyngeus.

„ 21. „ „ stomach, dorsal view.

„ 22. „ „ vagina, after treatment with caustic potash, x 15.

,, 23. „ „ opposite plates from vagina, x 60.

PLATE III.

Fig. 1. Plaiydoris spinulosa, n. sp., lateral teeth from radula, 36th and 37th from middle, x 300.

2. „ „ innermost 4 teeth, a = innermost, x 300.

3. „ „ outermost 8 teeth, x 300.

4. Halyerda punctata, n. sp., lateral view of animal, x 2.

5. „ ,, outermost 4 teeth from radula. x 300.

6. „ ,, arrangement of branchiae.

7. ,, „ innermost teeth of radula. « = rachis. x 300.

8. Plaiydoris inframaculata (Abr.), innermost teeth of radula. a = rachis. x 100.

9. ,, ,, 3 outermost teeth, x 300.

10. Plaiydoris speciosa (Abr.), 4 outermost teeth, x 300.

11. Chromodoris termilinearis, n. sp., animal, dorsal view.

12. ,, ,, median teeth from radula. « = rachis. x 480.

13. ,, ,, 3 outermost teeth from radula. a = outermost, x 480.

14. ,, ,, 3 teeth from middle of lateral row. x 480.

15. ,, ,, elements of jaw plate. a = side view of one. x 480.

16. Phyllidia nobilis, Bgh. f?), animal, dorsal view, x 2.

17. „ „ head and front of foot.

18. Mgires viUosus, n. sp., animal, lateral view from right, x 7.

19. „ „ tubercles surrounding branchial pore, from left side.

20. „ ,, rhinophore sheath.

21. „ „ buccal collar. x31.

22. ,, ,, lateral teeth from radula. x 310.

23. Kalinga ornaia, A. and EL, median teeth. « = rachis. x 310.

24. „ ,, lateral teeth. a = innermost, x 310.

PLATE IV.

Fig. 1. Philine aperta, LlNN., shell, x 2.

,, 2. Aplysia cornigera, Sow., teeth from radula, median and 2 lateral, x 127.

„ 3. „ „ 5th and 6th teeth from centre of radula. x 127.

„ 4. ,, ,, shell, x 2.

,, 5. „ „ teeth from radula, 12th from centre to margin, x 127.

„ 6. ,, „ section of hyaline gland, x 52.

„ 7. „ ,, gill, x 2.

„ 8. Aplysia intermedia, n. sp., median tooth from radula. x 340.

„ 9. Aplysia elongaia, Pse., median tooth from radula. x 340.

,, 10. ,, ,, lateral view of head.

„ 11. „ ,, shell, ventral, x 7.

„ 12. ,, ,, shell, lateral, x 7.

,, 13. Phyllaph/sia alboniarulala, n. sp., teeth from radula, median to 6th lateral, x 310.

>, 14. „ ,, 6 outermost teeth from radula. x 310.

i, 1"J- ,, „ shell, ventral x7'5.

i> 16. ,, „ apex of shell showing spire.

. 17.
18.

»•

19.

11

20.

J)

21.

11

22.

J)

23.

11

24.

)>

25.

»)

26.

OPISTHORRANCHTATE MOLLUSCS.. 363

PhyUaplysia pelhusida, a. sp., 19th to 22nd lateral teeth from radula. x 320.
„ „ rods from jaw plates, x 310.

,, ,, median and 2 lateral teeth from radula. x 320.

„ „ 9th to 12th lateral teeth from radula. x 320.

„ 7 outermost teeth from radula. x 310.
Aph/sielhi mollis, n. sp., lateral teeth from radula, 6th and 7th from margin, x 310.
„ „ rods from jaw plates.

„ ,, 3rd to 5th teeth from centre of radula. x 310.

,, „ 32nd to 35th teeth from centre to radula. x 310.

„ ,, median and 1st lateral teeth from radula. x 310.

PLATE V.

Phijllaplysia pellueida, n. sp., shell, dorsal view, x 7.
Aplysidla mollis, n. sp., shell, ventral view. x 7 "5.

,, „ calcareous layer of shell, dorsal view. x7-5.

Dolabrifera maillardi, Desh., animal, dorsal view, x 2.

„ ,, , seen from right side, x 2.

„ ,, teeth from radula, middle of lateral row. x 315.

„ „ lateral tooth from radula, 19th from centre, x 315.

„ „ apex of shell showing spire.

„ „ shell, ventral view, x 15.

„ „ median and 1st lateral teeth from radula. x 315.

Dolabrifera marginata, n. sp., animal, dorsal view, x 2.

„ ,, median and 1st to 3rd lateral teeth from radula. x 310.

„ „ lateral tooth from radula, 3rd from margin, x 310.

„ „ 22nd and 23rd lateral teeth from radula. x 310.

,, ,, shell, x 15.

Dolabdla scapula, Mart., lateral teeth from radula. x 74.

„ „ median and 1st to 3rd lateral teeth from radula. x 74.

Notarchus ceylonicus, n. sp., animal, dorsal view, x 2.

„ „ rods from jaw plates, x 130.

„ „ 9th and 10th lateral teeth from radula. x 114.

„ „ median and 1st and 2nd lateral teeth from radula. x 114.

„ „ 24th and 25th lateral teeth from radula. x 114.

„ „ outermost 3 teeth from radula. x 310.

Phvmbranchuu hrocki, Bgh., animal, dorsal view, x 2.

„ „ ,, from left side, x 2.

„ „ 5 innermost teeth from radula. a = innermost. xll6

„ „ rods from margin of jaw plate, x 320.

„ „ 12th and 13th lateral teeth from radula. x 116.

PLATE VI.

Fig. 1. Pleurobranchus hornelli, n. sp., shell, membranous margin not shown, x 13.
„ 2. „ „ 9 outermost teeth from radula. x 320.

„ 3. „ „ 7 innermost teeth from radula. x 320.

„ 4. ,, „ portion of jaw plate, x 300.

„ 5. „ „ section of mantle.

3 A 2

Fig.

1.

M

2.

11

3.

il

4.

l»

5.

»»

6.

Jl

7.

il

8.

11

9.

11

10.

11

11.

)!

12.

n

13.

11

14.

11

15.

11

16.

11

17.

11

18.

11

19.

11

20.

11

21.

1)

22.

11

23.

>)

24.

1)

25.

11

26.

1)

27.

11

28.

364

CEYLON PEARL OYSTER REPORT.

Fig. 6.

„ 7.

„ 8.

„ 9.

„ 10.

„ 11.

,. 12.

„ 13.

„ 14.

,. 15-

„ 16.

„ 17.

„ 18.

„ 19.

„ 20.

„ 21.

„ 22.

„ 23.

„ 24.

„ 25.

„ 26.

Phylla/plysia nlbomactthrta, n. sp., dorsal view, x 2.

I'll urohranchua (BerfheJIa) ri/rinus (R. and L.), shell, x 7-5.

,, „ „ sculpture of shell.

lateral teeth from radula. x 320.
portion of jaw plate, x 320.

Aplysiella mollis, n. sp., dorsal view, x 1 -5.
Phyllaplysia pelhmda, n. sp., dorsal view, x 2.
Onchidium vemwulatum, Cuv., dart, x 31.

papilla of dart sac.

apex of dart, x 78.

Marsenia perspicua, Linn., shell, x -j.

„ „ 2 rows of teeth from radula.

,, „ apex of lateral tooth.

,, ,, 2 median teeth.

another view of same, x 78.

male gonads, a, ampulla of dart sac ; b, dart sac ; c, rod of penis ;

d, vas deferens ; ?, retractor penis ; /, adductors, x 28.

hooks from glans of penis, x 116.

teeth from radula, 53rd and 54th from centre, x 300.

side view of tooth from near margin of radula. x 300.

5 outermost teeth from radula. x 310.

median and 2 lateral teeth from radula. x 300.
o

CEYLON PEAR] 0 I REPORT,

OPISTHOBRANCHIA PLATE I

G.P. F. del.

. .

1-5. HSRVJA CEYLONICA.n.sp. b-9.GALVINAPPODUCTA.nsp. LO-13, LtNGUELLA ' CINEREA.n.sp.

14-16 Pleurophyllidia Formosa 17. 18. Discodoris labifera, (Abr.)

CEYJ PEARL OYSTK.K h'Kl'uK I

2

OPISTHOBRANCHIA, PLATE II

g. p. i a«i.

1. PLEUROPHYLLIDIA FORMOSA. Kel.
1-3-17, Platydoris HERDMANI. n.sp

M*?a,rl«j»e iErsKind liU

2-7. Platydoris inframaculata 8-12, Platydoris speciosa. lAbr.)

18,19, Thordisa CAUDATA.n.sp 20-23, Platydoris spinulosa, n. sp.

CEYLON PEARL OYSTER REPOR1

OPISTHOBRANCHIA PLATE III.

G P F Sel

K'Fa.rla.ne 1 Erslnnp Lith Edinr

1-3. PLATYD ORIS SPINULOSA lisp 4-7, HALGERDA PUNCTATA, n.sp. 8,9, PLATYDORIS INFRAMACULATA.'.Abr.)

lO.PLATYDOHIS SPECIOSA.fAbr.; 11-15. CHROMODORIS TENUILINEARIS.n sp 16~17, PHYLLIDIA NOBILIS.Bgh.

18-22. Aegirus villosus, n sp 2-3.24, Kalinga ornata, A.&H.

CEYLON PEARL OYSTER REPORT,

OPISTIIi'.. CHIA, PLATE IV.

■

1 - . INI AP
9-L' JNGATA.

Sow. 8, Aply.: asp.

PHYLLAi 17- 21, PHYLLAI

22-26, Aply i i sp.

CEYLON t'KARL OYSTER RF.11' ■

OPISTHOBRANCHIA, PLATE Y
4 / ~\ ~L~s 6.

H'F«wia.u* 4 -T± i "■■■■

id.

HYLLAPLYSIA PE] n. sp IPLYSIELLA MOLLIS ,n.sp. 4 -10. DoLABRlFERA MAILLARDT. Desh.

IRIFERA MA 16,17, DOLABELLA SCAPULA. Mar. 18-23, NOTARCHUS GbYI NI<

24 28 :■'.[ WROBHANCWEA BROCKI. Bgh

CEYLON PEARL OYSTER REPI iRT

OPISTHOBRANCHIA, PLATE VI

HfarUne ASrikinr

-■JROBRANCHUS HOR.v [YLl.APLYSIA ALBOKACULATA n. sp. 7-10, PlEUF IBRANI

i iYLLAPLYSIA PE LU1CIDA. n. sp . 13 ~22 , ( I uv .

J6. KABSENIA PERSPICUA.(Liim)

[CEYLON PEARL OYSTER FISHERIES— 1905— SUPPLEMENTARY REPORTS, No. XXII.]

REPORT

ON THE

OSTRACODA

COLLECTED ISY

Professor HERDMAN, at CEYLON, in 1902.

BY

ANDREW SCOTT, A.L.S.,

RESIDENT FISHERIES ASSISTANT IN THE MARINE LABORATORY AT PIEL, BARROW-IN-FURNESS.

[With TWO PLATES.]

The Ostracoda forming the subject of this report were, with a few exceptions, found
amongst the sand and debris washed out of the vessels containing other groups of the
Ceylon collection. Pelagic species were occasionally taken in the tow-nettings made
during the examination of the pearl banks, and also in the series of plankton
collections taken during the outward and homeward journeys. Samples of mud and
sand from various pearl banks were also examined, but these yielded no members of
this group. Some of the material had been preserved in formol, which had apparently
a bad effect on the delicate shells of these creatures. In some cases the lime salts
had been partly or entirely dissolved out, making the identification a work of much
difficulty.

The Ostracoda are represented by 77 species belonging to 22 genera. Thirty-five
of the above number appear to be undescribed, and are now added to the Ceylon
Fauna, and figured in the plates.

A considerable amount of work has already been done and a good deal of information
published relating to the Ostracoda of Ceylon. Professor Brady has one paper in
' The Journal of the Linnean Society ' (vol. xix., No. 114, 1886) which deals entirely
with Ceylonese Entomostraca. It contains descriptions of new and other marine
species dredged in 2 fathoms off Kalpentyn, in the Gulf of Manaar. Descriptions of
other species are given by Professor Brady in reports published in the ' Transactions

366 CEYLON PEARL OYSTER REPORT.

of the Zoological Society.' The valuable work on the Ostracoda collected by the
" Challenger" is indispensable in dealing with tropical forms. 'The Monographs on
the North Atlantic and North- Western European Ostracoda,' by Professor Brady and
Dr. Norman, are also of much service. The splendid work of Dr. G. W. Muller,
published by the Naples Zoological Station, is a mine of information, as the anatomical
details of the animals are fully illustrated, and the positions of many hitherto
incompletely described species are thus firmly established. Finally, every paper
dealing with marine Ostracoda from foreign localities requires to be consulted, and
this has been done as far as possible in the present case ; but it is not considered
necessary to give references under the species, or to add a list of the literature, as the
authors named and the papers made use of are well known to all workers at the
group.

I am much indebted to Professor Brady, F.R.S., for looking over the identifications
made and for assistance given with some doubtful species. My father, Dr. T. Scott,
has also helped me greatly with the work of classifying the undescribed forms.
Owing to the absence of the appendages, the exact positions of one or two of the new
species, which happened to be only empty shells, are at present uncertain.

It is almost impossible to find good descriptive names in certain genera of Ostracoda
which are not pre-occupied, and geographical terms are apt to prove misleading. So
I have named the new species of Cythere and Cytherella in honour of some of the
officials and naturalists connected with Ceylon and its pearl fisheries who were
mentioned in Professor Herdman's Introduction to these Reports.

Section I. : MYODOCOPA.
Family: ASTEROPID^E.

Asterope oculata, Brady.

In washings from young pearl oysters collected from Cheval Paar, February and
March, 1902, and from Muttuvaratu Paar, November 19th, 1902; also in general
washings from invertebrates from Gulf of Manaar. Ten females and four males were
found altogether. Professor Brady records it from the surface at Trincomalee, and
from Cruz Bay.

Asterope quadrata, Brady.
Three females were present in the washings from invertebrates dredged on the
pearl banks. This species was described from specimens collected at Lyttelton
Harbour, New Zealand.

Asterope arthuri, Stebbing.
Specimens of this fine species were taken by the tow-net in 6 to 10 fathoms at
Karativo Paar on March 10th, 1902, and at 9 fathoms at 9 P.M. on Vankali Paar,

OSTRACODA. 367

March 13th, 1902. On the latter date there was much phosphorescence in the water.
Stebbing describes it in Dr. Arthur Willey's " Results," Crustacea, Part v.

Cyclasterope similis, Brady.
Two females in general washings from dredged invertebrates from Gulf of Mauaar.
Java Sound is the only previous locality given for the species.

Family: CYPRIDINID.E.

Cypridina faveolata, Brady.

One specimen, from the pearl banks in the Gulf of Manaar. It has hitherto been
recorded only from the China Sea, where a single example was found.

Pyrocypris chierchiae, Muller.

This species appeared to be generally distributed throughout the area investigated
and has been found in the following places : — Muttuvaratu, West Cheval, South-
south-west of Silavaturi, off Kalpentyn Island, at Galle, in washings from young pearl
oysters, and from the general washings from dredged invertebrates from the Gulf of
Manaar.

According to Professor Brady, this and other species of Pyrocypris occur in
immense numbers in tropical seas, and seem to contribute a very large share to the
phosphorescence of these regions. Muller states that as many as twenty thousand
of the above species have been taken in a single haul, and attributes their light-
producing power to the labial papillae which so far appear to be peculiar to the genus.

Codonocera cruenta, Brady. — Plate II., figs. 43 to 45.

A single specimen of this peculiar ostracod was taken in the collection made by
Professor Herdman on the homeward journey when west of Minikoi, in the Indian
Ocean. In Professor Brady's description of the animal it is stated that the post-
abdominal lamina? have only three ungues. In the present specimen there are
distinctly four, the fourth, however, being very small (Plate II., fig. 43). The
specimen was a male. The peculiar filaments on the antennules ending in bell-shaped
disks (fig. 44) and the muscular hand of the antenna (fig. 45) are noteworthy
characters; size, 173 millims.

Professor Brady's single specimen was taken at Pulo Penang.

Family: SARSIELLIDyE.

Sarsiella ornithoides, Brady.
One specimen of this distinct form was found in the collection made at Karativo
Paar on March 10th, 1902.

The only other locality for the species is Trincomalee.

3(58 CEYLON PEARL OYSTER REPORT.

Sarsiella carinata, n. sp. — Plate I., figs. 1 and 2 ; Plate II., figs. 40 and 41.

Shell membranous, thin and flexible, seen from the side subrhomboidal, height
equal to about two-thirds of the length. Anterior extremity truncate, and provided
with a prominent protuberance at the beginning of the dorsal curve ; posterior
narrower, deeply excavated, bounded above and below by prominent projections ;
dorsal margin boldly arched, highest behind the middle, whence it slopes rapidly
towards the large projection ; ventral margin nearly flat, except at the posterior end,
where it is deeply sinuated ; seen from above elongated, widest in the middle, width
considerably less than half the length ; anterior extremity narrow, emarginate ;
posterior wide, terminating in one median and two lateral protuberances. The dorsal
and ventral margins are in the form of a continuous ridge with radiating lines, two
longitudinal ribs of a similar structure run parallel to, but at a considerable distance
within the margins, both terminating posteriorly in protuberances ; ventral margin
adorned with four protuberances of various sizes ; in addition to these, there are two
between the ventral margin and the lower ridge and one on the surface of the lower
posterior projection. Surface of the shell marked with numerous circular depressions,
which have four or five short setae round the outside ; dorsal and ventral margins
ciliated; size, l'l millims.

Three specimens, all males, were found in washings from deep-water dredgings off
Galle. The antennules have the characteristic dense brush of long fine hairs ; post-
abdomen with five marginal ungues, increasing in length from the first to the last,
which is about five times as long as the first ; margins spinulose.

Sarsiella gracilis, n. sp. — Plate I., figs. 3 and 4 ; Plate II., fig. 37.

Shell membranous, thin and flexible ; seen from the side subcircular, height equal
to fully two-thirds of the length ; anterior extremity broadly rounded, crenate, with
a distinct notch near the middle ; posterior truncate, with a wide triangular process
above and below ; dorsal and ventral margins rounded, the former sloping rapidly at
its posterior end ; seen from above, oblong subovate, widest near the posterior end,
width slightly less than one-third of the length ; anterior extremity obtusely rounded,
posterior truncated, with a median process ; surface of the shell slightly ciliated,
covered with small impressed puncta ; anterior and ventral margins ciliated ; size,
l-34 millims.

Two specimens were found in a dredging from 100 fathoms, off Galle, and one in
the general washings from invertebrates from Gulf of Manaar. All were mature
females with ova. The postabdomen has three stout ungues with spinulose margins,
and three short plumose setae.

Sarsiella similis, n. sp. — Plate I., figs. 5 and 6 ; Plate 11., fig. 38.

Shell thin and flexible, subcircular, height fully two-thirds of the length ; seen
from the side anterior extremity rounded, crenate, with a slight notch in the centre,

OSTRACODA! 369

posterior truncate, with distinct projections above and below ; dorsal and ventral
margins boldly convex; seen from above subovate, fully twice as long as broad,
anterior extremity narrow, posterior widely truncate, with a small median process ;
for the greater part the sides arc nearly parallel, converging rapidly towards the
anterior end: surface of the shell slightly ciliated, covered with small impressed
puncta, the valves have four or five longitudinal ribs at the posterior and a number
of radiating lines af the anterior extremity; anterior and ventral margins ciliated;
size, 0'93 millim.

A number of ova-bearing females were found in washings from Muttuvaratu Paar
and in the general washings from Gulf of Manaar. The species resembles the last in
general shape, but is distinguished by the longitudinal ribs. The postabdomen
(Plate II., rig. 38) is much narrower, and the ungues more spinulose.

Sarsiella crispata, n. sp. — Plate I., figs. 7 and 8 ; Plate II., fig. 39.

Shell thin and flexible; seen from the side subcircular, broadly rounded in front,
slightly narrowed and produced into a wide ciliated beak posteriorly ; seen from above
broadly ovate, widest just behind the middle, twice as long as broad ; anterior
extremity narrow, posterior truncate, with one median and two lateral projections ;
surface covered with large impressed puncta, valves with a number of conspicuous
ridges, the two near the posterior end of the dorsal margin forming distinct projec-
tions, anterior and ventral margins crenate and ciliated, with a corrugated line a little
within ; size, 0*8 millim.

About a dozen mature females were obtained in the same material as the previous
species. Postabdomen (fig. 39) with four stout spinulose ungues of different lengths.

Sarsiella tumida, n. sp. — Plate I., figs. 9 and 10 ; Plate II., fig. 42.

Shell membranous, thin and flexible. Seen from the side subcircular, slightly
longer than wide; anterior extremity broadly rounded, with a small beak in the
centre ; posterior truncate and produced into a wide beak at the lower angle ; dorsal
and ventral margins boldly convex, the dorsal forming a distinct hump in the centre ;
seen from above broadly ovate, widest slightly in front of the middle, width equal to
less than two-thirds of the length, narrow in front, rectangularly truncate behind.
with a distinct median projection ; surface of the shell devoid of sculpture, but having
a large fold near the posterior end of the dorsal margin : anterior and ventral margins
ciliated: size, 1/26 millims.

Two mature females of this species were found in the general washings from
dredged invertebrates. The postabdomen is narrow and furnished with five spinulose
ungues, the longest being about five aud a half times the length of the shortest.

Future investigation may show this form to be the female of Scwsiella carinata,
as the comparative lengths of the ungues on the postabdomen of both species art
nearly alike, but in the meantime it is thought best to keep them separate.

3 E

370 CEYLON PEARL OYSTER REPORT.

Family: HALOCYPRIDiE.

Conchcecia magna, Claus.
Two specimens belonging to this species were taken in the tow-netting collected in
the Suez Canal, between Port Said and Suez, on the outward journey.

Conchcecia clausii, G. 0. Sars (?).
Two specimens apparently belonging to this species occurred in the above collection,
but their very poor condition makes their identity uncertain.

Conchoecetta acuminata, Claus.
One specimen in the above collection, one from near Perim Island in the Red Sea,
and a third from the middle of the Indian Ocean.

Halocypris concha, Claus.
This species occurred in a number of the collections taken between the Mediter-
ranean and Ceylon on both journeys, and also on the pearl banks in the Gulf of
Manaar.

Section II.: PODOCOPA.
Family: CYPRID^.

Macrocypris decora (Brady).
In general washings from invertebrata and in washings from sponges dredged in
the Gulf of Manaar.

Macrocypris orientalis, Brady.
In general washings from invertebrata from the Gulf of Manaar.

Macrocypris similis, Brady.
Also in general washings from invertebrata from the Gulf of Manaar.

Macrocypris maculata (Brady).
In the general washings from the Gulf of Manaar and in washings from Cheval
pearl oysters.

Pontocypris robusta, n. sp. — Plate L, figs. 17 and 18.
Shell seen from the side subtriangular, the height being equal to slightly more
than half the length, anterior extremity moderately broad and obliquely rounded,
posterior attenuated and almost acuminate ; dorsal margin boldly arched, highest in
the middle, sloping with a steep curve backwards, and more gently towards the front ;
ventral margin deeply sinuated well in front of the middle. Seen from above the

OSTRACODA. 371

outline is ovate, acuminate in front and rounded behind, greatest width in the middle
and equal to fully two-fifths of the length ; shell white, polished, with minute setae
and impressed puncta ; the valves are marked with five lucid spots arranged in a
semicircular manner; size, 0-82 millim.

A few specimens from the general washings from Gulf of Manaar ; from Cheval
pearl oyster washings ; and from Gulf of Manaar sponges.

At first it was thought this might be a form of Pontocypris trigonella, G. 0. Sars,
but a comparison with the figures given in the ' " Challenger" Report,' in the report
' On the Entomostraca from the Gulf of Guinea,' by my father, and in the ' Mono-
graph on the North Atlantic and North-western European Ostracoda,' by Brady and
Norman, shows that it is distinct.

Pontocypris elegans, n. sp. — Plate I., figs. 19 and 20.

Shell seen from the side oblong, compressed, subreniform, the height being equal
to rather less than one-third of the length ; anterior extremity obliquely rounded,
posterior produced and subacute at the ventral angle ; dorsal margin moderately
arched, greatest height slightly in front of the middle, and sloping gently towards
each extremity ; ventral margin deeply sinuated in the middle ; seen from above
compressed, ovate, widest in the middle, width about one-fourth of the length,
acuminate in front and slightly rounded behind ; shell white, polished, with a few
minute impressed puncta ; the valves are marked with a circular patch of lucid spots
and three smaller detached ones; size, 0-6 millim.

A few specimens were found in the same washings as the previous species.

Pontocypris rostrata, n. sp. — Plate I., figs. 21 and 22.

Shell seen from the side subtriangular, greatest height nearly equal to half the
length ; anterior extremity broadly rounded, almost truncate, posterior produced,
subacute in the centre ; dorsal margin boldly arched, highest slightly in front of the
middle, sloping very gently towards the front, but rapidly to the posterior ; ventral
margin sinuated in front of the middle, becoming convex near the posterior end where
it turns up to join the extremity ; seen from above ovate, greatest width in the
middle and about equal to one-third of the length, acuminate in front and behind,
sides distinctly compressed in front, forming a beak-like process ; shell white and
polished, valves unequal, the right being slightly smaller than the left ; size,
0-64 millim.

In washings from sponges dredged in the Gulf of Manaar.

Pontocypris tumida, n. sp. — Plate I., figs. 30 and 31.

Shell seen from the side subovate, the height being rather more than half the
length ; anterior extremity rounded, somewhat depressed, posterior subacute ; dorsal
margin boldly arched, highest in the middle, ventral slightly sinuated in front of the
middle ; seen from above ovate, widest behind the middle, length equal to two and

3 B 2

372 CEYLON PEARL OYSTER REPORT.

a half times the width, right valve smaller than the left ; shell white, smooth, and
polished ; size, 0"97 millim.

In general washings and in washings from Gulf of Manaar sponges.

Erythrocypris herdmani, n. sp. — Plate I., figs. 15 and 16.

Shell seen from the side suhtriangular, the height being about equal to half of the
length ; anterior extremity broadly rounded, posterior attenuated and subacuminate ;
dorsal margin strongly arched, almost angular at its highest point, forming a distinct
hump, greatest height considerably in front of the middle, and sloping with a steep
curve to both extremities, ventral margin slightly sinuated ; seen from above ovate,
width fully two-fifths of the length, greatest width about one-third from the anterior
extremity, extremities obtusely pointed, rather more acute in front than behind ;
valves yellowish, smooth, and shining, covered with numerous short rigid hairs, the
left valve has a distinct tooth at its posterior end ; size, 0'85 millim.

Specimens were found in general washings from Gulf of Manaar, in washings from
Cheval pearl oysters, and Manaar sponges, and on Karativo Paar. This species,
which is easily distinguished from any of those already described, is named in
compliment to Professor Herdman, whose labours on the Ceylon pearl banks have
added an extensive chapter to our knowledge of the tropical marine fauna.

Bairdia villosa, Brady.
In general washings from invertebrates from the Gulf of Manaar.

Bairdia attenuata, Brady.
In general washings from invertebrates from the Gulf of Manaar.

Bairdia woodwardiana, Brady.
In general washings from invertebrates from the Gulf of Manaar.

Eairdia amygdaloides, Brady.
In general washings from invertebrates from the Gulf of Manaar.

Bairdia faveolata, Brady.

In the general washings and in washings from Cheval pearl oysters.

Bairdia hirsuta, Brady.

In general washings, in washings from Cheval pearl oysters, and from Karativo
Paar.

All these species of Bairdia are new to the fauna of Ceylon.

Bairdia inomata, n. sp. — Plate I., figs. 11 and L2.
Shell seen from the side subreniform, height equal to more than half the length ;

OSTRACOPA. 373

anterior extivmil y obliquely rounded, posterior produced in the middle into an
obtusely angular beak: dorsal margin boldly arched, highest in the middle and
sloping steeply towards the extremities, slightly sinnated at the anterior and posterior
ends, ventral margin deeply sinuated in the middle ; seen from above compressed,
ovate, widest in front of the middle and nearly three times longer than wide,
extremities obtuse, subtruncate ; colour almost black, with one or two lighter hands ;
surface of the valves covered with closely set, short, stiff setae ; posterior end of the
ventral margin finely serrate ; right valve smaller than the left ; size, 0'74 millim.
A few specimens in general washings from invertebrates from Gulf of Manaar.

Bairdia robusta, n. sp. — Plate I., figs. 13 and 14.

Shell seen from the side subtriangular, height equal to fully two-thirds of the
length; anterior extremity obliquely rounded, posterior produced in the middle into
an obtusely angular beak ; dorsal margin greatly arched, highest slightly in front of
the middle, ventral nearly straight ; seen from above broadly ovate, length f'ullv
one and a half times the width, extremities subacute, posterior more so than the
anterior; shell white, polished, closely beset with black setae; anterior and posterior
extremities with a dense fringe of setaa ; size, 0'87 millim.

In general washings from invertebrates from Gulf of Manaar. This species
resembles Bairdia villosa, Brady, but is much more tumid and obtuse.

Anchistrocheles bradyi, n. sp. — Plate I., figs. 34 and 35.

Shell seen from the side reniform, height slightly less than half the length ;
anterior extremity wide, obliquely subtruncate. posterior broadly rounded, slightly
produced in the middle ; dorsal margin gently and evenly curved, highest behind the
middle, ventral sinuated in front of the middle ; seen from above the outline is
elongated, with parallel sides, nearly four times longer than wide ; both extremities
acute, anterior more so than the posterior; shell smooth, thin and fragile; size,
0-9 millim.

In dredged material from 6 fathoms to 10 fathoms, Karativo Paar.

The species resembles Anchistrocheles fumata, Brady, in general appearance, but
is easily recognised by the more acute extremities, as seen from the dorsal aspect. I
have great pleasure in naming it after Professor Brady, to whom we remain indebted
for much help both in this group and in the Copepoda.

Pseudocythere minuta, n. sp. — Plate I., figs. 28 and 29.

Shell seen from the side oblong, subquadrate, slightly higher in front than behind,
height slightly less than half the length ; anterior extremity rounded, with one or
two toothdike projections near the middle ; posterior oblique, subtruncate, much
compressed, produced at the upper angle into a broad, blunt beak, lower angle also
produced, but the beak is much smaller ; dorsal margin almost flat, highest near the

374 CEYLON PEARL OYSTER REPORT.

anterior .end, sloping very slightly towards the posterior, ventral margin slightly
sinuated in front of the middle ; seen from above ovate, widest in the middle,
tapering to the extremities, both of which are acuminate, the posterior much
attenuated ; width equal to less than half the length ; shell marked with numerous
puncta arranged in fairly regular, slightly curved, longitudinal rows. Size, 0'35
millim.

A single specimen in general washings from dredged invertebrates from Gulf of
Manaar.

Paradoxostoma cingalense, Brady.
One or two specimens in general washings from dredged invertebrates from Gulf of
Manaar. Off Kalpentyn is the only previous locality given lor the species.

Paradoxostoma attenuatum, n. sp.- — Plate I., figs. 32 and 33.

Seen from the side pear-shaped, about two and a half times longer than broad ;
anterior extremity very narrow, rounded, posterior broad and truncate, sloping
inwards, forming a distinct tooth at the junction with the lower margin ; dorsal
margin boldly arched, highest considerably behind the middle, ventral margin sloping
outwards from the anterior extremity ; very slightly sinuated, convex at the widest
part ; seen from above much compressed, with acute extremities, greatest width in
the middle, length four and a half times the width ; shell amber coloured, smooth and
transparent; size, 0'6 millim.

In washings from Gidf of Manaar sponges.

Paradoxostoma stebbingi, n. sp. — Plate II., figs. 1 and 2.

Shell seen from the side pear-shaped, two and a half times longer than broad ;
anterior extremity narrow, obliquely rounded, posterior subacute in the centre ;
dorsal margin boldly and evenly arched, highest slightly behind the middle ; ventral
margin sinuated near the anterior extremity ; seen from above compressed, ovate,
widest in the middle, and nearly four times longer than broad, extremities acute ;
shell smooth, white and semi-transparent ; size, 0'6 millim.

In general washings from dredged invertebrates.

I name this species after the Rev. T. R. R. Stebbing.

Xestolebris margaritea, Brady.
In the general washings, in washings from Cheval pearl oysters and Gulf of Manaar
sponges, and in a dredging from Karativo Paar.

Xestolebris tumefacta, Brady.
In the general washings from Gulf of Manaar.

Xestolebris aurantia, Baird.
Also in the general washings. The specimens were first identified as Xestolebris

OSTRACODA. 375

margaritea, but Professor Brady says he cannot distinguish them from our native
Xestolebris aurantia.

Xestolebris varieguta, Brady.
In the general washings, and in washings from Cheval pearl oysters.

Xestolebris faveolata, Brady.
In a dredging from Karativo Paar, 6 to 10 fathoms.

Xestolebris squamigera, n. sp. — Plate I., figs. 23 to 25 ; Plate II., figs. 28 and 29.

Shell seen from the side somewhat siliquose, two and a half times longer than
broad. Anterior extremity very narrow, rounded, posterior subacute ; dorsal margin
boldly arched, highest a little behind the middle, sloping rapidly and evenly to the
anterior end ; posteriorly the slope is more abrupt and almost truncate ; ventral
margin nearly straight, slightly sinuated in front of the middle ; seen from above the
outline is broadly ovate ; width about three-fifths of the length, and widest behind
the middle ; anterior extremity subacute, distinctly sinuated, and expanding rapidly ;
posterior extremity broadly rounded ; seen from the posterior end the valves are
deeply concave ; valves slightly unequal ; shell smooth, white and polished, with
numerous whitish spots scattered over the surface ; size, 0-52 millim.

In the general washings and in washings from Gulf of Manaar sponges. The
drawings on Plate I. represent an ova-bearing female ; the antennule and antenna are
shown on Plate II., figs. 48 and 49.

Xestolebris irrasa, n. sp. — Plate II., figs. 5, 6 and 46, 47.

Shell seen from the side ovate, height fully two-thirds of the length ; anterior
extremity narrow, posterior broad, both well rounded ; dorsal margin boldly arched,
highest in the middle, ventral very slightly convex, with slight sinuations at the
extremities ; seen from above ovate, widest behind the middle, width fully two-thirds
of the length, extremities obtusely pointed in front, wider and more rounded behind ;
surface of the shell marked with numerous concentric spinulose ridges, and clothed
with short stiff setae ; colour yellowish ; size, 0*53 millim.

In the general washings and in washings from Gulf of Manaar sponges. The
figures represent a female shell, the antennule and antenna being shown by figs. 46
and 47.

Xestolebris tumida, n. sp. — Plate II., figs. 3 and 4.
Shell seen from the side broadly pear-shaped, scarcely one and a half times longer
than broad, extremities well rounded, anterior much narrower than the posterior ;
dorsal margin greatly arched, highest behind the middle ; ventral slightly sinuated in
front and convex behind ; seen from above broadly ovate, rather longer than broad,
compressed and pointed in front, broadly rounded behind ; greatest width behind the

9,76 CEYLON PEAEL OYSTKK EEPOPT.

middle; shell smooth, white, with numerous whiter spots on its surface; size,
0'6 millim.

In general washings from the Gulf of Manaar.

The species resembles Xestolebris variegata, but is much more tumid.

Paracytheridea perplexa, n. sp. — Plate L, figs. 26 and 27.

Shell seen from the side subrhomboidal, two and a half times longer than broad ;
anterior extremity obliquely rounded, with three median teeth ; posterior produced
into a median triangular beak ; dorsal margin highest in front, sloping downwards
towards the posterior in an irregularly sinuous line ; ventral margin slightly convex
in front, rising with a gentle slope posteriorly, deeply notched at the hinder end, and
then suddenly produced into a curved triangular tooth-like process ; seen from above
the outline is much like that of a Trilobite, very wide and rounded in front ; narrowing
slightly posteriorly, then rapidly converging to form a wide triangular end ; greatest
width much in front of the middle, and equal to three-fourths of the length ; surface
of the shell marked with large puncta, and in side view with obliquely transverse
ridges; size, 0'65 millim.

A few specimens in the general washings from dredged invertebrates from Gulf of
Manaar.

The species resembles Paracytheridea depressa, G. W. Muller, but differs in the
termination of the lateral margins as seen from above. In P. depressa the margins
end in a distinct tooth projecting at nearly right angles to the sides.

Cytherura concinna, n. sp.- — Plate II., figs. 7 and 8.

Shell seen from the side subrhomboidal, with a well-marked dorsal ridge, height
equal to half the length ; anterior extremity obliquely rounded, posterior broadly
beaked above the middle ; dorsal margin rugged, broken by a few small blunt
projections, sloping gently upwards towards the posterior, ventral convex in the
middle ; seen from above broadly triangular, greatest width much behind the middle
and equal to two-thirds of the length ; anterior extremity rounded and produced into
a blunt median beak; posterior very wide, subtruncate, and produced into a large
median process, its margin forming a distinct rlexuous ridge coursing between the
outer and inner margins of each valve ; surface of the shell marked with numerous
puncta, and with a distinct depression in the centre of each valve; size, 0"53 millim.

One or two specimens in general washings from dredged invertebrates from Gulf of
Manaar.

Loxoconcha anomala, Brady.

In general washings, in washings from Cheval pearl oysters, and in a dredging
from Karativo Paar.

OSTKACODA. 377

Loxoconcha alata, Brady.
From the same material as the previous species.

Loxoconcha papillosa, Brady.
In general washings from Gulf of Manaar.

Loxoconcha sculpta, Brady.
In washings from Gulf of Manaar sponges.

Loxoconcha australis, Brady.
In the general washings and in washings from Gulf of Manaar sponges.

Cythere bimammillata, Brady.
In washings from Gulf of Manaar sponges.

Cythere darwini, Brady.

In the general washings, in washings from Gulf of Manaar sponges, and in a
dredging from Karativo Paar.

Cythere inconspicua, Brady.
In washings from Cheval pearl oysters.

Cythere ovalis, Brady.

In the general washings, in washings from Cheval pearl oysters, and in a dredging
from Karativo Paar.

Cythere polytrema, Brady.
In general washings from Gulf of Manaar.

Cythere rectangularis, Brady.
In general washings from Gulf of Manaar.

Cythere ruperti, Brady.
In the general washings and in washings from Gulf of Manaar sponges.

Cythere stimpsoni, Brady.
In a dredging from Karativo Paar.

Cythere subcuneata, Brady.
In general washings from Gulf of Manaar.

Cythere knoxi, n. sp. — Plate II., figs. 9 and 10.

Shell seen from the side elongated, subsigmoid, height equal to half the length,

3 c

378 CEYLON PEARL OYSTER REPORT.

extremities toothed ; anterior extremity broad and obliquely rounded, posterior
narrower, rounded off below, and obscurely angulated above ; dorsal margin sinuated
in the centre, highest in front and sloping towards the posterior, ventral sinuated in
the middle ; seen from above pear-shaped, greatest width near the posterior end,
and equal to more than half the length, the outline between the widest point and
the anterior end shows two distinct constrictions ; shell surface marked with coarse
impressed puncta ; each valve seen from the side shows two distinct grooves ; size,
0-56 millim.

In the general washings and in washings from Cheval pearl oysters.

Named after Captain Robert Knox, 20 years a captive in Ceylon in the
seventeenth century.

Cythere chalmersi, n. sp.— Plate II., figs. 11 and 12.

Shell seen from the side subsigmoid, height equal to half the length, extremities
corrugated but not toothed ; anterior extremity very obliquely rounded, posterior
narrower, and broadly rounded ; dorsal margin highest at the anterior end, nearly
flat, with a gentle slope towards the posterior ; ventral sinuated in the middle ; seen
from above pear-shaped, outline irregular, and marked by three constrictions, widest
near the posterior end, greatest width equal to about half the length ; anterior and
posterior ends with strong thickened lips, the posterior forming a distinct pro-
tuberance ; surface of the shell marked with moderately coarse impressed puncta,
and near the extremities with a row of circular depressions ; seen from the side, the
valves show three distinct grooves ; size, 07 millim.

In the general washings and in washings from Gulf of Manaar sponges.

Named after Dr. A. J. Chalmers, Professor in the M edical College, Colombo.

Cythere imthurni, n. sp. — Plate II., figs. 13 and 14.

Shell seen from the side oblong, subquadrangular, height slightly less than halt
the length ; extremities smooth, anterior extremely broad and obliquely rounded,
with an internal row of subcircular markings ; posterior narrower, rounded above and
truncate below ; dorsal margin highest at the anterior end, nearly flat, with a gentle
slope towards the posterior, ventral deeply sinuated near the middle ; seen from
above rather wedge shaped, with nearly parallel sides, deeply constricted near the
posterior end, length about two and a half times the width ; anterior margin
acuminate with projecting thickened lips, posterior almost truncated, projecting
slightly in the centre ; surface of the shell marked with large irregular impressed
puncta, and, when • viewed from the side, with a deep hollow near the posterior ;
size, 0-5 millim.

In the general washings, in washings from Cheval pearl oysters, and from Gulf ot
Manaar sponges.

Named after the Honourable E. F. im Thurn, Lieutenant-Governor of Ceylon
during the pearl oyster investigation in 1902.

OSTRACODA. 379

Cythere thompsoni, 11. sp. — Plate II., figs. 15 and 16.

Shell seen from the side ohlong, subquadrangular, height rather more than half
the length ; anterior extremity wide and obliquely rounded, with about a dozen short
thick teeth, posterior slightly narrower, almost truncate, with five or six short stout
teeth ; dorsaJ margin deeply sinuated in the middle, highest at the anterior end,
ventral margin slightly sinuated near the anterior end ; seen from above broadly
ovate, with irregular outline, width rather more than half the length, widest near the
posterior end ; anterior extremity obtuse, with two mucronate processes, posterior
wide, truncated, with broad, tooth-like median projections ; surface of the shell
marked with large regular ridges and deep depressions, which become very conspicuous
in dead shells; size, 0'73 millim.

In the general washings and in washings from Gulf of Manaar sponges.

This well-marked species is named in memory of my friend and fellow-worker, the
late Mr. Isaac C. Thompson, F.L.S., who was jointly responsible with me in the
preparation of the " Report on the Copepoda," published in the first volume of this
work.

Cythere donnani, n. sp. — Plate II., figs. 17 and 18.

Shell seen from the side subsigmoid, height equal to half the length ; anterior
extremity wide and very obliquely rounded, with about fourteen small teeth on the
lower margin, posterior narrow, slightly produced in the centre, with four small teeth
below the middle ; dorsal margin sinuated in the middle, highest at the anterior and
rounded off towards the posterior, ventral sinuated, convex in front and behind ; seen
from above broadly ovate, width slightly less than half the length, with rounded
extremities, slightly produced into thickened lips, widest near the posterior, and
slightly sinuated in the middle ; surface of the shell marked with a number of fairly
regular longitudinal ridges and rows of impressed puncta ; size, 0'(J8 millim.

In washings from Cheval pearl oysters, and from Gulf of Manaar sponges.

Named after Captain Donnan, Inspector of the Pearl Banks in 1902.

Cythere willeyi, n. sp. — Plate II., figs. 19 and 20.

Shell seen from the side oblong, subquadrangular, height equal to rather more than
half the length ; anterior extremity wide and very obliquely rounded, with about
seventeen short stout teeth, posterior much narrower, bluntly rounded, with four
short teeth on the lower margin ; dorsal margin very slightly sinuated in the middle,
highest in front, and rounding oft1 posteriorly, ventral sinuated slightly in front of
the middle ; seen from above broadly ovate, with irregular outline, width rather
more than half the length, widest in front "f the middle ; anterior extremity obtuse,
with two mucronate processes, posterior widely triangular ; surface of the shell
marked with large irregular puncta, and a distinct dorsal ridge ; size, 0"8 millim.

In general washings from Gulf of Manaar.

Named after Dr. Arthur Willey, Director of the Museum at Colombo.

3 c 2

380 CEYLON PEARL OYSTER REPORT.

Cythere hornelli, n. sp. — Plate II., figs. 21 and 22.

Shell seen from the side siliquose, height rather less than half the length ; anterior
extremities narrowed and rounded, smooth ; dorsal margin boldly arched, highest in
the centre, ventral sinuated in the middle ; seen from above wedge-shaped, width less
than half the length, widest near the posterior ; anterior extremity produced, with two
mucronate processes, posterior margins truncate, and produced into a bi'oad triangular
projection ; surface of the shell marked with a number of curved ridges and rows of
impressed puncta ; size, 0-53 millim.

In general washings from invertebrata from Gulf of Manaar. Named after
Mr. James Hornell, now Inspector of the Pearl Banks.

Cythere halyi, n. sp. — -Plate II., figs. 23 and 24.

Shell seen from the side oblong, quadrangular, height slightly less than half the
length, widest in the middle ; anterior extremity broadly rounded, with one tooth
below the centre ; posterior produced into a wide triangular beak ; dorsal margin flat,
with a corrugated margin, highest at the extremities, ventral slightly convex and
irregular ; seen from above narrowly ovate, with produced and thickened extremities,
widest behind the middle, width considerably less than half the length ; surface of
the shell marked with numerous impressed puncta and studded with bluntly pointed
spines ; size, 0"55 millim.

In general washings from invertebrata from Gulf ot Manaar. Named after
Mr. Haly, a former Director of the Colombo Museum.

Cythere kelaarti, n. sp. — Plate II., figs. 25 and 26.

Shell seen from the side oblong, subquadrangular, height scarcely equal to half
the length ; anterior extremity broadly and obliquely rounded, with about 21 short,
stout teeth ; posterior narrow and subtruncate, with seven short, thick spines ;
dorsal margin slightly sinuated, highest in front and sloping gently backwards, ventral
nearly straight, much contracted at the posterior ; seen from above doubly triangular,
widest much behind the middle, where the outline is produced into a blunt tooth ;
width slightly more than half the length ; anterior extremity with two, and the
posterior with four, spines ; surface of the shell marked with numerous impressed
puncta, very rough and studded with short, aculeate spines, one triangular tooth near
the posterior; size, 0'65 millim.

In general washings from dredged invertebrates from Gulf of Manaar. Named
after Dr. Kelaart, who investigated the pearl banks in 1857.

Cythere willisi, n. sp. — Plate II., figs. 27 and 28.

Shell seen from the side subquadrangular, with compressed margins, height slightly
less than two-thirds of the length ; anterior extremity broad and very obliquely rounded,

OSTRACODA. 381

with seven short, stout teeth on the lower portion ; posterior almost rectangularly
truncated, produced below the middle, where it bears two teeth and one or two
crenulations ; dorsal margin highest in front, sloping with a sinuous curve gently
backwards ; ventral sinuated in front and convex behind ; seen from above ovate,
with produced emarginate extremities, sharply constricted in the middle and near the
posterior end, width equal to half the length ; surface of the shell covered with large
irregularly angulated fossse, marked with a strong marginal ridge and an obliquely
transverse one, coursing from the posterior towards the anterior extremity ; highest
at the posterior end ; size, 073 millim.

In the general washings, in washings from Gulf of Manaar sponges, and in a
dredging from Karativo Paar. Named after Dr. J. C. Willis, Director of the
celebrated Botanic Gardens at Peradeniya, Ceylon.

Cythgre colletti, n. sp. — Plate II., figs. 29 and 30.

Shell seen from above subquadrangular with compressed margins, height equal to
slightly less than two-thirds of the length ; anterior extremity broad and obliquely
rounded, with 10 short, stout teeth on the lower portion ; posterior narrowed,
truncated, produced near the middle into a quadri-crenulate projection ; dorsal margin
highest in front, sloping rather steeply to the posterior ; ventral slightly sinuated in
front, convex behind the middle, rising quickly near the posterior, where there is a
strong, stout incurved tooth ; seen from above doubly triangular, widest slightly
behind the middle, width rather more than half the length, the margins near the
posterior end are produced into strong teeth ; anterior extremity blunt and produced
into the mucronate processes, posterior truncate ; surface of the shell marked with
curved ridges and rows of impressed puncta ; each valve has a strong tooth near the
lower margin of the posterior end ; size, 0"5 millim.

In washings from Gulf of Manaar sponges.

Named after the late Mr. Oliver Collett, a well-known naturalist in Ceylon.

Cythere holds worthi, n. sp. — Plate II., figs. 31 and 32.

Shell seen from the side subrhomboidal, height equal to fully two-thirds of the
length ; anterior extremity broad and obliquely rounded, posterior truncated, with a
small projection in the centre ; dorsal margin boldly arched, highest in the middle ;
ventral convex in the middle ; seen from above broadly ovate, widest near the
middle, width equal to two-thirds of the length ; anterior extremity narrowed and
bluntly rounded ; posterior broadly rounded, almost truncate ; surface of the shell
marked with concentric rows of impressed puncta ; size, 0*5 millim.

In general washings from dredged invertebrata from Gulf of Manaar.

Named after Mr. Holds worth, the naturalist who investigated the pearl banks in
1868.

382 CEYLON PEARL OYSTER REPORT.

Section III. : PLATYCOPA.
Family : CYTHERELLID^E.

Cytherella ondaatjei, n. sp. — Plate II., tigs. 33 and 34.

Shell seen from the side subquadrangular, height rather less than two-thirds of
the length ; anterior extremity broadly rounded, posterior obliquely rounded, sloping
steeply inwards ; dorsal margin slightly arched, highest behind the middle, ventral
deeply sinuated in the middle ; seen from above elongate narrow, widest behind the
middle, width equal to one-third of the length ; anterior extremity bluntly rounded,
posterior almost truncated, margins much hollowed above the middle ; surface of the
shell marked with numerous shining spots, and a conspicuous triangular groove ;
covered with fine spinules, anterior extremity and ventral margin finely spinulate ;
size, 0"6 millim.

In a dredging from Karativo Paar.

Named after Dr. Ondaatje, a former naturalist and collector in Ceylon.

Cytherella vraspillaii, n. sp. — Plate II., figs. 35 and 36.

Shell seen from the side subelliptical, height less than two-thirds of the length ;
valves unequal, left valve larger than the right ; anterior extremity rounded,
posterior rounded and slightly produced in the centre ; dorsal margin slightly arched,
highest in the centre ; ventral flattened, slightly sinuated ; seen from above pear
shaped, widest behind the middle, width slightly less than two-thirds of the length,
obtusely pointed in front, broadly rounded behind, lateral margins boldly convex ;
surface of the shell smooth and polished; size, 0'55 millim.

In general washings from dredged material from Gulf of Manaar.

This species is named in honour of Mr. V. Vbaspillai, the well-known Adigar of
Musali, who has rendered able service in connection with the pearl banks for many
years.

OSTRACODA. 383

EXPLANATION OF PLATES.

PLATE I.

Fig. I. Sarsiella carinata, n. sp., from right side.

„ 2. „ „ „ above, x 45.

,, 3. Sarsiella gracilis, a. sp., from right side.

„ i. ,, „ „ aliove. x 36.

„ 5. Sarsiella similis, n. sp., from right aide.

„ 6. „ „ „ above, x 45.

,, 7. Sarsiella crispata, n. sp., from right side.

,, 's. „ „ „ above, x 57.

„ 9. Sarsielln timtidu, n. sp., from right size.

,, 10. „ „ „ above, x 37.

„ 11. Pmirdia inornata, n. sp., from right side.

,, 12. „ „ „ above, x 62.

., 13. Bairdia robusta, n. sp., from right side.

,, 14. „ ,, ,, above, x 44.

,, 15. Erythrocypris herd/mam, n. sp., from right side.

„ 16. ,, ,, ,, above. x50'8.

,, 17. l'<inio(ii[nU ivbusta, n. sp., from right side.

„ 18. „ ,, ,, above. x55-5.

„ 19. Pontocypris elegans, n. sp., from right side.

„ 20. ,, ,, „ above, x 74.

„ 21. Pontocypris rostrata, n. sp., from right side.

,, 22. „ „ ,, above, x 71.

,, 23. Xestolebris squamigera, n. sp., from right side.

„ 24. „ ,, ,, above, x 90.

„ 25. „ ,, ,, posterior end.

., 26. Paracytlieridaza perplexa, n. sp., from above.

„ 27. „ „ „ right side, x 71.

,, 28. Pseudocythere minuta, n. sp., from right side.

,, 29. ,, „ ,, above, x 134.

,, 30. Pontocypris tumida, n. sp., from right side.

,, 31. „ „ „ above, x 45.

,, 32. Pamdvroxtotm titkauiitinii, n. sp., from right side.

,, 33. ,, „ „ above, x 74.

,, 34. Anchistrocheles bradyi, n. sp., from right side.

„ 35. ,, ,, „ above, x 45.

PLATE II.

Fig. 1. Paradoxostoma stebbingi, n. sp., from right side.

„ 2. ,, „ „ above, x 74.

,, 3. Xestolebris tumida, n. sp., from right side.

„ 4. „ ,, ,, above, x 72.

,, 5. Xestolebris irrasa, n. sp., from right side.

„ 6. „ „ „ above, x 72.

384 CEYLON PEARL OYSTER REPORT.

Fig. 7. Cytherwa amcinna, n. sp., from above.
,, 8. „ ,, ,, right side, x 72.

,, 9. Cythere hum, n. sp., from right side.

„ 10. „ „ „ above, x 72.

,, 11. Cythere chalmersi, n. sp., from right side.

„ 12. „ „ „ above, x 61.

„ 13. Cythere iiiUhurni, n. sp., from right side.

„ 14. „ „ „ above, x 76.

,, 15. Cythere thompsoni, n. sp., from right side.

„ 16. „ „ „ above, x 60.

,, 17. Cythere donnani, n. sp., from right side.

„ 18. „ „ „ above. x37-5.

,, 19. Cythere willeyi, n. sp., from right side.

,, 20. „ ,, ,, above, x 46.

,, 21. Cythere hornelli, n. sp., from right side.

„ 22. „ ,, „ above, x 72.

,, 23. Cythere halyi, n. sp., from right side.

„ 24. ,, ,, „ above, x 60.

,, 25. Cythere kelaarti, n. sp., from right side.

„ 26. ,, „ „ above, x 60.

,, 27. Cythere vMlisi, n. sp., from right side.

„ 28. „ „ „ above, x 47.

,, 29. Cythere colletti, n. sp., from right side.

„ 30. „ ,, „ above, x 72.

„ 31. Cythere holdsworthi, n. sp., from right side.

„ 32. ,, ,, „ above, x 70.

„ 33. ( 'ythirella ondaatjei, n. sp., from right side.

,, 34. ,, ,, ,, above, x 60.

,, 35. Cytherella waspillaii, n. sp., from right side.

,, 36. ,, „ ,, above, x 60.

„ 37. Sarsu lla gracilis, postabdomen. x 70.

„ 38. Sarsiella similis, „ x 120.

„ 39. Sarsiella crispata, „ x 70.

,, 40. Sarsiella carinata, „ x 70.

„ 41. ,, „ antennule. x 70.

„ 42. Sarsiella tumida, postabdomen. x 70.

„ 43. Codonocera cruenta, postabdomen. x 70.

,, 44. „ modified seta of antennule. xllO.

„ 45. ,, prehensile branch of antenna, x 70.

,, 46. Xestolebris irrasa, antennule. x 260.

,, 47. „ antenna, x 156.

„ 48. Xestolebris squamigera, antennule. x 260.

,, 49. ,, antenna, x 156.

HARRISON AND SONS, PRINTERS IN ORDINARY TO HIS MAJESTY, ST. MARTIN'S LANE, LONDON.

CEYLON PEAK I. ( »\ STER REPOR1

OSTRACODA PLATE I.

Eott del

■ ■ T.i-.h Ldin

CEYLON PE \M, OYSTER REPOR1

OSTRACODA, PLATE II.

5

■ •-• • isi::-::-.!;..^-;-^

S---->v- -i

?cott «el

f'I'I irii^t ■ .-

MBL WHOI 1.IBBAKY

II

UH IflTY Z

ioq r

' r

Jt'

0

"■ [

1

: C

VI

•:' '3

V

ft

" j

B

-

:- :2

it

',.;■; I

;' ' '■;■